Diabetes-specific quality of life but not healthstatus is independently associated with glycaemiccontrol among patients with type 2 diabetes: Across-sectional analysis of the ADDITION-Europetrial cohort

Laura Kuznetsov a, Simon J. Griffin a, Melanie J. Davies b,Torsten Lauritzen c, Kamlesh Khunti b, Guy E.H.M. Rutten d,Rebecca K. Simmons a,*aMRC Epidemiology Unit, University of Cambridge, Cambridge, United KingdombDiabetes Research Unit, Leicester Diabetes Centre, University of Leicester, Leicester, United KingdomcDepartment of Public Health, Section of General Practice, University of Aarhus, Aarhus, Denmarkd Julius Center, Department of General Practice, University Medical Center Utrecht, Utrecht, Netherlands

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 1 0 4 ( 2 0 1 4 ) 2 8 1 – 2 8 7

a r t i c l e i n f o

Article history:

Received 25 June 2013

Received in revised form

25 November 2013

Accepted 21 December 2013

Available online 15 January 2014

Keywords:

Health status

Diabetes-specific quality of life

ADDQoL

SF-36

HbA1c

Type 2 diabetes

a b s t r a c t

Aims: To examine the association between health status, diabetes-specific quality of life

(QoL) and glycaemic control among individuals with type 2 diabetes.

Methods: 1876 individuals with screen-detected diabetes and a mean age of 66 years under-

went assessment of self-reported health status (SF-36), diabetes-specific QoL (the Audit of

Diabetes Dependent Quality of Life (ADDQoL19)) and glycated haemoglobin (HbA1c) at five

years post-diagnosis in the ADDITION-Europe trial. Multivariable linear regression was used

to quantify the cross-sectional association between health status, diabetes-specific QoL and

HbA1c, adjusting for age, sex, education, alcohol consumption, physical activity, BMI, intake

of any glucose-lowering drugs, and trial arm.

Results: The mean (SD) SF-36 physical and mental health summary scores were 46.2 (10.4)

and 54.6 (8.6), respectively. The median average weighted impact ADDQoL score was �0.32

(IQR �0.89 to �0.06), indicating an overall negative impact of diabetes on QoL. Individuals

who reported a negative impact of diabetes on their QoL had higher HbA1c levels at five

years after diagnosis compared with those who reported a positive or no impact of diabetes

(b-coefficient [95% CI]: b = 0.2 [0.1, 0.3]). Physical and mental health summary SF-36 scores

were not significantly associated with HbA1c in multivariable analysis.

Conclusions: Diabetes-specific QoL but not health status was independently associated with

HbA1c. Practitioners should take account of the complex relationship between diabetes-

specific QoL and glucose, particularly with regard to dietary behaviour. Future research

should attempt to elucidate via which pathways this association might act.

# 2013 Elsevier Ireland Ltd. All rights reserved.

* Corresponding author at: MRC Epidemiology Unit, University of Cambridge, Box 285, Hills Road, Cambridge CB2 0QQ, United Kingdom.Tel.: +44 0 1223 330315; fax: +44 0 1223 330316.

Contents available at ScienceDirect

Diabetes Researchand Clinical Practice

journal homepage: www.elsevier.com/locate/diabres

E-mail address: rebecca.simmons@mrc-epid.cam.ac.uk (R.K. Simmons).

0168-8227/$ – see front matter # 2013 Elsevier Ireland Ltd. All rights reserved.http://dx.doi.org/10.1016/j.diabres.2013.12.029

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 1 0 4 ( 2 0 1 4 ) 2 8 1 – 2 8 7282

1. Introduction

Type 2 diabetes is associated with short-term and long-term

complications which can negatively affect patients’ well-

being, health status and quality of life (QoL) in many ways –

physically, psychologically and socially [1,2]. Impaired

health status may lead to impaired QoL in some but not

all cases [3]. Effective management of diabetes includes

adoption of healthy lifestyle behaviours and an often

complex medication regimen. Both approaches have been

shown to lower and stabilise glucose levels [4,5]. Evidence

suggests that good glycaemic control reduces the risk of

long-term micro- and macro-vascular complications [6,7].

Assessment of health status and diabetes-specific QoL is

important because individuals with diabetes often have to

cope with a variety of advice, recommendations and

medications which may be burdensome. Even if these

interventions improve glucose levels, pharmacological

treatment might not improve health status or diabetes-

related QoL, and may even reduce them. The results of

studies examining the relationship between glucose and QoL

in patients with established diabetes are inconsistent: some

support the association of tighter glycaemic control with

improved QoL [1,8], while others do not [9]. Further, there are

no studies in patients with screen-detected diabetes, in

whom the balance of benefits and harms of treatment of

hyperglycaemia may be different from those further along

the disease trajectory.

Previous research has identified a number of factors

associated with good glycaemic control in people with

diabetes, including higher socioeconomic status [10], treat-

ment with fewer oral glucose-lowering drugs [11], and healthy

lifestyle behaviours [12]. However, to our knowledge, no

studies have considered health status or diabetes-specific QoL

as independent explanatory variables for glucose control. We

hypothesised that better health status and higher QoL scores

would be associated with lower glucose (HbA1c) levels. Using

data from the ADDITION-Europe trial of screen-detected type 2

diabetes patients, we examined the association between

health status, diabetes-specific QoL and HbA1c at five years

post-diagnosis.

2. Materials and methods

The study design and rationale for the ADDITION-Europe

study have been reported [13]. In brief, the Anglo-Danish-

Dutch Study of Intensive Treatment in People with Screen

Detected Diabetes in Primary Care is a pragmatic, cluster-

randomised, parallel-group trial in Denmark, the Netherlands,

and the UK. 343 general practices were randomly assigned to

screening of registered patients aged 40–69 years (50–69 years

in the Netherlands) without known diabetes followed by

routine care of diabetes (n = 176) or screening followed by

intensive treatment of multiple risk factors (n = 167). Screen-

ing was undertaken between April 2001 and December 2006,

and identified 3233 patients with type 2 diabetes, of whom

3057 agreed to participate in the treatment trial. Patients were

excluded if they had an illness with a life expectancy of less

than 12 months, a psychological disorder, were housebound,

pregnant or lactating. The study was approved by the ethics

committee local to each study centre, and all participants

provided informed consent. ADDITION-Europe is registered as

NCT00237549.

3. Measures

ADDITION-Europe health assessments included physiological

and anthropometric measurements, venesection and the

completion of questionnaires. Data collection methods have

been described previously [13]. This analysis includes data

taken exclusively from five-year follow-up. Anthropometric

and clinical measurements were undertaken by trained staff

following standard operating procedures. HbA1c was analysed

by DCCT aligned ion-exchange high-performance liquid

chromatography using Menarini 8160 in the Netherlands,

Bio-Rad Variant II in Leicester, and Tosoh G7 machines in

Denmark and Cambridge. An HbA1c value of <7% (53 mmol/

mol) was defined as good glycaemic control [14]. Socio-

demographic information, lifestyle behaviours (smoking

status, alcohol consumption) and intake of glucose-lowering

drugs, was collected using standardised self-report question-

naires. Physical activity was assessed using the validated

International Physical Activity Questionnaire (IPAQ) and

coded into low, medium and high categories according to

published guidelines [15].

Health status was assessed using the 36-Item Short Form

Health Survey (SF-36) [16] and diabetes-specific QoL using

the Audit of Diabetes Dependent Quality of Life (ADDQoL19)

[17]. We included both a generic and a disease-specific

instrument in our analysis because each method comprises

different information which may be differentially sensitive

to clinically relevant issues [18]. The SF-36 consists of 36

items which form eight subscales: physical functioning,

role-physical, bodily pain, general health, vitality, social

functioning, role-emotional, and mental health ranging

from 0 to 100 with higher scores indicating better health.

From these eight subscales two summary measures [the

Physical Component Summary (PCS) and the Mental

Component Summary (MCS)] can be computed [16]. Cron-

bach’s a indicated satisfactory reliability of all SF-36 health

domain scales (ranging from 0.8 to 0.92). The ADDQoL

measures an individual’s perception of the impact of

diabetes on various aspects of their QoL and the importance

of these aspects [2]. Patients rate the impact of diabetes on

different domains on a scale from �3 (maximum negative

impact) to +1 (maximum positive impact), and then rate the

importance of the domain for their QoL on a scale from 3

(very important) to 0 (not at all important) [19]. The weighted

impact score for each domain is computed by multiplying

the unweighted rating by the importance rating, and ranges

from �9 (maximum negative impact) to 3 (maximum

positive impact) [2]. To calculate an overall Average

Weighted Impact (AWI) score, the weighted ratings of

applicable domains are summed and divided by the number

of applicable domains. A negative AWI score reflects a

negative impact of diabetes on QoL. The Cronbach’s a of the

ADDQoL unweighted items was 0.92.

Table 1 – Participant characteristics including healthstatus and diabetes-specific quality of life in the ADDI-TION-Europe cohort at five years post diagnosis(n = 1876).

Variables Mean (SD) or n (%)

Socio-demographic variables

Age (years) 65.6 (6.9)

Male sex 1145 (61)

Full-time education completed

at �17 years

1105 (58.9)

Caucasiana 1744 (95.6)

Clinical variables

BMI (kg/m2) 31 (5.49)

HbA1c (%) 6.7 (0.92)

HbA1c (mmol/mol) 50

Intake of glucose-lowering drugs 1134 (60.4)

Number of glucose-lowering drugs,

median (range)b1 (0–4)

Self-reported lifestyle behaviours

Current smoker 359 (19.1)

Current alcohol user 1316 (70.1)

Physical activityc

low 431 (23.0)

moderate 643 (34.3)

high 802 (42.8)

Health status & diabetes-specific

quality of life

SF-36 PCS 46.2 (10.37)

SF-36 MCS 54.6 (8.63)

Median ADDQoL (IQR)d �0.32 (�0.89 to �0.06)

Participants reporting a negative impact

of diabetes (ADDQoL AWI score < 0)d1294 (78.6)

Values are means (SD) or % (n) unless stated otherwise. BMI, body-

mass index. HbA1c, glycosylated haemoglobin. SF-36 PCS, Physical

Component Summary measure. SF-36 MCS, Mental Component

Summary measure. ADDQoL AWI, the overall average weighted

impact score of the ADDQoL (range �9 to 3). IQR, interquartile range.a Values are based on 1824 participants.b Values are based on 1715 participants.c Physical activity was assessed using the validated International

Physical Activity Questionnaire (IPAQ).d Values are based on 1646 participants.

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 1 0 4 ( 2 0 1 4 ) 2 8 1 – 2 8 7 283

4. Statistical analyses

Descriptive characteristics were summarised using means,

standard deviations, or frequencies. In order to examine

differences between participants with and without complete

data for analysis, we compared characteristics between

groups using the chi-square test for categorical data, and the

t-test or Mann–Whitney U test for continuous data. Due to

the highly skewed distribution of the ADDQoL AWI score it

was collapsed into a dichotomous variable: patients who

reported a negative impact of diabetes (values �8.83 to

�0.05) (‘1’), and those who reported no or a positive impact of

diabetes (values 0 to 1.27) (‘0’). Univariable linear regression

was used to quantify the crude association between the SF-

36 subscales and summary scores, the ADDQoL, other

covariates and continuous HbA1c (Model 1). Next, multi-

variable linear regression analysis was performed adjusting

for age, sex, age when completed full-time education (<17

years or �17 years of age), alcohol consumption (yes/no),

physical activity (low/moderate/high), body mass index

(BMI), and trial arm (Model 2), and additionally adjusting

for intake of any glucose-lowering drugs (yes/no) (Model 3).

Smoking status (non-smoker/ex-smoker or current smoker)

was not significantly associated with HbA1c in univariable

analysis and therefore was not included in multivariable

analysis. Regression results are presented as unstandar-

dised b-coefficients with their 95% confidence intervals (95%

CI). The residuals of regression models were examined to

ensure that they were approximately normally distributed.

We also examined mean SF-36 and ADDQoL scores among

individuals with HbA1c values above and below 7%

(53 mmol/mol). When stratified for sex and trial group a

similar pattern of results was found. As such, we pooled

both sexes and trial groups and conducted analyses

adjusting for sex and trial group differences. Statistical

significance was set at p < 0.05. Data were analysed using

SPSS for Windows 19.0 (SPSS, Inc., Chicago, IL, USA) and

Stata/SE 12.0 (Stata-Corp, College Station, TX, USA).

5. Results

Of the 2859 patients still alive at 5 years, 2400 (84%) returned

to a clinical research facility for follow-up health assess-

ments and 1876 (66%) had complete data for analysis. There

were no significant differences between participants

included in the analysis and those who were not included

for age, education, smoking, BMI, and HbA1c. However, those

who were not included were more likely to be men

( p < 0.001), to consume alcohol ( p < 0.001), and to be less

physically active ( p = 0.003) compared to those who were

included.

The mean (SD) age of ADDITION-Europe participants at

the five-year health assessment was 66 (7) years and 61%

were male (Table 1). On average, the cohort was obese (mean

BMI 31 (5.49) kg/m2), with good glycaemic control (HbA1c:

6.7% (0.92) (50 mmol/mol), and 60% were on glucose-low-

ering drugs. The median (range) number of glucose-lowering

drugs was 1 (0–4).

6. Distribution of SF-36 and ADDQoL scores

Mean SF-36 PCS and MCS scores were 46.2 (10.37) (range 7.4–

66.7) and 54.6 (8.63) (range 8.8–72.4), respectively (Table 1). The

lowest reported subscale mean score from the SF-36 ques-

tionnaire was for ‘‘vitality’’ (mean = 65.9), while the highest

score was for ‘‘social functioning’’ (mean = 89.5) (Fig. 1). The

majority of participants (78.6%) reported a negative impact of

diabetes on their QoL, and the median ADDQoL AWI score was

�0.32 (IQR �0.89 to �0.06) (Table 1). The ADDQoL domain with

the greatest negative impact on QoL was ‘‘freedom to eat’’

(mean = �2), and with the least impact was ‘‘society reaction’’

(�0.24) (Fig. 2).

7. Association between health status,diabetes-specific QoL and HbA1c

Younger age, lower educational status, no alcohol consump-

tion, low levels of physical activity, higher BMI, and intake of

65.9

66.1

74.4

75.2

77.3

82

84.4

89.5

0 20 40 60 80 100

Vitality

General hea lth

Role-physical

Bodily pai n

Physical fun ctionin g

Ment al healt h

Role-emot ional

Social func tionin g

SF-36 scores

Fig. 1 – The eight mean SF-36 dimension scores of the

ADDITION-Europe cohort at five years post diagnosis

(range 0-100, higher scores indicating better health)

(n=1876).

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 1 0 4 ( 2 0 1 4 ) 2 8 1 – 2 8 7284

glucose-lowering drugs were associated with higher HbA1c

values. There was a significant univariate association between

the SF-36 summary scores, the SF-36 sub-scales (data not

shown) and the ADDQoL measures with HbA1c (Table 2). After

adjustment for the aforementioned covariates (Model 2), the

association between the SF-36 scores and HbA1c became non-

significant. Adjustment for intake of glucose-lowering drugs

(Model 3) slightly attenuated but did not change the significant

association between the ADDQoL AWI score and HbA1c (b = 0.2

[0.10, 0.31]).

Mean SF-36 PCS and MCS scores were higher in participants

with HbA1c < 7% (53 mmol/mol) compared to those with

HbA1c�7% (53 mmol/mol) (PCS: 46.7 (10.12) vs. 44.7 (10.9),

p < 0.001, MCS: 54.9 (8.46) vs. 53.8 (9.01), p = 0.014). Median

ADDQoL AWI scores were also higher in those with

HbA1c < 7% (53 mmol/mol) (�0.26 vs. �0.5, p < 0.001).

--0-0

-0.5-0.5

-0.57-0.65

-0.68-0. 7

-0.71-0.73

-0.76-0.79

-1.15-1.18

-1.21-2

-2. 5 -2 -1. 5 -1 mean weighted impact

Fig. 2 – Mean weighted impact (ADDQoL score) of diabetes on in

years post diagnosis (range -9 to +3, negative scores indicating

8. Discussion

We observed a small independent association between

diabetes-specific QoL and glycaemic control. Individuals

who reported a negative impact of diabetes on their QoL

had higher HbA1c levels at five years post diagnosis compared

with those who reported a positive or no impact of diabetes.

The SF-36 PCS and MCS scores were not independently

associated with HbA1c.

The mean SF-36 PCS [46.2 (10.4)] and MCS [54.6 (8.6)] scores

of ADDITION-Europe participants were similar to those

reported in previous studies among individuals with a

diabetes duration of approximately five years. For example,

in a Canadian cohort of individuals (mean age 54 years), the

SF-36 PCS and MCS scores were 49.2 (7.4) and 51 (9.5),

respectively [20]. In Dutch diabetes patients [mean age 64.4

(8.8)], the SF-36 PCS score was 48.3 and the MCS score was 54.4

[21]. These scores compare favourably to other chronic

conditions, where lower health status was observed. For

example, among patients with knee or hip osteoarthritis

[mean age 67 years, disease duration 5.7 (4.9) years], the PCS

and MCS scores were 31.9 (8.4) and 47.0 (11.0), respectively [22].

The mean ADDQoL AWI score was higher in the ADDITION-

Europe cohort [�0.75 (1.17)] compared with other populations,

suggesting that ADDITION participants report a higher

diabetes-specific QoL five years after diagnosis. For example,

in an Australian cohort (mean age 61 years) with established

diabetes [mean diabetes duration 7.6 (8.1) years] the mean

ADDQoL AWI score was �1.66 [23]. The higher ADDQoL AWI

score in our cohort may be explained by a shorter disease

duration, better controlled HbA1c (76% achieved HbA1c�7%

(53 mmol/mol) in our cohort vs. 49% in the Australian cohort),

and a low percentage of our patients requiring insulin

treatment [23]. We observed a large variation in the sig-

nificance attached to different ADDQoL domains. The greatest

-0.24-0.420.45.48.4933

-0. 5 0

Society re acti onFinan cesLiving co ndition sDependen ceLocal or long-distance journeysPhys ical appe aranc eFriendships and soci al lif eSelf- confidenceFamily li feLeisure acti vit iesMotivatio nClosest person al relat ionshi pHolidaysWorking li feDo phys icall ySex li feFreedom to drinkFeelin gs about the futureFreed om to eat

dividual life domains in the ADDITION-Europe cohort at five

lower quality of life).

Table 2 – Crude and adjusted associations betweenhealth status, diabetes-specific quality of life and HbA1c

in the ADDITION-Europe cohort at five years postdiagnosis (n = 1876).

Variables Unstandardisedb-coefficient (95% CI)

Model 1

Age �0.02 (�0.03, �0.02)***

Sex (men = 0) �0.07 (�0.15, 0.02)

Full-time education completed

at �17 years (at �17 years = 0)

0.21 (0.13, 0.29)***

Smoking status (non-smoker/

ex-smoker = 0)

0.04 (�0.06, 0.15)

Alcohol consumption (no = 0) �0.21 (�0.3, �0.12)***

Physical activity (high = 0)

low 0.14 (0.04, 0.24)**

moderate �0.01 (�0.10, 0.08)

BMI (kg/m2) 0.04 (0.03, 0.05)***

Intake of any glucose-lowering

drug (no intake = 0)

0.50 (0.41, 0.57)***

SF-36 PCS �0.01 (�0.1, �0.003)***

SF-36 MCS �0.01 (�0.01, �0.003)**

ADDQoL AWI (no impact/positive

impact of diabetes on QoL = 0)a0.28 (0.17, 0.39)***

Model 2

SF-36 PCS �0.002 (�0.07, 0.002)

SF-36 MCS �0.003 (�0.08, 0.002)

ADDQoL AWI (no impact/positive

impact of diabetes on QoL = 0)a0.26 (0.15, 0.36)***

Model 3

SF-36 PCS �0.001 (�0.01, 0.003)

SF-36 MCS �0.002 (�0.01, 0.002)

ADDQoL AWI (no impact/positive

impact of diabetes on QoL = 0)a0.20 (0.10, 0.31)***

Values are unstandardised b-coefficients (95% confidence interval).

Model 1: crude. Model 2: adjusted for trial arm, age, sex, education,

alcohol consumption, physical activity, and BMI. Model 3: addi-

tionally adjusted for intake of glucose-lowering drugs. BMI, body

mass index. HbA1c, glycosylated haemoglobin. SF-36 PCS, Physical

Component Summary measure. SF-36 MCS, Mental Component

Summary measure. ADDQoL AWI, the overall average weighted

impact score of the ADDQoL.a Values are based on 1646 participants.* p < 0.05,** p < 0.01,*** p < 0.001.

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 1 0 4 ( 2 0 1 4 ) 2 8 1 – 2 8 7 285

negative impact on diabetes-specific QoL in our cohort was

‘‘freedom to eat’’ (�2) followed by ‘‘feelings about the future’’

(�1.21), while ‘‘finances’’ (�0.42) and ‘‘society reaction’’ (�0.24)

were the least significant domains. Singaporean diabetes

patients [mean age 58 (8.8)] with established type 2 diabetes

also reported that ‘‘freedom to eat’’ had the greatest negative

impact on their QoL [24].

Although there was no association between health status

and HbA1c in ADDITION-Europe participants five years after

diabetes diagnosis, we observed a significant association

between disease-specific QoL and HbA1c. The ADDQoL score

was designed to assess the impact of diabetes-related

complications rather than other co-morbidities; in contrast,

the SF-36 is a generic health status measure which is less likely

to detect differences due to treatment regimen and more likely

to detect differences due to other non-diabetes-related

comorbidities. Additionally, the lack of association between

the SF-36 and HbA1c may be explained by the relatively low

mean and standard deviation of HbA1c [6.7% (0.92) (50 mmol/

mol)] and the small proportion of patients (4.7%) prescribed

insulin, suggesting that the patients were well controlled, a

factor that has previously been associated with QoL. Our

results highlight the importance of assessing disease-specific

QoL in diabetes patients.

Some studies have reported that increasing treatment

intensity in patients with diabetes was associated with

worsening health or QoL. For example, in patients with type

2 diabetes (mean age 60 years, mean diabetes duration 12

years) patients on insulin reported greater impact of diabetes

on QoL compared with those on oral hypoglycaemic agents or

diet alone [18]. The findings from a Finnish cohort (mean age

63 years) with established type 1 and type 2 diabetes (mean

diabetes duration 10 years) showed that the diet treatment

group had a significantly better QoL level compared to the

tablet or combined treatment (patients treated with tablets

and insulin) groups [25]. When we looked at the association

between health status, diabetes-specific QoL and HbA1c by

trial group, the results revealed no significant association

between the SF-36 summary scores and HbA1c, and the

association between the ADDQoL measure and HbA1c was

similar in both groups. This suggests that increases in the

intensity of treatment using oral medication did not adversely

affect participant’s health or diabetes-specific QoL in the first

five years after diagnosis.

We observed small differences in the SF-36 PCS, MCS and

ADDQoL AWI scores in participants with normal and elevated

HbA1c levels. However, while the difference in HbA1c between

those who reported a negative impact of diabetes and those

who reported no impact or a positive impact of diabetes on

QoL was small (0.2%), the difference between the 25th and 75th

percentiles was 0.9%. Given that a reduction of 0.5% in HbA1c is

considered a clinically significant improvement [26], and a 1%

reduction in HbA1c is associated with 21% reduction in risk for

any diabetes-related endpoint [27], our finding suggests that

diabetes-specific QoL may be an important and potentially

modifiable risk factor for improving glycaemic control. It has

been shown that people with diabetes experience improved

QoL from participation in diabetes self-management training

programmes [28]. Therefore interventions aiming to develop

diabetes self-management skills might impact on diabetes-

specific QoL and improve HbA1c. Further, as a restricted diet

was a major driver of poor diabetes-specific QoL, perhaps a

patient-centred, holistic approach to improving QoL is needed

alongside advice concerning diet, physical activity, medica-

tion and adherence in trying to lower HbA1c. When assisting

patients in the management of their diabetes, practitioners

should consider both the illness experience (which impacts on

QoL) and the progression of the disease (focusing on clinical

outcomes such as HbA1c), and be able to reconcile these two

aspects. However, a recent paper demonstrated the dilemma

confronting practitioners given that healthy eating may

impact negatively on QoL and yet in the longer term is

associated with better glycaemic control [29].

The relationship between QoL and glucose is clearly

complex. QoL refers to the physical, psychological, and social

domains of health that are influenced by a person’s experi-

ences, beliefs, expectations, and perceptions [30]. QoL is

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 1 0 4 ( 2 0 1 4 ) 2 8 1 – 2 8 7286

related to both diabetes and its risk factors, for example, BMI

[31]. Given that ADDITION-Europe participants were obese,

this characteristic may reduce QoL and lie on the pathway

between QoL and HbA1c. It is also feasible, for example, that

QoL may impact an individual’s ability to engage in, and

maintain healthy behaviours, such as physical activity, diet,

and medication adherence, which in turn might affect HbA1c.

Furthermore, depression, which is known to be more

prevalent in patients with diabetes compared to those without

[32], may also contribute to reduced QoL [33], and forms part of

the complex relationship between diabetes, dietary behaviour

and QoL. The relationship between QoL and glucose is also bi-

directional: QoL may affect diabetes self-efficacy, self-care

behaviours, glucose and complications, just as each of these

variables may affect each other and QoL [1]. While no studies

have explored QoL as an explanatory variable for glycaemic

control, research suggests that increases in HbA1c of 1% or

greater are associated with substantial decreases in QoL, while

decreases of the same magnitude are associated with smaller,

but clinically important improvements in QoL [8].

The strengths of our study include a large sample size,

which was drawn from a representative, population-based

sample in three different European countries. We included

both a general health and a disease-specific QoL measure, and

used a validated physical activity questionnaire. However, due

to the cross-sectional nature of our analysis the association

between diabetes-specific QoL and HbA1c must be interpreted

with caution. We cannot infer a causal relationship and,

although we adjusted for several variables related to QoL and

HbA1c, there may be residual confounding. Furthermore, the

study sample was largely Caucasian and middle-aged, which

restricts generalisability to different populations. Lifestyle

behaviours and drug intake were measured by self-report

which may be subject to recall and social desirability bias. It

would have been interesting to include dietary and medication

adherence behaviours to examine the pathways through

which QoL impacts on HbA1c levels but they were not available

in the whole cohort. It has been shown that some psycholo-

gical factors such as health-related beliefs, social support,

coping style and personality type may affect QoL [1]; therefore,

further research should include these factors when exploring

the association between diabetes-specific QoL and HbA1c as

they may act either as predictors, as confounders or both.

In conclusion, we observed a small independent association

between diabetes-specific QoL and glycaemic control. Indivi-

duals who reported a negative impact of diabetes on their QoL

had higher HbA1c levels at five years post diagnosis compared to

those who reported a positive or no impact of diabetes. To help

patients to reduce their HbA1c levels, practitioners should take

account of the complex relationship between diabetes-specific

QoL and glucose, particularly with regards to dietary behaviour.

Future research should attempt to elucidate via which path-

ways and in which direction the association between diabetes-

specific QoL and HbA1c might act.

Conflicts of interests

LK, SJG, KK and RKS declared that they have no conflicts of

interests.

Acknowledgements

Author contributions: LK, RKS and SJG conceived the study

question. SJG, MJD, TL, KK, and GEHMR are ADDITION-Europe

PIs. LK analysed and interpreted the data. LK, SJG, MJD, TL, KK,

GEHMR and RKS drafted the manuscript. All authors critically

revised the manuscript for important intellectual content and

approved the final version.

Funding: The ADDITION-Europe trial was funded by

National Health Service Denmark, Danish Council for Strategic

Research, Danish Research Foundation for General Practice,

Danish Centre for Evaluation and Health Technology Assess-

ment, Danish National Board of Health, Danish Medical

Research Council, Aarhus University Research Foundation,

Wellcome Trust, UK Medical Research Council, UK NIHR

Health Technology Assessment Programme, UK National

Health Service R&D, UK National Institute for Health Research,

Julius Center for Health Sciences and Primary Care, University

Medical Center, Utrecht, Novo Nordisk, Astra, Pfizer, Glax-

oSmithKline, Servier, HemoCue, Merck. MJD is an NIHR Senior

Investigator. LK was supported by the German Research

Foundation (DFG) Grant KU 3056/1-1.

TL has received unrestricted grants for the ADDITION study

from public foundations and the Medical Industry: Novo

Nordisk AS, Novo Nordisk Scandinavia AB, ASTRA Denmark,

Pfizer Denmark, GlaxoSmithKline Pharma Denmark, SERVIER

Denmark A/S and HemoCue Denmark A/S. TL has held three

lectures for the medical industry within the past 2 years. TL

hold shares in Novo Nordisk. MJD has acted as consultant,

advisory board member and speaker for Novartis, Novo

Nordisk, Sanofi-Aventis, Lilly, Merck Sharp & Dohme, Boeh-

ringer Ingelheim and Roche. She has received grants in

support of investigator and investigator initiated trials from

Novartis, Novo Nordisk, Sanofi-Aventis, Lilly, Pfizer, Merck

Sharp & Dohme and GlaxoSmithKline. GEHMR has acted as

consultant and advisory board member for Novo Nordisk,

Merck Sharp & Dohme and Astra-Zeneca. He has received a

grant in support of investigator initiated trials from Merck

Sharp & Dohme.

r e f e r e n c e s

[1] Rubin RR, Peyrot M. Quality of life and diabetes. Diab MetabRes Rev 1999;15:205–18.

[2] Bradley C, Speight J. Patient perceptions of diabetes anddiabetes therapy: assessing quality of life. Diab Metab ResRev 2002;18(Suppl. 3):S64–9.

[3] Bradley C. Importance of differentiating health status fromquality of life. Lancet 2001;357:7–8.

[4] Psaltopoulou T, Ilias I, Alevizaki M. The role of diet andlifestyle in primary, secondary, and tertiary diabetesprevention: a review of meta-analyses. Rev Diab Stud2010;7:26–35.

[5] Esposito K, Chiodini P, Bellastella G, Maiorino MI, GiuglianoD. Proportion of patients at HbA1c target <7% with eightclasses of antidiabetic drugs in type 2 diabetes: systematicreview of 218 randomized controlled trials with 78,945patients. Diab Obes Metab 2012;14:228–33.

d i a b e t e s r e s e a r c h a n d c l i n i c a l p r a c t i c e 1 0 4 ( 2 0 1 4 ) 2 8 1 – 2 8 7 287

[6] Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HA. 10-Year follow-up of intensive glucose control in type 2diabetes. N Engl J Med 2008;359:1577–89.

[7] Group AC, Patel A, MacMahon S, Chalmers J, Neal B, Billot L,et al. Intensive blood glucose control and vascularoutcomes in patients with type 2 diabetes. N Engl J Med2008;358:2560–72.

[8] Testa MA, Simonson DC. Health economic benefits andquality of life during improved glycemic control in patientswith type 2 diabetes mellitus: a randomized, controlled,double-blind trial. JAMA 1998;280:1490–6.

[9] Goddijn PP, Bilo HJ, Feskens EJ, Groeniert KH, van der ZeeKI, Meyboom-de Jong B. Longitudinal study on glycaemiccontrol and quality of life in patients with type 2 diabetesmellitus referred for intensified control. Diab Med1999;16:23–30.

[10] Chaturvedi N, Stephenson JM, Fuller JH. The relationshipbetween socioeconomic status and diabetes control andcomplications in the EURODIAB IDDM ComplicationsStudy. Diab Care 1996;19:423–30.

[11] Chan JC, Gagliardino JJ, Baik SH, Chantelot JM, Ferreira SR,Hancu N, et al. Multifaceted determinants for achievingglycemic control: the International Diabetes ManagementPractice Study (IDMPS). Diab Care 2009;32:227–33.

[12] Thomas DE, Elliott EJ, Naughton GA. Exercise for type 2diabetes mellitus. Cochrane Database Syst Rev2006;CD002968.

[13] Griffin SJ, Borch-Johnsen K, Davies MJ, Khunti K, Rutten GE,Sandbaek A, et al. Effect of early intensive multifactorialtherapy on 5-year cardiovascular outcomes in individualswith type 2 diabetes detected by screening (ADDITION-Europe): a cluster-randomised trial. Lancet 2011;378:156–67.

[14] American Diabetes Association. Standards of medical carein diabetes–2007. Diab Care 2007;30(Suppl. 1):S4–1.

[15] Craig CL, Marshall AL, Sjostrom M, Bauman AE, Booth ML,Ainsworth BE, et al. International physical activityquestionnaire: 12-country reliability and validity. Med SciSports Exer 2003;35:1381–95.

[16] Ware JE, Snow KK, Kosinski M, Gandek B. SF-36 HealthSurvey Manual and Interpretation Guide. second ed.Boston, MA: New England Medical Center, The HealthInstitute; 2000.

[17] Bradley C, Todd C, Gorton T, Symonds E, Martin A,Plowright R. The development of an individualizedquestionnaire measure of perceived impact of diabetes onquality of life: the ADDQoL. Qual Life Res 1999;8:79–91.

[18] Jacobson AM, de Groot M, Samson JA. The evaluation of twomeasures of quality of life in patients with type I and type IIdiabetes. Diab Care 1994;17:267–74.

[19] Bradley C. ADDQoL19. Supplement to ADDQoL18 UserGuidelines. Retrieved February 2013, from URL: http://www.healthpsychologyresearch.com/Admin/uploaded/Guidelines/addqol18_userguidelines_rev24jan05a.pdf.

[20] Reid RD, Tulloch HE, Sigal RJ, Kenny GP, Fortier M,McDonnell L, et al. Effects of aerobic exercise, resistance

exercise or both, on patient-reported health status andwell-being in type 2 diabetes mellitus: a randomised trial.Diabetologia 2010;53:632–40.

[21] Wermeling PR, Gorter KJ, van Stel HF, Rutten GE. Bothcardiovascular and non-cardiovascular comorbidity arerelated to health status in well-controlled type 2 diabetespatients: a cross-sectional analysis. Cardiovasc Diabetol2012;11:121.

[22] Rannou F, Boutron I, Jardinaud-Lopez M, Meric G, Revel M,Fermanian J, et al. Should aggregate scores of the MedicalOutcomes Study 36-item Short Form Health Survey be usedto assess quality of life in knee and hip osteoarthritis? Anational survey in primary care. Osteoarthr Cartilage2007;15:1013–8.

[23] Ostini R, Dower J, Donald M. The Audit of Diabetes-Dependent Quality of Life 19 (ADDQoL): feasibility,reliability and validity in a population-based sample ofAustralian adults. Qual Life Res 2012;21:1471–7.

[24] Shim YT, Lee J, Toh MP, Tang WE, Ko Y. Health-relatedquality of life and glycaemic control in patients with type 2diabetes mellitus in Singapore. Diab Med 2012;29(8):e241–8.

[25] Keinanen-Kiukaanniemi S, Ohinmaa A, Pajunpaa H,Koivukangas P. Health related quality of life in diabeticpatients measured by the Nottingham Health Profile. DiabMed 1996;13:382–8.

[26] Farmer AJ, Perera R, Ward A, Heneghan C, Oke J, BarnettAH, et al. Meta-analysis of individual patient data inrandomised trials of self monitoring of blood glucose inpeople with non-insulin treated type 2 diabetes. BMJ2012;344:e486.

[27] Stratton IM, Adler AI, Neil HA, Matthews DR, Manley SE,Cull CA, et al. Association of glycaemia with macrovascularand microvascular complications of type 2 diabetes(UKPDS 35): prospective observational study. BMJ2000;321:405–12.

[28] Cochran J, Conn VS. Meta-analysis of quality of lifeoutcomes following diabetes self-management training.Diab Educ 2008;34:815–23.

[29] Donald M, Dower J, Coll JR, Baker P, Mukandi B, Doi SA.Mental health issues decrease diabetes-specific quality oflife independent of glycaemic control and complications:findings from Australia’s living with diabetes cohort study.Health Qual Life Outcomes 2013;11:170.

[30] Testa MA, Simonson DC. Assesment of quality-of-lifeoutcomes. N Engl J Med 1996;334:835–40.

[31] Centers for Disease Control and Prevention. MeasuringHealthy Days. Atlanta, Georgia: CDC; November 2000.

[32] Ali S, Stone MA, Peters JL, Davies MJ, Khunti K. Theprevalence of co-morbid depression in adults with type 2diabetes: a systematic review and meta-analysis. Diab Med2006;23:1165–73.

[33] Ali S, Stone M, Skinner TC, Robertson N, Davies M, KhuntiK. The association between depression and health-relatedquality of life in people with type 2 diabetes: a systematicliterature review. Diab Metab Res Rev 2010;26:75–89.