table of contents 1. revised criteria for insomnia...

Insomnia Disorder

Task Force Review

Table of Contents

1. REVISED Criteria for Insomnia Disorder – REVISED proposed criteria for DSM-5 as edited by APA staff, and Dr. Michael First, compared to DSM-IV criteria

2. Addendum to CPHC Cover Memo – Additional memo from the CPHC, giving the Insomnia proposal the new score of 2 with the addition of the specifiers for co-morbid psychiatric and medical disorders in the criteria 3. Dr. Reynolds' Response to CPHC – The current proposed criteria for DSM-5 include the

recommendations from the CPHC

4. Cover Memo from the CPHC – Cover memo provided by the CPHC after review of the

proposal (see 6 below)

5. Report from the CPHC – Detailed report provided by the CPHC after review of the proposal (see 6 below)

6. CPHC Submission

Cover Memo - CPHC Review of Insomnia Disorder SRC Proposal for Insomnia Disorder -1 Edinger et al_68(10)992 Ohayon and Reynolds, 2009 Ohayon et al., 2011 Sleep 2005 (28) 1049-1057 SRC's Summary Sheet - Proposal 1 SRC Proposal 2 - Insomnia Disorder – Response SRC Summary Sheet - Proposal 2 SRC Scoring Sheet - Proposal 2 SRC Proposal 3 - Insomnia Disorder – Response SRC Summary Sheet - Proposal 3 SRC Scoring Sheet - Proposal 3

The following criterion F from the previous criteria version has been deleted as recommended by Michael First, and the criterion G in the previous version is retained as recommended by the CPHC and is now the new criterion F in this revised version: “The disturbance is not attributable to the physiological effects of a substance or another medical condition, and is not better explained by another mental disorder”

Insomnia Disorder

Sleep-Wake Disorders

Proposed Criteria for DSM-5 Revised September-2012

DSM-IV Criteria

Insomnia Disorder

A. The predominant complaint is dissatisfaction with sleep quantity

or quality, associated with one or more of the following

symptoms:

1. Difficulty initiating sleep (In children, this may be

manifested as difficulty initiating sleep without caregiver

intervention.)

2. Difficulty maintaining sleep, characterized by frequent

awakenings or problems returning to sleep after

awakenings (In children, this may be manifested as

difficulty returning to sleep without caregiver

intervention.)

3. Early-morning awakening with inability to return to

sleep

4. In children, prolonged resistance to going to bed and/or

bedtime struggles

B. The sleep disturbance causes clinically significant distress or

impairment in social, occupational, educational, academic,

behavioral, or other important areas of functioning.

Primary Insomnia

A. The predominant complaint is difficulty initiating or maintaining

sleep, or nonrestorative sleep, for at least 1 month.

B. The sleep disturbance (or associated daytime fatigue) causes

clinically significant distress or impairment in social, occupational, or

other important areas of functioning.

C. The sleep disturbance does not occur exclusively during the course

of Narcolepsy, Breathing-Related Sleep Disorder, Circadian Rhythm

Sleep Disorder, or a Parasomnia.

D. The disturbance does not occur exclusively during the courseof

another mental disorder (e.g., Major Depressive Disorder, Generalized

Anxiety Disorder, a delirium).

E. The disturbance is not due to the direct physiological effects of a

substance (e.g., a drug of abuse, a medication) or a general medical

condition.

The following criterion F from the previous criteria version has been deleted as recommended by Michael First, and the criterion G in the previous version is retained as recommended by the CPHC and is now the new criterion F in this revised version: “The disturbance is not attributable to the physiological effects of a substance or another medical condition, and is not better explained by another mental disorder”

C. The sleep difficulty occurs at least 3 nights per week.

D. The sleep difficulty is present for at least 3 months.

E. The sleep difficulty occurs despite adequate opportunity for

sleep.

F. Frequently coexisting mental and medical disorders do not

adequately explain the predominant complaint of insomnia.

Specify if:

With non-sleep disorder psychiatric comorbidity, including substance

use disorders (Note: code relevant mental disorder)

With other medical comorbidity (Note: code relevant non-mental

disorder)

With other sleep disorder (Note: code relevant sleep disorder)

To: Jack McIntyre, M.D. and Joel Yager, M.D. Chairs, DSM-5 Clinical and Public Health Review Committee (CPHC) FROM: Charles F. Reynolds III, M.D. Chair, DSM-5 Sleep Wake Disorders Workgroup RE: thanks for your review of Insomnia Disorder; and supplemental memorandum, proposed diagnostic criteria set for Hypersomnolence Disorder (aka Hypersomnolence Disorder NEC) DATE: August 24, 2012 CC: David Kupfer, M.D. and Darrel Regier, M.D. First, please accept my thanks, on behalf of the Sleep Wake Disorders workgroup, for the four recommendations offered by the CPHC on 8/23/12 with respect to the proposed diagnostic criteria for Insomnia Disorder. In my opinion your recommendations strengthen the diagnostic criteria for Insomnia Disorder proposed by the workgroup, their clinical utility, and their coherence with the text. Thus, in response to the four CPHC recommendations and following discussion with Drs Kupfer and Regier, I have now:

(1) included specifiers for co-morbid psychiatric and medical disorders in the criteria (and not just in the accompanying text),

(2) moved “descriptive” specifiers (co-morbid conditions) to first place in the list of specifiers to highlight their importance,

(3) changed “may” to “should” in the Descriptive specifier (thus, …”comorbid conditions should be used as descriptive features specifiers…”), and

(4) expanded “Psychiatric disorders” to “Psychiatric disorders, including substance use disorders.”

I think that similar changes would strengthen the proposed diagnostic criteria set for Hypersomnolence Disorder, now under review by the CPHC, because the clinical issues are exactly the same with respect to the importance of comorbid psychiatric and medical disorders. Thus, in order to acknowledge the importance of comorbid psychiatric and medical disorders, we would add a final diagnostic criterion to the proposed criteria set, exactly analogous to our amendment of the criteria set for Insomnia Disorder, based upon the CPHC recommendation:

Drs. McIntyre and Yager August 24, 2012 Page 2 “Frequently co-existing psychiatric and medical disorders do not adequately explain the predominant complaint of excessive sleepiness (or “insomnia,” in the case of the now modified criteria set for insomnia disorder). Similarly, we would amend the accompanying text to better highlight their clinical importance (by moving this portion of the text up in the list of specifiers), and we would expand “Psychiatric disorders” to “Psychiatric disorders, including substance use disorders” Thanks, again, for your helpful reviews of both Insomnia Disorder and Hypersomnolence Disorder (aka HD NEC). Your comments provide a way for us to move forward that is both clinically useful and reflective of scientific progress in sleep disorders medicine.

CPHC Memo: Insomnia. Addendum report

From: Drs. McIntyre and Yager on behalf of the DSM-5 Clinical and Public Health

Review Committee (CPHC)

To: Dilip Jeste, M.D. President APA

DATE: 8/23/2012

Attached is the cover letter for our report on Insomnia. (It is dated 8/1/2012, but actually that is incorrect, it should read 8/12/2012 and it was e-mailed to you on 8/13/2012.) You will note that the major reason this proposal was scored 3.5 (Fair to Poor) by CPHC (even though 3 of our 4 external reviewers scored the proposal “1”) was the absence in the criteria of any acknowledgement of comorbid psychiatric or medical disorders. (Members of CPHC felt that the external reviewers did not pay sufficient attention to this important issue.) Subsequent to our report, we were provided with the proposed DSM5 text for this diagnosis and in the text are listed specifiers that include co-morbid psychiatric and medical disorders.

We revisited this proposal on 8/21/2012 and, after considerable discussion we now rate this proposal as:

2 = GOOD, IF the specifiers for co-morbid psychiatric and medical disorders are included in the criteria (not just in the accompanying text.) This score was the unanimous vote of CPHC members.

We offer the following additional comments:

We suggest that the “Descriptive” specifiers (co-morbid conditions) be moved up in the list of specifiers to highlight their importance.

We suggest that the word “may” be changed to “should” in the Descriptive specifier “…comorbid conditions should be used as descriptive features specifiers..

We suggest that “Psychiatric disorders” be expanded to “Psychiatric disorders, including substance use disorders.” (If it is the convention in DSM5 to always assume “including substance use disorders” whenever “Psychiatric disorders” is used then this suggestion is not needed.

If our intent in this addendum is not clear please let us know.

CPHC Memo: Insomnia Disorder

From: Drs. McIntyre and Yager on behalf of the DSM-5 Clinical and Public

Health Committee (CPHC)

To: Dilip Jeste, M.D. President APA

DATE: August 1, 2012

Enclosed please find the CPHC report on Insomnia, reviewed at our meetings of July

31st and August 7, 2012.

Overall, our committee found that the concerns raised by the SRC and reviewer 2 to be

more compelling than the support offered by the other three of our external reviewers.

We noted that the three positive reviewers did not address major issues that concerned

the SRC or that concerned our committee, which included the work group’s proposal to

eliminate criteria D and E from the DSMIV criteria without in our view providing

adequate evidence or reasoning. We understand the WG’s desire to not impute

causality by listing “comorbid” conditions, but we do not understand the WG’s

reluctance to even acknowledge the co-existence of potentially important psychiatric,

substance abuse and/or general medical associated conditions as specifiers in the

diagnosis. Without such mention, we agree with the SRC and reviewer 2 that based on

the proposed DSM5 criteria a large number of patients risk being diagnosed with

Insomnia Disorder with insufficient or no attention to the co-existing disorders that

appear in conjunction with Insomnia. The considerable prevalence of these co-existing

conditions is clearly pointed out in the epidemiological data provided by the Work Group

itself.

Based on our external reviews and discussions in the CPHC, we rate this proposal as:

3.5 = Poor to Fair

We offer the following additional comments:

We generally concur with the SRC’s division of the proposal into several specific

components and with their relative assessments for each of those component parts.

CPH Committee Report to BOT MJV 8/7/12

Proposed Change(s): Insomnia Disorders

The Work Group has proposed seven specific changes, the first of which is considered

“Substantial”, change from DSM-IV: replacing “Primary Insomnia” and “Insomnia related to

another mental/medical disorder” with “Insomnia Disorder” and concurrent specification of

clinically comorbid conditions.

SRC Assessment: Score/Recommendations/Comments SRC has now completed three reviews of this proposal and is recommending that the DSM-IV

classification be retained, rather than accept the Workgroup proposal.

The SRC reviewed the Work Group’s proposals three times and was not supportive of the

recommended changes, feeling that there was inadequate data for the majority of changes

requested and that there were internal tautologies that would allow individuals to be diagnosed

who did not fulfill the intended diagnostic requirements. Specifically, the SRC rated requests

1,2,3 and 6 as a “5” or “poor”, request 4 as a “2” or “moderate support”, and request 5 as “4.33”,

or “limited to poor” support. One additional reviewer rated change in frequency request as a “4”

and the addition of sleep dissatisfaction as a “3”

PROPOSED SRC RECOMMENDATION:

MODEST (3) support (questionable) for #4 requiring sleep dissatisfaction

LIMITED (4) support (probably not-justified) for the rest of the proposal

Additional Comments:

The removal of the comorbidity requirement is still not well justified. The Eddinger article

referred to states that comorobidity between sleep disorder is problematic. The SRC is

concerned with comorbidty with other psychiatric illnesses, namely depression

The Sleep Disorders Workgroup recommends seven changes to the Insomnia

Disorder diagnostic criteria for consideration in DSM-5, including one substantial change

(#1), three changes of modest magnitude (#2, 4, 5), and three changes that represent simple

criterion clarification (#3, 6, 7).

1. Replacing “Primary Insomnia” and “Insomnia related to another mental/medical

disorder” with Insomnia Disorder, with specification of clinically comorbid conditions

• Substantial change (3)

• Potential loss of specificity in terms of phenotypes among those dx with insomnia

• “most cases of insomnia are comorbid”

• SRC Rating #5…POOR

2. Integrating the construct of sleep dissatisfaction to the definition of insomnia

• Modest (2)

• Likely to improve diagnostic specificity.

• Negative….those who minimize level of dissatisfaction may remain

undiagnosed/untreated


3. Adding early morning awakening as a separate symptom

• Level 1 change

• This criterion clarification may enhance specificity of

insomnia diagnosis and, potentially, improve treatment outcome.

• Negative…maybe confusing in elderly who retire earlier…


4. Adding a minimum frequency criterion (i.e., 3 nights per week) with sleep disturbance

• Modest (2)

• The proposed cut-point of three nights per week would help

differentiate individuals with occasional (sub threshold) insomnia from those with more

clinically meaningful insomnia

• Negative…those with 1 or 2 nights insomnia undiagnosed

• SRC Rating #2….MODERATE Support

5. Raising the minimum duration threshold from 1 to 3 months for chronic insomnia

• 2 (modest)

• Insomnia lasting only one month might be better conceptualized as an

episode rather than a disorder. Very few mental/medical disorders are considered chronic

before they exceed 6- or 12-month durations

• Negative…some people with insomnia of shorter durations could remain

undiagnosed/untreated

• Recommend using specifiers for acute, short term, chronic

• SRC Rating #4.33…LIMITED Support

6. Providing specific examples of daytime impairments

• Some individuals with persistent insomnia (e.g., older adults) tend to

minimize or underestimate the impact of insomnia on their daytime functioning

• Negative consequences…none

• Magnitude of change 1


7. Specifying that sleep disturbance occurs despite adequate opportunity for sleep

• This specification would help distinguish clinical insomnia from

volitional sleep deprivation.

• Negative consequences…none

• Magnitude of change 1

Arguments for:

o Consistent with 2005 NIH State of the Science conference on insomnia

The paradigm shifts from the need for clinicians to make causal attributions between

coexisting disorders and requires only specifying clinical comorbidities.

o The clinical utility of this approach is two-fold: (1) it acknowledges the bidirectional and

interactive effects between sleep disorders like insomnia and co-existing medical and

psychiatric disorders. As such, (2) the proposed change highlights the need for

independent clinical attention to the patient’s sleep disorder

o more conservative scientifically than DSM-IV because it does not

involve making potentially unwarranted and clinically inappropriate causal attributions

o more comprehensive clinically, decreasing the possibility of overlooking a

clinically relevant coexisting condition.

o More conservative and comprehensive, in that it eliminates imputation of causality,

which may be difficult to do or erroneous in nature.

o *DSM IV definition of “Primary Insomnia” performs poorly, in reliability and validity.

o *Consistent with 2005 NIH Conference recommendations; Would lead to increased

recognition of insomnia as a separate condition that has its own treatment needs.

Arguments Against:

o Potential loss of specificity and connection to phenotypes, such as depression

o Little empirical data to warrant such a substantial change, or more moderate changes in

frequency

o Could lead to significant increase in diagnosis, while at the same time lead to decreased

recognition of important comorbid disorders, such as depression with suicidal risk.

o Insomnia secondary to another mental/medical disorder actually has good reliability and has the

best epidemiologic data supporting its usage

o Changes in childhood requirements may pathologize normal pediatric behavior

CPH reviewers scores and discussion points

Reviewer 1 Support….STRONG (1) include

o Distinguishing between primary / secondary doesn’t make sense

o Supporting papers indicate that the current diagnostic nosology is inadequate

o Reason for change…clinical utility

o Not likely to decrease diagnosis, but will impart clarity

o Negatives…covered fully and minimal

Reviewer 2 Support…POOR (5) do not include

o Both the primary insomnia diagnosis which is low in reliability and validity, and the insomnia

related to another mental disorder, which is much higher in reliability and validity, are eliminated in

the new proposal.

o This in turn, is likely to result in an over-diagnosis of the new category.

o Thus, the lumping together of these categories will result in lower discriminative power which is

counter to the goals of the DSM

Reviewer 3 Support….STRONG (1) to include workgroups recommendations

Major Change: Negative consequences of not accepting this recommended change

include:

1. retention of a criterion set with poor performance characteristics

2. exposing patients with insomnia disorder to the increased risk that coexisting conditions

warranting independent treatment may be overlooked in diagnostic assessment and

treatment planning

3. forcing the clinician to make causal attributions between co-existing disorders

The Major Change would result in overwhelming improvements

Minor Changes: Omission of DSM-IV exclusion criterion D would be necessary to reflect the

proposed Major Change. The other Minor Changes would not result in "overwhelming"

improvements, but would clearly be an improvement.

Acknowlegement of the bidirectionality of insomnia and many other conditions is of the

utmost importance. The DSM-IV currenty forces clinicians to make a causal attribution

between co-existing disorders. This could lead to the underdiagnosis and undertreatment

of insomnia. The appreciation that two disorders may exist simultaneously and therefore

warrant simultaneous treatment has been clearly demostrated in the setting of depression

and insomnia: the simultaneous treatment of both conditions results in better outcomes

than the treatment of either condition alone

Reviewer 4 Support…..STRONG (1) to included workgroup recommendations

1) Thank you for asking me to comment on proposed changes to the insomnia section for DSM-V. I

agree strongly with the proposal to replace primary insomnia and insomnia related to mental

disorder with insomnia disorder. This “paradigm shift” removes the temptation to make a priori

causal attributions between co-existing disorders and requires a full diagnostic evaluation of all

insomnia complaints.

2) I would have preferred the term co-existing disorders, each requiring a full diagnostic

evaluation, to co-morbid. The term “co-morbid” muddies the diagnostic waters since it may

(colloquially) imply a relationship between disorders before a relationship is established. After all

we assume that people who co-habit or co-sleep don’t just co-exist. (Ann Fam Med 2009;7:357-

363)

3) We know physicians receive little training on how to sort out sleep complaints and insomnia

complaints are particularly difficult to sort out.

4) Therefore, this section needs to be clear and concise since it has an important “teaching”

function.

5) Insomnia is tricky because:

• The definition of insomnia is subjective

• like a fever, insomnia can have a variety of “causes” and, therefore, a very wide differential

• Insomnia may just be a symptom of another “true” sleep disorder e.g. delayed sleep phase

syndrome

• Or be a symptom of a insomnia disorder e.g. conditioned, learned, psychophysiologic insomnia

• The term co-morbid inadvertently adds to the difficulty because colloquially it has varying

meanings built around the core concept of more than 1 condition existing in an individual. (Ann

FAM Med 2009; 7:357-363). Often, co morbidity incorrectly implies a special relationship. It is a

rare physician who automatically thinks of a differential when a depressed patient complains of

insomnia even if we now call insomnia “co-morbid” and needing “attention”.

6) Therefore, at least in the insomnia section of the DSM V, it would be helpful to emphasize

• That co morbid means “co-existing”

• When first encountered, all sleep complaints require full diagnostic evaluation.

• Therefore, insomnia complaints require more than “attention”

• Adequate evaluation requires a discussion of true sleepiness (i.e. high Epworth) v fatigue

(inability to sleep day or night; the result of hyperarousal)

• And, how hyperarousal can become “co-morbid” with initial precipitants of insomnia

• One very useful way to help clinicians through this thicket is to suggest they use Speilman’s 3 Ps

as the rubric when evaluating ALL insomnia complaints

7) Finally,

• Adding early morning awakening as a separate symptom is misleading. The differential diagnosis

of early morning awakening is the same as the differential for problems initiating sleep and similar

to the differential for problems with sleep maintenance, e.g. obstructive sleep apnea, substance

use, circadian rhythm issues (Larks and Owls).

• DSM IV Criterion D is also misleading since all insomnia complaints need full differential attention

whether they occur with OSA or Circadian Rhythm disorder. Presumptive diagnoses are not

helpful.

• A minimum frequency criterion (3 nights per week) is not helpful clinically; minimal duration

threshold from 1-3 months for chronic insomnia is also artificial.

• Adding “sleep dissatisfaction” likely will improve case finding and is therefore useful.

CPHC discussion/remaining points of dispute:

* While the Work Group and 3 of the 4 CPHC reviewers supported the changes outlined, the CPHC

discussion did not find sufficient support the changes proposed. There was concern that specificity

would be lost and insomnia would be the only symptom treated. By not recognizing causality,

psychiatric conditions underlying insomnia would be overlooked.

Level of CPHC Support

1. __ Excellent

2. __Good

3. __Fair

_xx__Fair to Poor CPHC Score 3.5

4. __ Poor

CPHC Comments and Recommendations

* Little empirical data to warrant such a substantial change, or more moderate changes in frequency

*Could lead to significant increase in diagnosis, while at the same time lead to decreased recognition of

important comorbid disorders, such as depression with suicidal risk.

*Insomnia secondary to another mental/medical disorder actually has good reliability and has the best

epidemiologic data supporting its usage

*Changes in childhood requirements may pathologize normal pediatric behavior

• The CPHC felt that the Work Group and 3 of the 4 CPHC reviewers did not address the concerns

repeatedly raised by the SRC and within discussion at CPHC. CPHC does recommend consideration

of recommendation #4 to add a minimum frequency criterion with sleep disturbance such as 3

nights per week.

MOEC TO: Jack McIntyre, M.D., (Chair, DSM-5 CPHC Committee) Joel Yager, M.D. (Vice-Chair, CPHC Committee) FROM: Charles F. Reynolds III, M.D. (Chair, DSM-5 Sleep Wake Disorders Workgroup) RE: request for CPHC review of Insomnia Disorder DATE: May 18, 2012 I am writing on behalf of the DSM-5 Sleep Wake Disorders workgroup and on Dr. Kupfer’s referral to request CPHC review of proposed criteria for “Insomnia Disorder.” Copies of the SRC reviews and our responses to them are attached to this memorandum. Summary: our proposal for “Insomnia Disorder” represents a substantial change from DSM-IV: replacing “Primary Insomnia” and “Insomnia related to another mental/medical disorder” with “Insomnia Disorder” and concurrent specification of clinically comorbid conditions. In other words, we are advocating a non-hierarchical approach. The SRC has now completed three reviews of our proposal and is recommending that the DSM-IV classification be retained, rather than our proposal. We have an honest difference of opinion with our colleagues on the SRC and have been unable to resolve the differences. Reasons for and Consequencs of the Proposed Change: Our proposal reflects a change of paradigm now widely accepted in the field of sleep disorders medicine, articulated in a 2005 NIH State of the Science conference on insomnia. In essence, the paradigm shifts from the need for clinicians to make causal attributions between coexisting disorders and requires only specifying clinical comorbidities. The clinical utility of this approach is two-fold: (1) it acknowledges the bidirectional and interactive effects between sleep disorders like insomnia and co-existing medical and psychiatric disorders. As such, (2) the proposed change highlights the need for independent clinical attention to the patient’s sleep disorder. For example, one may treat a patient with a major depressive disorder to the point where most depressive symptoms have remitted, but with persistence of complaints of disturbed sleep. Unless the sleep disorder is also treated appropriately, the patient will be at higher risk for a relapse of major depressive disorder. Our proposal is more conservative scientifically than DSM-IV because it does not involve making potentially unwarranted and clinically inappropriate causal attributions. It is also more comprehensive clinically, decreasing the possibility of overlooking a clinically relevant coexisting condition. The Edinger et al. report in the October, 2011 Archives of General Psychiatry showed that DSM-IV criteria for “Primary Insomnia”

have poor performance characteristics. Our proposal addresses the deficiency highlighted by the Edinger field trial by removing the requirement that clinicians make causal attributions. Negative media attention: none Recommendations provided by forensic advisors: none with respect to Insomnia Disorder Thank you for your consideration of our proposal.

1

Memo Outlining Evidence for Change Insomnia Disorder

(December 13, 2011)

NOTE: Relevant sections of the attached references are identified below in bold text.

Summary: The Sleep Disorders Workgroup recommends seven changes to the Insomnia Disorder diagnostic criteria for consideration in DSM-5, including one substantial change (#1), three changes of modest magnitude (#2, 4, 5), and three changes that represent simple criterion clarification (#3, 6, 7). Below, we outline the proposed changes along with justifications, supporting evidence, and potential negative consequences.

1. Replacing “Primary Insomnia” and “Insomnia related to another mental/medical disorder” with Insomnia Disorder, with specification of clinically comorbid conditions

2. Integrating the construct of sleep dissatisfaction to the definition of insomnia 3. Adding early morning awakening as a separate symptom 4. Adding a minimum frequency criterion (i.e., 3 nights per week) with sleep disturbance 5. Raising the minimum duration threshold from 1 to 3 months for chronic insomnia 6. Providing specific examples of daytime impairments 7. Specifying that sleep disturbance occurs despite adequate opportunity for sleep

____________________________

Change #1. Replace the diagnoses of “Primary Insomnia” and “Insomnia related to another (psychiatric, medical) disorder” with “Insomnia Disorder”. Reasons for the change: The proposed change in paradigm moves away from the need to make causal attributions between co-existing disorders. It also acknowledges the bi-directional or interactive effects between sleep disorders and co-existing medical/psychiatric disorders. Another reason for this change is that the construct of “secondary insomnia” may lead to under treatment; this new conceptualization underscores that the patient has a sleep disorder in its own right that warrants independent clinical attention. For these reasons, we recommend using “Insomnia Disorder” whenever diagnostic criteria are met, whether or not there is a co-existing psychiatric, medical, or another sleep disorder; the presence of any of these disorders should be listed separately under comorbid conditions, but without such terms as “related to, due to, or associated with”, which often cannot be established. Potential negative consequences: While clinically appealing, a potential disadvantage of this simplified classification is the lost of specificity in terms of phenotypes among patients diagnosed with an Insomnia Disorder. The use of dimensional measures to delineate severity and other contributing factors might compensate for this loss. Magnitude of the change: 3 (substantial) Evidence for the change: This proposed amendment is supported by several lines of evidence. First, epidemiological data indicate that insomnia typically presents (> 50% of all cases) with another psychiatric or medical disorder (Ohayon & Reynolds, 2009; see p. 958, Table 5; Pearson et al., 2006; see p. 1779, Table 1; Roth and Jaeger, 2006; see p. 1368, Table 3), as opposed to a disorder on its own (i.e., Primary Insomnia). Recent data also show that the diagnostic reliability of insomnia, particularly Primary Insomnia, is relatively poor (Edinger

2

et al., 2011; see p. 996, Table 2). This may be no surprise because establishing a reliable differential diagnosis between “Primary Insomnia” and “Insomnia related to another disorder” implies that a clinician can identify the cause and the consequence of the main condition, a determination that is often difficult, if not impossible to make. In fact, there is increasing evidence that insomnia can be both cause and consequence of another mental disorder. The proposed change, which represents an important shift in paradigm, was first recommended at the NIH State-of-the-Science Conference on the Manifestation and Management of insomnia (NIH, 2005; see text below) and is already widely accepted in the field. Adopting this new paradigm/terminology would simplify differential diagnosis, improve diagnostic accuracy, and eliminate the need for criteria C, D, E from DSM-IV. Edinger JD, Wyatt JK, Stepanski EJ et al. (2011). Testing the reliability and validity of DSM-

IV-TR and ICSD-2 insomnia diagnoses. Arch Gen Psychiatry 68:992-1002. National Institutes of Health (2005). National Institutes of Health State of the Science

Conference statement on Manifestations and Management of Chronic Insomnia in Adults. Sleep 28:1049-57. “… most cases of insomnia are comorbid with other conditions. Historically, this has been termed “secondary insomnia”. However, the limited understanding or mechanistic pathways in chronic insomnia precludes drawing firm conclusions about the nature of these associations or direction of causality. Furthermore, there is concern that the term secondary may promote under treatment. Therefore, we propose that the term comorbid insomnia may be more appropriate.” (NIH, 2005)

Ohayon MM, Reynolds CF, 3rd (2009). Epidemiological and clinical relevance of insomnia

diagnosis algorithms according to the DSM-IV and the International Classification of Sleep Disorders (ICSD). Sleep Med 10:952-60.

Pearson NJ, Johnson LL, Nahin RL. (2006). Insomnia, trouble sleeping, and complementary and alternative medicine: Analysis of the 2002 national health interview survey data. Arch Intern Med 166:1775-82. Roth T, Jaeger S, et al (2006). Sleep problems, comorbid mental disorders, and role functioning

in the national comorbidity survey replication. Biol Psychiatry 60:1364-71. Change #2. Add the construct of “sleep dissatisfaction” to the definition of insomnia. Reasons for the change: Adding the construct of sleep dissatisfaction is likely to improve diagnostic specificity. This change could improve detection of clinically significant insomnia among subgroups of individuals (e.g., elderly) who typically report little impairment or distress associated with insomnia symptoms, but who are otherwise dissatisfied with their sleep. Potential negative consequences: One potential shortcoming of this change, although an unlikely one, is that some individuals with clinically significant insomnia symptoms who minimize their level of dissatisfaction may remain undiagnosed and untreated. Magnitude of the change: 2 (modest) Evidence for the change: Epidemiological surveys indicate that substantial proportions (8%-13%) of individuals with insomnia symptoms (i.e., difficulty falling or staying asleep) are not necessarily dissatisfied with their sleep and, conversely, some individuals dissatisfied with

3

sleep do not report specific insomnia symptoms (Morin et al., 2006; see p. 126, Figure 1). Other evidence suggests that the presence of sleep dissatisfaction in addition to insomnia symptoms increases considerably the proportion of individuals with daytime impairments relative to those with insomnia symptoms alone (Ohayon et al., 2011; see p. 4, Table 2). Hence, adding sleep dissatisfaction to the definition of insomnia is likely to improve diagnostic specificity. Morin CM, LeBlanc M, Daley M, Grégoire JP, Mérette C. Epidemiology of insomnia:

Prevalence, self-help treatments, consultations, and determinants of help-seeking behaviors. Sleep Med 7: 123-130.

Ohayon MM, Riemann D, Morin CM, Reynolds CF, 3rd. (2011). Hierarchy of insomnia criteria based on daytime consequences. Sleep Med 12: xxx-xxx.

Change #3. Add “Early Morning Awakening” as a separate symptom of insomnia. Reasons for the change: The early morning awakening symptom has traditionally been included under the terminology of “difficulty maintaining sleep”. However, this symptom (i.e., short sleep duration caused by an inability to resume sleep once awake) does not have the same clinical presentation or significance as nocturnal awakenings with difficulty in resuming sleep after awakening. This criterion clarification may enhance specificity of insomnia diagnosis and, potentially, improve treatment outcome. Potential negative consequences: One potential drawback of this change is that some individuals (mostly older people) who tend to have very early bedtime (e.g., 8:30pm), and consequently early morning wake up time (e.g., 3:30am), could be misdiagnosed with an Insomnia Disorder. To minimize this possibility, the narrative will be explicit to distinguish clinically significant early morning awakening and early morning awakening due to an early sleep-wake schedule. Magnitude of change: 1 (criterion clarification) Evidence for the change: Early morning awakening is a prevalent insomnia symptom, either as the only presenting symptom (2.1%) or mixed with other difficulty initiating or maintaining sleep (7.5%-12.3%) (Morin et al., 2006; see p. 126, Table 1; Ohayon & Reynolds, 2009, see p. 956, Table 4) and it may help improve specificity of insomnia diagnosis. Hohagen F, Kappler C, et al. (1994). Sleep onset insomnia, sleep maintaining insomnia and

insomnia with early morning awakening--temporal stability of subtypes in a longitudinal study on general practice attenders. Sleep 17:551-54.

Morin CM, LeBlanc M, et al. (2006). Epidemiology of insomnia: prevalence, self-help treatments, consultations, and determinants of help-seeking behaviors. Sleep Med 7: 123-30.

Ohayon MM, Reynolds CF, 3rd. (2009). Epidemiological and clinical relevance of insomnia diagnosis algorithms according to the DSM-IV and the International Classification of Sleep Disorders (ICSD). Sleep Med 10:952-60.

Change #4. Addition of a minimum frequency criterion (i.e., at least three nights per week) to make the diagnosis of insomnia.

4

Reasons for the change: The proposed cut-point of three nights per week would help differentiate individuals with occasional (sub threshold) insomnia from those with more clinically meaningful insomnia. Also, this cut-point is consistent with that used in ICD-10 and with current research practices in the field (Lichstein et al., 2003, see p. 430, Table 2). This change would then contribute to harmonizing criteria across diagnostic nosologies. Potential negative consequences: One potential disadvantage of increasing this threshold is to leave some people experiencing insomnia one or two nights per week undiagnosed and untreated. A dimensional classification could incorporate a sub threshold category for individuals who do not (yet) meet the frequency or duration criterion to receive an insomnia diagnosis. This latter classification would be helpful to study the natural history of insomnia (i.e., whether individuals with occasional insomnia symptoms eventually develop a more chronic Insomnia Disorder). Magnitude of the change: 2 (modest) Evidence for change: The proposed change is supported by evidence based on receiver-operating curve (ROC) analyses showing that sensitivity and specificity indices are maximized (i.e., correct identification of true insomnia cases and correct excluson of false positives) with a frequency of occurrence of insomnia falling between 3 and 4 nights per week (Lineberger et al., 2006; see p. 482, Figure 1 and Table 2; Lichstein et al., 2003; see p. 439; Figure 2, Table 3). This optimal frequency is achieved with severity criteria varying between 20 and 30 minutes for sleep onset latency and wake after sleep onset symptoms, with a reduced frequency achieved with with more severe symptoms (i.e., SOL/WASO > 60 min). There is also some evidence suggesting that the frequency of occurrence of insomnia symptoms is an important determinant of morbidity and impairment. Lichstein KL, Durrence HH, et al. (2003). Quantitative criteria for insomnia. Behav Res Ther

41:427-45. Lineberger MD, Carney CE, et al. (2006). Defining insomnia: Quantitative criteria for

insomnia severity and frequency. Sleep 29:479-85. Ohayon MM, Riemann D, Morin CM, Reynolds CF. (2011). Hierarchy of insomnia criteria

based on daytime consequences. Sleep Med 12:xxx-xxx Change #5. Raising the minimal duration threshold from 1 to 3 months for chronic insomnia. Reasons for the change: One month is a very short period to define insomnia as a chronic condition. Very few mental/medical disorders are considered chronic before they exceed 6- or 12-month durations. Insomnia lasting only one month might be better conceptualized as an episode rather than a disorder. Potential negative consequences: One potential disadvantage of increasing threshold for chronic insomnia to three months is that some people with insomnia of shorter durations could remain undiagnosed and untreated. To minimize this possibility, we recommend using specifiers for acute (< 1 month duration), short-term (between 1 and 3 months), and chronic insomnia (longer than 3 months). Magnitude of the change: 2 (modest)

5

Evidence for the change: Epidemiological data suggest a bimodal distribution of insomnia durations with about 45% of respondents reporting insomnia of less than three months duration and the remaining reporting insomnia lasting more than three months, including half with insomnia persisting over one year (Roth et al., 2006; see p. 1366, Figure 1). One longitudinal study showed that 74% of individuals with insomnia at baseline continued reporting insomnia up to one year later and 46% up to three years later (Morin et al., 2009; see p. 449, Table1). Most individuals seeking treatment in clinical settings or in the context of clinical trials exceed this threshold. Morbidity may also increase with insomnia persisting longer than three months (Ohayon et al., 2011). Ohayon MM, Riemann D, Morin CM, Reynolds CF, 3rd. (2011). Hierarchy of insomnia

criteria based on daytime consequences. Sleep Med 12: xxx-xxx. Morin CM, Bélanger L, et al. (2009). The natural history of insomnia: a population-based 3-

year longitudinal study. Arch Intern Med 169:447-53. Roth T, Jaeger S, et al. (2006). Sleep problems, comorbid mental disorders, and role

functioning in the national comorbidity survey replication. Biol Psychiatry 60:1364-71. Change #6. Add specific examples of daytime impairments (e.g., fatigue or low energy; sleepiness; cognitive impairments; mood disturbance). Reasons for the change: Some individuals with persistent insomnia (e.g., older adults) tend to minimize or underestimate the impact of insomnia on their daytime functioning, partly due to the lack of clear indicators of such impairments. This is likely to lead to under diagnosis and absence of treatment. The addition of specific examples of impairments may facilitate assessment and recognition of the impact of insomnia on daytime functioning. Potential negative consequences: None. Magnitude of the change: 1 (criterion clarification) Evidence for the change: The addition of specific examples of daytime impairments would be consistent with those used in the Research Diagnostic Criteria for insomnia and endorsed by the American Academy of Sleep Medicine (Edinger et al., 2004, see p. 1580). Edinger JD, Bonnet MH, et al. (2004). Derivation of research diagnostic criteria for insomnia:

report of an American Academy of Sleep Medicine Work Group. Sleep 27:1567-96. Change #7. Add the following specification “the sleep difficulty occurs despite adequate opportunity for sleep” Reasons for the change: This specification would help distinguish clinical insomnia from volitional sleep deprivation. Such change would be consistent with the Research Diagnostic Criteria (Edinger et al., 2004). Potential negative consequences: None Magnitude of change: 1 (criterion clarification) Evidence for change: See Research Diagnostic Criteria (Edinger et al., 2004, p. 1580).

6

Edinger JD, Bonnet MH, et al. (2004). Derivation of research diagnostic criteria for insomnia: report of an American Academy of Sleep Medicine Work Group. Sleep 27:1567-96.

DSM-IV Criteria ■ 307.42 Primary Insomnia A. The predominant complaint is difficulty initiating or maintaining sleep, or nonrestorative sleep, for at least 1 month. B. The sleep disturbance (or associated daytime fatigue) causes clinically significant distress or impairment in social, occupational, or other important areas of functioning. C. The sleep disturbance does not occur exclusively during the course of Narcolepsy, Breathing-Related Sleep Disorder, Circadian Rhythm Sleep Disorder, or a Parasomnia. D. The disturbance does not occur exclusively during the course of another mental disorder (e.g., Major Depressive disorder, Generalized Anxiety Disorder, a delirium). E. The disturbance is not due to the direct physiological effects of a substance (e.g., a drug of abuse, a medication) or a general medical condition. Proposed DSM-5 Criteria ■ 307.42 Insomnia Disorder (Highlighted text reflect changes from DSM-IV)

A. The predominant complaint is dissatisfaction with sleep quantity or quality made by the patient (or by a caregiver or family in the case of children or elderly).

B. Report of one or more of the following symptoms:

• Difficulty initiating sleep; in children this may be manifested as difficulty initiating sleep without caregiver intervention

• Difficulty maintaining sleep characterized by frequent awakenings or problems returning to sleep after awakenings (in children this may be manifested as difficulty returning to sleep without caregiver intervention)

• Early morning awakening with inability to return to sleep • Non restorative sleep (adults) • Resistance going to bed (children)

C. The sleep complaint is accompanied by significant distress or impairment in daytime functioning as indicated by the report of at least one of the following:

• Fatigue or low energy • Daytime sleepiness • Cognitive impairments (e.g., attention, concentration, memory) • Mood disturbance (e.g., irritability, dysphoria) • Behavioral problems (e.g., hyperactivity, impulsivity, aggression) • Impaired occupational or academic function • Impaired interpersonal/social function • Negative impact on caregiver or family functioning (e.g., fatigue, sleepiness)

D. The sleep difficulty occurs at least three nights per week. E. The sleep difficulty is present for at least three months. F. The sleep difficulty occurs despite adequate opportunity for sleep.

ONLINE FIRST

ORIGINAL ARTICLE

Testing the Reliability and Validityof DSM-IV-TR and ICSD-2 Insomnia DiagnosesResults of a Multitrait-Multimethod Analysis

Jack D. Edinger, PhD; James K. Wyatt, PhD; Edward J. Stepanski, PhD; Maren K. Olsen, PhD;Karen M. Stechuchak, MS; Colleen E. Carney, PhD; Ambrose Chiang, MD; M. Isabel Crisostomo, MD;Margaret D. Lineberger, PhD; Melanie K. Means, PhD; Rodney A. Radtke, MD;William K. Wohlgemuth, PhD; Andrew D. Krystal, MD

Context: Distinctive diagnostic classification schemesfor insomnia diagnoses are available, but the optimal in-somnia nosology has yet to be determined.

Objectives: To test the reliability and validity of in-somnia diagnoses listed in the American Psychiatric As-sociation’s DSM-IV-TR and the International Classifica-tion of Sleep Disorders, second edition (ICSD-2).

Design: Multitrait-multimethod correlation design.

Setting: Two collaborating university medical centers,with recruitment from January 2004 to February 2009.

Participants: A total of 352 adult volunteers (235 ofwhom were women) who met research diagnostic crite-ria for insomnia disorder.

Main Outcome Measures: Goodness-of-fit ratings of10 DSM-IV-TR and 37 ICSD-2 insomnia diagnoses for eachpatient.Ratingswereprovidedby3clinicianpairswhouseddistinctiveassessmentmethodstoderivediagnostic impres-sions. Correlations computed within and across clinicianpairs were used to test reliability and validity of diagnoses.

Results: Findings suggested that the best-supported DSM-IV-TR insomnia categories were insomnia related to an-

other mental disorder, insomnia due to a general medi-cal condition, breathing-related sleep disorder, andcircadian rhythm sleep disorder. The category of pri-mary insomnia appeared to have marginal reliability andvalidity. The best-supported ICSD-2 categories were theinsomnias due to a mental disorder and due to a medi-cal condition, obstructive sleep apnea, restless legs syn-drome, idiopathic insomnia, and circadian rhythm sleepdisorder–delayed sleep phase type. Psychophysiologi-cal insomnia and inadequate sleep hygiene received muchmore variable support across sites, whereas the diagno-sis of paradoxical insomnia was poorly supported.

Conclusions: Both the DSM-IV-TR and ICSD-2 provideviable insomnia diagnoses, but findings support se-lected subtypes from each of the 2 nosologies. Nonethe-less, findings regarding the frequently used DSM-IV-TRdiagnosis of primary insomnia and its related ICSD-2 sub-types suggest that their poor reliability and validity areperhaps due to significant overlap with comorbid insom-nia subtypes. Therefore, alternate diagnostic paradigmsshould be considered for insomnia classification.

Arch Gen Psychiatry. 2011;68(10):992-1002.Published online June 6, 2011.doi:10.1001/archgenpsychiatry.2011.64

T HERE ARE SEVERAL DIAGNOS-tic nosologies for insom-nia1-6 designed to system-atizedescriptionsofpatients,facilitate communication

amonghealthcarepractitioners,guide treat-ment choices, predict clinical course, andstandardize research.7 These nosologies dif-fer markedly in their complexity and reli-ance on information external to the clini-cal interview. The American PsychiatricAssociation’sDSM (DSM-III-R,DSM-IV, andDSM-IV-TR)3,4 describes a few global in-somnia diagnoses and relies primarily onclinical interview. In contrast, the Interna-tionalClassificationofSleepDisorders(ICSD)and the International Classification of Sleep

Disorders, second edition (ICSD-2)5,8 de-lineate numerous primary and secondaryinsomnia subtypes and incorporate find-ings from interview and laboratory tests.

These divergent classification schemesare products of discrepant views about howmany subtypes are needed to describe allindividuals with insomnia. Proponents ofthe DSM-IV-TR nosology argue that manyICSD-2 insomnia subtypes have little em-pirical substantiation and should be sub-sumed within broader categories.9-11 How-ever, the DSM-IV-TR system allows forconsiderable heterogeneity within

See also page 990

Author Affiliations are listed atthe end of this article.

ARCH GEN PSYCHIATRY/ VOL 68 (NO. 10), OCT 2011 WWW.ARCHGENPSYCHIATRY.COM992

©2011 American Medical Association. All rights reserved. at University of Florida, on December 13, 2011 www.archgenpsychiatry.comDownloaded from

http://www.archgenpsychiatry.com

diagnostic categories and may not provide optimal dis-crimination among distinctive insomnia disorders.12-18 Whatis clear is that the 2 nosologies result in markedly discor-dant classifications when applied to the same sample ofpatients with insomnia.19 This state of affairs creates costlyvariability in assessment and management of patients withinsomnia and needless disunity within the insomnia re-search literature.

Whether the DSM-IV-TR or ICSD-2 offers a more ac-curate scheme for insomnia classification and diagnosisremains unknown. The few studies20-24 that assessed re-liability of DSM-IV-TR or ICSD-2 diagnostic categorieshave found only modest reliability for the insomnia sub-types they evaluated. Reliability data are unavailable formany of the diagnoses in each system. Moreover, avail-able literature21,25-27 provides indirect and limited sup-port for a selected subset of DSM-IV-TR and ICSD-2 di-agnoses. For example, one study26 showed that treatmentrecommendations of clinicians vary as a function of theDSM-IV and ICSD diagnoses they assign. Other stud-ies21,25 comparing patient groupings resulting from stan-dard clinical classification with the groupings resultingfrom statistical clustering procedures have shown some(albeit minimal) congruence between these 2 classifica-tion approaches. Whereas such studies represent prox-ies for testing the validity of insomnia diagnoses, formalomnibus empirical tests of these nosologies have not yetbeen conducted to our knowledge.

Nonetheless, the DSM-IV-TR and ICSD-2 have contin-ued to enjoy widespread clinical and research use. Giventhe existence of these 2 discordant nosologies, insomniadiagnosis and treatment remain a hit-and-miss processguided more by clinicians’ instincts and beliefs about vari-ous insomnia subtypes than by a well-validated diagnos-tic system. Clearly, research to ascertain the most viableinsomnia nosology is sorely needed. This dual-site studywas conducted with the following aims: (1) to determineand compare the reliabilities of DSM-IV-TR and ICSD-2insomnia subtypes; and (2) to derive and compare con-vergent validity (CV) and discriminant validity (DV) in-dices for the DSM-IV-TR and ICSD-2 insomnia diagno-ses. Our overarching goal was to ascertain the optimalscheme for insomnia classification.

METHODS

DESIGN

This study was conducted at Duke University Medical Centerand Rush University Medical Center using a multitrait-multimethod28 design. Multiple insomnia diagnoses were as-sessed in a research cohort using multiple assessment meth-ods. Within each study site, 6 sleep specialists were groupedto form 3 clinician pairs. Each pair was then assigned 1 of 3assessment approaches to use throughout the study for dis-cerning the insomnia diagnosis (or diagnoses) of each partici-pant they interviewed. One pair used solely a structured sleepinterview for discerning diagnoses; the second pair used stan-dard unstructured clinical interviews; and the third pair reliedon unstructured clinical interviews combined with polysom-nographic (PSG) data. The latter 2 pairs were also given ac-cess to information from sleep diaries and sleep history ques-tionnaires completed by study participants. Each clinician

formulated impressions independently without knowledge ofthe other clinicians’ impressions. The multiple traits consid-ered were 10 DSM-IV-TR and 37 ICSD-2 insomnia diagnoses,which represent all DSM-IV-TR and ICSD-2 insomnia diagno-ses that can be ascertained via interview. Following their in-terviews, clinicians rated how well each of these 47 insomniadiagnoses fit each patient. A standard multitrait-multi-method28 correlational analysis was then applied to these rat-ings to test the reliability and validity of the insomnia diagno-ses considered. The institutional review boards of thecollaborating medical centers reviewed and approved the studyprotocol. Participants provided written informed consent andreceived parking expenses plus a maximum $400 payment forparticipation.

PARTICIPANTS

Recruitment occurred between January 2004 and February 2009through posted announcements and physician referrals. In-cluded individuals (1) met Research Diagnostic Criteria26 forinsomnia disorder, (2) were aged 18 years or younger, and (3)spoke English fluently. Excluded individuals (1) had an un-stable or life-threatening medical condition, (2) were immi-nently suicidal, (3) scored 24 or lower on the Mini-Mental StateExamination, or (4) were previously evaluated by any studyclinicians. Of the 425 individuals enrolled, 8 were removed be-cause they did not meet study selection criteria, 50 withdrewbefore beginning any study interviews, and 15 failed to com-plete all interviews. The remaining 352 participants (201 fromDuke University Medical Center and 151 from Rush Univer-sity Medical Center) composed the final sample. Table1 showsdemographic characteristics for the sample.

POLYSOMNOGRAPHY

Participants underwent 2 consecutive nights of PSG with a moni-toring montage consisting of 2 channels of electroencephalog-raphy (C3-M2, Oz-Cz), 1 chin electromyography channel, 2

Table 1. Demographic Characteristics of Study Sample

CharacteristicOverall

(N=352)

DukeUniversityMedicalCenter

(n=201)

RushUniversityMedicalCenter

(n=151)

Age, mean (SD), y 46.44 (14.40) 49.25 (14.83) 42.70 (12.95)Sex, No. (%)

Male 117 (33.24) 67 (33.33) 50 (33.11)Female 235 (66.76) 134 (66.67) 101 (66.89)

Race, No. (%)White 212 (60.23) 140 (69.65) 72 (47.68)African American 112 (31.82) 56 (27.86) 56 (37.09)American Indian or

Alaskan1 (0.28) 1 (0.50) 0

Asian or Pacific Islander 13 (3.69) 2 (1.00) 11 (7.28)Other 14 (3.98) 2 (1.00) 12 (7.95)

Education, mean (SD), ya 15.08 (3.10) 15.54 (2.77) 14.47 (3.41)Marital status, No. (%)a

Single 135 (38.46) 53 (26.37) 82 (54.67)Married or living as

married142 (40.46) 99 (49.25) 43 (28.67)

Divorced, separated, orwidowed

74 (21.08) 49 (24.38) 25 (16.67)

aOne subject is missing education and 1 subject has missing data for maritalstatus. Percentage calculations exclude subjects with missing data.




channels of electro-oculography (left eye to M1, right eye toM2), 1 channel of airflow (nasal-oral thermistor), 2 channelsof respiratory effort (thoracic and abdominal impedance), 1channel of pulse oximetry, 2 channels of anterior tibialis elec-tromyography (right and left legs), and 1 channel of body po-sition monitoring. Participants followed their customary bed-times and rising times on PSG nights. Those who occasionallyused hypnotics underwent PSG sessions without such medi-cations, whereas those who used hypnotics 3 or more nightsper week or were taking antidepressants and/or anxiolytics un-derwent PSG sessions with these medications.

All PSG sessions were scored using traditional scoring cri-teria for sleep stages, apneas and hypopneas, periodic limb move-ments, and related arousals.29-31 Summary data including re-spiratory parameters (apnea-hypopnea index, desaturation index,etc), periodic limb movement indices (number of movementswith and without arousals), and types and dosages of medica-tions taken on PSG nights were included in a report made avail-able to 1 clinician pair at each study site.

ELECTRONIC SLEEP DIARY

Participants recorded sleep data for 2 weeks using a handheldcomputer. Those who had difficulty using this device com-pleted paper sleep diaries. The computer presented questionsabout each night’s bedtime, sleep onset latency, number and lengthof nocturnal awakenings, time of final awaking, rising time, andsleep medication and alcohol use. Also, respondents’ ratings (10-point scale) of sleep quality and how rested they felt on arisingwere acquired. Diary data were downloaded (or hand-enteredfor paper diaries) into a PC computer, and a printout of dailyvalues and 2-week averages was generated showing the follow-ing: bedtime, sleep onset latency, number of nocturnal awaken-ings, time awake after sleep onset and prior to final awakening,time of final awakening, time of rising out of bed, total sleep time,total time awake, time in bed, sleep efficiency (total sleep time/time in bed�100), sleep quality, restedness on arising, medi-cation and alcohol use, and the time when diary entries were made.

SLEEP HISTORY QUESTIONNAIRE

Participants completed a 10-page questionnaire. This solicitedinformation about their demographic characteristics, sleep com-plaints, medical and psychiatric history, and treatment history.

STRUCTURED SLEEP INTERVIEW

The Duke Structured Interview for Sleep Disorders (DSISD) wasused by 1 clinician pair at each site to derive participants’ insom-nia diagnoses. The DSISD incorporates criteria for ascertainingDSM-IV-TR and ICSD-2 sleep disorder diagnoses and is dividedinto 4 modules: insomnia-related disorders, excessive daytimesleepiness–related disorders, sleep/wake schedule disorders, andparasomnias. The insomnia module assesses insomnias relatedto other mental disorders, general medical disorders, substanceabuse, circadian rhythm disorders, restless legs syndrome, inad-equate sleephygiene, etc. Screeningquestionsallowsectionswithina module to be skipped contingent on definitive negative an-swers from a respondent. However, interviewers may continuesections when answers to screening questions are ambiguous. Pre-vious studies show that the DSISD has acceptable reliability andvalidity for DSM-IV-TR and ICSD-2 insomnia diagnoses.32,33

ELECTRONIC DIAGNOSTIC RATING FORMS

The rating forms consisted of the series of 10 DSM-IV-TR and37 ICSD-2 diagnoses presented on the screen of a specially pro-

grammed handheld computer. Each diagnosis appeared indi-vidually accompanied by a 100-pixel visual analog scale (VAS)labeled “doesn’t fit at all” at its left extreme and “fits extremelywell” at its right extreme. Clinicians considered each diagno-sis separately and decided how well it fit the participant in ques-tion. Clinicians moved a pointer on the VAS to indicate the good-ness of fit for each DSM-IV-TR and ICSD-2 insomnia diagnosislisted. These ratings were converted into numeric values re-flecting their locations on the 100-pixel VAS and used as theprimary data for our multitrait-multimethod analyses.

INTERVIEWER PAIRS AND INTERVIEW PROCESS

Six clinicians at each study site were stratified by sleep medi-cine experience (�10 years vs �10 years) and by professionaldegree (MD vs PhD). They were then randomly paired withinstrata to form 3 pairs who were reasonably similar in their ex-perience and mix of clinical specialties. They were then as-signed their respective assessment method: (1) solely the DSISD;(2) a combination of an unstructured clinical interview, sleephistory questionnaires, and sleep diaries; or (3) a combinationof an unstructured clinical interview, sleep history question-naires, sleep diaries, and PSG information. All clinicians re-ceived training in the use of the computerized diagnostic ratingforms, and clinicians using the DSISD also were given trainingin its administration. Each participant underwent 4 interviews(2 structured and 2 unstructured interviews). Clinicians usingthe structured sleep interview method conducted separate in-terviews because the DSISD required independent administra-tion and interpretation. The 2 remaining clinician pairs each con-ducted a joint interview with each participant. During jointinterviews, one clinician interviewed the participant while theother clinician remained silent. When the initial interviewer gainedsufficient information to formulate diagnostic impressions, heor she exited the interview room and the second clinician inter-viewed the participant further if desired. The pair using the un-structured clinical interview and PSG method also reviewed PSGresults. A randomization procedure was used so that each clini-cian served as the initial interviewer for a randomly determined50% of the participants interviewed.

PROCEDURE

Study candidates first underwent telephone screening with thesite’s project coordinator. Those passing this screen next metwith the project coordinator to provide informed consent andundergo a Mini-Mental State Examination.34 Those who passedthe Mini-Mental State Examination screening were enrolledand completed the following: (1) the Structured Clinical In-terview for DSM-IV Disorders; (2) the sleep history question-naire and sleep diary monitoring; and (3) the PSG sessions.Participants then were stratified by sex and age group (aged18-39, 40-59, and �60 years) and randomized to 1 of the 6 pos-sible orders of interviews within strata. After completing all in-terviews, participants chose an in-person or telephone debrief-ing with the principal investigator ( J.D.E. at Duke UniversityMedical Center) or co–principal investigator (E.J.S. or J.K.W.at Rush University Medical Center). During debriefing, the PSG-informed final insomnia diagnosis (or diagnoses) were sharedand a treatment referral was made if desired.

ANALYTIC APPROACH

We followed traditional analytic guidelines28 for multitrait-multimethod research designs. Analyses entailed computing cor-relations among the clinicians’ VAS ratings across methods anddiagnoses within each diagnostic system (DSM-IV-TR and




ICSD-2) separately. Resulting correlation matrices were usedto evaluate reliability and validity for each diagnosis. Reliabil-ity connotes the degree of agreement between clinicians whouse the same assessment method; this is commonly called in-terrater reliability. The correlations between ratings made bythe 2 clinicians within each pair for each diagnosis served asreliability indices. With 3 clinician pairs at each of 2 study sites,a total of 6 reliability correlations were derived for each diag-nosis. Convergent validity connotes how well clinicians usingdifferent assessment methods agree in their diagnoses. The CVindices were those correlations reflecting the level of agree-ment shown for each diagnosis between clinician pairs usingdiffering assessment methods. Concordant with a method de-scribed by Campbell and Fiske,28 the ratings of paired clini-cians were first averaged for each diagnosis. We then com-puted correlations of the resultant averaged ratings of eachdiagnosis produced by the 3 distinctive clinician pairs. By usingthis method, we derived 3 CV correlations per diagnosis at eachsite for a total of 6 such indices for each diagnosis.

Discriminant validity implies that diagnoses are distinctiveand can be discriminated. This construct connotes that agree-ment between clinicians rating the same diagnosis should benotably greater than the agreement observed between clini-cians rating distinctive diagnosis. Hence, DV required consid-eration of correlations of the averaged diagnostic ratings for dis-crepant diagnoses (within and between clinician pairs). The DVwas supported when there was greater correlation between rat-ings of the same diagnosis derived by different assessment meth-ods than was found for (1) different diagnoses derived by dif-ferent methods and (2) different diagnoses assessed by the samemethod. Because the data were not normally distributed, weused Spearman correlation coefficients in all of these analyses.

RESULTS

CLINICIAN TURNOVER

At Duke University Medical Center, the pair using theunstructured clinical interview with access to PSG in-formation experienced staff turnover. At Rush Univer-sity Medical Center, each clinician pair changed mem-bership; however, 2 of the pairs (the pair using thestructured sleep interview method and the pair using theunstructured clinical interview with access to PSG in-formation) retained one of the clinicians for the entirestudy period. Correlation analyses ignored staffing changesfor the following reasons: (1) the information availableto the clinicians (eg, PSG or not) was unchanged; (2) oneor both members of the pair remained unchanged for 2clinician pairs at each site; and (3) clinician character-istics remained reasonably stable.

FREQUENCY ANALYSES

We first examined the percentage of cases wherein all 6interviewers rated each diagnosis as a possible fit (rat-ing �0) as well as the percentage of cases wherein all 6clinicians viewed each diagnosis as nonapplicable (rat-ing=0). Primary insomnia, insomnia related to anothermental disorder, breathing-related sleep disorder, and in-somnia due to a general medical condition were the mostfrequently selected DSM-IV-TR diagnoses. The remain-ing DSM-IV-TR diagnoses were rated less frequently butmost were assigned ratings higher than 0 by 1 or more

clinicians for at least 20% of the cases. Only the diagno-sis of no sleep disorder was so infrequent that it wasdropped from our reliability and validity analyses.

Many ICSD-2 categories were rated infrequently andhence were excluded from analyses. Diagnoses retainedwere psychophysiological insomnia, paradoxical insom-nia, idiopathic insomnia, inadequate sleep hygiene, in-somnia due to a mental disorder, insomnia due to a medi-cal condition, insomnia due to a drug or substance,obstructive sleep apnea, circadian rhythm sleep disorder–delayed sleep phase type, restless legs syndrome, peri-odic limb movement disorder, environmental sleep dis-order, and other sleep disorder. These were all assigneda rating higher than 0 by 1 or more clinicians in morethan 29% of the cases evaluated.

DIAGNOSTIC RELIABILITY

Table 2 and Table 3 show the reliability indices ob-tained. The DSM-IV-TR categories with the highest in-terrater reliability were the insomnias related to anothermental disorder or due to a medical condition, breathing-related sleep disorder, and circadian rhythm sleep dis-order. More modest reliability estimates were noted foralcohol-related sleep disorder and substance-induced sleepdisorder. Results for primary insomnia, dyssomnia nototherwise specified, and other sleep disorder were mixedwith lower reliability estimates found at Rush Univer-sity Medical Center.

The ICSD-2 diagnoses showing the greatest inter-rater agreement included insomnia due to a mental dis-order, insomnia due to a medical condition, periodic limbmovement disorder, restless legs syndrome, obstructivesleep apnea, and circadian rhythm sleep disorder–delayed sleep phase type. More modest reliability indi-ces were obtained for the diagnosis of insomnia due to adrug or substance. The Rush University Medical Centersite showed lower reliability estimates than the Duke Uni-versity Medical Center site for psychophysiological in-somnia and idiopathic insomnia using the clinical inter-view method. The Duke University Medical Center sitehad lower reliability estimates than the Rush UniversityMedical Center site for inadequate sleep hygiene withinthe clinician pair using the structured sleep interviewmethod and the clinician pair using the unstructured clini-cal interview with access to PSG information. Interview-ers across both sites showed better agreement for the para-doxical insomnia diagnosis when given sleep historyquestionnaire and diary data to review compared withthe DSISD only. The category of other sleep disordershowed much lower reliability at Rush University Medi-cal Center than at Duke University Medical Center.

VALIDITY ANALYSES

For our validity analyses, we retained diagnoses that showedat least modest reliability (mean r� .20) and/or had highendorsement rates (rated �0 by �1 clinician for �50%of all cases across sites). Accordingly, we eliminated thecategory of other sleep disorder listed in the DSM-IV-TRand ICSD-2 because of poor reliability at the Rush Uni-versity Medical Center site and infrequent use overall.




Table 2. Interrater Reliability Indices for DSM-IV-TR Insomnia Diagnoses Across Study Sites and Assessment Methodsa

DSM-IV-TR Insomnia Diagnosis

Interrater Reliability, % (95% CI)

Structured Interview Clinical Interview Clinical Interview�PSG

Duke Rush Duke Rush Duke Rush

Primary insomnia 0.44(0.32 to 0.55)b

0.08(−0.08 to 0.24)

0.43(0.31 to 0.54)b

0.14(−0.02 to 0.29)

0.28(0.14 to 0.40)b

0.30(0.14 to 0.44)b

Breathing-related sleep disorder 0.76(0.69 to 0.81)b

0.35(0.20 to 0.48)b

0.66(0.57 to 0.73)b

0.61(0.49 to 0.70)b

0.70(0.62 to 0.76)b

0.62(0.51 to 0.71)b

Circadian rhythm sleep disorder 0.44(0.32 to 0.54)b

0.50(0.37 to 0.61)b

0.71(0.64 to 0.77)b

0.46(0.32 to 0.58)b

0.43(0.31 to 0.54)b

0.39(0.24 to 0.51)b

Dyssomnia NOS 0.40(0.27 to 0.51)b

0.33(0.18 to 0.46)b

0.53(0.42 to 0.62)b

−0.004(−0.17 to 0.16)

0.26(0.12 to 0.38)b

0.08(−0.08 to 0.24)

Related to a mental disorder 0.59(0.50 to 0.68)b

0.51(0.38 to 0.62)b

0.62(0.52 to 0.69)b

0.60(0.49 to 0.70)b

0.55(0.45 to 0.64)b

0.71(0.62 to 0.78)b

Due to a medical condition 0.43(0.31 to 0.54)b

0.38(0.23 to 0.51)b

0.58(0.48 to 0.66)b

0.47(0.33 to 0.58)b

0.47(0.35 to 0.57)b

0.57(0.45 to 0.67)b

Alcohol-related sleep disorder 0.31(0.18 to 0.43)b

0.39(0.24 to 0.52)b

0.57(0.47 to 0.66)b

0.42(0.27 to 0.54)b

0.21(0.07 to 0.34)c

0.38(0.23 to 0.51)b

Due to a substance 0.24(0.10 to 0.36)b

0.25(0.10 to 0.40)c

0.29(0.16 to 0.41)b

0.40(0.26 to 0.53)b

0.27(0.14 to 0.39)b

0.31(0.16 to 0.45)b

Other sleep disorder 0.22(0.09 to 0.35)c

0.04(−0.13 to 0.19)

0.23(0.10 to 0.36)b

−0.09(−0.24 to 0.08)

0.24(0.10 to 0.36)b

0.05(−0.11 to 0.21)

Abbreviations: CI, confidence interval; NOS, not otherwise specified; PSG, polysomnography.aDuke indicates Duke University Medical Center; Rush, Rush University Medical Center.bP� .001.cP� .01.

Table 3. Interrater Reliability Indices for ICSD-2 Insomnia Diagnoses Across Study Sites and Assessment Methodsa

ICSD-2 Insomnia Diagnosis

Interrater Reliability, % (95% CI)

Structured Interview Clinical Interview Clinical Interview�PSG

Duke Rush Duke Rush Duke Rush

Psychophysiological 0.51(0.40 to 0.60)b

0.34(0.19 to 0.47)b

0.52(0.41 to 0.61)b

0.12(−0.04 to 0.27)

0.27(0.14 to 0.39)b

0.55(0.42 to 0.65)b

Paradoxical 0.15(0.01 to 0.28)c

0.12(−0.04 to 0.28)

0.51(0.40 to 0.61)b

0.39(0.24 to 0.52)b

0.50(0.39 to 0.60)b

0.41(0.26 to 0.53)b

Idiopathic 0.72(0.65 to 0.78)b

0.57(0.45 to 0.67)b

0.79(0.73 to 0.84)b

0.16(0.0001 to 0.31)c

0.41(0.29 to 0.52)b

0.43(0.28 to 0.55)b

Due to a mental disorder 0.59(0.49 to 0.67)b

0.53(0.40 to 0.63)b

0.63(0.54 to 0.71)b

0.61(0.50 to 0.70)b

0.55(0.44 to 0.64)b

0.67(0.57 to 0.75)b

Due to a medical condition 0.48(0.37 to 0.58)b

0.42(0.28 to 0.55)b

0.63(0.53 to 0.70)b

0.54(0.42 to 0.65)b

0.51(0.40 to 0.61)b

0.57(0.45 to 0.67)b

Due to a drug or substance 0.28(0.15 to 0.40)b

0.30(0.14 to 0.43)b

0.36(0.24 to 0.48)b

0.34(0.19 to 0.47)b

0.31(0.18 to 0.43)b

0.40(0.26 to 0.53)b

Obstructive sleep apnea 0.76(0.70 to 0.81)b

0.49(0.35 to 0.60)b

0.68(0.59 to 0.74)b

0.64(0.54 to 0.73)b

0.66(0.57 to 0.73)b

0.64(0.53 to 0.72)b

Delayed sleep phase syndrome 0.64(0.55 to 0.72)b

0.27(0.12 to 0.42)b

0.75(0.68 to 0.80)b

0.53(0.40 to 0.63)b

0.38(0.26 to 0.49)b

0.40(0.26 to 0.53)b

Restless legs syndrome 0.70(0.62 to 0.76)b

0.41(0.27 to 0.54)b

0.79(0.73 to 0.83)b

0.44(0.30 to 0.56)b

0.69(0.61 to 0.76)b

0.61(0.50 to 0.70)b

Periodic limb movement disorder 0.56(0.45 to 0.64)b

0.56(0.43 to 0.66)b

0.51(0.40 to 0.61)b

0.40(0.25 to 0.52)b

0.54(0.43 to 0.63)b

0.57(0.45 to 0.67)b

Inadequate sleep hygiene 0.13(−0.01 to 0.26)

0.32(0.17 to 0.46)b

0.47(0.35 to 0.57)b

0.31(0.16 to 0.45)b

0.03(−0.11 to 0.17)

0.24(0.08 to 0.38)d

Environmental 0.22(0.09 to 0.35)d

0.34(0.19 to 0.47)b

0.34(0.21 to 0.46)b

0.41(0.27 to 0.53)b

0.19(0.05 to 0.32)d

0.46(0.32 to 0.57)b

Other sleep disorder 0.43(0.31 to 0.53)b

0.10(−0.06 to 0.26)

0.45(0.33 to 0.55)b

0.12(−0.04 to 0.27)

0.46(0.34 to 0.56)b

−0.01(−0.17 to 0.15)

Abbreviations: CI, confidence interval; ICSD-2, International Classification of Sleep Disorders, second edition; PSG, polysomnography.aDuke indicates Duke University Medical Center; Rush, Rush University Medical Center.bP� .001.cP� .05.dP� .01.




Table 4 and Table 5 show CV and DV indices de-rived for the DSM-IV-TR and ISCD-2 insomnia diagnosesexamined. A diagnosis is considered valid when the CVvalues are statistically significant (with higher values con-

noting greater CV) and the DV values are consistently lowerthan the CV values and, preferably, nonsignificant. TheCV and DV values are not reported when poor reliabilitywas found for that diagnosis within a study site.

Table 4. Validity Indices for Selected DSM-IV-TR Insomnia Subtypesa

Insomnia Diagnosis

SI CI CI�PSG

SI CI CI�PSG SI CI CI�PSG SI CI CI�PSG

Duke University Medical Center(A) Primary

CV .57b .33b .33b

DV −0.03 (C) 0.01 (G) 0.07 (C) −0.01 (F) 0.02 (G) 0.08 (F) −0.01 (E) 0.01 (G) −0.04 (F)(B) BRSD

CV .77b .46b .43b

DV 0.09 (E) 0.05 (F) 0.18 (H) 0.09 (E) 0.04 (G) 0.16 (E) 0.11 (F) 0.16 (F) 0.08 (F)(C) CRSD

CV .66b .56b .52b

DV −0.03 (A) −0.02 (A) 0.10 (H) −0.01 (A) −0.01 (D) 0.10 (G) 0.07 (A) 0.03 (A) 0.02 (G)(D) Mental disorder

CV .69b .64b .68b

DV 0.17 (E) 0.23 (E) 0.11 (E) 0.09 (F) 0.18 (E) 0.11 (E) 0.11 (F) 0.14 (E) 0.11 (E)(E) Medical condition

CV .56b .62b .57b

DV 0.17 (D) 0.18 (H) 0.09 (G) 0.23 (D) 0.18 (D) 0.14 (D) 0.12 (B) 0.19 (H) 0.30 (H)(F) Alcohol induced

CV .72b .49b .52b

DV 0.13 (G) 0.09 (D) 0.11 (B) 0.24 (G) 0.10 (E) 0.16 (B) 0.20 (G) 0.09 (E) 0.10 (G)(G) Substance induced

CV .34b .35b .21c

DV 0.13 (F) 0.24 (F) 0.20 (F) 0.07 (D) 0.07 (E) 0.06 (B) 0.09 (E) 0.10 (C) 0.13 (E)(H) Dyssomnia NOS

CV .43b .38b .31b

DV 0.11 (D) 0.07 (D) 0.12 (E) 0.18 (E) 0.16 (E) 0.19 (E) 0.18 (B) 0.11 (E) 0.30 (E)

Rush University Medical Center(A) Primary NR NR NR(B) BRSD

CV .58b .51b .34b

DV 0.03 (F) 0.02 (F) 0.09 (E) 0.12 (E) 0.08 (F) 0.15 (E) 0.09 (E) 0.05 (E) 0.11 (E)(C) CRSD

CV .45b .50b .39b

DV 0.15 (G) 0.14 (G) 0.05 (B) 0.10 (G) 0.16 (F) 0.07 (G) 0.09 (G) 0.14 (D) 0.05 (G)(D) Mental disorder

CV .61b .73b .64b

DV 0.22 (G) 0.18 (G) 0.23 (G) 0.23 (F) 0.12 (F) 0.14 (C) 0.20 (F) 0.20 (G) 0.23 (G)(E) Medical condition

CV .50b .58b .45b

DV 0.19 (G) 0.12 (B) 0.13 (D) 0.07 (G) 0.05 (F) 0.05 (B) 0.11 (G) 0.15 (B) 0.18 (G)(F) Alcohol induced

CV .36b .30b .42b

DV 0.20 (D) 0.38 (G) 0.20 (D) 0.11 (C) 0.26 (G) 0.15 (D) 0.19 (D) 0.25 (G) 0.19 (G)(G) Substance induced

CV .31b .49b .39b

DV 0.22 (D) 0.10 (C) 0.17 (D) 0.39 (F) 0.26 (F) 0.25 (F) 0.23 (D) 0.12 (F) 0.23 (D)(H) Dyssomnia NOS

CV NR NR NR

Abbreviations: BRSD, breathing-related sleep disorder; CI, unstructured clinical interview method; CI�PSG, unstructured clinical interview with access topolysomnography information; CRSD, circadian rhythm sleep disorder; CV, convergent validity; DV, discriminant validity; NOS, not otherwise specified; NR, notreported owing to low reliability at study site indicated; SI, structured sleep interview method.

aThe CVs for the SI vs CI, SI vs CI�PSG, and CI vs CI�PSG methods are listed only once for each diagnosis to eliminate redundancy. The DVs shown are thehighest correlations derived from the monomethod-heterotrait comparisons (ie, comparisons for differing diagnoses within methods of assessment) and theheteromethod-heterotrait comparisons (ie, comparisons for differing diagnoses across differing methods of assessment). For the DVs, adjacent letters inparentheses connote the diagnosis most strongly correlated with the diagnosis listed in the first column. For example, “−0.03 (C)” shown in the third column ofthe first row of the Duke University Medical Center data indicates the correlation between the target diagnosis (primary insomnia) and CRSD derived within thestructured interview method.

bP� .001.cP� .01.




The best-supported DSM-IV-TR diagnosis wasinsomnia associated with another mental disorder(Table 4). Insomnia due to a medical condition,breathing-related sleep disorder, and circadian rhythmsleep disorder also showed reasonable validity. Alcohol-related sleep disorder showed more modest validityindices, whereas other substance-induced insomniareceived less support. Primary insomnia and dyssomnianot otherwise specified were least supported. Their CVindices were in the low to medium range across meth-ods at Duke University Medical Center; validity correla-tions for primary insomnia and dyssomnia not other-wise specified were not calculated at Rush UniversityMedical Center owing to low reliability indices notedthere.

The best-supported ICSD-2 diagnoses were insomniadue to a mental disorder and insomnia due to a medicalcondition (Table 5); their CV indices mainly fell in thelarge range and the DV indices were generally in theinsignificant range. Obstructive sleep apnea was alsoreasonably supported, but its pattern of correlationssuggested that clinicians who had access to PSG infor-

mation differed from those who did not. Restless legssyndrome, circadian rhythm sleep disorder–delayedsleep phase type, and idiopathic insomnia receivedmore modest support: the CV indices for these catego-ries fell in the medium to large range and were gener-ally larger than their related DV indices. Insomnia dueto a drug or substance and environmental sleep disor-der received less support, with the CV indices falling inthe small to medium range. At Duke University MedicalCenter, the validity indices for psychophysiologicalinsomnia ranged from small to large, but poor reliabilityfor this diagnosis at Rush University Medical Centerobviated its validity testing there. A comparison of CVand DV indices acquired for inadequate sleep hygienesuggested reasonable validity for this diagnosis at theRush University Medical Center site, but its poor reli-ability at Duke University Medical Center preventedassessing its validity there. The validity indices werevariable for periodic limb movement disorder acrosssites and methods, whereas paradoxical insomniareceived little support overall.

Table 5. Validity Indices for Selected ICSD-2 Insomnia Subtypesa

Insomnia Diagnosis

SI CI CI�PSG


Duke University Medical Center(A) Psychophysiological

CV .50b .42b .27b

DV 0.03 (L) 0.06 (H) 0.11 (K) 0.10 (D) 0.11 (K) 0.18 (K) 0.08 (F) 0.04 (K) 0.06 (C)(B) Paradoxical

CV .24b .14 .22c

DV 0.22 (C) 0.21 (C) 0.28 (C) 0.12 (C) 0.15 (C) 0.11 (C) 0.24 (K) 0.12 (K) 0.15 (C)(C) Idiopathic

CV .72b .58b .68b

DV 0.24 (H) 0.26 (H) 0.19 (H) 0.31 (H) 0.30 (H) 0.21 (H) 0.28 (B) 0.11 (B) 0.15 (B)(D) Mental disorder

CV .69b .66b .66b

DV 0.18 (F) 0.19 (F) 0.12 (F) 0.09 (I) 0.12 (F) 0.08 (F) 0.08 (I) 0.13 (F) 0.10 (I)(E) Inadequate sleep hygiene

CV NR NR NR(F) Medical disorder

CV .57b .65b .57b

DV 0.18 (I) 0.20 (I) 0.17 ( J) 0.19 (D) 0.29 (K) 0.13 (D) 0.25 (I) 0.30 (I) 0.23 (I)(G) OSA

CV .74b .45b .38b

DV 0.21 (L) 0.14 (L) 0.13 (L) 0.13 (F) 0.13 (L) 0.18 (F) 0.10 (L) 0.09 (I) 0.12 (K)(H) DSPS

CV .72b .59b .53b

DV 0.24 (C) 0.31 (C) 0.08 (I) 0.26 (C) 0.30 (C) 0.06 ( J) 0.19 (C) 0.21 (C) 0.13 (C)(I) RLS

CV .63b .72b .68b

DV 0.43 (L) 0.41 (L) 0.26 (L) 0.34 (L) 0.50 (L) 0.30 (F) 0.42 (L) 0.41 (L) 0.23 (F)( J) Due to drug or substance

CV .47b .37b .37b

DV 0.14 (F) 0.18 (K) 0.15 (F) 0.05 (D) 0.12 (K) 0.13 (L) 0.17 (F) 0.10 (F) 0.17 (F)(K) Environmental

CV .30b .25b .31b

DV 0.14 (L) 0.14 (I) 0.24 (B) 0.18 ( J) 0.29 (F) 0.12 (F) 0.11 (A) 0.18 (A) 0.12 (G)(L) Periodic limb movements

CV .55b .29b .23c

DV 0.43 (I) 0.34 (I) 0.42 (I) 0.41 (I) 0.50 (I) 0.41 (I) 0.26 (I) 0.13 ( J) 0.14 (I)

(continued)




DATA SYNTHESIS

Because the study design and data obtained may be novelto many readers, we conducted an additional classifica-tion analysis to summarize results and place them in apractical context. Since to our knowledge there are nostandardized methods for classifying outcomes of mul-titrait-multimethod studies, we offer the rationally de-rived classification rules in Table 6. These rules con-sider the size and significance of the reliability and validitycorrelations obtained to gauge the acceptability of eachdiagnosis. The correlations themselves were appraisedusing Cohen’s guidelines,35 wherein correlation coeffi-cients in the order of 0.10 are regarded as small, those of

0.30 are medium, and those of 0.50 or higher are large.These cutoffs could be considered arbitrary when ap-plied to individual reliability or validity indices, but ourclassification approach arguably provides a practical man-ner for synthesizing our results.

Table 7 shows how the diagnoses rate when applyingthese classification rules. Within DSM-IV-TR, insomnia re-lated to another mental disorder is rated as highly accept-able; breathing-related sleep disorder, insomnia due to amedical condition, and circadian rhythm sleep disorder areacceptable; alcohol-related sleep disorder is marginally ac-ceptable; and the remaining diagnoses are unacceptable.Within ICSD-2, insomnia due to a mental disorder is highlyacceptable; obstructive sleep apnea and insomnia due to a

Table 5. Validity Indices for Selected ICSD-2 Insomnia Subtypesa (continued)

Insomnia Diagnosis

SI CI CI�PSG


Rush University Medical Center(A) Psychophysiological

CV NR NR NR(B) Paradoxical

CV .07 .18d .19d

DV 0.16 (C) 0.31 (C) 0.18 (C) 0.12 (F) 0.02 (F) 0.13 (C) 0.05 (E) 0.08 (H) 0.16 (C)(C) Idiopathic

CV .33b .57b .32b

DV 0.16 (B) 0.21 (D) 0.20 (H) 0.31 (B) 0.25 (H) 0.20 (L) 0.18 (B) 0.20 (I) 0.18 (I)(D) Mental disorder

CV .66b .69b .64b

DV 0.25 ( J) 0.18 (I) 0.18 ( J) 0.21 (C) 0.14 (L) 0.16 (H) 0.20 ( J) 0.17 ( J) 0.28 ( J)(E) Inadequate sleep hygiene

CV .34b .28b .37b

DV 0.31 (G) 0.13 (K) 0.13 (G) 0.19 ( J) 0.17 ( J) 0.16 ( J) 0.20 (K) 0.26 ( J) 0.16 (K)(F) Medical disorder

CV .53b .57b .46b

DV 0.29 (I) 0.20 (I) 0.17 ( J) 0.20 (I) 0.32 (L) 0.25 (I) 0.09 (K) 0.17 (G) 0.17 ( J)(G) OSA

CV .53b .50b .30b

DV 0.31 (E) 0.13 (E) 0.23 ( J) 0.16 (I) 0.32 (L) 0.40 (L) 0.13 (E) 0.12 (B) 0.10 ( J)(H) DSPS

CV .37b .50b .38b

DV 0.24 ( J) 0.19 (C) 0.16 (C) 0.13 ( J) 0.25 (C) 0.17 (E) 0.20 (C) 0.16 (D) 0.17 (C)(I) RLS

CV .36b .34b .50b

DV 0.29 (F) 0.20 (F) 0.13 (D) 0.20 (F) 0.54 (L) 0.36 (L) 0.15 (F) 0.25 (F) 0.24 (L)( J) Due to drug or substance

CV .24c .47b .22c

DV 0.25 (D) 0.19 (E) 0.20 (H) 0.13 (E) 0.17 (E) 0.26 (E) 0.23 (G) 0.16 (E) 0.28 (D)(K) Environmental

CV .31b .39b .39b

DV 0.15 (H) 0.23 (L) 0.21 (L) 0.13 (E) 0.12 (E) 0.19 (E) 0.04 (B) 0.09 (L) 0.16 (E)(L) Periodic limb movements

CV .27b .10 .48b

DV 0.28 (I) 0.13 (G) 0.09 (F) 0.23 (K) 0.54 (I) 0.21 (I) 0.21 (K) 0.40 (G) 0.24 (I)

Abbreviations: CI, unstructured clinical interview method; CI�PSG, unstructured clinical interview with access to polysomnography information; CV,convergent validity; DSPS, delayed sleep phase syndrome; DV, discriminant validity; ICSD-2, International Classification of Sleep Disorders, second edition; NR,not reported owing to low reliability at study site indicated; OSA, obstructive sleep apnea; RLS, restless legs syndrome; SI, structured sleep interview method.

aThe CVs for the SI vs CI, SI vs CI�PSG, and CI vs CI�PSG methods are listed only once for each diagnosis to eliminate redundancy. The DVs shown are thehighest correlations derived from the monomethod-heterotrait comparisons (ie, comparisons for differing diagnoses within methods of assessment) and theheteromethod-heterotrait comparisons (ie, comparisons for differing diagnoses across differing methods of assessment). For the DVs, adjacent letters inparentheses connote the diagnosis most strongly correlated with the diagnosis listed in the first column. For example, “0.03 (L)” shown in the third column of thefirst row of the Duke University Medical Center data indicates the correlation between the target diagnosis (psychophysiological insomnia) and periodic limbmovements derived within the structured interview method.

bP� .001.cP� .01.dP� .05.




medical condition are acceptable; restless legs syndrome,circadian rhythm sleep disorder–delayed sleep phase type,and idiopathic insomnia are marginally acceptable; and theremaining diagnoses are unacceptable.

COMMENT

The DSM-IV-TR and ICSD-2 sleep disorders nosologiesare used widely, but few studies have tested their reliabil-ity and validity. Results of this trial show that each sys-tem includes diagnoses with acceptable reliability and va-lidity. Within DSM-IV-TR, insomnia related to anothermental disorder, insomnia due to a general medical con-dition, breathing-related sleep disorder, and circadianrhythm sleep disorder were best supported. Alcohol-related sleep disorder was more marginally supported,whereas other substance-induced sleep disorder was notwell supported. Least supported were dyssomnia not oth-erwise specified, other sleep disorder, and, surprisingly,primary insomnia. Although primary insomnia was fre-quently rated, our data call into question its reliability andvalidity.

Within ICSD-2, insomnia due to a mental disorder, in-somnia due to a medical condition, and obstructive sleepapnea were well supported. Circadian rhythm sleep dis-order–delayed sleep phase type and idiopathic insomniaalso received reasonable support and fell in the margin-ally acceptable classification. Restless legs syndrome wasalso classified as marginally acceptable rather than accept-able largely owing to its overlap or correlation with peri-odic limb movement disorder. However, it is well recog-nized that periodic limb movements are highly prevalent

among patients with restless legs syndrome.36 Given thisconsideration, perhaps the restless legs syndrome diagno-sis should be classified as acceptable. In contrast, insom-nia due to a drug or substance and environmental sleep dis-order appeared much more marginal. Support for theremaining ICSD-2 diagnoses was more variable and gen-erally poor.

Table 7 suggests that an optimal insomnia nosologyderives from a melding of DSM-IV-TR and ICSD-2 diag-noses. The categories of insomnia related or due to a men-tal disorder, insomnia due to a medical condition, andbreathing-related sleep disorder or obstructive sleep ap-nea seemingly merit strong consideration for inclusion.These categories occur within both nosologies but haveslightly different labels in each. Restless legs syndromealso seems to be a viable diagnosis, as does DSM-VI-TRcircadian rhythm sleep disorder, which seems favored overthe more specific ICSD-2 delayed sleep phase type. Fi-nally, DSM-IV-TR alcohol-related sleep disorder along withICSD-2 idiopathic insomnia merit consideration as well.

The DSM-IV-TR diagnosis of primary insomnia andmost related ICSD-2 subtypes were rated frequently bystudy clinicians but garnered minimal support. The cor-relations obtained suggest that the addition of PSG datato interview findings seemingly complicates diagnosticascertainment and reduces reliability and validity for mostof these categories. Because cognitive and behavioralmechanisms are thought to be important perpetuatingmechanisms in primary insomnia and its related ICSD-2subtypes, assignment of these diagnoses is largely de-pendent on ascertaining the presence of such mecha-nisms as the likely cause of the insomnia disorder ob-

Table 6. Classification Systema

Classification Reliability CV DV

Highly acceptable All r values are in large range (�0.50) All CV r values are in large range (�0.50) All DV r values are in small range (�0.30)or insignificant range

Acceptable All r values are in medium range (�0.30)or high range (�0.50)

All CV r values are in medium range(�0.30) or high range (�0.50)

�80% of DV r values are in small range(�0.30) or insignificant range

Marginally acceptable All r values are significant and �1 r valuefalls in small range (�0.30); remainderfall in medium or large range

All CV r values are significant and �1r value falls in small range (�.30);remainder fall in medium or large range

DV r values are consistently lower than CVr values with �1 DV greater than orequal to site-specific CV values obtained

Unacceptable Does not meet any above criteria Does not meet any above criteria Does not meet any above criteria

Abbreviations: CV, convergent validity; DV, discriminant validity.aThe classification system for correlation coefficients is by Cohen.35

Table 7. Classification Results

Classification DSM-IV-TR Diagnoses ICSD-2 Diagnoses

Highly acceptable Related to a mental disorder Due to a mental disorderAcceptable Breathing-related sleep disorder; due to a medical

condition; circadian rhythm sleep disorderObstructive sleep apnea; due to a medical condition

Marginally acceptable Alcohol-related sleep disorder Restless legs syndrome; delayed sleep phase type;idiopathic

Unacceptable Primary insomnia; dyssomnia NOS; substance-inducedsleep disorder; other sleep disorder

Psychophysiological; paradoxical; due to a drug orsubstance; periodic limb movement disorder;inadequate sleep hygiene; environmental sleepdisorder; other sleep disorder

Abbreviations: ICSD-2, International Classification of Sleep Disorders, second edition; NOS, not otherwise specified.




served. However, patients presumed to have primaryinsomnia often have some symptoms of depression and/oranxiety, whereas those presumed to have comorbid formsof insomnia often manifest the cognitive-behavioral ab-errations thought to perpetuate primary insomnia.37

Hence, primary vs comorbid insomnia distinctions areoften subtle and perhaps even arbitrary. Perhaps refine-ment of the definitional criteria for primary insomnia andits related ICSD-2 subtypes could improve their reliabil-ity and validity, but efforts to do so will be complicatedby their degree of overlap with the comorbid insomniasubtypes. Ultimately, it may be necessary to adopt a dif-ferent paradigm for insomnia diagnosis.

Perhaps the primary vs comorbid insomnia distinc-tion could be abandoned in favor of a more inclusive di-agnostic term such as insomnia disorder. This is the ap-proach being taken in the upcoming DSM-5. Patients whomeet the insomnia criteria outlined in the Research Diag-nostic Criteria26 for insomnia disorder are to be assignedthis diagnosis regardless of any coincident sleep-disrup-tive comorbidities. However, diagnosticians are encour-aged to also diagnose coexisting psychiatric and medicalcomorbidities. Such an approach should simplify clini-cians’ diagnostic task with patients who have insomnia.Of course, use of the global insomnia disorder diagnosiscould also encourage a generic one-size-fits-all approachto insomnia treatment. Whether this concern is war-ranted will only be determined once the DSM-5 is placedinto use.

We could also consider abandoning the sorting of pa-tients into diagnostic “bins” but instead using dimen-sional measures of insomnia symptoms for patient char-acterization. This approach is advocated by the developersof Profile Analysis via Multidimensional Scaling.38,39 Thismethod applies a multidimensional scaling analysis to pa-tients’ scores on syndrome-relevant questionnaires to iden-tify core symptom profiles. Profile Analysis via Multidi-mensional Scaling then locates each patient’s symptompattern by assigning patient-specific weights for each pro-file. These weights connote the degree of match with eachcore profile and designate the patient’s exact diagnosticlocation in relation to these profiles. Patients are not forcedinto single, often poorly fitting diagnostic categories butrather are characterized by their overall symptom ar-rays. This method may more accurately characterize eachpatient and support efforts toward the development ofindividualized therapies. The DSM-5 advocates use of di-mensional measures, so perhaps future nosologies willincorporate this method.

In deciding on future changes to our insomnia diag-nostic systems, it may be useful to consider this study’sfindings in conjunction with other projects such as thefield trials for the DSM-5 sleep disorders nosology. Ourdata provide information about interrater reliability andthe validity and credibility of the insomnia diagnosesexamined in the eyes of our study clinicians. However,the DSM-5 trials will include many more study sitesthan used in this project and will produce the test-retestreliability data that were not provided with our method.Such additional findings in conjunction with our resultsshould guide future improvements to our insomnianosologies.

It is important to consider the potential effects of ourstudy method on the findings obtained. Given the statis-tical demands of the multitrait-multimethod research de-sign, it was necessary to have clinicians rate the goodnessof fit of each insomnia diagnosis on VASs for each patientthey evaluated. Such ratings may reveal clinicians’ sub-jective decision-making processes when assigning diag-noses, but they do not replicate diagnostic outcomes inclinical situations wherein clinicians assign diagnoses inan all-or-none fashion. Cross-validation of our results usingmethods that replicate real-world diagnostic practices maytherefore be warranted.

Admittedly, this study had several limitations. Oursample included mainly research volunteers. Whethersimilar results would have been obtained with clinicalpatients or those with insomnia who are selected fromthe community remains unknown. Furthermore, datawere obtained from only 2 study sites. Replication of thisstudy across additional sites would have been desirable,albeit quite costly. Because many of the subtypes con-sidered were seldom or never rated by study clinicians,the study may have benefited from a much larger and di-verse sample. Additionally, the greater turnover in ourclinician raters at Rush University Medical Center mayhave contributed to greater variability in results notedthere. Finally, our findings suggested notable differ-ences in the reliability and validity indices obtained acrosssites for selected diagnoses. Whether such differencesshould be attributed to demographic differences in theirstudy samples, general site-specific biases, and/or idio-syncratic diagnostic propensities of clinicians cannot bedetermined given our study design. Nonetheless, few pre-vious studies have assessed reliability and validity of in-somnia diagnoses routinely used in practice. Thus, re-sults presented herein fill a void and provide guidancefor revisions of our insomnia classification schemes.

Submitted for Publication: October 29, 2010; final re-vision received March 15, 2011; accepted March 31, 2011.Published Online: June 6, 2011. doi:10.1001/archgenpsychiatry.2011.64Author Affiliations: Veterans Affairs Medical Center (DrsEdinger, Olsen, and Means and Ms Stechuchak) and DukeUniversity Medical Center (Drs Edinger, Olsen, Chiang,Lineberger, Means, Radtke, and Krystal), Durham, NorthCarolina; Sleep Disorders Service and Research Center,Department of Behavioral Sciences, Rush University Medi-cal Center, Chicago, Illinois (Drs Wyatt and Crisos-tomo); ACORN Research, LLC, Memphis, Tennessee (DrStepanski); Ryerson University, Toronto, Ontario, Canada(Dr Carney); and Veterans Affairs Medical Center, Mi-ami, Florida (Dr Wohlgemuth).Correspondence: Jack D. Edinger, PhD, Psychology Ser-vice (116B), Veterans Affairs Medical Center, 508 Ful-ton St, Durham, NC 27705 ([email protected]).Financial Disclosure: Dr Edinger has been a consultantfor Kingsdown and Philips/Respironics and has receivedresearch support from Philips/Respironics. Dr Wyatt hasreceived research support from Philips/Respironics. DrKrystal has received grants or research support from theNational Institutes of Health, sanofi-aventis, Cephalon,GlaxoSmithKline, Merck, Neurocrine, Pfizer, Sepracor,




Somaxon, Takeda, Transcept, Philips/Respironics, Neu-rogen, Evotec, Astellas, and Neuronetics and has been aconsultant for Abbott, Actelion, Arena, Astellas, Axiom,AstraZeneca, Bristol-Myers Squibb, Cephalon, Eli Lilly,GlaxoSmithKline, Jazz, Johnson & Johnson, King, Merck,Neurocrine, Neurogen, Neuronetics, Novartis, Orga-non, Ortho-McNeil-Janssen, Pfizer, Respironics, Roche,sanofi-aventis, Sepracor, Somaxon, Takeda, Transcept,and Kingsdown.Funding/Support: This research was supported by grantR01 MH67057 from the National Institute of MentalHealth.Role of the Sponsor: The funding agency had no role inthe design and conduct of the study; collection, man-agement, analysis, and interpretation of the data; or prepa-ration, review, or approval of the manuscript.Disclaimer: The views expressed in this article are thoseof the authors and do not necessarily represent the viewsof the Department of Veterans Affairs.Previous Presentation: This paper was presented in partat the 20th Congress of the European Sleep Research So-ciety; September 17, 2010; Lisbon, Portugal.Additional Contributions: Kevan VanLandingham, MD,PhD, Laurie Keefer, PhD, Andrea Canada, PhD, AimeeDanielson, PhD, Babak Mokhlesi, MD, Margaret Park, MD,Mike Summers, MD, and Tony Proske, MD, served asclinical interviewers and Marci Loiselle, Angela Kirby,Pamela Smith, Faye Knauss, Kathy Schelble, Laura Benson,and Lindsey Gluszek served as study coordinators.

REFERENCES

1. Association of Sleep Disorders Centers; Association for the Psychophysiologi-cal Study of Sleep. Diagnostic classification of sleep and arousal disorders: 1979first edition. Sleep. 1979;2(1):1-154.

2. World Health Organization. International Classification of Diseases, Ninth Revi-sion, Clinical Modification. Geneva, Switzerland: World Health Organization; 1992.

3. American Psychiatric Association. Diagnostic and Statistical Manual of MentalDisorders. 4th ed. Washington, DC: American Psychiatric Association; 1994.

4. American Psychiatric Association. Diagnostic and Statistical Manual of MentalDisorders. 4th ed, text revision. Washington, DC: American Psychiatric Asso-ciation; 2000.

5. American Sleep Disorders Association. International Classification of SleepDisorders. Rochester, MN: American Sleep Disorders Association; 1990.

6. American Sleep Disorders Association. International Classification of Sleep Dis-orders, Revised: Diagnostic and Coding Manual. Rochester, MN: American SleepDisorders Association; 1997.

7. Buysse DJ, Young T, Edinger JD, Carroll J, Kotagal S. Clinicians’ use of the In-ternational Classification of Sleep Disorders: results of a national survey. Sleep.2003;26(1):48-51.

8. American Academy of Sleep Medicine. International Classification of SleepDisorders. 2nd ed. Darien, IL: American Academy of Sleep Medicine; 2005.

9. Soldatos CR, Kales JD, Tan TL, Kales A. Classification of sleep disorders. Psy-chiatr Ann. 1987;17:454-457.

10. Buysse DJ, Reynolds CF III, Monk TH, Hoch CC, Yeager AL, Kupfer DJ. Quanti-fication of subjective sleep quality in healthy elderly men and women using thePittsburgh Sleep Quality Index (PSQI). Sleep. 1991;14(4):331-338.

11. Vgontzas AN, Kales A, Bixler EO, Vela-Bueno A. In discussion of: Nofzinger EA,Buysse DJ, Reynolds CF 3rd, Kupfer DJ. Sleep disorders related to another men-tal disorder (nonsubstance/primary): a DSM-IV literature review. J Clin Psychiatry.1993;54(7):244-259.

12. Hauri PJ, Olmstead E. Childhood-onset insomnia. Sleep. 1980;3(1):59-65.13. Nofzinger EA, Buysse DJ, Reynolds CF III, Kupfer DJ. Sleep disorders related to

another mental disorder (nonsubstance/primary): a DSM-IV literature review. J ClinPsychiatry. 1993;54(7):244-259.

14. Hauri PJ, Wisbey J. Wrist actigraphy in insomnia. Sleep. 1992;15(4):293-301.

15. Edinger JD, Krystal AD. Subtyping primary insomnia: is sleep state mispercep-tion a distinct clinical entity? Sleep Med Rev. 2003;7(3):203-214.

16. Krystal AD, Edinger JD, Wohlgemuth WK, Marsh GR. NREM sleep EEG fre-quency spectral correlates of sleep complaints in primary insomnia subtypes.Sleep. 2002;25(6):630-640.

17. Downey R III, Bonnet MH. Training subjective insomniacs to accurately per-ceive sleep onset. Sleep. 1992;15(1):58-63.

18. McCall WV, Edinger JD. Subjective total insomnia: an example of sleep statemisperception. Sleep. 1991;15(1):71-73.

19. Ohayon MM, Roberts RE. Comparability of sleep disorders diagnoses using DSM-IVand ICSD classifications with adolescents. Sleep. 2001;24(8):920-925.

20. Schramm E, Hohagen F, Grasshoff U, Riemann D, Hajak G, Weess HG, BergerM. Test-retest reliability and validity of the Structured Interview for Sleep Dis-orders According to DSM-III-R. Am J Psychiatry. 1993;150(6):867-872.

21. Edinger JD, Fins AI, Goeke JM, McMillan DK, Gersh TL, Krystal AD, McCall WV.The empirical identification of insomnia subtypes: a cluster analytic approach.Sleep. 1996;19(5):398-411.

22. Morin CM, Stone J, McDonald K, Jones S. Psychological management of in-somnia: a clinical replication series with 100 patients. Behav Ther. 1994;25:291-309.

23. Buysse DJ, Reynolds CF III, Hauri PJ, Roth T, Stepanski EJ, Thorpy MJ, BixlerEO, Kales A, Manfredi RL, Vgontzas AN, Stapf DM, Houck PR, Kupfer DJ.Diagnostic concordance for DSM-IV sleep disorders: a report from the APA/NIMH DSM-IV field trial. Am J Psychiatry. 1994;151(9):1351-1360.

24. Ohayon MM, Guilleminault C, Zulley J, Palombini L, Raab H. Validation of thesleep-EVAL system against clinical assessments of sleep disorders and poly-somnographic data. Sleep. 1999;22(7):925-930.

25. Hauri PJ. A cluster analysis of insomnia. Sleep. 1983;6(4):326-338.26. Edinger JD, Bonnet MH, Bootzin RR, Doghramji K, Dorsey CM, Espie CA, Ja-

mieson AO, McCall WV, Morin CM, Stepanski EJ; American Academy of SleepMedicine Work Group. Derivation of research diagnostic criteria for insomnia:report of an American Academy of Sleep Medicine Work Group. Sleep. 2004;27(8):1567-1596.

27. Buysse DJ, Reynolds CF III, Kupfer DJ, Thorpy MJ, Bixler E, Kales A, ManfrediR, Vgontzas A, Stepanski E, Roth T, Hauri P, Stapf D. Effects of diagnosis ontreatment recommendations in chronic insomnia: a report from the APA/NIMHDSM-IV field trial. Sleep. 1997;20(7):542-552.

28. Campbell DT, Fiske DW. Convergent and discriminant validation by the multitrait-multimethod matrix. Psychol Bull. 1959;56(2):81-105.

29. Rechtshaffen A, Kales A. A Manual of Standardized Terminology, Techniques,and Scoring Systems of Sleep Stages of Human Subjects. Bethesda, MD: USNational Institute of Neurological Diseases and Blindness, Neurological Infor-mation Network; 1968.

30. Coleman R. Periodic movements in sleep (nocturnal myoclonus) and restlesslegs syndrome. In: Guilleminault C, ed. Sleeping and Waking Disorders: Indica-tions and Techniques. Menlo Park, CA: Addison-Wesley; 1982:265-295.

31. American Sleep Disorders Association. EEG arousals: scoring rules and ex-amples: a preliminary report from the Sleep Disorders Atlas Task Force of theAmerican Sleep Disorders Association. Sleep. 1992;15(2):173-184.

32. Carney CE, Edinger JD, Olsen MK, Stechuchak KM, Krystal AD, Wyatt JK. Inter-rater reliability for insomnia diagnoses derived from the Duke Structured Inter-view for Sleep Disorders. Sleep. 2008;31(suppl):A250.

33. Edinger JD, Wyatt JK, Olsen MK, Stechuchak KM, Carney CE, Chiang A, KrystalAD, Lineberger MD, Means MK, Radtke RA, Knauss F. Reliability and validity ofinsomnia diagnoses derived from the Duke Structured Interview for SleepDisorders. Sleep. 2009;32(suppl):A265.

34. Folstein MF, Folstein SE, McHugh PR. “Mini-Mental State”: a practical methodfor grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12(3):189-198.

35. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hills-dale, NJ: Lawrence Erlbaum; 1988.

36. Becker P. Restless legs syndrome. In: Lee-Chiong T, ed. Sleep: A Comprehen-sive Handbook. Hoboken, NJ: John Wiley & Sons; 2006:473-481.

37. Nowell PD, Buysse DJ, Reynolds CF III, Hauri PJ, Roth T, Stepanski EJ, ThorpyMJ, Bixler E, Kales A, Manfredi RL, Vgontzas AN, Stapf DM, Houck PR, KupferDJ. Clinical factors contributing to the differential diagnosis of primary insom-nia and insomnia related to mental disorders. Am J Psychiatry. 1997;154(10):1412-1416.

38. Davison ML, Gasser M, Ding S. Identifying major profile patterns in a popula-tion: an exploratory study of WAIS and GATB patterns. Psychol Assess. 1996;8(1):26-31.

39. Kim SK, Frisby CL, Davison ML. Estimating cognitive profiles using Profile Analy-sis via Multidimensional Scaling (PAMS). Multivariate Behav Res. 2004;39(4):595-624.




Sleep Medicine 10 (2009) 952–960

Contents lists available at ScienceDirect

Sleep Medicine

j ou rnal homepage: www.else v ier .com/l oca te /s leep

Fast Track Article

Epidemiological and clinical relevance of insomnia diagnosis algorithms according to the DSM-IV and the International Classification of Sleep Disorders (ICSD)

Maurice M. Ohayon a,*, Charles F. Reynolds III b

a Stanford Sleep Epidemiology Research Center, Stanford University School of Medicine, Stanford, CA 94303, USA b Department of Psychiatry, University of Pittsburgh School of Medicine, Pittsburgh, PA 15213, USA

a r t i c l e i n f o

Article history: Received 29 May 2009 Received in revised form 12 July 2009 Accepted 15 July 2009

Keywords: Insomnia Epidemiology Classifications Mental disorders

a b s t r a c t Background: Although the epidemiology of insomnia in the general population has received considerable attention in the past 20 years, few studies have investigated the prevalence of insomnia using operational definitions such as those set forth in the ICSD and DSM-IV, specifying what proportion of respondents satisfied the criteria to reach a diagnosis of insomnia disorder. Methods: This is a cross-sectional study involving 25,579 individuals aged 15 years and over representative of the general population of France, the United Kingdom, Germany, Italy, Portugal, Spain and Finland. The participants were interviewed on sleep habits and disorders managed by the Sleep-EVAL expert system using DSM-IV and ICSD classifications. Results: At the complaint level, too short sleep (20.2%), light sleep (16.6%), and global sleep dissatisfaction (8.2%) were reported by 37% of the subjects. At the symptom level (difficulty initiating or maintaining sleep and non-restorative sleep at least 3 nights per week), 34.5% of the sample reported at least one of them. At the criterion level, (symptoms + daytime consequences), 9.8% of the total sample reported having them. At the diagnostic level, 6.6% satisfied the DSM-IV requirement for positive and differential diagnosis. However, many respondents failed to meet diagnostic criteria for duration, frequency and severity in the two classifications, suggesting that multidimensional measures are needed. Conclusions: A significant proportion of the population with sleep complaints do not fit into DSM-IV and ICSD classifications. Further efforts are needed to identify diagnostic criteria and dimensional measures that will lead to insomnia diagnoses and thus provide a more reliable, valid and clinically relevant classification.

© 2009 Elsevier B.V. All rights reserved.

1. Introduction

Insomnia has been frequently studied in the general population of Western Europe and North America [1]. Few studies, however, have used well-defined criteria to assess insomnia; its presence relies mainly on positive answers to general questions about difficulties in initiating or maintaining sleep. Sometimes frequency or severity gradations have additionally been used to determine the presence of insomnia. Some epidemiological studies have investigated the occurrence of mental disorders in relation to insomnia symptoms, mainly with the use of anxiety or depression scales [1]. A high co-occurrence of insomnia symptoms and mental disorders has been reported in the general population [2–5]. But few of these studies have applied an operational definition of insomnia, although current classifications, such as the DSM-IV [6], the

* Corresponding author. Address: Stanford Sleep Epidemiology Research Center, Stanford University School of Medicine, 3430 W. Bayshore Road, Palo Alto, CA 94303, USA. Tel.: +1 650 494 1137; fax: +1 650 947 9813, +1 650 493 1225.

E-mail address: [email protected] (M.M. Ohayon).

International Classification of Sleep Disorders [7] and the Interna- tional Classification of Diseases [8], provide guidelines to assess insomnia. Each classification uses a funnel-shaped structure from a huge to a narrow definition: criteria, syndromes, episodes and diagnoses. ‘‘Criteria” refer to a set of symptoms or a guideline in a given diagnosis.

‘‘Insomnia” (like ‘‘pain”) has different meanings depending on the clinical picture presented by the subject. It can be a complaint (related to sleep quantity or quality), a symptom (part of a sleep disorder or of a mental or organic disorder) or a sleep disorder diagnosis (primary or secondary) implying the need for a differential diagnosis process. This distinction has been attempted in some epidemiological studies [2–5,9]. The clinical symptoms of insomnia, as described by the classifications, apply to a large part of the general population. It is unlikely, however, that all of these individuals suffer from an insomnia disorder [1]. Previous studies [3,10] have indicated that sleep dissatisfaction could be a better indicator of sleep pathology than insomnia criteria used by classifications like DSM-IV and the ICSD. Consequently, this report aims to document the prevalence of insomnia not only at the criterion

1389-9457/$ - see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.sleep.2009.07.008

http://www.sciencedirect.com/science/journal/13899457

http://www.elsevier.com/locate/sleep

mailto:[email protected]

http://dx.doi.org/10.1016/j.sleep.2009.07.008

M.M. Ohayon, C.F. Reynolds III / Sleep Medicine 10 (2009) 952–960 953

level, but also at the complaint level and at the diagnosis level (positive and a differential diagnosis). The final objective will be to show how the differential diagnosis evaluation could be essential to recognizing the true prevalence of insomnia and to improving clinical utility of insomnia diagnoses.

2. Methods

2.1. Sample

Individuals from seven European countries (France, the United

Kingdom, Germany, Italy, Portugal, Spain and Finland) were interviewed by telephone about their sleeping habits, sleep and health (mental and physical). Information about current medications (psychotropics, prescribed medications and over-the-counter drugs) was also collected, along with the indications as reported by the subject. In each country, a representative sample was drawn from the non-institutionalized population (Table 1). The targeted population consisted of 255,451,106 inhabitants, as of the latest census figures for each country at the time of the study. The same two-stage design was used in every country. First, phone numbers were randomly pulled according to the geographic distribution of the population per official census data available at the time. Sec- ond, the Kish selection procedure [11], a controlled selection method, was applied to maintain the representation of the sample according to age and gender and allowed for the selection of one respondent in the household. If the household member chosen by the Kish procedure refused to participate, the household was classified among the refusals, replaced by another telephone number in the same area, and the process repeated.

Verbal consent was obtained before the subjects were interviewed. For subjects younger than 18 years of age, the verbal consent of the parent(s) was also secured. All the studies were approved and supported by an ethics and research committee in each surveyed country: the Imperial College (UK), the Regensburg University (Germany), the San Rafaele Hospital (Italy), the Santa Maria Hospital (Portugal), the Hospital General Univ. Vall d’Hebron (Spain), the Haaga Neurological Research Centre (Finland) and a research ethics committee in Montreal (Canada) where the principal investigator (M.M.O.) was located at the time. Individuals with insufficient fluency in the national language, with a hearing or speech impairment or with an illness precluding the feasibility of an interview were excluded. The overall participation rate was 80.3% (Specific participation rates by country are in Table 1). Over- all, 25,579 subjects were interviewed.

2.2. Instrument

Lay investigators performed the interviews using the Sleep-

EVAL system [12,13]. Sleep-EVAL is specially designed to conduct

epidemiological studies of sleep habits and sleep and mental disorders in the general population. Interviews typically begin with a standard questionnaire composed of sociodemographic information, sleep/wake schedule, physical health and a series of questions related to sleep symptoms and symptoms of mental disorders. From the answers provided, the system elicits a series of diagnostic hypotheses (causal reasoning process) that are confirmed or re- jected throughout further questioning and by deductions of the consequences of each answer (non-monotonic, level-2 feature). The system allows concurrent diagnoses in accordance with the Diagnostic and Statistical Manual of Mental Disorders, forth edition (DSM-IV) [6], the International Classification of Sleep Disorders (ICSD) [7] and the International Classification of Diseases (ICD- 10) [8]. The differential diagnosis process is based on a series of key rules allowing or prohibiting the co-occurrence of two diagnoses in accordance with the classifications implemented in the system. The interview ends once all diagnostic possibilities are exhausted.

The latest validation studies were conducted with 105 patients at the Sleep Disorders Centers at Stanford University (USA) and Regensburg University (Germany) [14]. Patients at the Sleep Disor- ders Centers were interviewed twice: (1) by physicians using the Sleep-EVAL system who did not know the diagnoses given by the Sleep-EVAL system; and (2) by a senior sleep specialist clinician using his/her expertise to give one to three main diagnoses and one to 10 main symptoms per patient. This specialist was also ignorant of the diagnosis made by the Sleep-EVAL system. The Sleep-EVAL diagnoses were later compared to those of the sleep specialists. A kappa of .93 was obtained between the Sleep-EVAL system and the sleep specialists on Obstructive Sleep Apnea Syn- drome. The sensitivity of Sleep-EVAL for OSAS was 92.5% and the specificity was 100%. For all types of sleep-disordered breathing, the sensitivity of Sleep-EVAL was 98.2% and the specificity was 95.1%. Agreement for insomnia diagnoses was obtained in 96.9% of cases (kappa 0.78).

The duration of interviews ranged from 28 to 150 min. The longest interviews involved subjects with sleep disorders associated with mental disorders. Interviews were completed over two or more sessions if the duration of a session exceeded 60 min.

2.3. Variables 2.3.1. At the complaint level

Complaints were assessed throughout a reported severity of sleep dissatisfaction (subjects moderately or completely dissatisfied with their sleep). For the perception of sleep as being too short or too light, use of sleep-enhancing medication was also used as an exploratory criterion for dissatisfaction with the quality or quantity of sleep. The reported duration of sleep was compared to sleep schedule variables collected for all subjects of the sample.

Table 1 Description of the samples in each country.

Country Target population

Number of subjects interviewed

Number of solicited participants

Participation rate (%)

Date of the study

France 45,914,509 5622 6966 80.8 April 4, to July 12, 1993 United Kingdomt

45,709,600 4972 6249 79.6 June 14, 1994 to October 20, 1994 Germanyt

66,649,000 4115 6047 68.1 January 16, 1996 to October 16, 1996 Italyt

46,332,282 3970 4442 89.4 December 17, 1996 to April 5, 1997 Portugal* 8,300,000 1,853 2234 82.9 June 8,1998 to September 5, 1998 Spaint

38,500,000 4065 4648 87.5 September 28, 1998 to April 1, 1999 Finland* 4,045,715 982 1256 78.2 May 18, to June 30, September

to October 22, 2000 Total 255,451,106 25,579 31,842 80.3 t The target population was P15 years old. * The target population was P18 years old.

954 M.M. Ohayon, C.F. Reynolds III / Sleep Medicine 10 (2009) 952–960

2.3.2. At the symptom level The symptoms of insomnia as described in DSM-IV and ICSD

classifications were also evaluated:

• Difficulty Initiating Sleep (DIS): being quite or completely dissatisfied with sleep latency (at least 30 min) or reporting long sleep latency at sleep-onset as a major sleep problem.

• Difficulty Maintaining Sleep (DMS): Difficulty in resuming sleep after awakening in the night or identification of disrupted sleep as being a major problem.

• Early Morning Awakenings (EMA): short nocturnal sleep duration caused by an inability to resume sleep once awake).

• Non-Restorative Sleep (NRS): sleep of normal duration associated with a complaint of being unrested upon awakening or identification of fatigue at awakening as a major problem.

All these symptoms were required to occur at a frequency of

three nights per week or more.

2.3.3. At the criteria level Daytime repercussions were assessed through 15 questions an-

swered on a 5-point scale ranging from no impact to severe impact. These items covered cognitive functioning (memory, concentration, and efficacy); affective tone (irritability, anxiety, and depression); sensory irritability; and difficulty completing daily tasks (work, study or household). Subsequently, subjects identified which repercussions had an impact on their health, work or occupational activities, family and social life.

2.3.4. At the diagnosis level

Diagnoses were based upon diagnostic criteria specified in DSM-IV and ICSD. The differential diagnosis procedure required by the classifications was also applied (see Fig. 2 for DSM-IV classification). ICD-10 was used for specification of organic medical diseases.

2.4. Data analyses

The data were weighted to compensate for disparities between

the sample and the national census figures for the non-institutionalized population. Descriptive and qualitative variables were ana- lyzed using the chi-squared statistic. Ninety-five percent confidence intervals were calculated for prevalence rates and odds ratios of associations. Reported differences were significant at .05 or less.

3. Results

3.1. Demographic characteristics

Demographic characteristics for the whole sample are pre-

sented in Table 2. Gender and age distribution were comparable among the countries studied. The Italian, Spanish and Portuguese samples had lower rates of separated or divorced subjects than in other countries. A higher proportion of homemakers was found in the Italian and Spanish samples and, consequently, a lower rate of daytime workers compared to the other samples.

3.2. At the complaint level

Surveyed subjects slept on average 7 h 9 min. The youngest

subjects slept on average 30 min more than the oldest individuals of the sample (Table 3). Dissatisfaction with sleep duration was expressed by 18.2% of the sample. There was little variation between the age categories with the exception of the youngest subjects,

who expressed less dissatisfaction than middle-aged individuals (35–64 years old). Similarly, the perception of sleep as being too short was reported by 20.2% of the sample. The prevalence significantly decreased with age.

The perception of having light sleep occurred in 16.6% of the sample and significantly increased with age. Global sleep dissatisfaction was reported by 8.2% of the sample. The prevalence was mostly stable across age groups; significant differences were observed only between subjects less than 35 years old and subjects 35 years and over. 3.3. At the symptom level

All subjects were queried about insomnia symptomatology according to DSM-IV and ICSD criteria, as defined by having either DIS, DMS, EMA or NRS at least 3 nights per week. These symptoms were frequently reported at a criterion level in the European sample studied. The median duration of insomnia symptoms was 60 months.

As seen in Table 4, DMS was the most commonly reported symptom with a prevalence of 23.1%. DIS, EMA and NRS were reported in similar proportions. DIS, DMS and EMA significantly increased with age, while NRS significantly decreased in the two older age groups compared to subjects aged less than 55 years.

3.4. At the diagnosis level

The DSM-IV describes the essential features of insomnia disorder as (1) being a predominant complaint of insomnia, (2) lasting for at least one month, and (3) causing significant distress or daytime impairments in social, occupational or other sectors of daily life. The presence of all three DSM-IV criteria was observed in 9.8% of the sample.

We found that about half of the subjects who met the essential DSM-IV criteria of insomnia also met criteria for a mental disorder. Anxiety disorders were the most frequent, with Generalized Anxi- ety Disorder accounting for about half. Mood disorders, mostly major depressive disorder, were the second most frequent disorders (Fig. 1). These subjects also had the highest rate of comorbid mood and anxiety disorders. About 25% of subjects who reported insomnia symptoms for less than one month or without significant impairment of functioning had a mental disorder (Fig. 1).

Subsequently, the differential diagnosis process as outlined by the DSM-IV was applied to subjects who met the three DSM-IV criteria cited above. The positive and differential diagnosis process is shown in Fig. 2. As can be observed, when insomnia was accompanied by a general medical condition or substance use (or abuse) responsible for the insomnia symptoms, subjects received a diagnosis of ‘‘Sleep Disorder Due to a General Medical Condition, Insomnia type,” or a diagnosis of ‘‘Substance-Induced Sleep Disor- der, Insomnia type.”

If a mental disorder was currently present, it was further determined if the insomnia complaint constituted the predominant characteristic of the symptomatology or was simply associated with it. This was ascertained using a series of additional questions assessing the importance of insomnia complaints compared with mental disorders. In the case of the former, subjects received a diagnosis of ‘‘Insomnia Related to Another Mental Disorder” and, in the case of the latter, the insomnia disorder diagnosis was ruled out and a mental disorder diagnosis was given.

Insomnia without a mental or general medical condition and without substance use (or abuse) led to a diagnosis of ‘‘Primary Insomnia.” Finally, subjects with characteristics of several forms of insomnia were classified in the ‘‘Dyssomnia not otherwise specified” category.


Table 2 Demographic characteristics of the sample.

N % s.e.

Gender Female 13,295 52.0 0.29 Male 12,284 48.0 0.23

Age groups

<25 4606 18.0 0.40 25–34 4794 18.7 0.41 35–44 4350 17.0 0.44 45–54 3739 14.6 0.51 55–64 3400 13.3 0.54 P65 4690 18.3 0.45

Marital status

Single 7360 28.8 0.31 Married-common law 14,284 55.8 0.25 Separated-divorced 1316 5.1 0.98 Widowed

Occupation

2616 10.2 0.66

Shift or night worker 2018 7.9 0.70 Daytime worker 10,578 41.4 0.27 Unemployed 1383 5.4 0.91 Student 2242 8.8 0.53 Retired 5142 20.1 0.43 Homemaker 4215 16.5 0.50

Education

No education 1650 6 4 0 70

9 years or less 8538 33.4 0.91 11–12 years 9054 35.4 0.53 13–14 years 3341 13.1 0.43 15+ 2352 9.2 0.50 Master-PhD 646 2.5 0.27

As can be observed in Table 5, Primary Insomnia and Insomnia Related to Another Mental Disorder were the two most frequent diagnoses, with a prevalence of 3% for Primary Insomnia, and 1.5% for Insomnia Related to Another Mental Disorder.

The International Classification of Sleep Disorders was used in all the countries except France. As shown in Table 6, subjects with DSM-IV Primary Insomnia split mainly among three diagnoses in the ICSD: Psychophysiological Insomnia, Idiopathic Insomnia, and Insufficient sleep syndrome.

The DSM-IV diagnosis ‘‘Insomnia Related to Another Mental Disorder” has no equivalent in the ICSD. The mental disorder diagnoses can be divided into three ICSD diagnoses: Mood disorder associated with sleep disturbances; Panic disorder associated with

sleep disturbances; and Anxiety disorder associated with sleep disturbances. Thus, the prevalences between the two classifications cannot be readily compared because the diagnoses are not defined similarly. 3.5. Reported consequences of insomnia

For an insomnia diagnosis, the DSM-IV and ICSD classifications require that the subject have daytime repercussions in different areas of life (health, familial, social or work functioning) due to a sleep disturbance.

About 40% of subjects reported at least mild consequences of their insomnia symptoms. Among subjects reporting daytime consequences: • 44% reported that insomnia affected their health. Nearly two-

thirds (61.6%) said the impact on health was due to physical fatigue or weariness.

• 30% of subjects reported an impact on their work and/or daily activities. Most frequently reported causes of the impact on work or activities were physical fatigue or weariness (53.7%), mental or intellectual fatigue (49.5%), irritability (39.3%) and decreased efficiency (33.8%).

• 31.3% of subjects mentioned effects on family relationships. The most frequently reported causes were irritability (58.5%), physical fatigue (49.2%), mental fatigue (36.7%) and depressive mood (34.8%).

• Finally, 24% said insomnia caused problems in social relationships. Main reported causes affecting social relationships were irritability (55.4%), physical fatigue (53.6%), mental fatigue (36.5%) and depressive mood (35.2%).

3.6. Association with organic (medical) diseases

Insomnia symptomatology is often exacerbated as a result of an organic disease. Musculskeletal or articular diseases (arthritis, backache, pain in limbs, etc.) were the most frequently associated: 14.6% of subjects meeting insomnia criteria suffered from such a disease as compared to 3.4% non-insomnia subjects (OR 4.9 [95% CI: 4.3–5.7]). Heart diseases ranked second, with 5.4% of insomnia subjects having a cardiovascular disease compared to 2.7% in the non-insomnia subjects (OR 2.1 [95% CI: 1.7–2.6]). Gastrointestinal diseases and pulmonary diseases were third in prevalence, with 2.3% of insomnia subjects for each disease compared to 1.0% and

Table 3 Prevalence of sleep complaints and sleep duration.

Age groups

<25 25–34 35–44 45–54 55–64 P65

% [95% Cl] % [95% Cl] % [95% Cl] % [95% Cl] % [95% Cl] % [95% Cl]

Total

% [95% Cl]

Sleep duration <5:00 1.0 [0.7–1.2] 2.0 [1.6–2.3] 2.2 [1.8–2.7] 3.0 [2.5–3.6] 4.6 [3.9–5.3] 6.2 [5.5–6.9] 3.1 [2.9–3.3] 5:00–5:59 3.9 [3.4–4.5] 5.3 [4.6–5.9] 6.7 [6.0–7.4] 8.1 [7.3–9.0] 10.0 [9.0–11.0] 10.9 [10.0–11.8] 7.3 [7.0–7.7] 6:00–6:59 12.4 [11.4–13.3] 19.7 [18.6–20.8] 22.7 [21.4–23.9] 23.8 [22.5–25.2] 22.0 [20.6–23.4] 21.0 [19.8–22.2] 20.0 [19.5–20.5] 7:00–7:59 28.3 [27.0–29.6] 36.1 [34.8–37.5] 37.6 [36.1–39.0] 35.2 [33.6–36.7] 31.3 [29.7–32.8] 25.7 [24.4–26.9] 32.3 [31.7–32.8] 8:00–8:59 39.0 [37.6–40.4] 31.0 [29.7–32.3] 26.8 [25.5–28.1] 24.9 [23.5–26.3] 25.1 [23.7–26.6] 26.4 [25.1–27.7] 29.2 [28.7–29.8] P9:00 15.5 [14.4–16.5] 5.9 [5.2–6.6] 4.1 [3.5–4.7] 5.0 [4.3–5.7] 7.0 [6.2–7.9] 9.9 [9.0–10.7] 8.1 [7.7–8.4] Mean (s.d.)a 7:30 (1:10) 7:12 (1:06) 7:03 (1:04) 6.59 (1:10) 6:56 (1:19) 6:58 (1:30) 7:09 (1:15)

Dissatisfaction 16.1 [14.8–17.3] 17.3 [16.0–18.5] 19.1 [17.7–20.5] 20.3 [18.8–21.8] 19.4 [17.9–21.0] 17.9 [16.6–19.2] 18.2 [17.6–18.7] duration sleepa

Too short durationa 21.3 [19.9–22.7] 24.4 [23.0–25.8] 23.9 [19.4–22.4] 20.9 [19.4–22.4] 16.1 [14.6–17.5] 14.0 [12.8–15.1] 20.2 [19.6–20.8] Light sleepa 9.9 [8.9–10.9] 13.8 [12.7–14.9] 16.2 [17.8–20.6] 19.2 [17.8–20.6] 19.9 [18.4–21.4] 21.5 [20.2–22.9] 16.6 [16.0–17.1] Global sleep 5.8 [5.1–6.5] 6.4 [5.7–7.1] 8.6 [8.8–10.7] 9.7 [8.8–10.7] 10.0 [9.0–11.0] 9.7 [8.8–10.5] 8.2 [7.9–8.6]

dissatisfactiona

Use of sleep medicationa 1.0 [0.7–1.3] 1.8 [1.5–2.2] 3.1 [5.4–7.0] 6.2 [5.4–7.0] 9.7 [8.7–10.7] 14.6 [13.6–15.6] 5.9 [5.6–6.2]

a p < 0.0001 between age groups.


Table 4 Prevalence of insomnia symptoms by age groups.

Difficulty initiating sleepa

Difficulty maintaining sleepa

Early morning awakeninga

Non-restorative sleepa

Total insomnia symptomsa

Age groups <25 25–34 35–44 45–54 55–64 P65 Total % [95% Cl] % [95% Cl] % [95% Cl] % [95% Cl] % [95% Cl] % [95% Cl] % [95% Cl] 10.4 [9.5–11.3] 7.9 [7.2–8.7] 8.5 [10.0–12.0] 11.0 [10.0–12.0] 13.1 [11.9–14.2] 14.9 [13.9–16.0] 10.9 [10.5–11.2] 12.4 [11.5–13.4] 15.2 [14.2–16.3] 18.8 [21.8–24.5] 23.1 [21.8–24.5] 33.3 [31.7–34.9] 38.7 [37.3–40.1] 23.1 [22.6–23.7] 9.1 [8.3–10.0] 8.6 [7.8–9.4] 10.4 [13.3–15.5] 14.4 [13.3–15.5] 16.3 [15.1–17.6] 16.7 [15.6–17.8] 12.3 [11.9–12.7] 12.1 [11.1–13.0] 11.5 [10.6–12.4] 11.5 [10.9–13.0] 11.9 [10.9–13.0] 9.9 [8.9–10.9] 9.4 [8.5–10.2] 11.1 [10.7–11.5] 26.6 [25.3–27.9] 27.2 [25.9–28.4] 29.6 [28.2–30.9] 34.4 [32.9–36.0] 42.0 [40.3–43.7] 47.7 [46.3–49.1] 34.5 [33.7–34.8]

a p < 0.0001 between age groups.

0.9% respectively for the non-insomnia subjects (OR 2.30 [95% CI: 1.81–2.92] for gastrointestinal diseases and OR 2.41 [95% CI: 1.89–3.09] for pulmonary diseases).

4. Discussion

4.1. Insomnia diagnosis (disorder) prevalence has been under- investigated in the general population

Epidemiological studies in the general population have assessed

mainly insomnia symptomatology and, on rare occasions, have used sets of criteria to determine the prevalence and severity of insomnia diagnoses (disorders) [2,3,9]. Furthermore, there has been no attempt to frame a differential diagnostic process to arrive at reliable insomnia diagnoses. Therefore, this large-scale study is the first to explore insomnia criteria and sleep complaints with respect to the positive and differential diagnostic processes used by two classifications (DSM-IV and ICSD).

Fig. 1. Distribution of mental disorders among insomnia subjects. These rates represent the association between mental disorders and insomnia complaints regardless of the final diagnostic issue (positive diagnosis).

4.2. How was DSM-IV classification able to identify candidate subjects for insomnia?

The DSM-IV classification process resulted in about 19% of the population being identified as having either an insomnia diagnosis or a mental disorder diagnosis with insomnia criteria and/or sleep complaints (Diagram 1). About 9% of subjects complained about their sleep and met insomnia criteria. This stratum of the population is of particular interest because they are potential candidates to seek help for their insomnia. Even when accounting for other sleep disorder diagnoses that need to be ruled out, we still have about 15% of the population that does not fit into a diagnostic category. Furthermore, only a third of elderly subjects with a predominant complaint of insomnia have obtained an insomnia diagnosis, in contrast to half of the younger subjects. This is likely due to a failure of the elderly to report the daytime consequences of insomnia and/or for the classifications to propose adequate criteria for this age category. It is possible that since most of them are retired, widowed or living alone, there is a reduced probability that they or others have noticed the consequences of insomnia. A possibility, offered by the DSM-IV classification, would have been to classify these subjects into the category of ‘‘Dyssomnia not otherwise specified.” But this solution is not acceptable—how indeed can we classify close to half of these subjects with complaints in a ‘‘catchall” category? At the very least, this observation suggests that further refinement is needed in the classification to better fit the reality of the population. In this context, we believe that the addition of a dimensional evaluation to each of the traditional categorical diagnoses, measuring severity, distress or impairment, would enhance the quality (reliability and validity) of the insomnia diagnosis for both clinical and research applications. For example, to assess whether a patient is responding to treatment, clinicians need the information provided by dimensional measures of severity, distress and impairment [15]. 4.3. How valid is the ICSD classification in identifying insomnia subjects?

The excessive number of diagnoses in the ICSD (about 20 different diagnoses having insomnia as a main or an associated feature) resulted in a lower prevalence of these diagnoses and, therefore, a less clinically convenient (or easy to use) procedure to recognize subjects with insomnia diagnosis. Many of the diagnoses have a prevalence of less than 1% in the general population. Such is the case for idiopathic insomnia; environmental sleep disorder; sleep-onset association disorder; nocturnal eating (drinking) syndrome; hypnotic, stimulant or alcohol-dependent


Fig. 2. DSM-IV Sleep-EVAL decision tree.

sleep disorders. While such nuanced classification may have its utility in specialty sleep disorder clinics, it hardly fits the reality of the general population and of general medical practice. This can be observed in Diagram 2, where 8.8% of the sample have a sleep complaint and have insomnia criteria without having a diagnosis. Furthermore, the approach to find a diagnosis of ‘‘mental disorder associated with sleep disturbances” is very different from that proposed by the DSM-IV. In the ICSD, the sleep disturbances must be caused by the mental disorder, i.e., the sleep disturbances occurred at the same time or after the mental disorder and relief is observed with the amelioration of the mental disorder. In the DSM-IV, these subjects receive a mental disorder diagnosis because the sleep disturbance is part of the mental disorder. A diagnosis of Insomnia Related to Another Mental Disorder is

given only when insomnia is the predominant complaint. Usually, the subject reported that his/her insomnia bothered him/her much more than the depressive or anxious mood and that better sleep would bring an amelioration of the mood. DSM-IV begs the question whether insomnia persisting beyond the resolution of other symptoms of depression or anxiety should be classified as Primary Insomnia or as insomnia related to the (prior) mental disorder.

Almost half of our insomnia subjects reported that this sleep problem had an impact on their health mainly in the form of fatigue or weariness resulting from poor sleep. Another third reported a detrimental effect on their work or interpersonal relationships (social, friendship, and family), with the two most frequently reported complaints being fatigue and irritability. Again,


Table 5 Prevalence of DSM-IV insomnia disorder diagnoses.

Women (n = 13,295) Men (n = 12,284) Total (n = 25,579) Insomnia criteria resulted in: % [95% CI] % [95% CI] % [95% CI]

Insomnia disorder diagnoses Primary Insomniaa

3.9

[3.6–4.2]

2.0

[l.8–2.3]

3.0

[2.8–3.2]

Substance-Induced Sleep Disorder, Insomnia type 1.0 [0.9–1.2] 0.8 [0.6–1.0] 0.9 [0.8–1.0] Insomnia Related to Another Mental Disordera

2.0 [1.7–22] 1.0 [0.8–1.2] 1.5 [1.4–1.7] Sleep disorder due to a general medical condition, insomnia typea

1.5 [1.3–1.7] 0.7 [0.6–0.9] 1.1 [1.0–1.2] Dyssomnia not otherwise specified 0.2 [0.1–0.3] 0.1 [0.0–0.2] 0.1 [0.1–0.1] Total insomnia diagnosesa 8.6 [8.1–9.1] 4.6 [4.2–5.0] 6.6 [6.3–6.9]

Psychiatric Diagnoses Mood disorders*,a

3.8

[3.5–4.1]

1.4

[2.1–2.7]

3.1

[2.9–3.3]

Anxiety disorders*,a 8.2 [7.7–8.6] 5.2 [4.8–5.6] 6.8 [6.4–7.1]

Adjustment disorders*,a 0.5 [0.4–0.6] 0.3 [02–0.4] 0.4 [0.3–0.5]

* Prevalence excludes positive cases without a complaint of insomnia. a p < 0.0001 between gender.

Table 6 Prevalence of selected ICSD sleep disorder diagnoses.

Women (n = 10,371) Mean (n = 9590) Total (n = 19,961) Insomnia criteria resulted in: % [95% CI] % [95% CI] % [95% CI]

Psychophysiological Insomnia 1.7 [1.5–2.0] 1.1 [0.9–1.3] 1.4 [1.2–1.6] Circadian rhythm disorders* 1.0 [0.8–1.1] 0.7 [0.6–0.9] 0.8 [0.7–1.0] Idiopathic Insomnia 0.6 [0.5–0.8] 0.5 [0.4–0.6] 0.6 [0.5–0.7] Insufficient sleep syndrome* 05 [0.7–1.1] 0.8 [0.6–1.0] 0.9 [0.7–1.0] Obstructive sleep apnea syndrome*,a

16 [1.4–1.9] 2.6 [2.3–2.9] 2.1 [1.9–2.3] Mood disorder associated with sleep disturbances*,a

3.1 [2.7–3.4] 1.6 [1.4–1.9] 2.4 [2.2–2.6] Panic disorder associated with sleep disturbances* 0.9 [0.7–1.1] 0.4 [0.2–0.5] 0.6 [0.5–0.7] Anxiety disorder associated with sleep disturbances*,a

1.5 [1.3–1.8] 0.7 [0.5–0.8] 1.1 [1.0–1.3]

* Prevalence excludes positive cases without a complaint of insomnia. a p < 0.0001 between gender.

Diagram 1. Overlapping between sleep complaints, insomnia symptoms and DSM-IV diagnoses in the general population. DSM-IV Insomnia diagnoses: Primary Insomnia; Substance-induced sleep disorder, insomnia type; Insomnia Related to Another Mental Disorder; Sleep disorder due to a general medical condition, insomnia type; Dyssomnia not otherwise specified.

these issues have been poorly addressed in previous epidemiological studies, although they are well-documented consequences of insomnia in clinical studies. Daytime functioning of the insomnia patient is characterized by tiredness, without necessarily a shorter sleep-onset delay at the MSLT [16,17], and concentration and memory problems [18,19].

4.4. Links with consultations and treatments

A way of knowing whether an epidemiological study is describing the reality of a complaint is to explore how many individuals are seeking medical help or are using medications to correct their problems. The use of sleep-enhancing medications varied greatly


Diagram 2. Overlapping between sleep complaints, insomnia symptoms and ICSD diagnoses in the general population. ICSD Diagnoses used: idiopathic insomnia; Psychophysiological Insomnia; Insufficient sleep syndrome; Mood disorder associated with sleep disturbances; Panic disorder associated with sleep disturbances; Anxiety disorder associated with sleep disturbances; Circadian rhythm disorders; Obstructive sleep apnea syndrome; Periodic limb movement disorder; Restless legs syndrome; environmental sleep disorder; sleep-onset association disorder; nocturnal eating (drinking) syndrome; hypnotic, stimulant or alcohol-dependent sleep disorders.

among the countries studied. Generally speaking, Mediterranean countries (France, Portugal, and Italy) used such medications more frequently than did the United Kingdom and Germany. But the usage of sleep-enhancing medications increased with age in all countries and was higher in women than in men. Medical consultations for insomnia were requested by only one-fourth of insomnia diagnosis subjects. When the insomnia was related to another mental disorder, the rate of consultation increased to more than 40%. Among these subjects, 29.3% received medication, typically an antidepressant or an anxiolytic. For the 60% of subjects who did not report their sleep disturbances to their physician but who consulted him/her in the previous year, only 7.2% got a medication. This indicates that if patients do not take the initiative to let their physician know about their sleep disturbances, there is little chance that the physician will explore sleep disturbances or mental disorders in the patient. Studies conducted with general practitioners have documented this phenomenon on many occasions [19–21]. The latter point and the findings of this study emphasize that educational efforts are needed to increase the recognition of sleep disorders.

4.5. Finally, given the number of people reporting chronic sleep complaints, how many are not recognized (diagnosed) by the classifications?

According to our data, a lot of them are not recognized and stay

out of reach of a proper identification. This shortcoming could have several implications. First, some criteria may be missing in the classification: the four major insomnia symptoms are insufficient to explain the prevalence of sleep dissatisfaction. For example, nocturnal awakening is the most prevalent symptom of insomnia and is frequently associated with the classic insomnia symptoms [22]. Second, measures of the dimensional aspects of the symptomatology are weak: more descriptive dimensions must be used. For example, classifications could use dimensions such as chronicity, severity, comorbidity, age, etc., but the most prevalent and specific dimensions are still to be defined for insomnia. Third, the primary or secondary etiology of insomnia and its comorbidity must be described and more carefully documented. Fourth, chronicity, duration and frequency are still underestimated by the two classifications: 60% of all the insomniacs of the general population have their problem for more than five years.

4.6. Limitations

Obviously, this study is not without limitations. First, lay interviewers were used to read the questions during the interviews. This situation, however, is not different from nearly all the large epidemiological studies. The main difference with other surveys is that the decision of the presence or absence of a symptom, criteria or diagnosis was taken by the Sleep-EVAL system, not the interviewer, whose only task was to enter the answers of the participant. Also, the questions were formulated in a simple man- ner, using everyday words, in order to be understood by individuals with little education. There are many advantages using computerized diagnostic tool, such as Sleep-EVAL, e.g., shorter assessment times, customized and standardized questionnaire administration, elimination of separate manual data entry stage, and considerable reduction in missing data.

Second, the total sample is a combination of samples across surveys from seven countries. While separate analyses of the data from these national surveys indicate similar results for sleep disorders (and therefore the data can be combined), the fact remains that the sample represents a composite. Still, we believe the results indicate the strategy was defensible.

Third, the question of the reliability of sleep data collected by telephone could be raised, but the literature suggests telephone interviews in general are appropriate and yield results comparable to other strategies. Studies report survey modes appear to influence respondents equally [23–25] whether the outcome of interest is alcohol [26] or substance abuse [27]. Rohde et al. [28] report good inter rater reliability between face to face and telephone interviews assessing DSM-IV psychiatric disorders with adults.

Fourth, in this study, sleep disorder diagnoses were based solely on self-reported symptoms. Sets of minimum criteria, as described in the ICSD and DSM-IV classifications, were used. We did not have physiological parameters on sleep generated by procedures such as electroencephalography (EEG). While such measures are desirable for some diagnoses, to date they have not been regularly incorpo- rated into community-based, epidemiologic studies. Thus, while such data would be useful to have, self-reports and interview based measures remain the most widely used measures in community surveys. Our study was no exception. Furthermore, insomnia diagnoses do not require polysomographic results for the diagnosis.


5. Conclusions

Our data illustrate the importance of undertaking epidemiological studies of insomnia at several levels: complaint, criterion and diagnosis of a disorder. The distinction among these levels is of major interest:

(1) ‘‘Complaint” reflects population demand in term of needs

and can be used to evaluate the efficiency of health care providers in recognizing and meeting these needs. Study at the level of complaints may also aid in understanding the motives behind help-seeking and how help-seeking relates to symptom burden.

(2) Study at the criterion level gives an indication of the value of grouping symptoms to define clinical entities easily and reli- ably recognized by health care providers. In turn, this allows for efforts at validation of diagnostic entities as specified by the classifications. An investigation of criterion level also presents the opportunity to explore manifestations of illness for which subjects may ignore or not understand the relationship between their symptomatology and pathology or impairment.

(3) Study at the level of diagnoses allows for the possibility of evaluating the effectiveness of managing and treating disorders. The equivalent results observed between self-report complaints and classification criteria speak for validating a classification in its double role (identification, treatment), leading to better education and training of general practitioners, who are the most frequently consulted providers.

The difficulty of classification systems in correctly identifying

insomnia may reflect, in part, the dilemma of society concerning the need for sleep: lost sleep versus lost time sleeping. As empha- sized in the recent Institute of Medicine report, ‘‘Sleep Disorders and Sleep Deprivation: an Unmet Public Health Need” [29], our society needs to come to terms with the public health burden of insomnia and other sleep disorders. Doing so will be facilitated by the availability of measurement tools that accurately capture the true extent of the problem.

Acknowledgements

Funding/Support: This study was supported by National Insti- tutes of Health Grant R01NS044199 (Dr. Ohayon); UPMC (Univer- sity of Pittsburgh Medical Center) Endowment in Geriatric Psychiatry (Dr. Reynolds) and P30 MH71977 (Dr. Reynolds).

Role of the sponsor: The sponsors had no influence on the design, conduct, or analysis of this study.

Data access and responsibility: The first author (Dr. Ohayon) takes responsibility for the integrity of the data and the accuracy of the data analysis. All authors had full access to all the data in the study.

References

[1] Ohayon MM. Epidemiology of insomnia: what we know and what we still need to learn. Sleep Med Rev 2002;6:97–111.

[2] Ohayon MM. Prevalence of DSM-IV diagnostic criteria of insomnia: distinguishing insomnia related to mental disorders from sleep disorders. J Psychiatr Res 1997;31:333–46.

[3] Ohayon MM, Caulet M, Priest RG, Guilleminault C. DSM-IV and ICSD-90 insomnia symptoms and sleep dissatisfaction. Br J Psychiatry 1997;171:382–8.

[4] Breslau N, Roth T, Rosenthal L, Andreski P. Sleep disturbance and psychiatric disorders: a longitudinal epidemiological study of young adults. Biol Psychiatry 1996;39:411–8.

[5] Ford DE, Kamerow DB. Epidemiologic study of sleep disturbances and psychiatric disorders. An opportunity for prevention? JAMA 1989;262:1479–84.

[6] APA (American Psychiatric Association). Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV). Washington: The American Psychiatric Association, 1994.

[7] American Academy of Sleep Medicine. International Classification of Sleep Disorders: Diagnostic and Coding Manual (ICSD). Rochester, Minnesota: American Sleep Disorders Association, 1990 [revised in 1997; 2nd ed., 2005].

[8] World Health Organization. International Statistical Classification of Diseases and Related Health Problems (ICD). Tenth revision. 2nd ed. Geneva: World Health Organization; 1994.

[9] Leger D, Guilleminault C, Dreyfus JP, Delahaye C, Paillard M. Prevalence of insomnia in a survey of 12,778 adults in France. J Sleep Res 2000;9:35–42.

[10] Ohayon MM, Caulet M, Guilleminault C. Complaints about nocturnal sleep: how a general population perceives its sleep, and how this relates to the complaint of insomnia. Sleep 1997;20:715–23.

[11] Kish L. Survey sampling. New York: John Wiley & sons Inc; 1965. [12] Ohayon M. Knowledge based system Sleep-EVAL: decisional trees and

questionnaires. Ottawa: National Library of Canada; 1995. [13] Ohayon M. Improving decision making processes with the fuzzy logic

approach in the epidemiology of sleep disorders. J Psychosom Res 1999;47:297–311.

[14] Ohayon MM, Guilleminault C, Zulley J, Palombini L, Raab H. Validation of the Sleep-EVAL system against clinical assessments of sleep disorders and polysomnographic data. Sleep 1999;22:925–30.

[15] Kraemer HC. DSM categories and dimensions in clinical and research contexts. Int J Methods Psychiatr Res 2007;16(Suppl. 1):S8–S15.

[16] Stepanski E, Zorick F, Roehrs T, Young D, Roth T. Daytime alertness in patients with chronic insomnia compared with asymptomatic control subjects. Sleep 1988;11:54–60.

[17] Sugerman JL, Stern JA, Walsh JK. Daytime alertness in subjective and objective insomnia: some preliminary findings. Biol Psychiatry 1985;20:741–50.

[18] Mendelson WB, Garnett D, Gillin GC, Wengartner H. The experience of insomnia and night-time functioning. Psychiatry Res 1986;19:235–50.

[19] Hohagen F, Kappler C, Schramm E, Rink K, Weyerer S, Riemann D, et al. Prevalence of insomnia in elderly general practice attenders and the current treatment modalities. Acta Psychiatrica Scandinavica 1994;20(2):102–8.

[20] Everitt DE, Avron J, Baker MW. Clinical decision-making in the evaluation and treatment of insomnia. Am J Med 1990;89:357–62.

[21] Ohayon MM, Caulet M, Arbus L, Billard M, Coquerel A, Guieu JL, et al. Are prescribed medications effective in the treatment of insomnia complaints? J Psychosom Res 1999;47:359–68.

[22] Ohayon MM. Nocturnal awakenings and comorbid disorders in the American general population. J Psychiatr Res 2008;43:48–54.

[23] Crippa JA, de Lima Osório F, Del-Ben CM, Filho AS, da Silva Freitas MC, Loureiro SR. Comparability between telephone and face-to-face structured clinical interview for DSM-IV in assessing social anxiety disorder. Perspect Psychiatr Care 2008;44:241–7.

[24] Senior AC, Kunik ME, Rhoades HM, Novy DM, Wilson NL, Stanley MA. Utility of telephone assessments in an older adult population. Psychol Aging 2007;22:392–7.

[25] Hill J, McVay JM, Walter-Ginzburg A, Mills CS, Lewis J, Lewis BE, et al. Validation of a brief screen for cognitive impairment (BSCI) administered by telephone for use in the medicare population. Dis Manag 2005;8:223–34.

[26] Slutske WS, True WR, Scherrer JF, Goldberg J, Bucholz KK, Heath AC, et al. Long- term reliability and validity of alcoholism diagnoses and symptoms in a large national telephone interview survey. Alcohol Clin Exp Res 1998;22:553–8.

[27] Aktan GB, Calkins RF, Ribisl KM, Kroliczak A, Kasim RM. Test–retest reliability of psychoactive substance abuse and dependence diagnoses in telephone interviews using a modified Diagnostic Interview Schedule-Substance Abuse Module. Am J Drug Alcohol Abuse 1997;23:229–48.

[28] Rohde P, Lewinsohn PM, Seeley JR. Comparability of telephone and face to face interviews in assessing axis I and II disorders. Am J Psychiatry 1997;154:1593–8.

[29] Committee on Sleep Medicine and Research. Sleep Disorders and Sleep Deprivation: an Unmet Public Health Need. Washington, DC: Institute of Medicine: National Academies Press, 2006.

Sleep Medicine xxx (2011) xxx–xxx

Contents lists available at SciVerse ScienceDirect

Sleep Medicine

journal homepage: www.elsevier .com/locate /s leep

Original Article

Hierarchy of insomnia criteria based on daytime consequences

Maurice M. Ohayon a,⇑, Dieter Riemann b, Charles Morin c, Charles F. Reynolds III d

a Stanford Sleep Epidemiology Research Center, Stanford University School of Medicine, Stanford, CA, USAb Section for Clinical Psychology & Psychophysiology/Sleep Medicine, Department of Psychiatry & Psychotherapy, Freiburg University Medical Center, Germanyc Laval University, Quebec, Canadad Department of Psychiatry, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA

a r t i c l e i n f o

Article history:Received 8 November 2010Received in revised form 28 May 2011Accepted 24 June 2011Available online xxxx

Keywords:InsomniaDSM – 5 classificationEpidemiologyGlobal sleep dissatisfactionDaytime consequencesNon-restorative sleep

1389-9457/$ - see front matter � 2011 Elsevier B.V. Adoi:10.1016/j.sleep.2011.06.010

⇑ Corresponding author. Address: Stanford Sleep EpStanford University School of Medicine, 3430 W. B94303, USA. Tel.: +1 650 494 1137; fax: +1 650 947 9

E-mail address: [email protected] (M.M. Oha

Please cite this article in press as: Ohayon MMj.sleep.2011.06.010

a b s t r a c t

Objectives: To explore how insomnia symptoms are hierarchically organized in individuals reporting day-time consequences of their sleep disturbances.Methods: This is a cross-sectional study conducted in the general population of the states of California,New York, and Texas. The sample included 8937 individuals aged 18 years or older representative ofthe general population. Telephone interviews on sleep habits and disorders were managed with theSleep-EVAL expert system and using DSM-IV and ICSD classifications. Insomnia symptoms and globalsleep dissatisfaction (GSD) had to occur at least three times per week for at least three months.Results: A total of 26.2% of the sample had a GSD. Individuals with GSD reported at least one insomniasymptom in 73.1% of the cases. The presence of GSD in addition to insomnia symptoms considerablyincreased the proportion of individuals with daytime consequences related to insomnia. In the classifica-tion trees performed, GSD arrived as the first predictor for daytime consequences related to insomnia.The second predictor was nonrestorative sleep followed by difficulty resuming sleep and difficulty initi-ating sleep.Conclusions: Classification trees are a useful way to hierarchically organize symptoms and to help diag-nostic classifications. In this study, GSD was found to be the foremost symptom in identifying individualswith daytime consequences related to insomnia.

� 2011 Elsevier B.V. All rights reserved.

1. Introduction

For more than 30 years, the assessment of insomnia in epidemi-ologic studies was limited to the identification of insomnia symp-toms in terms of their prevalence and associations with mentaldisorders and medical conditions [1]. The DSM-III-R [2], its succes-sor the DSM-IV [3], and the second International Classification ofSleep Disorders (ICSD-II) [4] have represented important steps inthe recognition and placement of each symptom. Progressively,four main insomnia symptoms were retained: difficulty initiatingsleep (DIS), difficulty maintaining sleep (DMS), early morningawakenings (EMA), and nonrestorative sleep (NRS). The assess-ment of daytime consequences was also mandatory in the DSM-III-R and subsequently in the DSM-IV and the ICSD-II.

Meanwhile, several studies have demonstrated that the preva-lence of insomnia symptoms was dependent on their frequencyduring the week as opposed to the presence/absence of the

ll rights reserved.

idemiology Research Center,ayshore Road, Palo Alto, CA813/650 493 1225.yon).

et al. Hierarchy of insomnia

symptoms [1]. Symptom duration (for example, one month orlonger), however, was seldom investigated, although its impor-tance has been outlined in classifications for nearly three decades.Current classifications, DSM-IV and ICSD-II, have arbitrarily setone month as the criterion for defining insomnia disorder. Thisposes a difficulty, since it is unlikely that a 1-month symptomduration characterizes a chronic health problem.

Some researchers have argued that the problem of insomniaprevalence, as defined until now, has resulted from the absenceof crucial elements in the insomnia diagnosis. Indeed, some studieshave shown that many individuals have insomnia symptoms with-out complaining about their sleep, i.e., when asked about specificinsomnia symptoms they report having them but they remain sat-isfied with their sleep [5,6]. According to studies, a complaint ofsleep quality or quantity was the foremost condition for an insom-nia diagnosis and was prepotent in determining the hierarchy ofinsomnia symptoms [5–12]. Furthermore, some recent studieshave demonstrated that insomnia as a pathology exists only whendaytime consequences such as fatigue, excessive sleepiness, irrita-bility, mood swings, or cognitive difficulties are present [5,9,10].

Using a hierarchical classification method, we proposed, in thisstudy, to explore how insomnia symptoms are organized within

criteria based on daytime consequences. Sleep Med (2011), doi:10.1016/


mailto:[email protected]


http://www.sciencedirect.com/science/journal/13899457

http://www.elsevier.com/locate/sleep



2 M.M. Ohayon et al. / Sleep Medicine xxx (2011) xxx–xxx

individuals reporting daytime consequences related to sleep dis-turbances. Which symptoms were prepotent in determining a clin-ical picture and in driving help seeking behavior?

2. Methods

2.1. Sample

The study was performed between 2003 and 2005 [13,14]. Thetarget population was adults (18 years and older) living in thestates of California, New York, and Texas (USA). A total of 8937individuals aged 18 years or older, representative of the generalpopulation of these three states (3243 subjects in California,3445 subjects in New York, and 2249 subjects in Texas), wereinterviewed by telephone. They represented a total of 62.8 millioninhabitants. Rates of cooperation were 85.6% in California, 81.3% inNew York, and 83.2% in Texas using CASRO (Council of AmericanSurvey Research Organizations) standards.

2.2. Procedures

In the first stage, telephone numbers were randomly selectedproportional to the population size of each county in California,New York, and Texas. The selection was done within each stateusing a computerized residential phone book. In the second stage,during the telephone contact, the Kish method [15] was used to se-lect one respondent per household. This method allowed for theselection of a respondent based on age and gender to maintain asample representative of these two parameters. If the householdmember chosen declined to participate, the household wasdropped and replaced with another number from the same areaand the process was repeated.

Interviewers explained the goals of the study to potential par-ticipants. They requested verbal consent before conducting theinterview. The participants had the option of calling the principalinvestigator if they wanted further information. The study was ap-proved by the Stanford University Institutional Review Board (IRB).

Subjects who declined to participate or who gave up beforecompleting half the interview were classified as refusals. Excludedfrom the study were subjects who were not fluent in English orSpanish, who suffered from a hearing or speech impairment, orwho had an illness that precluded being interviewed. Phone num-bers were dropped and replaced only after a minimum of 10unsuccessful dial attempts were made at different times and ondifferent days, including weekends. An added-digit technique, thatis, increasing the last digit of a number by one, was employed tocontrol for unlisted telephone numbers. The final sample included21.4% unlisted telephone numbers.

The interviews lasted on average 74.5 (±37.8) minutes. An inter-view could be completed with more than one telephone call if long-er than 60 min or at the request of the participant. Participantsanswered an average of 308 questions. The shortest interviewsencompassed 110 questions and the longest 630 questions. The pro-ject manager or the team leaders also called nearly all the partici-pants who completed the interview. During this 6–8 min call, theyasked a series of random questions related to the interview and alsoasked the participants how satisfied they were with the interviewer.

It was required that all the interviewers have no specific back-ground in medicine and related sciences or in psychology. Theinterviewers were college students or had some college education.The training consisted of five 3-h sessions that covered the studyobjectives, ethics in research, use of the Sleep-EVAL software,and role-playing for interview situations. Interviewers were super-vised by two or three team leaders with a ratio of one team leaderfor six interviewers.

Please cite this article in press as: Ohayon MM et al. Hierarchy of insomniaj.sleep.2011.06.010

2.3. Instrument

Interviewers used the Sleep-EVAL knowledge-based expert sys-tem [16,17] to conduct the interviews. This computer software isspecially designed to administer questionnaires and conduct epi-demiological studies in the general population.

The system is composed of a nonmonotonic, level-2 inferenceengine, two neural networks, a mathematical processor, a knowl-edge base, and a base of facts. Simply put, the interview beginswith a series of questions asked of all the participants. It includes,in order of appearance: sociodemographic information, sleep/wake schedule, sleeping habits, sleep disturbance symptoms,medical and paramedical consultations, and hospitalizations inthe last 12-month period, physical diseases, use of prescribedand nonprescribed drugs, a health quality assessment scale,alimentation, fatigue scale, pain questionnaire, height and weightand, for women, questions on menopause. Questions were readout by the interviewer as they appeared on the screen. These ques-tions were either close-ended (e.g., yes/no, 5-point scale, multiplechoice) or open-ended (e.g., duration of symptom, description ofillness).

Once this information was collected, the system began the diag-nostic exploration of mental disorders. On the basis of responsesprovided by a subject to this questionnaire, the system formulatedan initial diagnostic hypothesis that it attempted to confirm or re-ject by asking supplemental questions or by deductions. Concur-rent diagnoses are allowed in accordance with the DSM-IV [3]and the International Classification of Sleep Disorders or ICSD [4].The system terminated the interview once all diagnostic possibili-ties were exhausted.

The differential process is based on a series of key rules allowingor prohibiting the co-occurrence of two diagnoses. The question-naire of the expert system is designed such that the decision con-cerning the presence of a symptom is based upon the interviewee’sresponses rather than the interviewer’s judgment. This approachhas proved to yield better agreement between lay interviewersand psychiatrists on the diagnosis of minor psychiatric disorders[18]. The system has been tested in various contexts in clinical psy-chiatry and sleep disorder clinics [19,20]. In psychiatry, kappashave ranged from 0.44 (schizophrenia disorders) to 0.78 (majordepressive disorder). Agreement for insomnia diagnoses was ob-tained in 96.9% of cases (kappa 0.78) [19].

2.4. Insomnia symptoms

Insomnia symptoms were defined as follows:

(1) Global sleep dissatisfaction (GSD): Dissatisfaction with thesleep quality or quantity.

(2) Difficulty initiating sleep (DIS): Difficulty falling asleep once inbed with the intention of sleeping.

(3) Difficulty maintaining sleep (DMS):

criteria

(a) Frequent nocturnal awakenings in the same night (atleast 3)

Or(b) difficulty resuming sleep after an awakening (DRS).

(4) Early morning awakenings (EMA): Premature awakening withan inability to resume sleep.

(5) Nonrestorative sleep (NRS): Feeling that the sleep is notrefreshing even if the sleep duration is normal.

Severity and duration of each symptom were set at three nightsor more per week and with a minimal duration of three months.We chose three months as the threshold criterion, rather thanone month, in order to capture chronicity.

based on daytime consequences. Sleep Med (2011), doi:10.1016/



Table 1Prevalence of insomnia complaints and symptoms within the population endorsingdaytime consequences of sleep disturbances.

Frequency (%) within individuals withdaytime consequences of sleepdisturbances

%withGSD

GSD only 10.1DIS only 3.5 55.8DMS onlyNA 2.9 28.4DRS 6.6 41.6EMA only 2.1 34.3NRS only 9.2 67.0DIS + DMS 3.2 75.5DIS + EMA 0.3 90.9a

DIS + NRS 3.5 87.6a

DMS + EMA 2.4 60.5DMS + NRS 5.0 77.2b

EMA + NRS 1.3 68.3DIS + DMS + EMA 3.0 87.0a

DIS + DMS + NRS 3.7 94.1a

DIS + EMA + NRS 1.1 97.1a

DMS + EMA + NRS 3.8 84.4a

DIS + DMS + EMA + NRS 4.7 96.0a

DIS: difficulty initiating sleep; DRS: difficulty resuming sleep; EMA: early morning

M.M. Ohayon et al. / Sleep Medicine xxx (2011) xxx–xxx 3

2.5. Daytime repercussions

Repercussions were assessed through 15 questions answeredon a 5-point scale ranging from no impact to severe impact. Theseitems covered cognitive functioning (memory, concentration, andefficacy), affective tone (irritability, anxiety, and depression), sen-sory irritability, fatigue, and excessive sleepiness.

2.6. Analyses

A classification tree was constructed based on the ExhaustiveCHAID method (Chi-squared Automatic Interaction Detector),which is used to study the relationships between a dependentmeasure (in this study: daytime consequences) and a large seriesof possible predictor variables that themselves may interact(insomnia symptoms and GSD). Predictor variables are chosenusing the Pearson’s chi-square test. The splitting point of the pre-dictor variables is defined using a quadratic discriminant analysis.Subsequently, the data are split into two subsets based on the splitpoint. This procedure is repeated for each of the two new subsets.The procedure stops when no significant predictor can be found.The significance level was p < 0.05.

awakenings; NA: nocturnal awakenings without difficulty resuming sleep) NRS:nonrestorative sleep; GSD: global sleep dissatisfaction.

a p < 0.01 with single symptom (DIS, DMS, EMA, and NRS).b p < 0.01 with single symptom (DIS, DMS, and EMA).

3. Results

The participants in the sample were aged between 18 and97 years; 54.1% of them were women. About half of the sample(51.4%) was married or living with a domestic partner and 30.4%of the participants were single. Retired individuals represented17.1% of the sample, students 6.8%, daytime workers 39.6%, andshift-workers 24.3%. Most of the participants were Caucasians(71.2%), 10.2% were Hispanics, 8% were Blacks, 3.5% were Asians,and 6.4% were of mixed races. Interviews were conducted inSpanish for about 1% of the sample.

3.1. Association between GSD, insomnia symptoms and daytimeconsequences

A total of 26.2% of the sample was dissatisfied with either sleepquality or with sleep quantity (or both) (GSD). More specifically,17.1% were dissatisfied with the quality of their sleep and 23.4%with the quantity of their sleep. As many as 73.1% of individualswith GSD reported at least one insomnia symptom. GSD withoutinsomnia symptoms was observed in 7.1% of the sample.

When at least two insomnia symptoms were present, more than75% of the individuals also reported GSD (see Table 1). The propor-tion of individuals with GSD was significantly higher between indi-viduals with only one insomnia symptom compared to those withtwo or more insomnia symptoms. Individuals with only nocturnalawakenings were less likely to report GSD.

As shown in Table 2, the presence of GSD considerably in-creased the proportion of subjects with daytime consequences re-lated to insomnia. In most cases, the association of one or severalinsomnia symptoms with GSD increased the proportion of individ-uals with daytime consequences to more than 80%.

3.2. Predictors of daytime consequences

An exhaustive CHAID tree method was used to examine therelational hierarchy between GSD and insomnia symptoms in pre-dicting daytime consequences associated with insomnia. The pre-dictor variables entered in the tree modeling were: DIS, DMS (NAand DRS), EMA, NRS, GSD, difficulty getting started in the morning,sleep duration, and sleep latency duration. The final result is pre-sented in Fig. 1. The diagram represents a tree with progressive


splits into smaller branches. The first split in the tree representsthe most significant predictor. Each branch is further split untilno significant difference is observed in the new split.

As it can be seen, the first split is with GSD. It is therefore themost important predictor for daytime consequences. As many as72.7% of individuals who were rather dissatisfied (node 3) and85.2% of those very dissatisfied (node 2) reported daytimeconsequences.

The second split is on NRS. Among individuals who were ratherdissatisfied with their sleep, the addition of NRS increased the pro-portion of individuals with daytime consequences to 84.4% (node9); being rather dissatisfied without NRS decreased the proportionto 66.2% (node 8).

In very dissatisfied individuals, the second split was also onNRS: very dissatisfied individuals who also had NRS reported day-time consequences in 91.1% of cases (node 7), while in those verydissatisfied but without NRS daytime consequences were reportedin 73% of cases (node 6).

Among individuals without GSD (satisfied), the proportion withdaytime consequences associated with insomnia was 30.2% (node2). The second split was again with NRS. No GSD but with NRS in-creased the proportion of individuals with daytime consequencesto 63.7% (node 5).

The third split on the tree branch involving individuals withoutGSD was with difficulty resuming sleep (DRS), which was thestrongest predictor of daytime repercussions in individuals satis-fied with their sleep (nodes 11 and 13).

For individuals very dissatisfied with their sleep, the third splitinvolved DIS and DRS. As it can be seen, in the absence of NRS, thepresence of DIS increased the proportion of individuals with day-time consequences to 87.9% (node 15). When NRS was present,the addition of DRS slightly increased the proportion of individualswith daytime consequences (node 17).

For individuals rather dissatisfied with their sleep, the thirdsplit occurs with DRS. As it can be seen, the presence of DRS con-siderably increased the number of subjects with daytime conse-quences among those without NRS (node 19).

The same tree method was done again, this time using GSD withquantity and GSD with quality as two separated variables. As seen




Table 2Prevalence and association between global sleep dissatisfaction, insomnia symptoms and daytime consequences for the whole sample.

Prevalence (%) in the sample % with GSD % of GSD with the symptom % with daytime consequences

DIS only Total 2.4 46.8 4.3 60.8With GSD 1.1 72.4a

No GSD 1.3 50.5NA only Total 2.7 19.9 2.0 46.1

With GSD 0.5 65.9No GSD 2.2 41.2

DRS only Total 5.3 31.5 6.4 52.3With GSD 1.7 69.0a

No GSD 3.7 44.6EMA only Total 1.7 25.2 1.6 51.1

With GSD 0.4 69.7b

No GSD 1.3 44.9NRS only Total 5.8 55.9 12.3 67.2

With GSD 3.2 80.6a

No GSD 2.5 50.3DIS + DMS Total 1.7 68.9 4.5 77.3

With GSD 1.2 84.6b

No GSD 0.5 61.0DIS + EMA Total 0.2 75.0 0.6 68.8

With GSD 0.2 83.3b

No GSD 0.1 25.0DIS + NRS Total 1.7 85.9 5.5 88.3

With GSD 1.4 90.0No GSD 0.2 77.8

DMS + EMA Total 1.7 44.4 2.9 57.1With GSD 0.8 78.0a

No GSD 1.0 40.5DMS + NRS Total 2.4 77.2 7.1 88.0

With GSD 1.9 88.0No GSD 0.5 88.1

EMA + NRS Total 0.7 67.3 1.7 78.8With GSD 0.5 80.0No GSD 0.2 76.5

DIS + DMS + EMA Total 0.9 81.8 2.7 81.8With GSD 0.7 87.0b

No GSD 0.2 58.3DIS + DMS + NRS Total 1.7 91.6 6.0 90.8

With GSD 1.6 93.3a

No GSD 0.1 63.7DIS + EMA + NRS Total 0.5 94.7 1.8 89.5

With GSD 0.5 91.7No GSD 0.0 50.0

DMS + EMA + NRS Total 1.7 85.5 5.6 93.1With GSD 1.5 92.0No GSD 0.2 100.0

DIS + DMS + EMA + NRS Total 2.2 94.0 7.8 90.4With GSD 2.1 92.4b

No GSD 0.1 60.0

DIS: difficulty initiating sleep; DRS: difficulty resuming sleep; EMA: early morning awakenings; NRS: nonrestorative sleep; GSD: global sleep dissatisfaction.a p < 0.001.b p < 0.01.


in Fig. 2, the results are not very different than the first tree. The firstsplit is with GSD with quantity and the second split is on GSD withquality.

4. Discussion

The main aim of this study was to explore how insomnia symp-toms were organized within a general population sample and todetermine the hierarchical importance of these symptoms to pre-dict daytime consequences of sleep disturbances. The resultsclearly show that global sleep dissatisfaction (GSD) with sleepquantity or sleep quality was the foremost symptom in identifyingindividuals with daytime consequences.

Are these findings surprising? At first glance, the answer is yes.However, many years before the appearance of the sleep disorderschapter in the DSM-III-R and the development of the InternationalClassification of Sleep Disorders, the American Institute of


Medicine, in 1979, had defined insomnia as being an unsatisfactorysleep [21]. Although the premises were good, such a definition wasimpractical and too vague; it is difficult to measure changes insymptoms when none are defined.

4.1. How the use of a general population sample enhances the designof a classification

Clinical samples are excellent to measure the efficacy of a treat-ment. However, as it comes to define symptoms of a disorder suchas insomnia, the problem is that their patients already share oneimportant characteristic: they are dissatisfied with their sleep.Therefore, actual definitions of classic insomnia symptoms arebased on the implicit assumption that people reporting an insom-nia symptom are dissatisfied with their sleep. It might be true in aclinical population, but in a general population sample it cannot beassumed: one of the consequences of such an assumption in the




Fig. 1. Hierarchical tree for insomnia complaints and symptoms: justification for global sleep dissatisfaction with sleep quantity or quality (grouped into a single variable).Legend: DIS: difficulty initiating sleep; DRS: difficulty resuming sleep; EMA: early morning awakenings; NRS: nonrestorative sleep; GSD: global sleep dissatisfaction.

Fig. 2. Hierarchical tree for insomnia complaints and symptoms: justification for global sleep dissatisfaction with sleep quantity or quality (treated as two separatedvariables) Legend: DIS: difficulty initiating sleep; DRS: difficulty resuming sleep; EMA: early morning awakenings; NRS: nonrestorative sleep; GSD: global sleepdissatisfaction.

M.M. Ohayon et al. / Sleep Medicine xxx (2011) xxx–xxx 5

general population is that a large proportion of individuals withinsomnia symptoms have no associated daytime consequences.This is clearly shown in our results (Table 2): in the absence ofGSD, less than half of individuals who reported insomnia symp-tom(s) had associated daytime consequences. This leads us to thequestion, ‘‘what are the differences between complaint, symptom,syndrome, and disorder in sleep disorders?’’

A complaint refers to a patient reporting a problem botheringhim or producing a handicap.

A symptom is a clinical fact recognized by the clinician based onanswers to a questionnaire or an examination.

A syndrome is a regrouping of symptoms and/or complaints interms of criteria that can have an intensity, a duration, a frequency,an evolution, and a specificity linked with age, gender, or othercategories.


One can already see a fundamental difference between epide-miological and clinical studies: epidemiological studies first lookfor symptoms identified a posteriori based on the answers to aquestionnaire. In clinical studies, a participant is primarily selectedbased on a complaint and only after that will symptoms be ex-plored. It is why it is so important to define our field very carefullyusing both general population and clinical samples. Our insomniamodel could help the field in defining more specific diagnosticalgorithms through the entire field of sleep.

The statistical method we have used may surprise many. Thedecision tree technique we used was developed in 1980 with theprimary purpose to help classifications in the field of medicaland psychiatric research [22]. The main problem is that it requiresa large number of subjects because it can have multiple splits andthe number of subjects can quickly become too small for reliable





analysis. However, when the prerequisites are respected, the meth-od is very valuable. It offers a useful way of summarizing data andshows the major natural divisions of the participants.

It should be kept in mind, however, that our results are based onsubjective reports. Therefore, symptoms are described in the man-ner that the participants experienced them and may not corre-spond exactly to polysomnographic measures. Since ours is anepidemiological study, we did not conduct laboratory testing withrespondents to confirm symptoms and diagnoses. In some cases,such as for disorders like Obstructive Sleep Apnea Syndrome, poly-somnographic recording (PSG) is needed to confirm the diagnosis.It has been documented that individuals with insomnia tend tomisjudge their sleep compared to PSG data [23,24].

5. Conclusions

To summarize, this study is the first one to underline and todocument the strength and impact of GSD in the diagnosis makingprocess of insomnia. GSD overrides all the other symptoms ofinsomnia: difficulty initiating sleep, nocturnal awakenings, earlymorning awakenings, and nonrestorative sleep. In the classificationtree, GSD emerges as the first symptom leading to the diagnosis ofinsomnia; associations with the other insomnia symptoms still re-main, but they appear only on second or third levels in the classi-fication tree.

We have studied how different parameters were playing in theidentification of different categories of insomnia in the generalpopulation. We have shown that using a more hierarchical proce-dure inside of the diagnosis of insomnia is a way to not only iden-tify people with insomnia more clearly but also to identify the oneswho need to be treated with a greater chance of compliance.

Scientifically, this approach is easily reproducible and couldlead to insomnia studies with greater comparability betweenthemselves and to fewer variations in the definitions of patientgroups.

The benefits in the appreciation of new medications could bedecisive in terms of efficacy and also in terms of follow-up andevaluation of unexpected adverse events. In terms of educationalpurposes, the interest of a classification corresponding to the real-ity of the medical community’s work is a major advantage towardencouraging general practitioners to become more interested inrecognizing and treating the sleep pathology of their patients.

Conflict of interest

The ICMJE Uniform Disclosure Form for Potential Conflicts ofInterest associated with this article can be viewed by clicking onthe following link: doi:10.1016/j.sleep.2011.06.010.


Acknowledgments

This study was support by Grants from the Bing Foundation(MMO), the Arrillaga Foundation (MMO), and the NIH#R01NS044199 (MMO).

References

[1] Ohayon M. Epidemiology of insomnia: what we know and what we still needto learn. Sleep Med Rev 2002;6:97–111.

[2] American Psychiatric Association (APA). Diagnostic and statistical manual ofmental disorders. 3rd ed. revised. Washington DC: APA; 1987.

[3] American Psychiatric Association (APA). Diagnostic and statistical manual ofmental disorders. 4th ed. revised. Washington DC: APA; 1994.

[4] AASM (American Academy of Sleep Medicine). International classification ofsleep disorders. 2nd ed. (ICSD-2). Westchester, IL; 2005.

[5] Ohayon MM, Roth T. What are the contributing factors for insomnia in thegeneral population? J Psychosom Res 2001;51:745–55.

[6] Ohayon MM, Caulet M, Guilleminault C. Complaints about nocturnal sleep:how a general population perceives its sleep, and how this relates to thecomplaint of insomnia. Sleep 1997;20:715–23.

[7] Ohayon MM, Zulley J. Correlates of global sleep dissatisfaction in the Germanpopulation. Sleep 2001;24:780–7.

[8] Jacobs JM, Cohen A, Hammerman-Rozenberg R, Stessman J. Global sleepsatisfaction of older people: the Jerusalem Cohort Study. J Am Geriatr Soc2006;54:325–9.

[9] Ohayon MM, Krystal A, Roehrs TA, Roth T, Vitiello MV. Using difficultyresuming sleep to define nocturnal awakenings. Sleep Med 2010;11:236–41.

[10] Ohayon MM, Caulet M, Priest RG, Guilleminault C. DSM-IV and ICSD-90insomnia symptoms and sleep dissatisfaction. Br J Psychiatry 1997;171:382–8.

[11] Ohayon MM, Paiva T. Global sleep dissatisfaction for the assessment ofinsomnia severity in the general population of Portugal. Sleep Med2005;6:435–41.

[12] Zilli I, Ficca G, Salzarulo P. Factors involved in sleep satisfaction in the elderly.Sleep Med 2009;10:233–9.

[13] Ohayon MM. Nocturnal awakenings and comorbid disorders in the Americangeneral population. J Psychiatr Res 2008;43:48–54.

[14] Ohayon MM, Krystal A, Roehrs TA, Roth T, Vitiello MV. Using difficultyresuming sleep to define nocturnal awakenings. Sleep Med 2010;11:236–41.

[15] Kish L. Survey sampling. New York: John Wiley & Sons Inc.; 1965.[16] Ohayon MM. Sleep-EVAL, knowledge based system for the diagnosis of sleep

and mental disorders. Ottawa: Industry Canada: Copyright Office, CanadianIntellectual Property Office; 1994.

[17] Ohayon MM. Improving decision-making processes with the fuzzy logicapproach in the epidemiology of sleep disorders. J Psychosom Res1999;47:297–311.

[18] Lewis G, Pelosi AJ, Araya RC, Dunn G. Measuring psychiatric disorder in thecommunity, a standardized assessment for use by lay interviewers. PsycholMed 1992;22:465–86.

[19] Ohayon MM, Guilleminault C, Zulley J, et al. Validation of the Sleep-EVALsystem against clinical assessments of sleep disorders and polysomnographicdata. Sleep 1999;22:925–30.

[20] Ohayon M. Validation of expert systems: examples and considerations.Medinfo 1995;8:1071–5.

[21] Institute of Medicine. Sleeping pills, insomnia and medical practice.Washington, DC: National Academy of Sciences; 1979.

[22] Kass GV. An exploratory technique for investigating large quantities ofcategorical data. Appl Stat 1980;29:119–27.

[23] Frankel BL, Coursey RD, Buchbinder R, Snyder F. Recorded and reported sleepin chronic primary insomnia. Arch Gen Psychiatry 1976;33:615–23.

[24] Rosa RR, Bonnet MH. Reported chronic insomnia is independent of poor sleepas measured by electroencephalography. Psychosom Med 2000;62:474–82.





SLEEP, Vol. 28, No. 9, 2005

INTRODUCTION

INSOMNIA IS THE MOST COMMON SLEEP COMPLAINT ACROSS ALL STAGES OF ADULTHOOD, AND FOR MIL-LIONS, THE PROBLEM IS CHRONIC. INSOMNIA can be a symptom of other disorders, like depression, or it can be a pri-mary disorder in itself. Whether it is the primary disorder or sec-ondary to some other condition, chronic insomnia is often asso-ciated with a wide range of adverse conditions, including mood disturbances; difficulties with concentration and memory; and some cardiovascular, pulmonary, and gastrointestinal disorders. Whether insomnia is the cause or result of associated problems is not always easily determined but is critical to treatment strategies for individual patients. A variety of behavioral and pharmacological approaches show promise for managing chronic insomnia symptoms. However, there has been limited guidance for clinicians in choosing the best treatment for chronic insomnia, due to the paucity of ran-domized clinical trials for many widely used treatments. Avail-able treatments include an array of behavioral or nonpharmaco-logic interventions; hypnotic medications; and antidepressant, antipsychotic, or antihistamine medications. As pointed out in the recent 2003 National Sleep Disorders Research Plan, published by the National Center on Sleep Dis-orders Research at the National Institutes of Health (NIH), there is great need for additional research to better define the nature of chronic insomnia and ways to characterize its detailed expression in diverse patients. Additional systematic research is also1greatly needed to provide a more thorough database from which clinicians and patients can make more informed choices about treatment options. To address these needs, the National Institute of Mental Health and the Office of Medical Applications of Research of the NIH sponsored a State-of-the-Science Conference on the Manifesta-tions and Management of Chronic Insomnia in Adults on June 13–15, 2005, in Bethesda, MD. During the first 2 days of the conference, experts presented the latest scientific knowledge about chronic insomnia and available treatments. After weighing all of the scientific evidence, an independent panel prepared and presented the following state-of-the-science statement. The panel was charged with answering five specific questions: • How is chronic insomnia defined, diagnosed, and classified, and what is known about its etiology? • What are the prevalence, natural history, incidence, and risk factors for chronic insomnia? • What are the consequences, morbidities, comorbidities, and

public health burden associated with chronic insomnia? • What treatments are used for the management of chronic in-somnia, and what is the evidence regarding their safety, efficacy, and effectiveness? • What are important future directions for insomnia-related re-search? The conference was intended for health care professionals, re-searchers, patients and their families, and members of the public interested in the nature of and available treatments for chronic insomnia. The conference included formal expert presentations focusing on the individual conference questions and oral and written input from professionals and members of the lay public. In addition, the independent panel benefited greatly from a com-prehensive systematic literature review, prepared by the Univer-sity of Alberta Evidence-based Practice Center.

1. How is chronic insomnia defined, diagnosed, and classified, and what is known about its etiology?

Definition

Insomnia has historically been defined by complaints of dis-turbed sleep in the presence of adequate opportunity and circum-stance for sleep. The disturbance may consist of one or more of three features: (1) difficulty in initiating sleep; (2) difficulty in maintaining sleep; or (3) waking up too early. A fourth charac-teristic, nonrestorative or poor-quality sleep, has frequently been included in the definition, although there is controversy as to whether individuals with this complaint share similar pathophys-iologic mechanisms with the others. The importance of sleep disruption often rests with its im-pact on the individual’s daytime function. Guidelines incorpo-rating impact on function along with the above features in the definition of insomnia have recently been published in an effort to standardize future insomnia research. In some populations, however, the impact of sleep disruption goes beyond the patient, such as the pediatric and the elderly (particularly nursing home residents), for whom the major impact may fall on the daytime function of parents and caregivers. To distinguish chronic from acute insomnia, which may occur in anyone at one time or another, varied definitions for chronic insomnia have been utilized from study to study, with minimum durations ranging from 30 days to as long as 6 months. Most cases of insomnia are comorbid with other conditions. Historically, this has been termed “secondary insomnia.” Howev-er, the limited understanding of mechanistic pathways in chronic

National Institutes of Health State of the Science Conference Statement

NIH consensus and state-of-the-science statements are prepared by independent panels of health professionals and public representatives on the basis of (1) the results of a systematic literature review prepared under contract with the Agency for Healthcare Research and Quality (AHRQ), (2) presentations by investigators working in areas relevant to the conference questions during a 2-day public session, (3) questions and statements from conference attend-ees during open discussion periods that are part of the public session, and (4) closed deliberations by the panel during the remainder of the second day and morning of the third. This statement is an independent report of the panel and is not a policy statement of the NIH or the Federal Government. The statement reflects the panel’s assessment of medical knowledge available at the time the statement was written. Thus, it provides a “snapshot in time” of the state of knowledge on the conference topic. When reading the statement, keep in mind that new knowledge is inevitably accumulating through medical research.

NIH Statement1049

NIH STATEMENT REGARDING THE TREATMENT OF INSOMNIA

Manifestations and Management of Chronic Insomnia in Adults June 13–15, 2005

SLEEP, Vol. 28, No. 9, 2005

insomnia precludes drawing firm conclusions about the nature of these associations or direction of causality. Furthermore, there is concern that the term secondary insomnia may promote under-treatment. Therefore, we propose that the term “comorbid insom-nia” may be more appropriate. Common comorbidities include psychiatric disorders, particularly depression and substance use disorders; cardiopulmonary disorders; and conditions resulting in chronic somatic complaints that may result in sleep disruption. Other associated sleep disorders can also contribute to insomnia, such as obstructive sleep apnea, restless legs syndrome, or peri-odic limb movement disorder. “Primary insomnia” implies that no other cause of sleep disturbance has been identified.

Diagnosis

Diagnosis is based primarily on patient-derived and family or caregiver complaints, as determined by the clinical interview. Medical history and physical examination are useful in establish-ing the presence of comorbid syndromes. Other tools have been used as an aid to diagnosis, with varying utility. Sleep diaries may help document sleep/wake cycles. Various questionnaires have been formulated but are limited by the lack of standardiza-tion. Actigraphy, in which a wrist-worn device measures move-ment to infer sleep and wake cycles, has not been fully validated in chronic insomnia. Finally, multichannel polysomnography, both in-lab and at home, is most often used in cases in which other sleep disorders, such as sleep apnea, are suspected.

Classification and Etiology

Insomnia may be classified based on specific symptoms (i.e., sleep onset or sleep maintenance) or the duration of symptoms. Etiology-based classification schemes have also been advocated. Evidence supports both psychological and physiological models in the etiology of insomnia. Models include the concepts of con-ditioning, hyperarousal, stress response, predisposing personali-ty traits, and attitudes and beliefs about sleep. Animal models are used to identify neural systems that regulate arousal and sleep. The precise relationship between physical illness and changes in brain function that result in insomnia remains to be further eluci-dated.

2. What are the prevalence, natural history, incidence, and risk fac-tors for chronic insomnia?

Prevalence

Although chronic insomnia is considered to be common, stud-ies on its actual prevalence have yielded variable estimates (i.e., the proportion of persons who have the disorder at a given point in time). Conclusive evidence from epidemiologic studies has been limited by their different definitions of chronic insomnia and by the lack of standardized diagnostic and screening meth-ods. Population-based studies suggest that about 30 percent of the general population has complaints of sleep disruption, while approximately 10 percent has associated symptoms of daytime functional impairment consistent with the diagnosis of insom-nia. Not surprisingly, prevalence appears to be greater in clinical practices, where about one-half of respondents report symptoms of sleep disruption.

Natural History

Few studies specifically describe the course and duration of insomnia. Unpublished data from a middle-aged population fol-lowed over 10 years describe a persistence of symptoms. A pro-spective study of outpatients of hospital-based populations with sleep complaints of at least a month’s duration showed that in the majority of insomniacs the symptoms are of long duration. The paucity of literature describing the natural course of insomnia un-derscores the need for large-scale longitudinal studies.

Incidence

Very little is known about chronic insomnia’s incidence, which is the frequency with which new cases of the disorder arise as time passes. Because incidence is affected only by the factors that cause and prevent a health outcome, it is the preferable frequency measure for etiologic research. Unfortunately, only a few stud-ies of chronic insomnia occurrence have examined its incidence. Increasing the number of studies of the incidence of chronic in-somnia is a clear research priority. In conjunction with studies of incidence, research on the dura-tion of chronic insomnia is also needed. The disorder can last for relatively short periods of time in some patients and for decades in others. Insomnia can also recur after a period of remission. When studies of chronic insomnia incidence are conducted, the newly ascertained cases can be followed longitudinally to describe the disorder’s natural history. In these studies, it will be possible to investigate factors that are suspected of affecting chronic insom-nia’s incidence, duration, or both. It will be particularly important to determine which therapies the treated patients receive and to distinguish patients whose chronic insomnia is produced by spe-cific, preexisting medical or psychological conditions.

Risk Factors

Several problems limit the ability to compare and integrate available information from existing observational studies on cor-relates of insomnia: (1) validated diagnostic instruments have not been applied in large, population-based studies; (2) the many comorbid physical and psychiatric conditions that precede or fol-low a diagnosis of insomnia challenge the ability to identify the independent relationships of related factors with insomnia; and (3) most observational research has used a cross-sectional design, so causal relationships cannot be inferred and only correlates of insomnia can be identified. Many studies have found greater prevalence of insomnia among older people, perhaps as a consequence of declining health and/or institutionalization. Whether rates of insomnia increase with age in healthy older people remains unclear. Most observational stud-ies of insomnia have found greater prevalence among women, especially in the postmenopausal years. Current evidence on differences among racial or ethnic groups in prevalence of insomnia within the United States is limited and inconclusive. Several studies have found higher prevalence of insomnia in divorced, separated, and widowed adults relative to married adults. Lower education and income have sometimes been associated with higher prevalence of insomnia. Several psychiatric and physical illnesses have strong and theoretically expected relationships with insomnia. Insomnia is a symptom of depression. Other medical conditions, including

NIH Statement 1050

SLEEP, Vol. 28, No. 9, 2005

arthritis, heart failure, pulmonary and gastrointestinal disorders, Parkinson’s disease, stroke, and incontinence, also affect sleep and increase the prevalence of insomnia. The extent to which treatment for these conditions ameliorates insomnia remains un-clear. Cigarette smoking, alcohol and coffee consumption, and nu-merous prescription drugs also affect sleep and are associated with increased prevalence of insomnia. Although modification of these behaviors has the potential to reduce the prevalence of in-somnia, evidence of this is limited.

Future Studies

Validated instruments with known psychometric properties are needed, with attention paid to ease of administration, cross-cultural applicability, and comparability to objective measures of sleep performance, both overall and within important subgroups. Attention is also needed concerning the reliable measurement of the degree of sleep disturbance and the severity of symptoms of insomnia. Another hypothesis relates to the possible genetic etiology of insomnia. Work is needed to quantify the importance of family history, along with a systematic search for specific genes. Correlates of insomnia should be explored for their relation-ships with the development of subsequent insomnia. For exam-ple, studies are needed of the impact on incidence of insomnia of divorce, separation and bereavement, polypharmacy, and major chronic diseases. Longitudinal observational studies are needed to identify fac-tors affecting incidence of and remission from insomnia. An ef-ficient approach would be to add validated questions on chronic insomnia to ongoing observational studies that assess the many potential determinants of insomnia incidence, persistence, and re-mission.

3. What are the consequences, morbidities, comorbidities, and pub-lic health burden associated with chronic insomnia?

Consequences, Morbidities, and Comorbidities

Some evidence suggests that insomnia may be associated with high health care utilization. The direct and indirect costs of chron-ic insomnia have been estimated as tens of billions of dollars an-nually. However, these estimates are highly variable and depend on many assumptions. In estimating the economic consequences of insomnia, it is difficult to separate the effects of insomnia from the effects of comorbid conditions. For example, a person with arthritis may have problems sleeping but may seek health care for the arthritis rather than for sleep problems. Only a few studies have examined the effects of insomnia on functioning in everyday life. These studies suggest that insomnia reduces quality of life and hinders social functioning. Two stud-ies have identified a relationship between chronic insomnia and work days missed. Furthermore, insomnia is related to impaired work performance. There is at least some evidence of a relation-ship between chronic insomnia and impaired memory, cognitive functioning, and depressed mood. Laboratory studies indicate that sleep loss results in impaired psychomotor and cognitive functioning. There is evidence that chronic insomnia or its treatment contributes to the increased number of falls in older adults.

Insomnia usually appears in the presence of at least one other disorder. Particularly common comorbidities are major depres-sion, generalized anxiety, substance abuse, attention deficit/hyper-activity in children, dementia, and a variety of physical problems. The research diagnostic criteria for insomnia recently developed by the American Academy of Sleep Medicine indeed share many of the criteria of major depressive disorder. Explaining this over-lap requires a study that determines how often insomnia precedes the disorders with which it is associated and continues to exist af-ter the other disorders are cured or go into temporary remission. A greater range of outcome measures is needed for future re-search. It may be valuable to validate patient self-reports of sleep latency, awakenings, and morning wakings. Both insomnia and its treatment may affect quality of life. Overall summary measures that simultaneously consider side ef-fects and benefits of treatments should be used. These measures can assess the outcomes of interventions using quality-adjusted life years. Costs of illness should also be assessed to allow for an analysis of the cost-effectiveness of treatments. The U.S. Depart-ment of Health and Human Services has developed guidelines for these assessments, and these should be consulted in the devel-opment of evaluation protocols. In addition to generic measures, insomnia-specific quality of life measures should also be used.

Public Health Burden

The focus of public health is on populations rather than solely on individuals. The public health consequences of insomnia are difficult to evaluate because the literature is not well developed at this time. Sleep research has focused on basic mechanisms and clinical studies. Relatively little attention has been paid to the public health burden of insomnia. To better understand the public health consequences of insomnia, several lines of research should be considered. The effects of insomnia upon premature death are not known. Separating the effects of insomnia from the effects of its comor-bidities will be a methodological challenge. Measures of sleep have been added to the National Health and Nutrition Examina-tion Survey and should be added to other major epidemiological studies, including the Behavioral Risk Factor Surveillance Sur-vey. The effect of insomnia on quality of life has been reported in few studies. Secondary analysis of data from major population studies that include both measures of sleep and measures of func-tioning and quality of life should be supported. New studies are needed to determine whether insomnia causes job-related disabil-ity. Furthermore, we need to support additional studies to deter-mine whether treatment for insomnia affects job performance and academic performance. The economic consequences of insomnia are not clearly under-stood. New studies are needed to estimate the direct and indirect costs of chronic insomnia and the potential societal benefits that might accrue from successful intervention programs. Finally, in-somnia has effects beyond individual patients. Families, caregiv-ers, and friends are also affected by the condition. More evidence is needed to document these effects.

4. What treatments are used for the management of chronic insom-nia, and what is the evidence regarding their safety, efficacy, and effectiveness?

NIH Statement1051

SLEEP, Vol. 28, No. 9, 2005

Epidemiological surveys have shown that the most common treatments used by people with chronic insomnia are over-the-counter (OTC) antihistamines, self-medication with alcohol, and prescription medications. The major forms of psychological treat-ments that have been systematically evaluated are the cognitive and behavioral therapies. Alternative and complementary treat-ments include melatonin and herbal remedies, such as valerian. Assessment of the efficacy of treatments for chronic insomnia is complicated by a number of factors. These include a lack of consistency in the criteria used to diagnose chronic insomnia, gaps in understanding the natural history of insomnia, and lack of clar-ity or consensus about the crucial outcomes of treatment. Further complicating the ability to assess treatments for chronic insom-nia is its overlap with many medical and psychiatric conditions, most notably depression. Although there have been randomized clinical trials (RCTs) for several treatments, there is inconsistency in applying rigorous methodology to the assessment of a number of currently used treatments. Additionally, most clinical trials are relatively short term. There is a paucity of information about the long-term effects on sleep, daytime functioning, and quality of life.

Behavioral and Cognitive Therapies

Behavioral and cognitive-behavioral therapies (CBTs) have demonstrated efficacy in RCTs. However, because few clinicians are experts in the use of CBT for the treatment of chronic insom-nia, these techniques are not in widespread use. The degree to which this treatment can be successfully disseminated to more diverse populations is unknown. Behavioral methods, which include relaxation training, stimulus control, and sleep restric-tion, were developed and first tested in the 1970s. More recently, cognitive therapy methods have been added to behavioral meth-ods. Cognitive therapy methods include cognitive restructuring, in which anxiety-producing beliefs and erroneous beliefs about sleep and sleep loss are specifically targeted. When these cogni-tive methods have been added to the behavioral methods to com-pose a cognitive-behavioral treatment package, it has been found to be as effective as prescription medications are for brief treat-ment of chronic insomnia. Moreover, there are indications that the beneficial effects of CBT, in contrast to those produced by medications, may last well beyond the termination of treatment. There is no evidence that such treatment produces adverse effects, but thus far, there has been little, if any, study of this possibility.

Prescription Medications

This section describes the use of two categories of medications, those that are approved by the U.S. Food and Drug Administra-tion (FDA) for the treatment of insomnia and those that are FDA-approved for the treatment of other disorders but are prescribed to treat insomnia. (The latter category is considered “off-label” us-age.) There are currently eight medications approved by the FDA for treatment of insomnia. Despite the fact that insomnia is often a chronic condition, only one of these medications (eszopiclone) has been approved for use without a specified time limit. The oth-er medications have approved use limited to 35 days or less.

Benzodiazepine Receptor Agonists

There are two broad groups of agents in this class of prescrip-

tion hypnotics: benzodiazepines (e.g., estazolam, flurazepam, quazepam, temazepam, and triazolam) and the more recently introduced agents that act as benzodiazepine receptors but have a nonbenzodiazepine structure (e.g., zaleplon, zolpidem, and eszopiclone). Results from clinical trials indicate that these agents are efficacious in the short-term management of insomnia. With the exception of eszopiclone, the benefits of these agents for long-term use have not been studied using RCTs. A recent clinical trial of eszopiclone provided evidence of sustained ef-ficacy for 6 months. Adverse effects associated with these medi-cations include residual daytime sedation, cognitive impairment, motor incoordination, dependence, and rebound insomnia. The frequency and severity of these adverse effects are much lower in the newer benzodiazepine receptor agonists, most likely because these agents have shorter half-lives. Although the available litera-ture suggests that, in the short term, tolerance and abuse of the benzodiazepine receptor agonists are not major problems in the general population with chronic insomnia, long-term use needs further study.

Antidepressants

Over the past 20 years, there has been a significant change in the use of prescription medications to treat chronic insomnia, with a decrease in the use of benzodiazepine-like agents and a substantial increase in the off-label use of antidepressants. Based on recent surveys, the antidepressant trazodone is now the most commonly prescribed medication for the treatment of insomnia in the United States. In short-term use, trazodone is sedating and im-proves several sleep parameters. These initial effects may not last beyond 2 weeks. However, there are no studies of long-term use of trazodone for treatment of insomnia. Another antidepressant, doxepin, has been found to have beneficial effects on sleep for up to 4 weeks for individuals with insomnia. Data on other antide-pressants (e.g., amitriptyline and mirtazepine) in individuals with chronic insomnia are lacking. All antidepressants have potentially significant adverse effects, raising concerns about the risk–ben-efit ratio. Moreover, there is a need to establish and communicate to prescribers dose-response relationships for all of these agents.

Other Prescription Medications

A number of other sedating medications have been used in the treatment of insomnia. These include barbiturates (e.g., pheno-barbital) and antipsychotics (e.g., quetiapine and olanzepine). Studies demonstrating the usefulness of these medications for either short- or long-term management of insomnia are lacking. Furthermore, all of these agents have significant risks, and thus their use in the treatment of chronic insomnia cannot be recom-mended.

OTC Medications

Antihistamines (H1 receptor antagonists, such as diphenhydr-amine) are the most commonly used OTC treatments for chronic insomnia, but there is no systematic evidence for efficacy and there are significant concerns about risks of these medications. Adverse effects include residual daytime sedation, diminished cognitive function, and delirium, the latter being of particular concern in the elderly. Other adverse effects include dry mouth, blurred vision, urinary retention, constipation, and risk of increased intraocular

1052 NIH Statement

SLEEP, Vol. 28, No. 9, 2005

pressure in individuals with narrow angle glaucoma.

Alcohol

There is no systematic evidence of benefit from alcohol as self-medication for chronic insomnia, despite its common usage. The numerous risks of excess alcohol consumption greatly outweigh any benefits for therapeutic use in insomnia. The adverse effects of alcohol on sleep quality are well documented.

Alternative and Complementary Therapies

Melatonin

Melatonin is a natural hormone produced by the pineal gland that plays a role in the control of circadian rhythms. Melato-nin is not regulated by the FDA, and preparations vary in their melatonin content, making comparisons across studies difficult. Although melatonin appears to be effective for the treatment of circadian rhythm disorders, little evidence exists for efficacy in the treatment of insomnia. Problems with melatonin include lack of a well-defined, effective dose. In short-term use, melatonin is thought to be safe, but there is no information about the safety of long-term use.

Herbal Therapies

Valerian is derived from the root of plants of the species vale-riana and is thought to promote sleep. Limited evidence shows no benefit beyond that provided by placebo. The FDA does not regulate valerian, and thus different preparations vary in valerian content. Safety data are minimal, but there have been case reports of hepatotoxicity in persons taking herbal products containing va-lerian. Many other herbal remedies have been promoted, but some are known to have toxicity and efficacy evidence is lacking.

L-tryptophan

L-tryptophan is an endogenous amino acid that has been used as a hypnotic. Systematic evidence supporting its use in the treat-ment of insomnia is extremely limited and based on studies with a small number of subjects. Concerns are also raised about possible toxic effects of L-tryptophan, particularly when used in combina-tion with certain psychiatric medications.

Other Treatments

There are a number of other alternative treatments, including tai chi, yoga, acupuncture, and light therapy, that have potential utility in the treatment of insomnia. These treatments have not been adequately evaluated at this time.

Research Recommendations

CBT and benzodiazepine receptor agonists have been shown to be beneficial in the acute management of chronic insomnia. Other therapies have also demonstrated some promise. However, little is known about the comparative benefits of these treatments, their generalizability, and their effects on understudied features of chronic insomnia. In order to address this lack of knowledge, RCTs will be required that: • Are large-scale and multisite.

• Compare at least two effective or promising treatments so that the comparative benefits of effective treatments can be evaluated. This should include comparisons among pharmacological agents, CBT, and combined treatment. • Evaluate the positive and adverse effects of treatments over longer timeframes, including the period after discontinuation of treatment.• Incorporate objective and subjective measures of daytime func-tion and quality of life in addition to traditional sleep parameters, such as sleep onset latency and total sleep time. • Systematically evaluate a variety of commonly used OTC and alternative remedies for insomnia that have not been formally evaluated. • Measure the costs and cost-effectiveness of treatments. The pharmaceutical industry is called upon to support com-parisons of its medications not only with placebo but also with other effective treatments, including CBT. Studies should be directed to important population subgroups, including children, nursing home residents, postmenopausal women, those with primary chronic insomnia, and those with in-somnia comorbid with other conditions. To overcome potential problems with reporting bias in clinical trials, the development of a central registry for all insomnia trials is recommended. This registry would allow a systematic synthe-sis of the available clinical trial data. As comparative efficacy data become available, it will be criti-cal to conduct effectiveness studies to determine generalizability to broader clinical populations in community settings. Surveys of physician prescribing behavior and decisionmaking are recommended to permit the anticipation and understanding of developments, such as the recent, dramatic increase in pre-scriptions for the antidepressant trazodone in the management of chronic insomnia.

5. What are important future directions for insomnia-related re-search?

Validated instruments are needed to assess chronic insomnia, with attention paid to the ease of administration and cross-cul-tural applicability. A greater range of outcome measures related to chronic insomnia and its consequences is also needed. Measures of sleep should be added to major epidemiologic studies. Studies are needed of the possible genetic etiology of chronic insomnia. The neural mechanisms underlying chronic insomnia are poorly understood. Studies aiming to identify neural mecha-nisms should use animal models and in vivo neural imaging ap-proaches in people with insomnia and in individuals with normal sleep. Work is needed to quantify the importance of family his-tory, along with a systematic search for specific genes. Longitudinal observational studies are needed to identify fac-tors affecting incidence of, natural history of, and remission from chronic insomnia. An efficient approach would be to add ques-tions on chronic insomnia to ongoing observational studies that assess the many potential determinants of insomnia incidence, persistence, and remission. The effects of insomnia on quality of life have been reported in few studies. Analyses of data from major population studies that include measures of sleep, measures of functioning, and quality of life should be supported. Studies are needed to determine whether insomnia causes job-related disability and whether treatment for

1053 NIH Statement

SLEEP, Vol. 28, No. 9, 2005

insomnia enhances job performance and academic performance. Studies are needed to estimate the direct and indirect societal costs of insomnia and the potential societal benefits that might accrue from successful intervention programs. Moreover, because chronic insomnia has effects that go beyond individual patients, more research is needed to quantify effects on families, friends, and caregivers. Cognitive-behavioral therapy (CBT) and benzodiazepine re-ceptor agonists have been shown to be beneficial in the acute management of chronic insomnia. Other therapies have also dem-onstrated some promise. However, little is known about the com-parative benefits of these treatments, their combination, and their effects on understudied features of chronic insomnia. To address this lack of knowledge, randomized controlled trials will be re-quired that are large scale and multisite and compare at least two effective or promising treatments. This should include compari-sons between pharmacological agents as well as between those agents and CBT. The pharmaceutical industry is called upon to compare its medications not only with placebo but also with other effective treatments, including CBT. Trials should include mea-sures of cost and cost-effectiveness. To overcome potential problems with reporting bias in clinical trials, the development of a central registry for all clinical trials is recommended. This registry would allow a systematic synthesis of the available clinical trial data. As comparative efficacy data become available, it will be criti-cal to conduct effectiveness studies to determine generalizability to broader clinical populations in community settings. Studies should be directed to important population subgroups, including children, nursing home residents, postmenopausal women, those with primary chronic insomnia, and those with in-somnia comorbid with other conditions.

CONCLUSIONS

Chronic insomnia is a major public health problem affecting millions of individuals, along with their families and communi-ties. Little is known about the mechanisms, causes, clinical course, comorbidities, and consequences of chronic insomnia. Evidence supports the efficacy of cognitive-behavioral therapy and benzo-diazepine receptor agonists in the treatment of this disorder. Very little evidence supports the efficacy of other treatments, despite their widespread use. Moreover, even for those treatments that have been systematically evaluated, the panel is concerned about the mismatch between the potential lifelong nature of this illness and the longest clinical trials, which have lasted 1 year or less. A substantial public and private research effort is warranted, includ-ing the development of research tools and the conduct of longi-tudinal studies and randomized clinical trials. Finally, there is a major need for educational programs directed at physicans, health care providers, and the public.

State-of-the-Science Panel

Alan I. Leshner, Ph.D. Conference and Panel Chairperson Chief Executive Officer American Association for the Advancement of Science Executive Publisher Science Washington, DC

Helen A. Baghdoyan, Ph.D. ProfessorDepartment of Anesthesiology University of Michigan Ann Arbor, Michigan

Susan J. Bennett, D.N.S., R.N., F.A.A.N. Professor Indiana University School of Nursing Indianapolis, Indiana

Sean M. Caples, D.O. Assistant Professor of Medicine Division of Pulmonary and Critical Care Medicine Mayo Clinic Rochester, Minnesota

Robert J. DeRubeis, Ph.D. Professor and Chair Department of Psychology University of Pennsylvania Philadelphia, Pennsylvania

Robert J. Glynn, Ph.D., Sc.D. Associate Professor of Medicine (Biostatistics) Harvard Medical School Brigham and Women’s Hospital Boston, Massachusetts

Robert Malcolm Kaplan, Ph.D. Professor and Chair Department of Health Services University of California, Los Angeles School of Public Health Los Angeles, California

James N. Kvale, M.D. Professor Department of Family and Community Medicine University of Texas Health Science Center at San Antonio San Antonio, Texas

Charles Poole, Sc.D., M.P.H. Associate Professor Department of Epidemiology University of North Carolina School of Public Health Chapel Hill, North Carolina

Lee N. Robins, Ph.D. Emeritus Professor of Social Science in Psychiatry Department of Psychiatry Washington University School of Medicine University Professor of Social Science, Emeritus Washington University St. Louis, Missouri

Catherine M. Waters, Ph.D., R.N. Associate Professor Department of Community Health Systems School of Nursing University of California, San Francisco San Francisco, California

Charles F. Zorumski, M.D. Samuel B. Guze Professor of Psychiatry Head Department of Psychiatry Washington University School of Med-icine St. Louis, Missouri 13

1054 NIH Statement

SLEEP, Vol. 28, No. 9, 2005

Speakers

Sonia Ancoli-Israel, Ph.D. Professor Department of Psychiatry University of California, San Diego School of Medicine San Diego, California

Ruth Benca, M.D., Ph.D. Professor Department of Psychiatry Psychiatric Institute and Clinics University of Wisconsin Medical School Madison, Wisconsin

Michael H. Bonnet, Ph.D. Professor Department of Neurology Wright State University School of Medicine Dayton, Ohio

Daniel J. Buysse, M.D. Professor of Psychiatry Western Psychiatric Institute and Clinic University of Pittsburgh Medical Center Pittsburgh, Pennsylvania

Jack Edinger, Ph.D. Clinical Professor Department of Psychiatry and Behavioral Sciences Duke University School of Medicine Durham, North Carolina

Daniel Foley, M.S. Epidemiologist Center for Mental Health Services Substance Abuse and Mental Health Services Administration U.S. Department of Health and Human Services Rockville, Maryland

Nalaka S. Gooneratne, M.D., M.S.C.E. Director Sleep Disorders Clinic for Seniors Division of Sleep Medicine University of Pennsylvania Medical School Philadelphia, Pennsylvania

Meir Kryger, M.D. Director Sleep Disorders Centre St. Boniface Hospital Research Centre Winnipeg, Manitoba, Canada

Kenneth Lichstein, Ph.D. Professor and Chair Department of Psychology University of Alabama Tuscaloosa, Alabama

W. Vaughn McCall, M.D., M.S. Professor and Chair Department of Psychiatry and Behavioral Medicine Wake Forest University Health Sciences Winston-Salem, North Carolina

Charles Morin, Ph.D. Professor École de Psychologie Université Laval Sainte-Foy, Québec, Canada

Judith Owens, M.D., M.P.H. Director

Pediatric Sleep Disorders Clinic Associate Professor of Pediatrics Brown Medical School Division of Pediatric Ambulatory Medi-cine Rhode Island Hospital Providence, Rhode Island

Gary S. Richardson, M.D. Senior Research Scientist Division of Sleep Research Henry Ford Hospital Detroit, Michigan

Timothy Roehrs, Ph.D. Director of Research Division of Sleep Research Henry Ford Hospital Detroit, Michigan

Thomas Roth, Ph.D. Chief Division of Sleep Research Henry Ford Hospital Detroit, Michigan

Clifford B. Saper, M.D., Ph.D. Professor and Chair Department of Neurology Harvard Medical School Beth Israel Deaconess Medical Center Boston, Massachusetts

James K. Walsh, Ph.D. Executive Director and Senior Scientist Sleep Medicine and Research Center St. John’s/St. Luke’s Hospitals Chesterfield, Missouri

Manisha Witmans, M.D., FRCPC, Dip. ABSM, F.A.A.P. Pediatric Respirologist and Sleep Medicine Specialist Assistant Professor University of Alberta Medical Director Pediatric Sleep Program and Pediatric Sleep Medicine Stollery Children’s Hospital University of Alberta Hospital Edmonton, Alberta, Canada

Terry B. Young, Ph.D., M.S. Professor Department of Population Health Sciences University of Wisconsin–Madison Madison, Wisconsin

Planning Committee

Israel Lederhendler, Ph.D. Planning Committee Chair National Institute of Mental Health National Institutes of Health Bethesda, Maryland

Sonia Ancoli-Israel, Ph.D. Professor Department of Psychiatry University of California, San Diego School of Medicine San Diego, California

1055 NIH Statement

SLEEP, Vol. 28, No. 9, 2005

David Atkins, M.D., M.P.H. Chief Medical Officer Center for Outcomes and Evidence Agency for Healthcare Research and Quality Rockville, Maryland

Daniel J. Buysse, M.D. Professor of Psychiatry Western Psychiatric Institute and Clinic University of Pittsburgh Medical Center Pittsburgh, Pennsylvania

Michael Coplen, M.A. Human Factors Research Program Office of Research and Development Federal Railroad Administration Washington, DC

Regina T. Dolan-Sewell, Ph.D.Coordinator for Sleep Research National Institute of Mental Health National Institutes of Health Bethesda, Maryland

Subash B. Duggirala, M.D., M.P.H., F.A.A.F.P. Senior Medical Advisor Office of Research, Development, and Information Centers for Medicare & Medicaid Services U.S. Department of Health and Human Services Baltimore, Maryland

Kenneth Fink, M.D., M.G.A., M.P.H. Director Evidence-based Practice Centers Program Center for Outcomes and Evidence Agency for Healthcare Re-search and Quality Rockville, Maryland

Harold Gordon, Ph.D. Program Official Clinical Neurobiology Branch Division of Treatment Research and Development National Institute on Drug Abuse National Institutes of Health Bethesda, Maryland

Lindsey Grandison, Ph.D. Division of Neuroscience and Behavior National Institute on Alcohol Abuse and Alcoholism National Institutes of Health Bethesda, Maryland

Robert Wolfe Greene, M.D., Ph.D. University of Texas Southwestern Medical Center Dallas, Texas

Eleanor Z. Hanna, Ph.D. Associate Director for Special Projects and Centers Office of Research on Women’s HealthOffice of the Director National Institutes of Health Bethesda, Maryland

Carl E. Hunt, M.D.

Director National Center on Sleep Disorders Research National Heart, Lung, and Blood Institute National Institutes of Health Bethesda, Maryland

Kathy Mann Koepke, Ph.D. Program Director Neurofunction and Sensory Conditions Division of Extramural Activities National Institute of Nursing Research National Institutes of Health Bethesda, Maryland

Barnett S. Kramer, M.D., M.P.H. Director Office of Medical Applications of Research Office of the Director National Institutes of Health Bethesda, Maryland

Kathryn A. Lee, Ph.D., R.N., F.A.A.N. Professor and Livingston Chair in Nursing Director Prenatal Clinical Nurse Specialist Program Department of Family Health Care Nursing School of Nursing University of California, San Francisco San Francisco, California

Alan I. Leshner, Ph.D. Conference and Panel Chairperson Chief Executive Officer American Association for the Advancement of Science Executive Publisher Science Washington, DC

Kelli K. Marciel, M.A. Communications Director Office of Medical Applications of Research Office of the Director National Institutes of Health Bethesda, Maryland

Elizabeth McNeil, M.D. Medical Officer Division of Anesthetic, Critical Care, and Addiction Drug Prod-ucts Center for Drug Evaluation and Research U.S. Food and Drug Administration Rockville, Maryland

Merrill M. Mitler, Ph.D. Program Director Systems and Cognitive Neuroscience National Institute of Neurological Disorders and Stroke National Institutes of Health Bethesda, Maryland

Andrew A. Monjan, Ph.D., M.P.H. ChiefNeurobiology of Aging Branch Neuroscience and Neuropsychology of Aging Program National Institute on Aging National Institutes of Health

1056 NIH Statement

SLEEP, Vol. 28, No. 9, 2005

Bethesda, Maryland

Ernestine Murray, R.N., M.A.S. Captain U.S. Public Health Service Senior Health Policy Analyst Senior Advisor on Tobacco Use Center for Outcomes and Evidence Agency for Healthcare Research and Quality Rockville, Maryland

Richard K. Nakamura, Ph.D. Deputy Director National Institute of Mental Health National Institutes of Health Bethesda, Maryland

Lata S. Nerurkar, Ph.D. Senior Advisor for the Consensus Development Program Office of Medical Applications of Research Office of the Director National Institutes of Health Bethesda, Maryland

Judith Owens, M.D., M.P.H. Director Pediatric Sleep Disorders Clinic Associate Professor of Pediatrics Brown Medical School Division of Pediatric Ambulatory Medi-cine Rhode Island Hospital Providence, Rhode Island

Rigoberto Roca, M.D. Deputy Director Division of Anesthetic, Critical Care, and Addiction Drug Prod-ucts Center for Drug Evaluation and Research U.S. Food and Drug Administration Rockville, Maryland

Susan Rossi, Ph.D., M.P.H. Deputy Director Office of Medical Applications of Research Office of the Director National Institutes of Health Bethesda, Maryland

Thomas Roth, Ph.D. Chief Division of Sleep Research Henry Ford Hospital Detroit, Michigan

Matthew V. Rudorfer, M.D. Acting Chief Adult Treatment and Preventive Interventions Research Branch Division of Services and Intervention Research National Institute of Mental Health National Institutes of Health Bethesda, Maryland

Michael Twery, Ph.D. Sleep and Respiratory Neurobiology Research Group Division of Lung Diseases National Heart, Lung, and Blood Institute National Institutes of Health

Bethesda, Maryland

James K. Walsh, Ph.D. Executive Director and Senior Scientist Sleep Medicine and Research Center St. John’s/St. Luke’s Hospitals Chesterfield, Missouri

David P. White, M.D. Director Sleep Disorders Program Division of Sleep Medicine Brigham and Women’s Hospital Boston, Massachusetts

Conference Sponsors

National Institute of Mental HealthThomas R. Insel, M.D. Director

Office of Medical Applications of Research Barnett S. Kramer, M.D., M.P.H. Director

Conference Cosponsors

National Center for Complementary and Alternative Medicine Stephen E. Straus, M.D. Director

National Heart, Lung, and Blood Institute Elizabeth G. Nabel, M.D Director

National Institute of Neurological Disorders and Stroke Story C. Landis, Ph.D. Director

National Institute of Nursing Research Patricia A. Grady, Ph.D., R.N., F.A.A.N. Director

National Institute on Aging Richard J. Hodes, M.D. Director

National Institute on Alcohol Abuse and Alcoholism Ting-Kai Li, M.D. Director

National Institute on Drug Abuse Nora D. Volkow, M.D. Director

Office of Research on Women’s Health Vivian W. Pinn, M.D. Director

U.S. Food and Drug Administration Lester Crawford, D.V.M., Ph.D. Acting Commissioner

1057 NIH Statement

SRC Memo (#16)

From: Drs. Kendler and Freedman on behalf of the DSM-5 Scientific Review

Committee (SRC)

To: John Oldham President APA, Dr. David Kupfer (DSM-5 Task Force Chair),

and Dr. Darrel Regier (DSM-5 Task Force Vice-Chair)

DATE: January 14, 2012

The SRC reviewed the following proposal (48 Insomnia) and requests more information before making the final evaluation:

48 Insomnia: The SRC reviewed the proposal on insomnia. We have outlined our understanding briefly of the proposal below:

The SRC appreciated the work group’s concerns regarding the division of insomnia disorder into primary and secondary types, given that the diagnosis of primary insomnia has poor reliability and convergent validity. Nonetheless, the data justifying this change is relatively scanty, given the magnitude of the change. Specifically, this proposal is likely to increase markedly both the prevalence of the insomnia diagnosis and the amount of comorbidity among psychiatric diagnosis. For example, many patients meeting criteria for MDD or GAD will now qualify for the additional diagnosis of insomnia. Given this, it is important to either provide data documenting that insomnia in association with another illness does not differ from insomnia presenting alone or depression or another mental disorder presenting alone in terms of antecedent, concurrent, and predictive validators or to clearly state that no substantial data exists which address this critical question.

Other changes of note are the addition of both early morning awakening (EMA) and, in children, resistance going to bed or “difficulty returning to sleep without caregiver intervention” as qualifying for the “B” criterion. With regard to the former, the disposition would to a large extent depend on the resolution of the issue above. With regard to the pediatric criteria, the work group does not provide justification for these additions, which seem to run the risk of pathologizing normal developmental variation. Therefore, it is important to provide data supporting the addition of these pediatric criteria.

DSM-IV included exclusion criterion D: The sleep disturbance does not occur exclusively during the course of Narcolepsy, Breathing-Related Sleep Disorder, Circadian Rhythm

Sleep Disorder, or a Parasomnia. This exclusion criterion is not in the DSM-5 proposal, and the change is not justified.

Regarding the recommended duration requirement of three months, the Work Group presents considerable evidence that insomnia typically lasts for more than three months. However, it would be important to see evidence that insomnia lasting more than three months differs from insomnia lasting less than three months (this would be needed to support the view that the latter warrants a diagnosis, whereas the former does not). In addition, this proposed change is somewhat confusing, in that the proposal states that three months should be required for “chronic insomnia,” but the proposed diagnosis is “insomnia disorder”. The work group suggests using specifiers for acute (< 1 month duration), short-term (between 1 and 3 months), and chronic insomnia (longer than 3 months),but a patient with less than three months duration wouldn’t meet criteria for the diagnosis of insomnia disorder, and therefore can’t meet criteria for a specifier for that diagnosis.

The work group proposes requiring a minimum frequency of sleep difficulties three times/week. A somewhat small ROC analysis is presented in support of this but it would be helpful to know if more supporting data exist, since again this is likely to change prevalence.

As a minor wording comment, the work group proposes adding the construct of “sleep dissatisfaction” to criterion A. But, in doing so, they have deleted “difficulty initiating or maintaining sleep, or nonrestorative sleep”. Under the proposal, a person who had difficulty initiating or maintaining sleep, or nonrestorative sleep, but was not dissatisfied with this (approximately 8-13% of individuals, according to the material provided, would not meet criteria for the diagnosis. Is that the work group’s intent?

MEMORANDUM

Date : March 2, 2012 From : DSM5 Sleep Disorders Work Group To : Scientific Review Group Re : DSM5 Response regarding the Insomnia Disorder Proposal Please find below the additional information requested to proceed with the evaluation of the proposed changes for Insomnia Disorder. 1. Primary vs. Secondary Insomnia. While we appreciate the SRC concern that the

proposal to drop the distinction between primary and secondary insomnia might increase prevalence rates, our work group believe this is an unlikely outcome. The proposed criteria for Insomnia Disorder are actually more stringent than those of DSM-IV with regard to the duration threshold, which would be increased from 1 to 3 months, and the addition of a minimum frequency of three nights per week with sleep difficulties. With regard to the potential increase in the prevalence of comorbid insomnia among psychiatric diagnoses, this is also an unlikely scenario because a separate/concurrent Insomnia Disorder would only be made (as for DSM-IV) when the sleep disturbance is sufficiently severe to warrant independent clinical attention. So, we would not expect clinicians to code Insomnia Disorder in the context of another psychiatric disorder, unless there is evidence of clinically significant insomnia (which is now better operationalized in DSM5).

While the evidence suggests that these two conditions (i.e., Insomnia Disorder with Comorbidity and Insomnia Disorder without Comorbidity) are different in terms of clinical presentation (more severe sleep disturbances in comorbid insomnia) and treatment response (worst outcome for comorbid insomnia), it is often difficult in clinical practice to make these distinctions accurately, hence the recommendation to eliminate these distinctions.

There is extensive evidence that insomnia is a significant and prevalent clinical condition in the psychiatric setting, yet it often remains under diagnosed given that some clinicians simply dismiss this complaint as an epiphenomenon. Therefore, even if the prevalence would actually increase with the proposed change, this might simply reflect the clinical reality and, with such a change, one would hope that it would get the clinical attention that it deserves.

2. The symptom of early morning awakening used to be subsumed under difficulties

maintaining sleep in DSM-IV. The Work Group believes it is important to distinguish this symptom from other forms of sleep maintenance difficulties as it may have different implications for diagnosis and may require different treatment.

The addition of pediatric criteria is based upon our concern to integrate a developmental perspective and the evidence that sleep disturbances occur at all ages, including in children. This criterion is not referring to the typical 3-year-old who stalls a bit at bedtime or the 4-month old infant who happens to wake once

per night for a feeding. Rather, it is meant for clinically significant sleep difficulties in children, consistent with the literature on the prevalence and impact of these behaviourally-based sleep disturbances and on standards of practice for the treatment of these difficulties. We are attaching a key paper on this issue providing evidence that these children are different from their “normal” developmentally-matched peers and that such condition is treatable (AASM, 2006, p. 1263-64, par. 1.0-1.2; Tables 3-4).

3. Criterion D. “The sleep disturbance does not occur exclusively during the course

of Narcolepsy, Breathing-Related Sleep Disorder, Circadian Rhythm Sleep Disorder, or a Parasomnia”. While we have omitted to address explicitly this exclusion, the rationale is the same as for medical and psychiatric disorders. The recommendation is that clinicians should no longer have to make a causal attribution as to whether insomnia is primary or associated with another sleep disorder. Any of these conditions could still be listed as a comorbid sleep disorder in as much as they are judge to have a clinical impact on insomnia.

4. Duration requirement. Although there is ample evidence (e.g., Ohayon et al.,

2012; Roth et al., 2006) of significant morbidity (e.g., role impairment, work absenteeism and disability, risks of depression and hypertension) associated with chronic insomnia (variously defined as greater than 3, 6, or 12 months), there has been no direct comparison of morbidity associated with insomnia lasting 1 month vs. 3, 6, or 12 months. Therefore, we cannot provide such evidence in support of our recommendation.

We agree with the SRC that there is some logical inconsistency in the semantic used to describe acute, short-term, and chronic insomnia. To avoid such confusion, we are now proposing to drop these specifiers and retain only the three-month cut-point to define an Insomnia Disorder. However, in order to prevent under diagnosis and treatment of persons with insomnia shorter than three months, we are proposing that a diagnosis of Insomnia Disorder (NEC) be made for individuals who meet all criteria (i.e., symptoms, burden, distress), except for the 3-month duration. We will address in the narrative these latter cases in the context of “sub threshold” insomnia. We believe it is important to recognize these cases with insomnia of shorter duration than three months in order to not withhold early treatment, which may be necessary to prevent chronicity and to minimize morbidity.

5. Minimum frequency. Although the evidence highlighted in our memo to the SRC

in support of this recommendation was restricted to a single study (Lichstein et al., 2003), mainly for sake of brevity, there is additional evidence to support this change. One additional paper is provided as supportive evidence (Lineberg et al., 2006; Figure 1 and Tables 2-4). In addition, there are several epidemiological studies (Morin et al., 2011; Ohayon, 2002; Ohayon & Reynolds, 2012, p. 954, par. 2.3.2) that have used this criterion (i.e., minimum frequency of 3 nights per week) to document the prevalence of insomnia; furthermore, most clinical trials (at least behavioral ones) typically use this criterion to operationalize the diagnosis of insomnia before enrolling patients. This criterion is already used by ICD-10 and its addition in DSM5 would harmonize criteria across diagnostic nosologies.

6. Sleep dissatisfaction. Our proposal is to add the construct of “sleep dissatisfaction” while retaining the other criteria about “difficulty initiating or maintaining sleep, or non restorative sleep”. The rationale for adding “sleep dissatisfaction” is that some individuals are dissatisfied with their sleep, yet do not report specific insomnia symptoms, whereas others may report insomnia symptoms without necessarily being dissatisfied with their sleep. Accordingly, our work group recommendation is that an Insomnia Disorder diagnosis should be reserved only for those individuals who are both dissatisfied with their sleep and present insomnia symptoms. Criteria A and B could be integrated into a single criterion if the SRC believes it would enhance clarity of diagnostic criteria.

SRC Memo (#21)


Committee (SRC)



DATE: March 29, 2012

We would like to add the following contextual comments on the attached report:

48R=Insomnia: The SRC voted on 6 separate elements of this proposal. Overall, it was felt that the group did not respond well in this revision. Details are given in the reviewers reports.

SRC Recommendation Report March 28, 2012

1

48R= Insomnia

There were two reviewers for these proposals. Their review summaries are below. The group voted on 6 parts of this proposal separately, see table below for the scores.

Scoring Summary: 1=strong support 2=moderate support (acceptable) 3=modest support (questionable) 4= limited support (probably not justified) 5= poor support (do not include) 6= insufficient data

48R Areas Scored Average Replacing “Primary Insomnia” and “Insomnia related to another mental/medical disorder” with Insomnia Disorder, with specification of clinically comorbid conditions

5

Adding early morning awakening as a separate symptom 5 The omission of DSM-IV exclusion criterion D, “The sleep disturbance does not occur exclusively during the course of Narcolepsy, Breathing-Related Sleep Disorder, Circadian Rhythm Sleep Disorder, or a Parasomnia.”

5

Raising the minimum duration threshold from 1 to 3 months for chronic insomnia 2 Adding a minimum frequency criterion (i.e., 3 nights per week) with sleep disturbance 4.33 Requiring sleep dissatisfaction 5


2

Proposal Title: Insomnia Disorder--Revised

Proposed change:

In their response to the SRC, the Work Group focuses on the following changes, which were highlighted as problematic in the SRC’s memo:


2. Adding early morning awakening as a separate symptom 3. The omission of DSM-IV exclusion criterion D, “The sleep disturbance does not occur exclusively

during the course of Narcolepsy, Breathing-Related Sleep Disorder, Circadian Rhythm Sleep Disorder, or a Parasomnia.”

4. Raising the minimum duration threshold from 1 to 3 months for chronic insomnia 5. Adding a minimum frequency criterion (i.e., 3 nights per week) with sleep disturbance 6. Requiring sleep dissatisfaction

Reviewer: 1

Date: March 18, 2012

Work group’s reply and reviewer’s response:

1) Primary vs. secondary insomnia: The SRC’s concern is that, under the DSM-5 proposal, a patient who had insomnia in the course of e.g., major depressive disorder or generalized anxiety disorder, would now meet criteria for both the non-insomnia psychiatric disorder and for insomnia disorder. This would increase the prevalence of both co-morbidity and insomnia disorder. In their response, the work group replies, “With regard to the potential increase in the prevalence of comorbid insomnia among psychiatric diagnoses, this is also an unlikely scenario because a separate/concurrent Insomnia Disorder would only be made (as for DSM-IV) when the sleep disturbance is sufficiently severe to warrant independent clinical attention.”

To this reviewer, it is unclear how the requirement for “independent clinical attention” is operationalized in the DSM-5 criteria. There is no language to this effect in the proposed criteria. Furthermore, the DSM-IV criterion D, which states “the disturbance does not occur exclusively during the course of another mental disorder (e.g., Major Depressive disorder, Generalized Anxiety Disorder, a delirium)”, is missing from the DSM-5 proposal. Therefore, the work group’s contention, that the insomnia diagnosis will only be assigned in the presence of another diagnosis if the former requires independent attention, is unwarranted. Rating: 5

2) Early morning awakening and the pediatric criteria:

For early morning awakening, the SRC’s concern was that EMA secondary to an episode of major depression would be viewed as a symptom of an insomnia disorder, perhaps leading to under-recognition of the depressive disorder. In its memo, the SRC stated that “the disposition [of this issue] would to a large extent depend on the resolution of the issue above,” and that appears to still be the case.


3

The SRC also asked the work group to provide data supporting the addition of pediatric criteria and expressed particular concern as to how pathological sleep difficulties would be differentiated from normative developmental variation. In response, the work group provides a very relevant and informative paper on standards of practice for pediatric insomnia (AASM 2006). As the work group notes, this paper provides evidence for the existence of “behavioral insomnia” in children that responds to behavioral intervention. However, it is also important to note that this publication includes diagnostic criteria for insomnia in children which are considerably more detailed and restrictive than those in the DSM-5 proposal (see Table 1). For example, the AASM manuscript provides specific criteria for two types of behavioral insomnia, sleep-onset association and limit-setting, with specific criteria such as “falling asleep is an extended process that requires specific conditions,” “the caregiver demonstrates insufficient or inappropriate limit setting to establish appropriate sleeping behavior in the child,” and others. Rating: 5

3) Eliminating criterion D: “The sleep disturbance does not occur exclusively during the course of Narcolepsy, Breathing-Related Sleep Disorder, Circadian Rhythm Sleep Disorder, or a Parasomnia.” The work group favors this deletion because clinicians would not have to make causal attributions. They note that these conditions could be listed as a comorbid sleep disorder “in as much as they are judged to have a clinical impact on insomnia.” In the case of Breathing-Related Sleep Disorder, Circadian Rhythm Sleep Disorder, given their diagnostic criteria, it appears that they would necessarily have to have a clinical impact on insomnia, and therefore the elimination of this criterion would seem to add to unnecessary comorbidity. Rating: 5

4) Duration requirement The work group now agrees with dropping the proposed specifiers (acute, short-term, and chronic). Instead, three months duration will be required. The Work Group recommends that patients who meet all criteria for the disorder except the duration criterion receive the diagnosis of insomnia disorder NEC, given that insomnia disorder lasting less than three months may warrant treatment. Rating: 2

5) Minimum frequency The proposed DSM-5 criteria require insomnia three times a week. In addition to the study that the work group provided in the initial proposal that included one ROC analysis (Lichstein et al, 2003), they now provide Lineberger et al. 2006. While this study (N=72) does include an ROC analysis, it does not seem to address the issue of frequency. The work group also notes that the 3x/week criterion has been used in several epidemiological studies and in most treatment trials. Furthermore, it is used in ICD-10, so that this would increase harmonization with ICD. Rating: 4

6) Sleep dissatisfaction The work group’s proposal is to require both sleep dissatisfaction and insomnia symptoms. The SRC noted that this would eliminate about 8-13% of patients, who have insomnia symptoms but do not express dissatisfaction with their sleep. In their response, the work group suggests integrating criteria A and B. Presumably, the work group’s specific proposal is that the “A” criterion would be eliminated, and “dissatisfaction with sleep” would become one of the “B”


4

criteria; of note, only one “B” criterion is required. Therefore, theoretically it would be possible for a patient to express dissatisfaction with sleep but not have difficulty initiating, EMA, etc. In that case, it is unclear how criterion E (“the sleep difficulty occurs at least three nights per week”) could be met. Rating: 5


5

Proposal Title: Insomnia

Proposed change: Elimination of secondary insomnia and additional of pediatric criteria

Reviewer: 2

Date: 3/22/2012

Part 1-Reasons and consequences of proposed change: Reasons for the change: Possible negative consequences of the proposed change: (For Parts 2 through 6 below check whether the element is addressed and all that apply)

Part 2 – Magnitude of the change: Criterion/criteria clarification

Modest change (Includes changes to a specifier, to the examples provided in an NOS category description, or to subtype criteria OR Includes the addition of a new specifier or subtype to a diagnosis that has not been widely studied or well validated

Substantial change: Refers to meaningful changes to the DSM-IV criteria of a diagnosis that has not been widely studied or well validated. Includes the addition of a new specifier or subtype to a well-validated diagnosis

Major change: Refers to meaningful changes to the DSM-IV criteria of a widely studied and well-validated diagnosis. Includes the addition of a new diagnosis to DSM-5

N/A Comments (optional): Part 3 – Evidence for the change:

Is the proposed set of criteria for disorder X superior to the DSM-IV criteria for disorder X?

Antecedent Validators

*Familial aggregation and/or co-aggregation (i.e., family, twin or adoption studies)

Socio-Demographic and Cultural Factors

Environmental Risk Factors

Prior Psychiatric History

N/A


6

Comments (optional):

Concurrent Validators [Note - while categories A and B would most typically be assessed after illness onset, they also could be assessed prior to illness onset as pre-morbid characteristics]

Cognitive, emotional, temperament, and personality correlates (unrelated to the diagnostic criteria)

Biological Markers, e.g., molecular genetics, neural substrates

Patterns of Comorbidity

N/A


Predictive Validators

*Diagnostic Stability

*Course of Illness

*Response to Treatment

N/A


Part 4 - For New Diagnostic Category:

Are these addressed?

A need for the category

Relationship with other DSM categories

Potential harm in adopting the category

Available treatments for diagnoses in the category

Diagnosis meets the criteria for a mental disorder

N/A


Part 5 - Deletion of Existing Diagnosis:

Clinical utility of the syndrome (the magnitude of adverse effects on our patients that would arise from the deletion of the syndrome):

Frequency of use


7

Importance in making treatment decisions

Role in stimulating the development of clinical program

Increasing diagnosis in professional and lay groups

N/A


Overall quality of information about the validity of the category (whether we have any confidence in the validity of this syndrome based on the set of validators outlined above):

Empirical support for their validity is minimal

Evidence for low reliability, however, would contribute to evidence in favor of exclusion

N/A Comments (optional): Part 6 - PROPOSED SRC RECOMMENDATION:

1) strong support 2) moderate support (acceptable) for the 3 months criteria 3) modest support (questionable) 4) limited support (probably not-justified) 5) poor support (do not include) for everything else 6) insufficient data

Additional Comments: The meta-structural problem is not adequately addressed in this re-submission. The work group claims that the secondary diagnosis is difficult and should be eliminated. However, the comorbidity is significant. According to the Ohayon and Reynolds epidemiological screening for insomnia identified a prevalence of 3% primary insomnia, 1.5% insomnia secondary to a psychiatric diagnosis, 3.1% mood disorder, and 6.8% anxiety disorder. Eliminating second insomnia does not take this marked overlap into account and even if treatment for the comorbidity involved addressing both issues, as it does for most comorbid problems, eliminating the secondary criteria does not add clarity to this issue nor is it supported by the data. The pediatric issue was addressed by the work group’s providing a review paper. It is not clear from reading the paper provided how limit setting bed time problems are distinguished from primary insomnia. Although the work group advocated for the diagnosis because treatment is mandated, the treatment in the article was behavioral and education for limits and elementary sleep hygiene. Given that the prevalence of resistance to going to bed is 20-30%, a clearly supported rationale for the diagnostic criteria seems warranted and is not yet achieved.

Memorandum of Response to SRC Review of 48R, Insomnia Disorder TO: DSM-5 Scientific Review Group FROM: Charles F. Reynolds III, M.D. Chair, DSM-5 Sleep Wake Disorders Workgroup RE: Response to SRC Review of 48R, Insomnia Disorder DATE: April 10, 2012 My colleagues on the DSM-5 Sleep Wake Disorders Workgroup join me in thanking the Scientific Review Committee (SRC) for their reviews (January 14, 2012; and March 28, 2012) of the proposed diagnostic criteria for Insomnia Disorder. We apologize for any lack of clarity in our responses, particularly our failure to explain the basic logic of the choices we have made. In this memo we shall try to explain the logic of critical choices involved in the five recommendations below, which were judged by the SRC to have poor support. Although we do not think there are any additional data per se that we can adduce, we do think that the clinical logic of our recommendations needs to be clarified. Such clarification is our sole purpose here. To recapitulate: the SRC Reviewer scored the Insomnia Disorder proposal at 5 (“poor support: do not include”) with respect to five components of the recommended change: (1) replacing “Primary Insomnia” (per DSM-IV) and “Insomnia related to another mental/medical disorder” with “Insomnia Disorder” and specification of clinically comorbid conditions; (2) adding early morning awakening as a separate symptom; (3) omission of DSM-IV exclusion criterion D (“The sleep disturbance does not occur exclusively during the course of Narcolepsy, Breathing-Related Sleep Disorder, Circadian Rhythm Sleep Disorder, or a Parasomnia.”); (4) adding a minimum frequency criterion (i.e., three nights per week) with sleep disturbance; and (5) requiring sleep dissatisfaction. The SRC scored at 2 (moderate support, acceptable) the workgroup’s recommendation to raise the minimum threshold duration from 1 to 3 months for chronic insomnia.

Response to SRC Memo 48 R (Insomnia Disorder) April 10, 2012 Page 2 Clinical logic of key recommendations:

(1) The proposed deletion of DSM-IV “Primary Insomnia” and “Insomnia related to…etc” reflects a change of paradigm that is now widely accepted in the field of sleep disorders medicine, and which was articulated in a 2005 NIH State of the Science conference on chronic insomnia (Sleep 2005; 28: 1049-1057). This paradigm moves away from the need for clinicians to make causal attributions between co-existing disorders and requires only specifying clinical comorbidities.

The clinical utility of this approach, we feel, is two-fold: (a) it acknowledges the bidirectional and interactive effects between sleep disorders like insomnia and co-existing medical/psychiatric disorders. As such, (b) we think that the proposed change emphasizes the likelihood of the patient’s having an independent sleep disorder warranting independent clinical attention (but which is often ignored in practice). For example, one can treat a patient with a major depressive episode to the point where most depressive symptoms have remitted, but with persistence of complaints of dissatisfying or poor sleep. Unless the patient’s insomnia disorder is also treated appropriately, the patient will be at higher risk for relapse and recurrence of a major depressive episode. In the process of formulating this recommendation, we were reminded by consulting methodologist, Dr. Helena Kraemer, that our proposed approach is actually more conservative scientifically because it does not involve making potentially unwarranted and clinically inappropriate causal attributions. In addition, Dr. Kraemer argued that our proposal to list coexisting disorders is more comprehensive clinically, thereby decreasing the probability of overlooking an important comorbidity. Moreover, Dr. Kraemer’s argument subsequently received further empirical support from the Edinger et al. report in the Archives of General Psychiatry (2011; 68(10): 992-1002), which showed that the DSM-IV diagnostic criteria for ”Primary Insomnia” have poor performance characteristics. Thus, the negative consequences of retaining DSM-IV criteria for “primary insomnia” would include retention of a criterion set with poor performance characteristics, as well as exposing patients with insomnia disorder to the increased risk that coexisting conditions warranting independent treatment may be overlooked in diagnostic assessment and treatment planning.

(2) The addition of early morning awakening as a separate symptom improves the specificity of symptoms. In other words, it alerts the clinician to the need to consider a set of differential diagnostic possibilities, such as major depression or such as age-dependent decreases in the ability to sleep, or of an age-dependent shift in the timing of the major sleep period (e.g., advanced sleep phase syndrome). The distinction being

Response to SRC Memo 48R (Insomnia Disorder) April 10, 2012 Page 3 made here is between clinically significant early morning awakening (as in depression) and early morning awakening due to an early sleep wake schedule. (3) Omission of DSM-IV exclusion criterion D: Our clinical reasoning here is the same as for our recommendation that coexisting medical and psychiatric disorders be separately specified. In other words, we think that clinicians should no longer have to make a causal attribution as to whether insomnia is “primary” or coexistent with another sleep disorder. Any of these conditions previously specified under the DSM-IV criterion D could, and indeed should, still be listed as a coexisting sleep disorder requiring clinical attention in their own right. (4) Adding a minimum frequency criterion of three nights weekly: our clinical rationale for specifying a frequency threshold is based empirically in the work of Stanford sleep epidemiologist Maurice Ohayon. Dr. Ohayon has found that thrice weekly occurrence of dissatisfying sleep is associated with reports of increased daytime distress and impairment in major role functioning (Ohayon MM, Riemann, Morin CM, and Reynolds CF, Hierarchy of insomnia criteria based on daytime consequences. Sleep Medicine, 2012 (1): 52-57). Most individuals seeking treatment in clinical settings or in the context of clinical trials exceed this threshold. The “bottom line” clinically is that a minimal insomnia frequency of three nights weekly (as well as a duration of three months or longer) represent important cutpoints for determining morbidity and impairment. Again, our methodology consultant, Dr. Helena Kraemer, reminded us that this recommendation is more conservative scientifically that the previous DSM-IV criteria because they are based not in expert opinion, but in actual data with demonstrated clinical relevance. (5) Requiring sleep dissatisfaction: Here our clinical logic is simply that the construct of sleep dissatisfaction is more strongly associated with daytime impairments than are insomnia symptoms alone and thus is more likely to be a driver of help seeking (Ohayon MM, Sleep Medicine Reviews, 2002 (6): 97-111). Summary: We thank the SRC for the opportunity to respond to both critiques. We hope that this presentation of our clinical logic will address the concerns of the SRC. Otherwise, we may be left with an honest difference of opinion about how best to translate the scientific advances in sleep medicine and sleep epidemiology since DSM-IV into more clinically useful but scientifically conservative criteria for Insomnia Disorder. It is our view that the proposed criteria serve the mission of DSM-5 to be a living, breathing document reflecting the best of the available science, while maximizing clinical utility for the

Response to SRC Memo 48R (Insomnia Disorder) April 12, 2012 Page 4 mental health and general clinician who is not expert in sleep disorders medicine. Conversely, retention of DSM-IV criteria for “primary insomnia” would be to ignore their poor performance characteristics and to expose patients to increased risk of less comprehensive diagnostic assessment (by not accounting for coexisting conditions). Thank you for your further consideration of our perspectives on this matter.

1

SRC Memo (#24)


Committee (SRC)



DATE: May 18, 2012

48R2: Insomnia Disorder: The SRC has reviewed this proposal three times and still gives it modest to limited support with inclusion in the DSM5 deemed unjustified. The removal of the comorbidity exclusion is still not adequately justified. Contrary to the claim of the Workgroup that their elimination of the comorbidity exclusion is justified by its low reliability, the Eddinger article provided by the Workgroup finds that comorbidity between insomnia and other psychiatric disorders is rated with higher reliability than primary insomnia itself. The SRC is concerned that comorbidty with other psychiatric illnesses, namely depression with its suicide risk, will be overlooked if this change is allowed to proceed.

SRC Recommendation Report May 18, 2012

1

48R2=Insomnia Disorder

There were two reviewers for these proposals. Their review summaries are below. The group voted on two parts of this proposal outlined in the table below.

Scoring Summary: 1=strong support 2=moderate support (acceptable) 3=modest support (questionable) 4= limited support (probably not justified) 5= poor support (do not include) 6= insufficient data

Area of Proposal Score Requiring Sleep Dissatisfaction 3 The rest of the proposal 4


2

Proposal Title: 48R2=Insomnia Disorder

Proposed change: This is the second revision (i.e., third submission) of the insomnia proposal. Points of contention include the following proposals:


2. Adding early morning awakening as a separate symptom 3. The omission of DSM-IV exclusion criterion D, “The sleep disturbance does not occur

exclusively during the course of Narcolepsy, Breathing-Related Sleep Disorder, Circadian Rhythm Sleep Disorder, or a Parasomnia.”

4. Adding a minimum frequency criterion (i.e., 3 nights per week) with sleep disturbance 5. Requiring sleep dissatisfaction

Reviewer: 1

Date: 4/30/2012

See prior reviews for ratings of magnitude, validators, etc. Part 6 - PROPOSED SRC RECOMMENDATION: SEE BELOW

1) strong support 2) moderate support (acceptable) 3) modest support (questionable) 4) limited support (probably not-justified) 5) poor support (do not include) 6) insufficient data


1) In his reply, Dr. Reynolds emphasizes that a major motivation for the proposed change is that the diagnosis of primary insomnia performed poorly in terms of reliability and validity. This is certainly a concern, and the SRC appreciates the Work Group’s desire to address it. However, it is important to note that, while primary insomnia fared poorly in Edinger et al (2011), insomnia related to another mental disorder was the best-supported diagnosis. In this proposal, both the poorly-performing primary insomnia and the well-performing insomnia related to another mental disorder are eliminated.

The Work Group proposes that any patient who meets criteria for insomnia would receive this diagnosis, even if the insomnia occurs in the context of another illness known to commonly include insomnia. In response to the two prior submissions, the SRC expressed concern that increased prevalence and comorbidity that would result, and did not feel that the evidence presented was sufficiently strong to warrant such a major change. In Dr. Reynold’s response, he notes that, if the new criteria are not adopted, there is a “risk that coexisting conditions warranting independent treatment may be overlooked in


3

diagnostic assessment and treatment planning.” However, in this proposal, there is no requirement that the insomnia warrant independent treatment. For example, insomnia occurring in the context of MDD, which did not require independent treatment, would still receive the diagnosis of comorbid insomnia. Indeed, in the Work Group’s previous response, they made a similar argument, arguing that prevalence and comorbidity would not increase because the diagnosis of insomnia would only be made when the insomnia “required independent treatment.” In response, the SRC noted the following (p.2): “…it is unclear how the requirement for “independent clinical attention” is operationalized in the DSM-5 criteria. There is no language to this effect in the proposed criteria. Furthermore, the DSM-IV criterion D, which states “the disturbance does not occur exclusively during the course of another mental disorder (e.g., Major Depressive disorder, Generalized Anxiety Disorder, a delirium)”, is missing from the DSM-5 proposal. Therefore, the work group’s contention, that the insomnia diagnosis will only be assigned in the presence of another diagnosis if the former requires independent attention, is unwarranted.” Unfortunately, the Work Group has not provided a direct response to this comment. It is unclear why the Work Group didn’t include as a criterion the requirement that the insomnia requires independent treatment, if indeed this is their intent. In any case, it would be very much appreciated if this aspect of our critique could receive a direct response. Also, as noted above, perhaps the Work Group could consider other revisions to the primary insomnia diagnosis that would rectify its poor performance, without eliminating the well-performing diagnosis of insomnia related to another mental disorder. Of note, Edinger et al (2011) state (p.1001) with regard to the Work Group’s proposal: “Of course, use of the global insomnia disorder diagnosis could also encourage a generic one-size-fits-all approach to insomnia treatment. Whether this concern is warranted will only be determined once the DSM-5 is placed into use.” Rating: 4

2) Early morning awakening as a separate criterion: The Work Group is concerned that important differential diagnostic considerations will be overlooked. Dr. Reynolds states, “[This revision] alerts the clinician to the need to consider a set of differential diagnostic possibilities, such as major depression or such as age-dependent decreases in the ability to sleep, or of an age-dependent shift in the timing of the major sleep period (e.g., advanced sleep phase syndrome). The distinction being made here is between clinically significant early morning awakening (as in depression) and early morning awakening due to an early sleep wake schedule.” Again, it appears that the Work Group’s intent is that, in a patient with EMA and MDD, that symptom should count toward both diagnoses. Conceptually, the SRC’s concern here is the same as in #1.


4

Rating: 4 Dr. Reynolds does not provide a response to the SRC’s critique of the pediatric criterion.

3) Again, the SRC’s concern is the same as in # 1. Rating: 4

4) and 5) The Work Group now provides further data in support of the 3x/week and sleep dissatisfaction criteria. Specifically, they provide a study (Ohayon 2012) that includes a large community based sample (N= 8937). Individuals were judged to have insomnia if they had sleep dissatisfaction and insomnia symptoms at least three times a week for at least three months. Of all insomnia symptoms, global sleep dissatisfaction showed the strongest association with daytime consequences. As noted, the frequency of 3x/week was not tested, but was posited a priori.

Adding minimum frequency: 4

Adding sleep dissatisfaction: 3


5

Proposal Title: 48R2=Insomnia Disorder

Proposed change: Removal of exclusion for comorbidity with other psychiatric illnesses

Reviewer: 2

Date: 4/30/2012

Part 1-Reasons and consequences of proposed change: Reasons for the change: To increase diagnosis of insomnia and thereby promote its treatment. Possible negative consequences of the proposed change: Diagnosis of insomnia and failure to appreciate that it is comorbid with other treatable and potentially fatal illness, namely depression. (For Parts 2 through 6 below check whether the element is addressed and all that apply)

Part 2 – Magnitude of the change: Criterion/criteria clarification

Modest change (Includes changes to a specifier, to the examples provided in an NOS category description, or to subtype criteria OR Includes the addition of a new specifier or subtype to a diagnosis that has not been widely studied or well validated

Substantial change: Refers to meaningful changes to the DSM-IV criteria of a diagnosis that has not been widely studied or well validated. Includes the addition of a new specifier or subtype to a well-validated diagnosis

Major change: Refers to meaningful changes to the DSM-IV criteria of a widely studied and well-validated diagnosis. Includes the addition of a new diagnosis to DSM-5

N/A Comments (optional): Part 3 – Evidence for the change:

Is the proposed set of criteria for disorder X superior to the DSM-IV criteria for disorder X?

Antecedent Validators

*Familial aggregation and/or co-aggregation (i.e., family, twin or adoption studies)

Socio-Demographic and Cultural Factors

Environmental Risk Factors

Prior Psychiatric History


6

N/A

Comments (optional): 0T

Concurrent Validators [Note - while categories A and B would most typically be assessed after illness onset, they also could be assessed prior to illness onset as pre-morbid characteristics]

Cognitive, emotional, temperament, and personality correlates (unrelated to the diagnostic criteria)

Biological Markers, e.g., molecular genetics, neural substrates

Patterns of Comorbidity

N/A


Predictive Validators

*Diagnostic Stability

*Course of Illness

*Response to Treatment

N/A


Part 4 - For New Diagnostic Category:

Are these addressed?

A need for the category

Relationship with other DSM categories

Potential harm in adopting the category

Available treatments for diagnoses in the category

Diagnosis meets the criteria for a mental disorder

N/A


Part 5 - Deletion of Existing Diagnosis:

Clinical utility of the syndrome (the magnitude of adverse effects on our patients that would arise from the deletion of the syndrome):


7

Frequency of use

Importance in making treatment decisions

Role in stimulating the development of clinical program

Increasing diagnosis in professional and lay groups

N/A


Overall quality of information about the validity of the category (whether we have any confidence in the validity of this syndrome based on the set of validators outlined above):

Empirical support for their validity is minimal

Evidence for low reliability, however, would contribute to evidence in favor of exclusion

N/A Comments (optional): Part 6 - PROPOSED SRC RECOMMENDATION:

1) strong support 2) moderate support (acceptable) 3) modest support (questionable) Requiring Sleep Dissatisfaction 4) limited support (probably not-justified) The rest of the proposal 5) poor support (do not include) 6) insufficient data


The removal of the comorbidity requirement is still not well justified. The Eddinger article referred to states that comorobidity between sleep disorder is problematic. The SRC is concerned with comorbidty with other psychiatric illnesses, namely depression.

The data from Eddinger are below:


8

On clinical interview, related to a mental disorder or related to a medical condition has higher irr than the primary diagnosis itself.

table of contents 1. revised criteria for insomnia...

Documents