psychometric evaluation of a chinese version of the lee fatigue scale-short form in women during...

PsFapo

Sha Deb By

� Fsy

International Journal of Nursing Studies xxx (2013) xxx–xxx

A R

Artic

Rece

Rece

Acce

Keyw

Acti

Fati

Post

Preg

Wom

Slee

*

Stat

fax:

(S.-C

(Y.-L

G Model

NS-2314; No. of Pages 9

PlShj.i

002

http

ychometric evaluation of a Chinese version of the Leetigue Scale-Short Form in women during pregnancy andstpartum

ao-Yu Tsai a, Shiow-Ching Shun a, Yeur-Hur Lai a, Ya-Ling Lee a, Shih-Yu Lee b,*

partment of Nursing, National Taiwan University, Taipei, Taiwan

rdine F. Lewis School of Nursing, Georgia State University, Atlanta, GA, USA

What is already known about the topic?

atigue is among the most prevalent and distressingmptoms in pregnant and postpartum women, with up

to 60–90% of women experiencing fatigue at some timeduring pregnancy or postpartum.� Fatigue is a significant health issue because it is associated

with increased maternal psychological symptoms duringpregnancy, increased risks for cesarean deliveries, and anew onset of postpartum depressive symptoms.

What this paper adds

� The Chinese version of the Lee Fatigue Scale-Short Form(C-LFS-SF) has sound psychometric properties and is apromising tool for assessing fatigue in a rapid and

T I C L E I N F O

le history:

ived 16 August 2012

ived in revised form 14 October 2013

pted 24 October 2013

ords:

graphy

gue

partum

nancy

en

p

A B S T R A C T

Background: Fatigue is among the most prevalent and distressing symptoms in pregnant and

postpartum women. Estimating fatigue severity with a psychometrically sound instrument

provides the most fundamental information for understanding women’s fatigue experience

and assessing the need for intervention to improve maternal and infant health outcomes.

Objectives: To evaluate the psychometric properties of a Chinese version of the 7-item Lee

Fatigue Scale-Short Form (C-LFS-SF) in pregnant and postpartum women.

Methods: The study was composed of two phases: translation of the scale into Chinese and

examination of content validity, and testing to establish the reliability and validity. A

convenience sample of 124 women completed health-related questionnaires, kept a

fatigue diary and wore a wrist actigraph for 7 days during the third trimester of pregnancy

and within three months postpartum.

Results: The C-LFS-SF showed satisfactory internal consistency (Cronbach’s alpha � .97)

and stability over the 7 days at each time point (intraclass correlation coefficient > .87).

Exploratory factor analysis showed that 88–94% of the total variance was explained by the

one-factor fatigue model. Significant associations among the C-LFS-SF and actigraphic

sleep quantity and quality variables supported adequate construct validity.

Conclusions: The C-LFS-SF has satisfactory psychometric properties and is an easy and

promising tool for assessing maternal fatigue during routine prenatal and postpartum

care. This scale needs to be further tested in a more diverse population of pregnant and

postpartum women, like women with high risk pregnancies or with medical conditions.

� 2013 Published by Elsevier Ltd.

Corresponding author at: Byrdine F. Lewis School of Nursing, Georgia

e University, P.O. Box 4019, Atlanta, GA, USA. Tel.: +1 404 413 1176;

+1 404 413 1205.

E-mail addresses: [email protected] (S.-Y. Tsai), [email protected]

. Shun), [email protected] (Y.-H. Lai), [email protected]

. Lee), [email protected] (S.-Y. Lee).

Contents lists available at ScienceDirect

International Journal of Nursing Studies

journal homepage: www.elsevier.com/ijns

ease cite this article in press as: Tsai, S.-Y., et al., Psychometric evaluation of a Chinese version of the Lee Fatigue Scale-ort Form in women during pregnancy and postpartum. Int. J. Nurs. Stud. (2013), http://dx.doi.org/10.1016/

jnurstu.2013.10.023

0-7489/$ – see front matter � 2013 Published by Elsevier Ltd.

://dx.doi.org/10.1016/j.ijnurstu.2013.10.023

http://dx.doi.org/10.1016/j.ijnurstu.2013.10.023

mailto:[email protected]







http://www.sciencedirect.com/science/journal/00207489


S.-Y. Tsai et al. / International Journal of Nursing Studies xxx (2013) xxx–xxx2

G Model


quantitative manner during routine prenatal and post-partum care.� The ability to detect changes suggests that potential uses

of the C-LFS-SF include assessments of fatigue before andafter interventions as an indication of the effectiveness ofa therapy or coping strategy.

1. Background

Fatigue is among the most prevalent and distressingsymptoms in pregnant and postpartum women, with upto 60–90% of women experiencing fatigue at some timeduring pregnancy or postpartum (McGovern et al., 2011;Reeves et al., 1991; Rodriguez et al., 2001). Fatigue duringpregnancy and postpartum is a significant health issuebecause it affects both the mother and the infant in short-term and long-term outcomes (Corwin et al., 2005).Outcomes include, but are not limited to, increasedpsychological symptoms during pregnancy, increasedrisks for cesarean deliveries, and a new onset ofpostpartum depressive symptoms (Chien and Ko, 2004;Dennis and Ross, 2005; Reeves et al., 1991). Rather thanself-limiting (Corwin and Arbour, 2007; Taylor andJohnson, 2010), fatigue can persist throughout pregnancyand continue up to 18 months after delivery (Elek et al.,1997; Lee and Zaffke, 1999; Parks et al., 1999). Prolongedand unresolved fatigue interferes with maternal capacityfor infant-care activities and has been associated withpoor maternal mental and physical health, delayed infantdevelopment, and more symptoms of infection in bothmothers and infants (Groer et al., 2005; Parks et al., 1999;Pugh and Milligan, 1993). Estimating fatigue severitywith a psychometrically sound instrument provides themost fundamental information for understandingwomen’s fatigue experience and assessing the need forintervention to improve maternal and infant healthoutcomes (Dodd et al., 2001; Fairbrother et al., 2008;Tsai et al., 2012).

Fatigue has been defined as an overwhelming unpleas-ant feeling reflecting physiological, psychological, andpotentially pathological state of exhaustion that decreasesan individual’s energy level and capacity for positive motorand cognitive performance and self-management abilities(Aaronson et al., 1999; Lerdal et al., 2011; Pugh andMilligan, 1993). Fatigue is a central component of nursingdiagnosis and nursing care plans. Existing literatureprovides a theoretical framework and empirical basis forunderstanding childbearing and childrearing fatigue inwomen’s lives (Lee et al., 1994; Pugh and Milligan, 1993;Runquist, 2007). In these models, fatigue develops wheninternal and external demand exceeds women’s resources.During pregnancy, labor and delivery, and postpartum,there are multiple physiological, psychological and situa-tional factors that predispose women to childbearing orchildrearing fatigue along a continuum ranging from mildtiredness to severe exhaustion (Hall et al., 2009; Lee et al.,1994; Pugh and Milligan, 1993). For example, sleepdisturbance is conceptualized as an internal environmen-tal demand and a situational factor causing maternalfatigue in these models. Available data also confirm thatfatigue during pregnancy and postpartum primarily results

from insufficient or poor sleep (Insana et al., 2011;Rychnovsky and Hunter, 2009; Tsai et al., 2012).

Among the self-report instruments used to assessmaternal fatigue, the Lee Fatigue Scale (LFS) is a reliableand validated instrument, and by far the most widely usedfor assessing diurnal fatigue patterns in both clinical andnon-clinical samples of pregnant and postpartum women(Elek et al., 1997; Gay et al., 2004; Maloni and Park, 2005).This instrument was originally designed as a 100 mmvisual analog scale with a 13-item fatigue subscale and 5-item energy subscale (Lee et al., 1991). For easieradministration and scoring when fatigue is a dynamicphenomenon assessed repeatedly in relation to an event ortime frame (Aaronson et al., 1999), the LFS was convertedto a short form (LFS-SF) numeric rating scale for fatiguecontaining 7 items with answers ranging from 0 (not at all)to 10 (extremely) (Gay et al., 2004; Lee and Gay, 2004; Leeand Lee, 2007). Our preliminary work and that of others inTaiwan and United States has demonstrated satisfactoryinternal consistency with the Chinese version of the LFS-SF(C-LFS-SF) (Lee et al., 2005; Tang et al., 2010; Tsai et al.,2012). However, validity of the C-LFS-SF has not beenreported. If the psychometric properties of the C-LFS-SFcan be fully documented, international research collabora-tions and Chinese speaking health care professionals, aswell as pregnant and postpartum women worldwide, canbenefit from this very efficient and practical instrument.

The purposes of this study were to (a) report thetranslation and back translation of the C-LFS-SF, and (b)examine the reliability, validity and factor structure of theC-LFS-SF in Taiwanese pregnant and postpartum women.

2. Methods

2.1. Design

A two-phase process was implemented in developingand evaluating the C-LFS-SF. First, the scale was translatedinto Mandarin Chinese and the content validity wasexamined, and second, testing was conducted to establishthe reliability and validity.

2.2. Phase 1: Translation, back-translation and evaluation of

content validity

The translation theory (Chau, 1983; Kraszewski, 1998)and the recommended procedures for cross-culturalresearch (Brislin, 1970; Brislin et al., 1973) were usedto forward-translate the LFS-SF into Mandarin Chinese andthen back-translate using the 3-point criteria (describedlater) developed by Flaherty and colleagues (1988) to testthe semantic and content equivalence of the original andtranslated version (Flaherty et al., 1988).

Two Chinese-American freelance translators first trans-lated the LFS-SF separately and then cross-examined andrevised it for the C-LFS-SF. This first version was evaluatedby a panel of six committee members (6th and 7th gradeChinese language arts teachers) for content equivalenceand the reading level of 5th to 6th grade. The final productfrom the committee’s review became the C-LFS-SF versiontwo. To ensure the semantic equivalence of the Mandarin

Please cite this article in press as: Tsai, S.-Y., et al., Psychometric evaluation of a Chinese version of the Lee Fatigue Scale-Short Form in women during pregnancy and postpartum. Int. J. Nurs. Stud. (2013), http://dx.doi.org/10.1016/j.ijnurstu.2013.10.023



ChifambacC-Lperbec

foumafreesocthelenme3 =

moentitemwoHowweparuniof fmetranmothecur

2.3.

colpreconBoa201for

struandassfatiHalsleevaltecwit199chaeve

2.4.

woaffiJanweabopar

S.-Y. Tsai et al. / International Journal of Nursing Studies xxx (2013) xxx–xxx 3

G Model


PlShj.i

nese translation, two Chinese-Americans who were notiliar with the original scale, but with biculturalkground, were asked to independently translate theFS-SF version two back to an English version. Eachson’s product was cross-examined and finalized toome the back translation English version.After the back-translation was completed, a panel ofr experts (one Chinese-American neonatologist, onester’s prepared Chinese-American neonatal nurse, onelance translator, and one retired Chinese-American

ial worker) examined the original English version and back translation English version for semantic equiva-ce. They rated each item on a 3-point scale recom-nded by Flaherty and colleagues (1988), with‘‘exactly the same meaning in both versions,’’ 2 = ‘‘al-st the same meaning in both versions,’’ and 1 = ‘‘differ- meaning in each version.’’ It was planned that any

s that were rated as ‘‘2’’ or lower by all examinersuld be re-evaluated and revised by a bilingual person.

ever, all items were rated as ‘‘3’’ and thus no changesre made. A pretest was conducted with two Chineseents who had a child hospitalized in the intensive caret, with no modification required. Additionally, a panelour bilingual Chinese immigrants then examined theaning of the items among the original LFS-SF, the backslation LFS-SF, and the C-LFS-SF version two. No

dification was required, and the final product became C-LFS-SF version three, which was adopted for therent study.

Phase II: Psychometric testing

A descriptive study of sleep in women with two datalection time points, one during the third trimester ofgnancy and a follow up three months postpartum, wasducted and approved by the Institutional Reviewrd of National Taiwan University Hospital (Tsai et al.,1, 2012). The reliability of the C-LFS-SF was assessedinternal consistency and test–retest reliability. Con-ct validity was evaluated by examining convergent

discriminative validity. Convergent validity wasessed by hypothesis testing of the relations betweengue and sleep quantity and efficiency (Gay et al., 2004;l et al., 2009; Lee and Zaffke, 1999) or daytimepiness (Lee et al., 1991), respectively. Discriminative

idity was examined by means of known group-hniques comparing the C-LFS-SF scores among womenh lower and higher levels of depression (Reeves et al.,1). The instrument’s sensitivity to intra-individualnge was also investigated by comparing morning andning fatigue levels.

Participants

A convenience sample of 124 healthy nulliparousmen attending the prenatal clinics at a university-liated hospital in Taipei, Taiwan was recruited betweenuary 2010 and June 2011. Women attending the clinicsre approached by a research assistant and informedut the study. If the women expressed an interest inticipating in the study, a screening interview was

performed and if eligible, the subject completed a consentform. Inclusion criteria were as follows: (a) �18 years ofage; (b) �29 weeks’ gestation; (c) expecting the first child;(d) experiencing no health complications; (e) able to speak,read, understand, and write Mandarin Chinese; and (f) nothaving past, current, or familial psychiatric illness or sleepdisorders. Women who had multiple gestations, workednight shifts, or took medications known to affect sleeppatterns were excluded. Of the 124 women who partici-pated in the study during pregnancy, 15 women did notparticipate in the postpartum study because of pretermlabor (n = 1), family relocation (n = 1), partner refusal(n = 1), loss of interest (n = 4), or time constraints (n = 8),resulting in a postpartum sample size of 109. Our samplesize was sufficient for psychometric testing of a 7-iteminstrument because guidelines suggest at least 10–15subjects per item for factor analysis (Pett et al., 2003).

2.5. Instruments

The instruments used in this study included: (1)actigraphy to monitor sleep patterns for 7 days, (2)fatigue-sleep diary (including the C-LFS-SF) to be keptfor 7 days in parallel to the actigraphy, and (3) self-administered questionnaires on daytime sleepiness anddepressive symptoms to be filled after the 7-day period ofactigraphy and diary.

Actigraphy. Pregnant and postpartum women wore anactigraphy monitor (Actiwatch-2, Philips-Respironics,Bend, OR, USA) on their non-dominant wrist for 7consecutive days and nights to objectively assess theirnighttime sleep quantity and quality. The actigraphymonitor is a wireless ambulatory device similar to awristwatch that records the degree of motion using anaccelerometer sensitive to body movements. Detectedphysical motion was stored as activity counts in 30-sepochs and converted to sleep parameters using Actiware-Sleep version 5.52 analysis software (Philips-Respironics,Bend, OR, USA). Actigraphy provides reliable and validobjective assessments of sleep when compared to poly-somnography with an 88% agreement rate between thetwo methods for sleep–wake identification (Cole et al.,1992). The sleep parameters derived from the actigraphyincluded: (1) sleep quantity as measured by total sleeptime at night, and (2) sleep quality as measured by sleepefficiency, defined as the ratio of total sleep time at night tototal time in bed multiplied by 100. Women were asked todocument their bedtimes and rise times in the daily diaryto facilitate the interpretation and analysis of the actigraphand C-LFS-SF data. Actigraphy data were analyzed by atrained, experienced research assistant for accuracy of thesleep parameters.

Chinese version of the 7-item Lee Fatigue Scale Short Form

(C-LFS-SF). While wearing the actigraphy monitor, preg-nant and postpartum women rated their level of fatigueeach morning, midday, afternoon, and evening for sevenconsecutive days using the C-LFS-SF included in a 7-daydaily diary (Lee and Gay, 2004; Lee et al., 1991). The itemsof the LFS are a word or a phrase based on the fatigueliterature and taking into account the content analyses ofpersonal interviews with patients complaining of fatigue


jnurstu.2013.10.023




G Model


(Lee et al., 1991). The LFS has excellent internalconsistency of .87–.97 (Elek et al., 1997; Gay et al.,2004; Lee et al., 1991). The validity of the LFS is supportedby satisfactory correlation with both the StanfordSleepiness Scale and Profile of Mood Status (Lee et al.,1991). The seven items from the C-LFS-SF were summedand averaged to provide a mean fatigue score, with higherscores indicating higher levels of fatigue. Because ourpreliminary analysis showed that maternal fatigue isprogressive in nature across the day (Tsai et al., 2012), onlythe morning and evening fatigue ratings in the 7-day dailydiary were used for psychometric assessments of the C-LFS-SF.

Epworth Sleepiness Scale (ESS). Pregnant and postpartumwomen completed the ESS (Johns, 1991), a self-adminis-tered questionnaire to assess their tendency to fall asleepin eight different daily-life situations in recent times usingan ordinal scale of 0 to 3 (never, slight, moderate, andhigh). The total ESS score ranging from 0 to 24 provides aglobal level of daytime sleepiness with a higher scoreindicating greater sleepiness. A total score of 8 or highersuggests clinically relevant daytime sleepiness (Rosenthaland Dolan, 2008). The ESS has high internal consistency(Cronbach’s alpha = 0.88), 5-month test–retest reliability,accuracy, and validity in adult men and women (Johns,1992, 1994, 2000, 2002). The Chinese version of the ESSwas shown to be valid and sensitive to clinical change(Chen et al., 2002). Cronbach’s alpha was .77 for pregnantwomen and .83 for postpartum women in the currentstudy.

Center for Epidemiologic Studies-Depression Scale (CES-

D). Pregnant and postpartum women completed the 20-item CES-D (Radloff, 1977) to assess maternal depressivesymptomatology within the last week. Each item of theCES-D describing how an individual felt or behaved wasrated on a four-point scale based on frequency ofoccurrence, ranging from 0 (rarely or none of the time)to 3 (most or all of the time). Scores range from 0 to 60,with higher scores indicating greater depressive symp-toms. A total score of 16 or more suggests clinicallydepressive symptoms. The CES-D has demonstrated goodtest–retest reliability and excellent construct validity, andhas been used in populations of childbearing women (Leeet al., 2000; Radloff, 1977). The Chinese version of the CES-

D is widely used among Taiwanese populations in theclinic and in the community and shows good validity andreliability (Chien and Cheng, 1985; Chou et al., 2003; Tsaiet al., 2011). Cronbach’s alpha was .85 for pregnant womenand .93 for postpartum women in the current study.

2.6. Data collection procedures

Table 1 summarizes the instruments used at twomeasuring points, including associated durations of datarecording. Two follow-up phone calls were made to thewomen during the 7-day study to answer questions andenhance procedure adherence. After seven days ofrecordings, women met with the research assistant toreturn the monitoring device and diaries and completedmood and daytime sleepiness questionnaires. Each womanwho participated in the study during pregnancy wascontacted two weeks after her due date to schedule thepostpartum follow-up study at her convenience withinthree months after birth. Postpartum women completedidentical monitoring procedures and questionnaires asduring pregnancy.

2.7. Data analysis

Data were analyzed using the Statistical Package for theSocial Sciences for Windows, version 17.0 (SPSS Inc.,Chicago, IL). Continuous variables were summarized usingmeans and standard deviations. Categorical variables weresummarized using frequency and percentages. The reli-ability and validity of the C-LFS-SF were evaluatedseparately for pregnant and postpartum women. Reliabili-ty was examined by computing the Cronbach’s alphacoefficient, performing item analysis, and calculating theintra-class correlation (ICC) among the 7 days of morningand evening fatigue ratings. The ICC coefficients wereestimated using a two-way random effect model. ACronbach’s alpha � .80 was considered acceptable for theinstrument’s internal consistency reliability, average inter-item correlations of .30–.70 were considered appropriateto measure its target construct, an ICC > .70 indicated goodtest–retest reliability and stability (Robinson et al., 1991;Scientific Advisory Committee of the Medical OutcomesTrust, 2002).

Table 1

Data collection schedule.

Measuring point #1

Third trimester of pregnancy

Measuring point #2

3 month-period postpartum

7-day period Subsequent survey 7-day period Subsequent survey

1 2 3 4 5 6 7 1 2 3 4 5 6 7

Actigraphy: total sleep time at night and

sleep efficiency

Fatigue-sleep diary: bedtimes, rise times,

and C-LFS-SF (morning, midday,

afternoon, evening)

Questionnaires: Epworth Sleepiness Scale,

Center for Epidemiologic Studies-

Depression Scale




nenstrudelof sbe

detploto <

usethenigmoSleewitcovandcovautbecindmemecortionbet(GaConingSF

(Ins

diffclinwithiglevet

sencha

3. R

3.1.

yea

Tab

Reli

Pr

Po

Note


G Model


PlShj.i

An exploratory factor analysis using principal compo-t analyses was performed to examine the factorialcture of the C-LFS-SF. Direct oblimin rotation with

ta set as zero was used because it satisfies the principlesimple structure and allows for the factors extracted tocorrelated. The number of factors extracted was

ermined based on an eigenvalue > 1.0 and the screet (Pett et al., 2003). Factor loadings were suppressed

.30 in absolute value.To evaluate convergent validity of the C-LFS-SF, wed generalized estimating equations (GEE) to examine

associations between morning fatigue and antecedentht sleep quality and quantity in pregnant (model 1 anddel 2) and postpartum women (model 3 and model 4).p parameters were considered the predictor variable,

h maternal age and gestational age at assessment asariates controlled in model 1 and model 2. Maternal age

postnatal weeks at assessment were considered theariates controlled in model 3 and model 4. GEE with anoregressive working correlation matrix was usedause sleep and fatigue data were collected within eachividual consecutively for seven days. Such statisticalthods take into account the dependency of repeatedasures within subjects and assume a decrease inrelation coefficients when the period between observa-s is greater. We expected an inverse association

ween fatigue and sleep according to previous reportsy et al., 2004; Lee and Zaffke, 1999; Tsai et al., 2012).vergent validity was additionally assessed by comput-

the Pearson correlation coefficient between the C-LFS-and the ESS, where a positive result was expectedana et al., 2011; Lee et al., 1991).

Discriminative validity was assessed by comparingerences in C-LFS-SF scores between women withically significant depressive symptoms and thosehout using unpaired t-tests, where individuals withh depressive symptoms were expected to have greaterels of fatigue (Lee et al., 1994; Reeves et al., 1991; Tsaial., 2012). Paired t-tests were used to evaluate thesitivity of the C-LFS-SF in detecting within-subjectnges in fatigue before and after a night’s sleep.

esults

Participant characteristics

A total of 124 women with a mean age of 31.76 � 4.36rs participated in the study during the third trimester of

pregnancy (Table 2). The majority of the women weremarried, had a college education, and were employed duringthe prenatal assessment conducted at 32.56 � 2.69 weeks’gestation. A total of 109 (87.9%) women continued partici-pating in the postpartum study. The postpartum assessmentwas conducted at 49.93 � 17.57 (range 14–84) days afterdelivery. Nine women who were employed during pregnancyhad returned to work by the postpartum assessment. Ninety-six (88%) of the postpartum women identified themselves asthe primary caregiver for their infant. Fifty-nine (47.6%)pregnant women and 54 (49.5%) postpartum women hadtotal ESS scores � 8 indicating clinically relevant daytimesleepiness. Twenty-six (21%) pregnant women and 36 (33%)postpartum women scored above the CES-D cutoff forpossible clinical depression. Notably the mean sleep efficien-cy for pregnant and postpartum women was lower than 85%,suggesting poor sleep quality.

3.2. Reliability

Cronbach’s alpha ranged from .97 to .99, suggestingsatisfactory internal consistency of the C-LFS-SF in thepopulation of pregnant and postpartum women. The itemanalysis showed adequate inter-item correlations suggest-ing that the 7 items were appropriate measures of fatigue(Table 3). The ICCs of the scale in the pregnant and

Table 2

Sample characteristics.

Characteristic Pregnant women

(N = 124)

Postpartum

women (N = 109)

Age (years) 31.76 � 4.36 31.48 � 4.32

Married 121 (97.6) 106 (97.2)

Education

High school 8 (6.5) 7 (6.4)

College 85 (68.5) 77 (70.6)

Graduate school 31 (25.0) 25 (22.9)

Employed 92 (74.2) 81 (74.3)

ESS scores (0–24) 7.17 � 4.08 7.55 � 4.95

CES-D scores (0–60) 10.20 � 7.02 11.87 � 9.69

Actigraphic total

sleep time (min)

390.10 � 51.42 338.34 � 68.82

Actigraphic sleep

efficiency (%)

80.52 � 6.02 74.79 � 8.61

Notes: Data are presented as mean � standard deviation scores or n (%);

ESS, Epworth Sleepiness Scale; CES-D, Center for Epidemiologic Studies

Depression Scale. The 7 day’s sleep was averaged to obtain the average

actigraphic total sleep time and the average sleep efficiency scores.

le 3

ability and item analysis.

Cronbach’s a ICC Item means Item variances Inter-item correlations

Mean (Range) Mean (Range) Mean (Range)

egnancy (N = 124)

Morning fatigue .98 .87 3.37 (2.89–3.63) 3.27 (2.79–4.13) .92 (.86–.97)

Evening fatigue .98 .89 5.91 (5.55–6.65) 3.35 (2.93–3.89) .89 (.82–.95)

stpartum (N = 109)

Morning fatigue .99 .95 5.02 (4.63–5.52) 5.26 (4.53–6.13) .94 (.88–.97)

Evening fatigue .97 .94 6.52 (6.01–7.37) 3.97 (2.89–5.35) .86 (.71–.97)

: ICC, intra-class correlation.


jnurstu.2013.10.023




G Model


postpartum women were all higher than .87, indicating theadequate stability of the C-LFS-SF over a 7-day period.

3.3. Construct validity

Factorial validity. Exploratory factor analysis showedthat 88–94% of the total variance was explained by theone-factor fatigue model, with the factor loadings rangingfrom .91 to .98 for pregnant women and .89 to .99 forpostpartum women, respectively (Table 4). The scree plotconfirmed a single factor solution. Bartlett’s test resultswere all statistically significant (p < .05), indicatingadequacy of data for factor analysis. The sample size forfactor analysis was sufficient, supported by the Kaiser–Meyer–Olkin measures of sampling adequacy all >.88.

Convergent validity. Morning fatigue as measured by theC-LFS-SF was significantly negatively associated withactigraphic total nighttime sleep (p < .001) and sleepefficiency (p < .05) of the previous night in pregnant and

postpartum women, suggesting that women with insuffi-cient sleep and poor sleep have more intense fatigue thenext morning (Table 5). Results remained statisticallysignificant even after controlling for maternal age andgestational age at assessment for pregnant women andpostnatal weeks at assessment for postpartum women.

Because the ICC for fatigue was high for both pregnantand postpartum women, the mean values for the 7 dayswere also used to evaluate the convergent validity. Bothhigher morning and evening fatigue as measured by the C-LFS-SF during pregnancy and postpartum periods weresignificantly correlated with higher ESS scores indicatingmore daytime sleepiness supporting convergent validity(Table 6).

3.4. Discriminative validity

Statistically significant differences in the morningfatigue levels as measured by the C-LFS-SF were foundbetween the pregnant and postpartum women withclinically significant depressive symptoms (CES-D � 16)and those without (CES-D < 16) (Table 6). Evening fatiguedid not significantly differ among women with higher andlower depression levels.

3.5. Sensitivity to change

Morning and evening fatigue severity ratings forpregnant women were 3.37 � 1.74 and 5.91 � 1.74, respec-tively. Morning and evening fatigue severity ratings forpostpartum women were 5.20 � 2.13 and 6.80 � 1.82,respectively. Paired t tests showed significant differences(p < .001) between morning and evening fatigue levels asmeasured by the C-LFS-SF in both pregnant and postpartumwomen, suggesting that the scale is sensitive in detectingwithin-subject changes in fatigue before and after a night’ssleep (Table 6). Postpartum women also reported higher

Table 4

Factor analysis for the Chinese version of the Fatigue Scale.

Items Pregnancy (N = 124) Postpartum

(N = 109)

Morning Evening Morning Evening

1. Tired .97 .96 .97 .95

2. Sleepy .98 .96 .98 .95

3. Drowsy .97 .96 .98 .94

4. Fatigued .98 .97 .99 .98

5. Worn out .96 .96 .98 .94

6. Concentrating .95 .95 .97 .93

7. Desire to lie down .94 .91 .95 .89

Eigenvalues 6.52 5.56 6.64 6.21

Total variance (%) 93.18 92.73 94.89 88.80

KMO .91 .88 .89 .88

Note: KMO, Kaiser–Meyer–Olkin measures of sampling adequacy.

Table 6

Correlation of fatigue levels with ESS scores and comparison of fatigue levels in women with high and low depressive symptoms as well as comparison of

morning and evening fatigue levels.

ESS CES-D � 16 CES-D < 16 Diurnal fatigue

r p-value Mean (SD) Mean (SD) t-test p-value Mean (SD) Paired t-test p-value

Pregnancy (N = 124) (n = 26) (n = 98)

Morning fatigue .30 <.01 4.41 (1.32) 3.10 (1.74) <.01 3.37 (1.74) <.001

Evening fatigue .22 .01 6.46 (1.44) 5.77 (1.79) .08 5.91 (1.74)

Postpartum (N = 109) (n = 36) (n = 73)

Morning fatigue .22 .01 5.71 (1.81) 4.63 (2.26) .04 5.20 (2.13) <.001

Evening fatigue .27 <.01 6.85 (1.99) 6.70 (1.90) .77 6.80 (1.82)

Table 5

Temporal associations between morning fatigue levels and previous night sleep.

Model Morning fatigue Independent variable b p-Value

1 Pregnancy (N = 124) Actigraphic previous night total sleep time �.006 <.001

2 Pregnancy (N = 124) Actigraphic previous night sleep efficiency �.029 .009

3 Postpartum (N = 109) Actigraphic previous night total sleep time �.008 <.001

4 Postpartum (N = 109) Actigraphic previous night sleep efficiency �.019 .022

Notes: Models 1 and 2 were adjusted for maternal age and gestational age at assessment. Models 3 and 4 were adjusted for maternal age and postnatal

weeks at assessment. All predictors were included as a linear variable and were modeled simultaneously.

Notes: ESS, Epworth Sleepiness Scale; CES-D, Center for Epidemiologic Studies Depression Scale.




fatipre

4. D

reliandintewhhig200itemwitdanstuhumadeexcwitpotstruthaintetherepcouDal

andOulenChifor

tranitemculequrate

thetimactmoaccallyposgenfatiactworepwitworepwitcrimconass(InsHu


G Model


PlShj.i

gue severity compared to their third trimester ofgnancy.

iscussion

We demonstrated that the C-LFS-SF is a valid andable instrument to assess fatigue in healthy pregnant

postpartum women. The C-LFS-SF had satisfactoryrnal consistency reliability and stability, all >.85,

ich is in accordance with previous studies reportingh internal consistency of this instrument (Gay et al.,4; Lee and Lee, 2007; Tsai et al., 2012). All 7 fatigues were moderately to strongly related to each other,

h few inter-item correlations > .90 implying a redun-cy issue, i.e., there may be overlapping items. A recent

dy using a four-item version of the LFS in adults withan immunodeficiency virus infection demonstrated

quate internal consistency with Cronbach’s alphaeeding .85 (Lerdal et al., 2011). Our results, alongh this recent study, suggest that the C-LFS-SF has theential to be even shorter than 7 items. The single-factorcture, combined with strong factor loadings, suggest

t the C-LFS-SF is a unidimensional measure asnded. The estimates found for several measures of

construct validity of the C-LFS-SF resembled thoseorted in previous studies conducted in Westernntries (Lee and Lee, 2007; Lee et al., 1991; Troy andgas-Pelish, 2003).Our results indicate that the C-LFS-SF is linguistically

culturally relevant to Chinese child-bearing women.r theory-based translation procedures ensure equiva-ce and prevent bias while using the C-LFS-SF innese women. Lee et al. (1991) used simple phrasesthe original LFS which decrease error during theslation process. As we translated the instrument, no

required great modification to fit into the Chinesetural context. The content validity and overallivalence of the Chinese version of the scale wered as excellent.

We used multiple methods to establish the validity of C-LFS-SF. We found that women’s antecedent night-e sleep quality and quantity as measured objectively byigraphy had a significant inverse association with therning fatigue of the subsequent day, demonstrating aneptable convergent validity of the C-LFS-SF. Addition-, morning and evening fatigue levels were significantlyitively correlated with the ESS, supporting the conver-t validity of the C-LFS-SF and the association betweengue severity and impact of daytime sleepiness onivities of daily living. Our study also showed thatmen with clinically significant depressive symptomsorted higher levels of morning fatigue than thosehout clinically significant depressive symptoms, andmen with clinically significant depressive symptomsorted similar levels of evening fatigue to womenhout these symptoms, demonstrating acceptable dis-

inative validity of the C-LFS-SF. These results weresistent with previous findings that maternal fatigue isociated with nighttime sleep quality and quantityana et al., 2011; Lee and Zaffke, 1999; Rychnovsky and

nter, 2009), daytime sleepiness (Lee et al., 1991), and

depressive symptoms (Dennis and Ross, 2005; Reeveset al., 1991; Tsai et al., 2012).

We also provided evidence that the C-LFS-SF can detectindividual changes in fatigue levels over time. Consistentwith previous studies (Gay et al., 2004; Lee and Zaffke,1999), we found that women reported higher levels offatigue during postpartum than pregnancy. The meanscore of the fatigue scale for women in our study wassimilar to that reported for pregnant and postpartumwomen in Western societies (Gay et al., 2004). The abilityto detect changes also suggests that the C-LFS-SF can beused to compare fatigue perceptions in women acrossdifferent populations. Others also found within-subjectchanges in which fatigue increased as pregnancy pro-gressed when fatigue was measured using the LFS (Elek etal., 1997).

The C-LFS-SF allows health care providers to assessfatigue in a rapid and quantitative manner, making itpossible when fatigue must be assessed repeatedly overtime. Fatigue has overwhelming implications for pregnan-cy, birth, parenting, and maternal health outcomes (Chienand Ko, 2004; Dennis and Ross, 2005; Reeves et al., 1991).Complications arising from maternal fatigue are potential-ly preventable if attention is paid to carefully assessing thissymptom with validated assessment tools. Althoughevidence-based treatment strategies do not exist forprenatal fatigue, studies with postpartum women haveshown exercise and telephone counseling to reduce fatiguelevels (Dritsa et al., 2008; Thome and Alder, 1999).

To our knowledge, we are the first to perform a fullpsychometric evaluation of the C-LFS-SF for the populationof pregnant and postpartum women. The strength of thestudy was that we had a sufficient sample size to performfactor analysis and to characterize changes in fatigue levelsfrom pregnancy to postpartum. We also used an objectivesleep measurement to establish the convergent validity ofthe C-LFS-SF. However, the present study has somelimitations. We chose the LFS-SF because of its brevityand frequent use for assessment fatigue patterns inchildbearing women in Western countries. The LFS-SFenables assessment of fatigue within 10–20 s, providing anefficient tool for early detection and intervention foroptimal management of fatigue during pregnancy andpostpartum period. The participants were a conveniencesample of healthy nulliparous women recruited from oneinstitution in university clinic settings. Different resultsmay be obtained if the sample included multiparouswomen who may have different experiences of fatigue dueto prior experiences of pregnancy or due to having morethan one child at home. Although our sample size issufficient for factor analysis, several of the statisticallysignificant results found in our study had rather smalleffect estimates. Due to the large number of statistical testsperformed in this study, our results needs to be interpretedwith caution.

In conclusion, we provided evidence that the C-LFS-SFhas satisfactory internal consistency, and content validityas well as construct validity. Given these adequatepsychometric properties, its brevity, and ease of adminis-tration, it is a valuable and promising tool for assessingfatigue during routine prenatal and postpartum care.


jnurstu.2013.10.023




G Model


Because the C-LFS-SF is sensitive to change, new knowl-edge can be obtained from using this instrument forresearch and clinical questions. Potential uses of the C-LFS-SF also include assessments of fatigue before and afterinterventions as an indication of the effectiveness of atherapy or coping strategy. The C-LFS-SF needs to befurther tested in a more diverse population of pregnant andpostpartum women, like women with high risk pregnan-cies or with medical conditions.

Conflict of Interest

None declared.

Funding

This work was funded by National Science Council,Taiwan (NSC 99-2314-B-002-002).

Ethical Approval

National Taiwan University Hospital Institutional Re-view Board (200911041R).

Acknowledgements

We thank all the women who kindly agreed to take partand share their fatigue experiences. We also thank theresearch assistants, Shan-Shan Yang, Wen-Hsun Chang,and Jin-Hui Yang who assisted in data collection.

Appendix A. Supplementary data

Supplementary data associated with this article can befound, in the online version, at http://dx.doi.org/10.1016/j.ijnurstu.2013.10.023.

References

Aaronson, L.S., Teel, C.S., Cassmeyer, V., Neuberger, G.B., Pallikkathayil, L.,Pierce, J., Press, A.N., Williams, P.D., Wingate, A., 1999. Defining andmeasuring fatigue. Image – The Journal of Nursing Scholarship 31 (1)45–50.

Brislin, R.W., 1970. Back-translation for cross-cultural research. Journal ofCross-Cultural Psychology 1, 185–216.

Brislin, R.W., Lonner, W.J., Thorndike, R.M., 1973. Cross-cultural ResearchMethods. John Wiley & Sons, New York.

Chau, S.S., 1983. How to translate ‘‘This is a red rose’’? In: Wilss, W., Thome,G. (Eds.), Translation Theory and its Implementation in the Teaching ofTranslating and Interpreting. GNV, Germany, pp. 124–134.

Chen, N.H., Johns, M.W., Li, H.Y., Chu, C.C., Liang, S.C., Shu, Y.H., Chuang,M.L., Wang, P.C., 2002. Validation of a Chinese version of the Epworthsleepiness scale. Quality of Life Research 11 (8) 817–821.

Chien, C.P., Cheng, T.A., 1985. Depression in Taiwan: epidemiologicalsurvey utilizing CES-D. Seishin Shinkeigaku Zasshi 87 (5) 335–338.

Chien, L.Y., Ko, Y.L., 2004. Fatigue during pregnancy predicts caesareandeliveries. Journal of Advanced Nursing 45 (5) 487–494.

Chou, F.H., Lin, L.L., Cooney, A.T., Walker, L.O., Riggs, M.W., 2003. Psycho-social factors related to nausea, vomiting, and fatigue in early preg-nancy. Journal of Nursing Scholarship 35 (2) 119–125.

Cole, R.J., Kripke, D.F., Gruen, W., Mullaney, D.J., Gillin, J.C., 1992. Auto-matic sleep/wake identification from wrist activity. Sleep 15 (5)461–469.

Corwin, E.J., Brownstead, J., Barton, N., Heckard, S., Morin, K., 2005. Theimpact of fatigue on the development of postpartum depression.Journal of Obstetric, Gynecologic, and Neonatal Nursing 34 (5)577–586.

Corwin, E.J., Arbour, M., 2007. Postpartum fatigue and evidence-basedinterventions. MCN: The American Journal of Maternal/Child Nursing32 (4) 215–220.

Dennis, C.L., Ross, L., 2005. Relationships among infant sleep patterns,maternal fatigue, and development of depressive symptomatology.Birth 32 (3) 187–193.

Dodd, M., Janson, S., Facione, N., Faucett, J., Froelicher, E.S., Humphreys, J.,Lee, K., Miaskowski, C., Puntillo, K., Rankin, S., Taylor, D., 2001.Advancing the science of symptom management. Journal of AdvancedNursing 33 (5) 668–676.

Dritsa, M., Da Costa, D., Dupuis, G., Lowensteyn, I., Khalife, S., 2008. Effectsof a home-based exercise intervention on fatigue in postpartumdepressed women: results of a randomized controlled trial. Annalsof Behavioral Medicine 35 (2) 179–187.

Elek, S.M., Hudson, D.B., Fleck, M.O., 1997. Expectant parents’ experiencewith fatigue and sleep during pregnancy. Birth 24 (1) 49–54.

Fairbrother, N., Hutton, E.K., Stoll, K., Hall, W., Kluka, S., 2008. Psycho-metric evaluation of the Multidimensional Assessment of Fatiguescale for use with pregnant and postpartum women. PsychologicalAssessment 20 (2) 150–158.

Flaherty, J.A., Gaviria, F.M., Pathak, D., Mitchell, T., Wintrob, R., Richman,J.A., Birz, S., 1988. Developing instruments for cross-cultural psychi-atric research. The Journal of Nervous and Mental Disease 176 (5)257–263.

Gay, C.L., Lee, K.A., Lee, S.Y., 2004. Sleep patterns and fatigue in newmothers and fathers. Biological Research for Nursing 5 (4) 311–318.

Groer, M., Davis, M., Casey, K., Short, B., Smith, K., Groer, S., 2005.Neuroendocrine and immune relationships in postpartum fatigue.MCN: The American Journal of Maternal/Child Nursing 30 (2) 133–138.

Hall, W.A., Hauck, Y.L., Carty, E.M., Hutton, E.K., Fenwick, J., Stoll, K., 2009.Childbirth fear, anxiety, fatigue, and sleep deprivation in pregnantwomen. Journal of Obstetric, Gynecologic, and Neonatal Nursing 38(5) 567–576.

Insana, S.P., Stacom, E.E., Montgomery-Downs, H.E., 2011. Actual andperceived sleep: associations with daytime functioning among post-partum women. Physiology & Behavior 102 (2) 234–238.

Johns, M.W., 1991. A new method for measuring daytime sleepiness: theEpworth sleepiness scale. Sleep 14 (6) 540–545.

Johns, M.W., 1992. Reliability and factor analysis of the Epworth sleepi-ness scale. Sleep 15 (4) 376–381.

Johns, M.W., 2000. Sensitivity and specificity of the Multiple SleepLatency Test (MSLT), the Maintenance of Wakefulness Test and theEpworth sleepiness scale: failure of the MSLT as a gold standard.Journal of Sleep Research 9 (1) 5–11.

Johns, M.W., 2002. Sleep propensity varies with behaviour and thesituation in which it is measured: the concept of somnificity. Journalof Sleep Research 11 (1) 61–67.

Johns, M.W., 1994. Sleepiness in different situations measured by theEpworth sleepiness scale. Sleep 17 (8) 703–710.

Kraszewski, C.S., 1998. Four Translation Strategies Determined by theParticular Needs of the Receptor. The Edwin Mellen Press, New York.

Lee, K.A., Gay, C.L., 2004. Sleep in late pregnancy predicts length of laborand type of delivery. American Journal of Obstetrics and Gynecology191 (6) 2041–2046.

Lee, K.A., Hicks, G., Nino-Murcia, G., 1991. Validity and reliability of a scaleto assess fatigue. Psychiatry Research 36 (3) 291–298.

Lee, K.A., Lentz, M.J., Taylor, D.L., Mitchell, E.S., Woods, N.F., 1994. Fatigueas a response to environmental demands in women’s lives. Image –The Journal of Nursing Scholarship 26 (2) 149–154.

Lee, K.A., McEnany, G., Zaffke, M.E., 2000. REM sleep and mood state inchildbearing women: sleepy or weepy? Sleep 23 (7) 877–885.

Lee, K.A., Zaffke, M.E., 1999. Longitudinal changes in fatigue and energyduring pregnancy and the postpartum period. Journal of Obstetric,Gynecologic, and Neonatal Nursing 28 (2) 183–191.

Lee, S.Y., Lee, K.A., 2007. Early postpartum sleep and fatigue for mothersafter cesarean delivery compared with vaginal delivery: an explor-atory study. The Journal of Perinatal & Neonatal Nursing 21 (2) 109–113.

Lee, S.Y., Lee, K.A., Rankin, S.H., Alkon, A., Weiss, S.J., 2005. Acculturationand stress in Chinese-American parents of infants cared for in theintensive care unit. Advances in Neonatal Care 5 (6) 315–328.

Lerdal, A., Gay, C.L., Aouizerat, B.E., Portillo, C.J., Lee, K.A., 2011. Patterns ofmorning and evening fatigue among adults with HIV/AIDS. Journal ofClinical Nursing 20 (15–16) 2204–2216.

Maloni, J.A., Park, S., 2005. Postpartum symptoms after antepartum bedrest. Journal of Obstetric, Gynecologic, and Neonatal Nursing 34 (2)163–171.

McGovern, P., Dagher, R.K., Rice, H.R., Gjerdingen, D., Dowd, B., Ukestad,L.K., Lundberg, U., 2011. A longitudinal analysis of total workload and




http://refhub.elsevier.com/S0020-7489(13)00331-3/sbref0005






































































































Park

Pett

Pug

Rad

Ree

Rob

Rod

Rose


G Model


PlShj.i

women’s health after childbirth. Journal of Occupational and Envi-ronmental Medicine 53 (5) 497–505.s, P.L., Lenz, E.R., Milligan, R.A., Han, H.R., 1999. What happens whenfatigue lingers for 18 months after delivery? Journal of Obstetric,Gynecologic, and Neonatal Nursing 28 (1) 87–93., M.A., Lackey, N.R., Sullivan, J.J., 2003. Making Sense of Factor Analy-sis. Sage, Thousand Oaks.h, L.C., Milligan, R., 1993. A framework for the study of childbearingfatigue. ANS: Advances in Nursing Science 15 (4) 60–70.loff, L.S., 1977. The CES-D scale: a self-report depression scale forresearch in the general population. Applied Psychological Measure-ment 1, 385–401.ves, N., Potempa, K., Gallo, A., 1991. Fatigue in early pregnancy.An exploratory study. Journal of Nurse-Midwifery 36 (5) 303–309.inson, J.P., Shaver, P.R., Wrightsman, L.S., 1991. Criteria for scaleselection and evaluation. In: Robinson, J.P., Shaver, P.R., Wrightsman,L.S. (Eds.), Measures of Personality and Social Psychological Attitudes.Academic Press, San Diego, pp. 1–15.riguez, A., Bohlin, G., Lindmark, G., 2001. Symptoms across pregnancyin relation to psychosocial and biomedical factors. Acta Obstetricia etGynecologica Scandinavica 80 (3) 213–223.nthal, L.D., Dolan, D.C., 2008. The Epworth sleepiness scale in the

identification of obstructive sleep apnea. The Journal of Nervous andMental Disease 196 (5) 429–431.

Runquist, J., 2007. Persevering through postpartum fatigue. Journal ofObstetric, Gynecologic, and Neonatal Nursing 36 (1) 28–37.

Rychnovsky, J., Hunter, L.P., 2009. The relationship between sleep char-acteristics and fatigue in healthy postpartum women. Women’sHealth Issues 19 (1) 38–44.

Scientific Advisory Committee of the Medical Outcomes Trust, 2002.Assessing health status and quality-of-life instruments: attributesand review criteria. Quality of Life Research 11 (3) 193–205.

Tang, W.R., Yu, C.Y., Yeh, S.J., 2010. Fatigue and its related factors in patientswith chronic heart failure. Journal of Clinical Nursing 19 (1–2) 69–78.

Taylor, J., Johnson, M., 2010. How women manage fatigue after childbirth.Midwifery 26 (3) 367–375.

Thome, M., Alder, B., 1999. A telephone intervention to reduce fatigue andsymptom distress in mothers with difficult infants in the community.Journal of Advanced Nursing 29 (1) 128–137.

Troy, N.W., Dalgas-Pelish, P., 2003. The effectiveness of a self-care inter-vention for the management of postpartum fatigue. Applied NursingResearch 16 (1) 38–45.

Tsai, S.Y., Kuo, L.T., Lai, Y.H., Lee, C.N., 2011. Factors associated with sleepquality in pregnant women: a prospective observational study. Nurs-ing Research 60 (6) 405–412.

Tsai, S.Y., Lin, J.W., Kuo, L.T., Thomas, K.A., 2012. Daily sleep and fatiguecharacteristics in nulliparous women during the third trimester ofpregnancy. Sleep 35 (2) 257–262.


jnurstu.2013.10.023


















































psychometric evaluation of a chinese version of the lee fatigue scale-short form in women during...

Documents