Journal of Psychosomatic Research 77 (2014) 174–179

Contents lists available at ScienceDirect

Journal of Psychosomatic Research

Relationships of diabetes-specific emotional distress, depression, anxiety,

and overall well-being with HbA1c in adult persons with type 1 diabetes
Ragnhild Bjarkøy Strandberg a,b,⁎, Marit Graue a,c, Tore Wentzel-Larsen d,e,f, Mark Peyrot g, Berit Rokne b,h
a
Centre for Evidence Based Practice, Bergen University College, Norway
b
Department of Global Public Health and Primary Care, Faculty of Medicine and Dentistry, University of Bergen, Norway
c
Department of Paediatrics, Haukeland University Hospital, Bergen, Norway
d
Norwegian Centre for Violence and Traumatic Stress Studies, Oslo, Norway
e
Centre for Child and Adolescent Mental Health, Eastern and Southern Norway, Norway
f
Centre for Clinical Research, Haukeland University Hospital, Norway
g
Department of Sociology, Loyola University MD, USA
h
Department for Research and Development, Haukeland University Hospital, Norway

a r t i c l e i n f o a b s t r a c t

Article history: Objective: Emotional problems are common in adults with diabetes, and knowledge about how different
Received 1 April 2014 indicators of emotional problems are related with glycemic control is required. The aim was to examine the
Received in revised form 27 June 2014 relationships of diabetes-specific emotional distress, depression, anxiety, and overall well-being with glycosylated
Accepted 30 June 2014 hemoglobin (HbA1c).
Methods: Of the 319 adults with type 1 diabetes attending the endocrinology outpatient clinic at a university
Keywords:
hospital in Norway, 235 (74%) completed the Diabetes Distress Scale, the Problem Areas in Diabetes Survey, the
Depression
Anxiety
Hospital Anxiety and Depression Scale, and the World Health Organization-Five Well-Being Index. Blood samples
Diabetes-specific emotional distress were taken at the time of data collection to determine HbA1c. Regression analyses examined associations of
HbA1c diabetes-specific emotional distress, anxiety, depression, and overall well-being with HbA1c. The relationship
Type 1 diabetes between diabetes-specific emotional distress and HbA1c was tested for nonlinearity.
Results: Diabetes-specific emotional distress was related to glycemic control (DDS total: unstandardized
coefficient = 0.038, P b .001; PAID total: coefficient = 0.021, P = .007), but depression, anxiety, and overall
well-being were not. On the DDS, only regimen-related distress was independently related to HbA 1c
(coefficient = 0.056, P b .001). A difference of 0.5 standard deviation of baseline regimen distress is associated
with a difference of 0.6 in HbA1c. No significant nonlinearity was detected in the relationship between diabetes-
specific distress and HbA1c.
Conclusions: To stimulate adequate care strategies, health personnel should acknowledge depression and diabetes-
specific emotional distress as different conditions in clinical consultations. Addressing diabetes-specific emotional
distress, in particular regimen distress, in clinical consultation might improve glycemic control.
© 2014 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-SA license
(http://creativecommons.org/licenses/by-nc-sa/3.0/).

Introduction to prevent late complications of diabetes, such as diabetic retinopathy,

nephropathy and neuropathy [7]. The American Diabetes Association
Type 1 diabetes is a chronic disease where the insulin-producing guidelines [8] recommend that persons with type 1 diabetes should
pancreatic beta cells are destroyed [1,2], which makes a lifetime have an HbA1c level of b7%. Despite the growing knowledge about what
treatment of exogenous insulin replacement necessary [3]. Diabetes is might improve glycemic control, many persons do not meet the treat-
a growing public health burden across the world [4], with consequences ment recommendation [9,10]. Emotional problems might complicate
for the individuals' daily lives [5], and substantial economic burden on the required self-management of the disease [11], and limit the persons'
the society [6]. Maintaining an appropriate glycemic control is important management of self-care activities necessary to achieve an adequate
glycemic control [12]. The recent DAWN2 study showed that emotional
problems are a challenge of concern in persons with diabetes across
cultures [13]. Further it was reported that although the health care
⁎ Corresponding author at: Centre for Evidence Based Practice, Bergen University
College, Post box 7030, N-5020 Bergen, Norway. Tel.: +47 55 58 72 43; fax: +47 55 29
providers acknowledged the importance of addressing emotional
83 64. problems in persons with diabetes [14], there was a gap between persons'
E-mail address: rob@hib.no (R.B. Strandberg). needs and current health care strategies [13].

http://dx.doi.org/10.1016/j.jpsychores.2014.06.015
0022-3999/© 2014 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-SA license (http://creativecommons.org/licenses/by-nc-sa/3.0/).
 R.B. Strandberg et al. / Journal of Psychosomatic Research 77 (2014) 174–179 175

The presence of depression or diabetes-specific emotional distress, or Table 1

a combination of these, might comprise barriers to adequate self- Person characteristics of the 235 persons with type 1 diabetes.

management in persons with type 1 diabetes [15]. The underrecognition Total, n 235
of emotional problems, such as depression, anxiety, and diabetes-specific
Male, n (%) 135 (57.4)
emotional distress, has been reported [16], and when such concerns are
recognized, problems might be identified as depression, even in patients Female, n (%) 100 (42.6)
whose problems are directly related to diabetes and its treatment [17]. Age
Diabetes-specific emotional distress can be defined as a range of Mean (SD) 39.0 (13.7)
emotional responses and reactions to life with diabetes, especially Min–max, years 18–69
those related to the treatment regimen and self-care demands. It is Diabetes duration, years
part of a person's experience of managing diabetes and its treatment Mean (SD) 18.6 (12.0)
in the social context of family and health-care personnel [18–20]. In Min–max 1–58
contrast, depression is more strongly related to an anhedonic state, in HbA1c, %a mean (SD) 8.1 (1.6)
which an individual is markedly affected by feelings of sorrow and Presence of one or more late complicationb yes (%) 81 (40.3)
Education, n (%)
hopelessness [21,22], and anxiety is predominantly related to fear,
University, more than 4 years 30 (13.2)
worry, and dread [21]. University, up to 4 years 67 (29.4)
Gonzales et al. [23] suggested that depression and diabetes-specific
College/high school 104 (45.6)
emotional distress are independent constructs in type 2 diabetes, and
Primary school, 9 years 27 (11.8)
later proposed that there can be confusion regarding what is actually DDS total meanc (SD) 19.5 (15.8)
addressed [17]. Hermanns et al. [24] showed that despite some overlap,
DDS EBd (SD) 26.3 (22.4)
people with depression and those with diabetes-specific emotional
distress did not constitute identical groups in patients within type 1 or DDS RDd (SD) 23.6 (20.9)
type 2 diabetes. It has been shown that depression and diabetes-specific d
DDS ID (SD) 12.9 (17.3)
emotional distress are differently associated with diabetes-specific
d
indicators, but this is mainly examined in persons with type 2 diabetes DDS PD (SD) 10.7 (15.8)
[25–27]. PAID total meanc (SD) 23.6 (18.6)
WHO-5 total meanc (SD) 60 (19.8)
Fisher et al. [18] found nonlinear relationships of diabetes-specific
HADS-A meanc (SD) 5.6 (3.7)
emotional distress with HbA1c, diet, self-efficacy, and physical activity HADS-D meanc (SD) 3.6 (3.5)
in two samples of persons with type 2 diabetes, with stronger relation- a
Mmol/mol: 65.
ships for lower levels of diabetes-specific emotional distress. The b
n = 201.
authors suggested that distress should be recognized at a lower level c
DDS, PAID and WHO-5: 0–100 scale, HADS-A and HADS-D: 0–21 scales.
d
than previously recommended, and further suggested that to split the The DDS subscales: the Emotional Burden subscale, the Regimen-related Distress
DDS scores into three groups (low, moderate and high distress) would subscale, the diabetes-related Interpersonal Distress subscale, and the Physician-related
Distress subscale.
better accommodate to the significant nonlinear relationship. Little
knowledge about a potential nonlinear association is available in
persons with type 1 diabetes, and it is not appropriate to assume that
emotional problems are similarly manifested and have the same clinical
consequences in persons with type 1 and type 2 diabetes. Therefore, the from 1 to 6 [19], but has been modified to a five-point Likert scale
main aim of this study was to examine the relationships of diabetes- ranging from 0 (not a problem) to 4 (a serious problem) [28]. The
specific emotional distress, anxiety, depression, and overall well-being questionnaire has been translated into Norwegian [29], and is consid-
with HbA1c, and to determine whether there is a nonlinear relationship ered internationally to have good psychometric properties [28–30]. A
between diabetes-related emotional distress and HbA1c in individuals total score of 0–100 was computed, where higher scores represent higher
with type 1 diabetes. levels of distress [28].
The Diabetes Distress Scale (DDS) was developed to address some of
Materials and methods the limitations in earlier instruments that measured disease-specific
emotional distress, and consists of 17 items divided into four subscales:
Sample and settings the Emotional Burden subscale (EB, five items), Physician-related
Distress subscale (PD, four items), Regimen-related Distress subscale
Of the 319 persons with type 1 diabetes, aged 18–69 years, attending (RD, five items), and diabetes-related Interpersonal Distress subscale
an adult outpatient clinic between October 2008 and January 2009 who (ID, three items). The DDS is based on a six-point Likert scale, ranging
were invited to participate in this study, 235 persons agreed to participate from 1 (no problem) to 6 (a serious problem) [20], and the measure
(74%). Some information was available to compare participants with has been translated into Norwegian [29]. It has shown good psychomet-
nonparticipants: age (39.0 versus 37.9 years, respectively; P = .535), ric properties across different countries and cultures [18,20,29]. A total
sex distribution (male 57% versus 66%, respectively; P = .244) and score of 0–100 was computed, where higher scores indicate greater
HbA1c level (8.1% (65 mmol/mol) versus 8.4% (68 mmol/mol), respec- emotional distress [20]. For the nonlinearity analyses, scales scored
tively; P = .285). Sociodemographic and clinical information about the 1–6 were computed to enable this part of the analysis to be more easily
study subjects are presented in Table 1. To determine HbA1c levels, compared with the Fisher et al. [18] results for nonlinearity in persons
blood samples were taken at the time of data collection, and analyzed with type 2 diabetes.
with DCA-2000 Analyzer (Bayer, Elkhart, IN, USA). The Hospital Anxiety and Depression Scale (HADS) was designed for
clinicians as a screening test for psychiatric disorder in non-psychiatric
Measures hospital departments [22]. It consists of two subscales, HADS-A
(anxiety) and HADS-D (depression), each with seven items with four-
The following questionnaires were included in the study. The point Likert scales, ranging from 0 to 3 [31], and 0–21 scales were
Problem Areas in Diabetes Survey (PAID) consists of 20 items and was computed for HADS-A and HADS-D, where higher score indicates
developed to gain insight into the breadth of emotional responses to worse anxiety or depression state. In a review study, Bjelland et al.
diabetes. It was initially based on a six-point Likert scale, ranging [32] found that the HADS performed well cross-culturally, and that its
176 R.B. Strandberg et al. / Journal of Psychosomatic Research 77 (2014) 174–179

validity was good to very good. The psychometric properties of the Ethical considerations
Norwegian version of the HADS showed to be satisfactory [33].
The World Health Organization-Five Well-Being Index (WHO-5) The study was approved by the Western Norway Committee for
was developed to measure well-being [34], and consists of five positive- Medical and Health Research Ethics (19580/865). Participants got
ly worded items that assess well-being during the preceding 14 days, written and oral information about the study, and were informed that
with six-point Likert scales ranging from 0 (not present) to 5 (constantly they could withdraw at any point of time.
present). A total score of 0 (worst thinkable well-being) to 100 (best
Results
thinkable well-being) was computed [35].
As recommended by Fayers and Machin [36], each scale score was Linear regression analysis
based on the mean of the valid items within each score if at least half
the items were valid, except for HADS, where at least 5 of 7 items had The following variables were significantly related with HbA1c in the bivariate
to be valid. Cronbach's alphas for the respondents with type 1 diabetes regression analysis (Table 2): PAID and DDS total scores (unstandardized coefficient
0.020, P = .001, and 0.033, P b .001, respectively); regimen-related distress DDS subscale
were DDS total 0.92, PAID total 0.95, WHO-5 0.89, HADS-A 0.81, HADS-D
and the emotional burden DDS subscale (0.039, P b .001 and 0.014, P = .005, respective-
0.81, DDS RD subscale 0.84, DDS EB subscale 0.88, DDS PD subscale 0.83, ly); HADS-A, HADS-D and WHO-5 were not. Further, the presence of one or more late
and DDS ID subscale 0.81. complications (0.621, P = .010) and lower level of education were significantly associated
with higher HbA1c (P = .025). Age and gender were not significantly related with HbA1c in
the bivariate regression.
When analyzing the indicators in separate analyses, all controlled for age, sex,
Statistical analysis education and late complications, results were similar to the bivariate regression results
for PAID total and DDS total (significant) and HADS-A, HADS-D and WHO-5 (not
Respondents that had any missing values on any of the explanatory significant). For the DDS subscales, the regimen-related distress and emotional burden
variables were excluded from regression analysis, giving a sample size were significant (0.038, P b .001 and 0.011, P = .036, respectively).
In the fully adjusted multiple regression analyses, both the DDS total score
of n = 185.
(0.038, P b .001) and PAID total score (0.021, P = .007) were significantly associated
HbA1c was the dependent variable in all analyses. To determine with glycemic control. In the model including the four DDS subscales, only the RD
whether there was any relationship of all indicators with HbA1c , subscale was significantly associated with HbA1c (0.056, P b .001). The model includ-
bivariate regression analyses of HbA 1c with DDS total, PAID total, ing DDS total explained 20.3% (R 2), and the model including PAID total explained
15.0% (R2), of the variation in HbA1c. The DDS subscale model explained 38.6% (R2)
each of the four DDS subscales, HADS-A, HADS-D, WHO-5 and each
of the variation in HbA1c. Overall well-being, depression or anxiety was not signifi-
of the adjustment variables age, sex, education and late complication cantly related to HbA1c in any of these models. The maximum VIF was 2.92, and the
were performed (results for adjustment variables not shown). Next, results from the sensitivity analysis using multiple imputation showed only minor
separate regression analyses were performed, one for each of the differences.
total diabetes distress scales, each of the two HADS scales and the Peyrot et al. [38] suggested the use of 0.5 standard deviation (SD) as the minimum
detectable difference (MDD), an estimate of the smallest change that can be subjectively
overall well-being total scale, in addition to each of the four DDS sub-
realized by individuals [39]. A difference of 0.5 SD (of baseline mean, Table 1) in PAID
scales, all adjusted for age, sex, education and late complications. In total and DDS total is associated with a difference of 0.2 and 0.3 in HbA1c, respectively. A
the last phase, three models were estimated with multiple regres- difference of 0.5 SD in the regimen-related emotional distress (RD subscale) is associated
sion analyses adjusted for age, sex, education, and late complications with a difference of 0.6 in HbA1c. Thus, persons who perceived that their regimen related
distress had noticeably increased might be expected to experience an increase of 0.6 in
as well as HADS-A, HADS-D and WHO-5. The DDS total and PAID total
their HbA1c.
scales were analyzed in separate regression models because these
instruments measure parallel constructs. Because the DDS subscales Nonlinear analyses
measure quite different areas of diabetes-specific emotional distress
[20] the four subscales were analyzed together in the last regression No significant nonlinear relationship was found between diabetes-specific emotional
distress and glycemic control; P for nonlinearity = 0.317 in the model based on the DDS
model. Multicollinearity in the multiple regression models was
total and P for nonlinearity = 0.309 in the model based on the PAID total. Also, graphs
checked by variance inflation factor (VIF). A sensitivity analysis of the estimated relationships did not indicate deviations from linearity. Moreover, no
using multiple imputation (200 imputed data sets) was performed significant nonlinear relationships were found between the four DDS subscales and
to test whether the results from the complete case regression analy- HbA1c; P for nonlinearity = 0.322 (RD), 0.464 (EB), 0.505 (ID) and 0.186 (PD). Results
ses described above were biased. from the supplementary analysis with quadratic regression analysis showed similar
results, where no significant nonlinear relationships between diabetes-specific emotional
The potential nonlinear relationship between diabetes-specific distress and HbA1c were apparent, with P for nonlinearity of 0.126 (DDS total), 0.112
emotional distress and glycemic control in the 185 persons with type (PAID total), 0.162 (RD subscale), 0.215 (EB subscale), 0.579 (ID subscale) and 0.060
1 diabetes was addressed using regression analysis with restricted (PD subscale).
cubic splines with four knots (requiring 3 degrees of freedom) to
incorporate possibly nonlinear relationships [37]. Because Fisher et al. Discussion
[18] used quadratic regression analyses in their study, supplementary
analyses of nonlinearity were performed with quadratic regression This study appears to be the first to demonstrate that among adults
analysis to assure that potential differences between our study and with type 1 diabetes, depression, anxiety, and overall well-being were
the Fisher et al. study were not a consequence of methodological not significantly related with glycemic control but there were
differences. As done in the linear analysis, the DDS total (1–6 scale) significant associations between diabetes-specific emotional distress
and PAID total were analyzed in separate regression models. Even and HbA1c. The Diabetes Distress Scale total score was more strongly
though the four DDS subscales were integrated into one model in the associated than the PAID total score, and regimen-related DDS subscale
linear analysis, the lower number of degrees of freedom in the nonlinear showed the strongest relationship with HbA1c (regression coefficients
analysis made it necessary to analyze the four DDS subscales in four and R2). There was no significant nonlinearity in the relationship
separate models when testing for nonlinearity. The nonlinear analyses between diabetes-specific emotional distress and HbA1c.
were adjusted for the WHO-5 score, HADS scores, sex, age, education, Fisher et al. [27] and Gonzales et al. [17] expressed concern that
and late complications. diabetes-specific emotional distress might be interpreted as depression
Significance was defined as P b .05 in all analyses. The linear regres- and addressed with care strategies based on the depression literature.
sions were analyzed with SPSS version 19/20 (IBM, Armonk, NY), the The associations of depression or diabetes-specific emotional distress
nonlinear analysis were performed in the R (The R Foundation for Statis- with glycemic control have been examined predominantly in persons
tical Computing, Vienna, Austria) package rms and multiple imputation with type 2 diabetes, where Fisher et al. [25,26] found a significant
in the R package mice. relationship of glycemic control with diabetes-specific emotional
 R.B. Strandberg et al. / Journal of Psychosomatic Research 77 (2014) 174–179 177

Table 2
Associations of diabetes-specific emotional distress, anxiety, depression and overall well-being with HbA1c.

Bivariate regression Regression with partial adjustmentc Regression with full adjustmentd

Coefficienta CI P Coefficienta CI P Coefficienta CI Coefficientb P

PAID total 0.020 0.008–0.032 .001 0.017 0.004–0.029 .009 0.021 0.006–0.036 0.244 .007
WHO-5 −0.003 −0.021–0.015 −0.035 .759
HADS-A −0.028 −0.115–0.059 −0.064 .527
HADS-D −0.032 −0.133–0.070 −0.069 .540
R2/adjusted R2 5.4%/4.9% 14.3%/11% 15.0%/10.1%
DDS total 0.033 0.019–0.047 b.001 0.030 0.015–0.044 b.001 0.038 0.021–0.055 0.374 b.001
WHO-5 0.001 −0.017–0.018 0.011 .920
HADS-A −0.048 −0.133–0.036 −0.112 .261
HADS-D −0.025 −0.123–0.073 −0.054 .618
R2/adjusted R2 10.8%/10.3% 18.6%/15.4% 20.3%/15.7%
DDS EB 0.014 4.2%/3.7%e 0.004–0.025 .005 0.011 13.2%/9.7%e 0.001–0.021 .036 −0.013 −0.027–0 .001 −0.188 .067
DDS RD 0.039 26.9%/26.5%e 0.030–0.049 b.001 0.038 32.6%/29.9%e 0.028–0.048 b.001 0.056 0.043–0.069 0.734 b.001
DDS ID 0.010 1.2%/0.6%e −0.003–0.023 .140 0.009 11.9%/8.4%e −0.004–0.022 .191 −0.010 −0.024−0.004 −0.108 .161
DDS PD 0.007 0.4%/–0.1%e −0.009–0.024 .370 0.006 11.3%/7.8%e −0.010–0.022 .431 0.000 −0.017–0.017 −0.002 .976
WHO-5 −0.010 1.4%/0.9%e −0.021–0.002 .106 −0.005 11.3%/7.8%e −0.017–0.008 .455 0.007 −0.009–0.023 0.085 .392
HADS-A 0.027 0.4%/−0.1%e −0.036–0.091 .392 0.014 11.1%/7.6%e −0.051–0.079 .673 −0.018 −0.094–0.058 −0.042 .635
HADS-D 0.014 0.1%/−0.4%e −0.052–0.081 .672 0.010 11.0%/7.5%e −0.058–0.079 .772 −0.012 −0.099–0.075 −0.026 .785
R2/adjusted R2 38.6%/34%
a
Unstandardized regression coefficients.
b
Standardized regression coefficients.
c
PAID total, DDS total, EB, RD, ID, PD, HADS-A, HADS-D and WHO-5 in separate regression models, each adjusted by age, sex, education and late complications.
d
Multiple regression analysis of PAID total, DDS total in separate models, adjusted for age, sex, education, late complications, HADS and WHO-5, then the four DDS subscales in one
separate model adjusted with the similar control variables.
e
R2 and adjusted R2 in the bivariate regression and the partly adjusted regression.

distress, but not with depression. The findings of our study with type 1 showed that positive affect was a unique predictor of mortality in
diabetes showing similar relationships support that diabetes-specific persons with diabetes, and argued for the value of addressing positive
emotional distress and depression should be recognized as different affect in clinical consultation. Indeed the literature review of Robertson
conditions in clinical consultation in type 1 diabetes [11]. Fisher et al. et al. [44] suggested that positive emotional health might facilitate bet-
[40] proposed that emotional distress is a core construct underlying ter self-management and improved health outcomes. A systematic re-
diabetes-specific emotional distress and depression (from depressive view of qualitative research studies of factors influencing ability to
symptoms to major depressive disorders), and emphasized that health self-management in type 1 and type 2 diabetes concluded that the
care providers should acknowledge the difference between the severity wider picture beyond the physical manifestation of diabetes must be
and content of emotional distress in clinical consultations. Our study taken into consideration [45]. Nevertheless, results from our study
cannot determine whether depression and diabetes-specific emotional suggest that a measure of overall well-being is too generic to reveal an
distress are different entities, or whether the differing relationships of association with the particular outcome of glycemic control, and that
glycemic control with depression or diabetes-specific emotional well-being must integrate some disease-specific elements if relation-
distress are due to differences in severity or content of emotional ships to specific biomedical outcomes are to be discovered. A recent
problems, but both possibilities are worth considering. study of severe hypoglycemia and psychological well-being supports
Aikens [11] suggested that diabetes-specific emotional distress this interpretation [46]. The authors found that neither generic overall
(measured with the PAID), rather than depression, might derive from well-being, nor diabetes-specific emotional distress (measured by the
activities strongly related to diabetes and its treatment, and disrupt PAID) was significantly related to hypoglycemia, whereas diabetes-
self-care activities that are directly linked to the disease, whereas specific positive well-being was significantly related to hypoglycemia.
depression might disrupt more lifestyle-oriented behaviors. Using the Moreover, Snoek et al. [47] found that individual care strategies for
DDS creates an opportunity to examine the role of specific domains of people with type 1 and type 2 diabetes improved the scale scores of
diabetes-related emotional distress. Regimen-related distress was diabetes-specific emotional distress but not overall well-being.
the only distress domain associated with HbA1c in the fully adjusted A significant nonlinear relationship between diabetes-specific
analysis. In the bivariate regression and the partly adjusted regression emotional distress and glycemic control was not found in this study,
analysis, the EB subscale was also significantly related to HbA1c although similar to results shown in the recent cross-sectional study of Joensen
not when controlling for the RD subscale and the other indicators. et al. [48]. That such a relationship was not identified in these studies
Therefore, we suggest that it may be distress related to the self-care of adults with type 1 diabetes, in conjunction with its presence in adults
demands of the treatment regimen that actually drives the relationship with type 2 diabetes [18], might indicate that emotional problems have
between diabetes-specific emotional distress and HbA1c. Hessler et al. different implications for persons with type 1 and type 2 diabetes. The
[41] showed that reductions in regimen distress were associated with findings of a significant relationship between distress and glycemic
improved glycemic control over time for persons with type 2 diabetes, control, and the lack of significant nonlinearity, suggest that interventions
and emphasized the importance of addressing regimen distress as part addressing diabetes distress might be applied at any non-zero level of
of diabetes care. Reddy et al. [42] suggested that the PAID might be diabetes distress, although this should be further investigated in larger
useful as a screening tool of diabetes-specific emotional distress in samples.
clinical consultation, but results from our study suggest that the DDS There were some limitations to this study. First, because it was a
might be more appropriate to capture distress regarding the cross-sectional study, no inferences about a causal effect between
self-management behaviors of diabetes, as the PAID does not provide diabetes-related emotional distress and HbA1c can be drawn. We do
a validated measure of regimen distress which seems to be the most not know whether the association between diabetes-specific emotional
important component of diabetes-specific distress. distress and HbA1c is a direct causal relationship, nor which direction a
In our study, the positively worded measure of overall well-being causal relationship might take, nor whether there might be an underly-
was not significantly related to glycemic control. Moskowitz et al. [43] ing mechanism that influences both diabetes-specific emotional
178 R.B. Strandberg et al. / Journal of Psychosomatic Research 77 (2014) 174–179

distress and HbA1c other than those controlled for in this study. Longitu- [7] The Diabetes Control and Complications Trial Research Group. The effect of intensive
treatment of diabetes on the development and progression of long-term complications
dinal studies of the relationship between glycemic control and diabetes- in insulin-dependent diabetes mellitus. N Engl J Med 1993;329:977–86.
specific emotional distress in type 1 diabetes are warranted. As the [8] American Diabetes Association. Standards of medical care in diabetes—2014. Diabetes
regimen-related emotional distress seems to be an active ingredient in Care 2014;37:S14–80.
[9] Cooper JG, Claudi T, Thordarson HB, Løvaas KF, Carlsen S, Sandberg S, Thue G. Treat-
the relationship between diabetes-specific distress and HbA1c, potential ment of type 1 diabetes in the specialist health service — data from the Norwegian
underlying mechanisms of this association (especially regimen Diabetes Register for Adults. Tidsskr Nor Laegeforen 2013;133:2257–62.
adherence behavior) need to be further examined in future studies. [10] Ali MK, Bullard KM, Saaddine JB, Cowie CC, Imperatore G, Gregg EW. Achievement of
goals in U.S. Diabetes Care 1999–2010. N Engl J Med 2013;368:1613–24.
A second limitation is that all the information about symptoms of [11] Aikens JE. Prospective associations between emotional distress and poor outcomes
depression and anxiety were self-reported because no diagnostic in type 2 diabetes. Diabetes Care 2012;35:2472–8.
information was available, and few respondents reported the highest [12] Peyrot M, McMurry JF, Krueger DF. A biopsychosocial model of glycemic control in
diabetes: stress, coping and regimen adherence. J Health Soc Behav 1999;40:141–58.
scores on the HADS. The study may therefore have underestimated
[13] Nicolucci A, Kovacs Burns K, Holt RIG, Comaschi M, Hermanns N, Ishii H, et al. Diabe-
the potential impact of these factors among persons with more severe tes Attitudes, Wishes and Needs second study (DAWN2TM). Cross-national
depression and anxiety. In addition, the HADS has been criticized as a benchmarking of diabetes-related psychosocial outcomes for people with diabetes.
measure of depression [49,50]. However, a study investigating the On behalf of the DAWN2 Study Group, Diabet Med 2013;30:767–77.
[14] Holt RIG, Nicolucci A, Kovacs Burns K, Escalante M, Forbes A, Hermanns N, et al.
cross-sectional and longitudinal relationship between depression and Diabetes Attitudes, Wishes and Needs second study (DAWN2TM). Cross-national
glycemic control using a diagnostic interview (CIDI) based on the comparisons on barriers and resources for optimal care–health care professional
DSM-IV criteria also did not find a significant relationship between perspective. On behalf of the DAWN2 Study Group, Diabet Med 2013;30:789–98.
[15] Lloyd CE, Pambianco G, Orchard TJ. Does diabetes-related distress explain the
depression and glycemic control [26]. presence of depressive symptoms and/or poor self-care in individuals with type 1
We have shown that glycemic control in adults with type 1 diabetes diabetes? Diabet Med 2010;27:234–7.
was not significantly associated with depression, anxiety, and overall [16] Pouwer F, Beekman ATF, Lubach C, Snoek FJ. Nurses' recognition and registration of
depression, anxiety and diabetes-specific emotional problems in outpatients with
well-being, but was significantly associated with diabetes-specific diabetes mellitus. Patient Educ Couns 2006;60:235–40.
emotional distress, especially that regarding the treatment regimen. [17] Gonzales JS, Fisher L, Polonsky WH. Depression in diabetes: have we been missing
Gonzales et al. [17] argue that the recognition of the content of something important? Diabetes Care 2011;34:236–9.
[18] Fisher L, Hessler DM, Polonsky WH, Mullan J. When is diabetes distress clinically
diabetes-specific emotional distress in clinical consultations might meaningful? Establishing cut points for the diabetes distress scale. Diabetes Care
require only a small shift in the perspective of the clinician. A recent 2012;35:259–64.
systematic review of emotional health and diabetes self-care emphasized [19] Polonsky WH, Anderson BJ, Lohrer PA, Welch G, Jacobsen AM, Aponte JE, et al.
Assessment of diabetes-related distress. Diabetes Care 1995;18:754–60.
that talking about the persons' thoughts and understanding of the disease
[20] Polonsky WH, Fisher L, Earles J, Dudl RJ, Lees J, Mullan J, et al. Assessing psychosocial
in clinical consultation might make it easier for health care providers to distress in diabetes: development of the diabetes distress scale. Diabetes Care
recognize those in poor emotional health [51]. Yet Beverly et al. [52] 2005;28:626–31.
showed that 30% of their sample of persons with type 1 and type 2 [21] Watson D, Clark LA, Weber K, Assenheimer JS, Strauss ME, McCormick RA. Testing a
tripartite model: I. Evaluating the convergent and discriminant validity of anxiety
diabetes were reluctant to discuss self-care in clinical consultation, and and depression symptom scales. J Abnorm Psychol 1995;104:3–14.
that reluctant persons reported less frequent self-care, higher diabetes- [22] Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr
specific emotional distress and more depressive- and anxiety symptoms; Scand 1983;67:361–70.
[23] Gonzales JS, Delahanty LM, Safren SA, Meigs JB, Grant RW. Differentiating symptoms
these findings illustrate the complexity of the interaction between the of depression from diabetes-specific distress: relationships with self-care in type 2
clinician and the persons with diabetes in clinical consultation. Our diabetes. Diabetologia 2008;51:1822–5.
study highlights that addressing distress related to the disease during [24] Hermanns N, Kulzer B, Krichbaum M, Kubiak T, Haak T. How to screen for depression
and emotional problems in patients with diabetes: comparison of screening charac-
clinical consultation would enable greater insight into whether such teristics of depression questionnaires, measurement of diabetes-specific emotional
distress is apparent, and what specifically this distress might be constituted problems and standard clinical assessment. Diabetologia 2006;49:469–77.
of for the individual person. In particular, addressing distress related to [25] Fisher L, Glasgow RE, Strycker LA. The relationship between diabetes distress and
clinical depression with glycemic control among patients with type 2 diabetes. Dia-
the treatment regimen and self-care demands might give health care betes Care 2010;33:1034–6.
providers information necessary to assist the person in bettering their [26] Fisher L, Mullan JT, Arean P, Glasgow RE, Hessler D, Masharani U. Diabetes distress
diabetes self-management. If change in glycemic control is targeted, but not clinical depression or depressive symptoms is associated with glycemic con-
trol in both cross-sectional and longitudinal analyses. Diabetes Care 2010;33:23–8.
focusing on diabetes-specific emotional distress may yield greater
[27] Fisher L, Skaff MM, Mullan JT, Arean P, Mohr D, Masharani U, et al. Clinical depression
improvement than focusing solely on attaining overall well-being. versus distress among patients with type 2 diabetes. Diabetes Care 2007;30:542–8.
[28] Welch G, Weinger K, Anderson B, Polonsky WH. Responsiveness of the Problem
Areas in Diabetes (PAID) questionnaire. Diabet Med 2003;20:69–72.
Conflict of interests [29] Graue M, Haugstvedt A, Wentzel-Larsen T, Iversen MM, Karlsen B, Rokne B.
Diabetes-related emotional distress in adults: reliability and validity of the Norwegian
The authors declare that they have no conflict of interest. versions of the Problem Areas in Diabetes Scale (PAID) and the Diabetes Distress Scale
(DDS). Int J Nurs Stud 2012;49:174–82.
[30] Snoek FJ, Pouwer F, Welch GW, Polonsky WH. Diabetes-related emotional distress in
Acknowledgments Dutch and U.S. diabetic patients: cross-cultural validity of the problem areas in
diabetes scale. Diabetes Care 2000;23:1305–9.
[31] Snaith RP. The hospital anxiety and depression scale. Health Qual Life Outcomes
The authors would like to thank the respondents in this study, as 2003;1.
well as the nurses and physicians at the university hospital endocrinology [32] Bjelland I, Dahl AA, Haug TT, Neckelmann D. The validity of the hospital anxiety and
unit, who participated in the data collection. depression scale. An updated review. J Psychosom Res 2002;52:69–77.
[33] Mykletun A, Stordal E, Dahl AA. Hospital Anxiety and Depression (HAD) scale: factor
structure, item analyses and internal consistency in a large population. Br J Psychiatry
References 2001;179:540–4.
[34] Bech P, Olsen LR, Kjoller M, Rasmussen NK. Measuring well-being rather than the
[1] Atkinson MA, Maclaren NK. The pathogenesis of insulin-dependent diabetes absence of distress symptoms: a comparison of the SF-36 Mental Health subscale
mellitus. N Engl J Med 1994;331:1428–36. and the WHO-Five Well-being Scale. Int J Methods Psychiatr Res 2003;12:85–91.
[2] JA Bluestone, K Herold, G Eisenbarth, Genetics, pathogenesis and clinical interventions [35] de Wit M, Pouwer F, Gemke BJ, Delemarre-van de Waal HA, Snoek FJ. Validation of
in type 1 diabetes, Nature; 464: 1293-1300. the WHO-5 Well-Being Index in adolescents with type 1 diabetes. Diabetes Care
[3] Atkinson MA, Eisenbarth GS, Michels AW. Type 1 diabetes. Lancet 2014;383:69–82. 2007;30:2003–6.
[4] Wild S, Roglig G, Green A, Sicree R, King H. Global prevalence of diabetes. Estimates [36] Fayers PM, Machin D. Quality of life: assessment, analysis and interpretation. West
for the year 2000 and projections for 2013. Diabetes Care 2004;27:1047–53. Sussex: John Wiley & Sons Ltd; 2000.
[5] Rubin RR, Peyrot M. Quality of life and diabetes. Diabetes Metab Res Rev [37] Harrell FE. Regression modeling strategies. New York: Springer; 2001.
1999;15:205–18. [38] Peyrot M, Rubin RR, Polonsky WH. Diabetes distress and its association with clinical
[6] American Diabetes Association. Economic costs of diabetes in the U.S. in 2012. outcomes in patients with type 2 diabetes treated with pramlintide as an adjunct to
Diabetes Care 2013;36:1033–46. insulin therapy. Diabetes Technol Ther 2008;10:461–6.
 R.B. Strandberg et al. / Journal of Psychosomatic Research 77 (2014) 174–179 179

[39] Cohen J. Statistical power analysis for the behavioral sciences. London: Academic adults with type 1 diabetes attending specialist tertiary clinics. Diabetes Res Clin
Press; 1969. Pract 2014;103:430–6.
[40] Fisher L, Gonzalez JS, Polonsky WH. The confusing tale of depression and distress in [47] Snoek FJ, Kersch NYA, Eldrup E, Harman-Boehm I, Hermanns N, Kokoszka A, et al.
patients with diabetes: a call for greater clarity and precision. Diabet Med Monitoring of Individual Needs in Diabetes (MIND)-2: follow-up data from the
2014;31:764–72. cross-national Diabetes Attitudes, Wishes, and Needs (DAWN) MIND study.
[41] Hessler D, Fisher L, Glasgow RE, Strycker LA, Dickinson LM, Arean PA, et al. Reductions in Diabetes Care 2012;35:2128–32.
regimen distress are associated with improved management and glycemic control over [48] Joensen LE, Tapager I, Willaing I. Diabetes distress in type 1 diabetes—a new
time. Diabetes Care 2014;37:617–24. measurement fit for purpose. Diabet Med 2013;30:1132–9.
[42] Reddy J, Wilhelm K, Campbell L. Putting PAID to diabetes-related distress: the potential [49] Doyle F. Letter to Editor. Why the HADS is still important: reply to Coyne & van
utility of the problem areas in diabetes (PAID) scale in patients with diabetes. Sondern. J Psychosom Res 2012;73:74.
Psychosomatics 2013;54:44–51. [50] Coyne JC, van Sonderen E. No further research needed: abandoning the Hospital and
[43] Moskowitz JT, Epel ES, Acree M. Positive affect uniquely predicts lower risk of Anxiety depression Scale (HADS). J Psychosom Res 2012;72:173–4.
mortality in people with diabetes. Health Psychol 2008;27:S73–82. [51] Hudson JL, Bundy C, Coventry PA, Dickens C. Exploring the relationship between
[44] Robertson SM, Stanley MA, Cully JA, Naik AD. Positive emotional health and diabetes cognitive illness representations and poor emotional health and their combined
care: concepts, measurement, and clinical implications. Psychosomatics 2012; association with diabetes self-care. A systematic review with meta-analysis. J
53:1–12. Psychosom Res 2014;76:265–74.
[45] Wilkinson A, Whitehead L, Ritchie L. Factors influencing the ability to self-manage [52] Beverly EA, Ganda OP, Ritholz MD, Lee Y, Brooks KM, Lewis-Schroeder NF, et al. Look
diabetes for adults living with type 1 or 2 diabetes. Int J Nurs Stud 2014;51:111–22. who's (not) talking: diabetic patients' willingness to discuss self-care with physicians.
[46] Hendrieckx C, Halliday JA, Bowden JP, Colman PG, Cohen N, Jenkins A, et al. Severe Diabetes Care 2012;35:1466–72.
hypoglycaemia and its association with psychological well-being in Australian