See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/7999072

Assessing Psychosocial Distress in Diabetes: Development of the Diabetes

Distress Scale

Article  in  Diabetes Care · April 2005

DOI: 10.2337/diacare.28.3.626 · Source: PubMed

CITATIONS READS
1,073 3,512

7 authors, including:

William H Polonsky robert James Dudl

University of California, San Diego Kaiser Permanente
113 PUBLICATIONS   10,370 CITATIONS    32 PUBLICATIONS   2,652 CITATIONS   

SEE PROFILE SEE PROFILE

Richard Alan Jackson

Grass Roots Diabetes
84 PUBLICATIONS   9,052 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Preventing CVD with a bundle View project

Right Care Initiative View project

All content following this page was uploaded by Richard Alan Jackson on 17 June 2014.

The user has requested enhancement of the downloaded file.

Epidemiology/Health Services/Psychosocial Research
O R I G I N A L A R T I C L E

Assessing Psychosocial Distress in

Diabetes
Development of the Diabetes Distress Scale
WILLIAM H. POLONSKY, PHD, CDE1 JOEL LEES, MD5

L
iving with diabetes can be tough. In
LAWRENCE FISHER, PHD2 JOSEPH MULLAN, PHD2 the face of a complex, demanding,
JAY EARLES, PSYD3 RICHARD A. JACKSON, MD6 and often confusing set of self-care
R. JAMES DUDL, MD4 directives, patients may become frus-
trated, angry, overwhelmed, and/or dis-
couraged. Diabetes-related conflict with
loved ones may develop, and relation-
ships with health care providers may be-
OBJECTIVE — The purpose of this study was to describe the development of the Diabetes
Distress Scale (DDS), a new instrument for the assessment of diabetes-related emotional distress,
come strained. The risk of depression is
based on four independent patient samples. elevated (1,2). As a result, motivation for
self-care may be impaired. To investigate
RESEARCH DESIGN AND METHODS — In consultation with patients and profes- the nature and breadth of such distress, a
sionals from multiple disciplines, a preliminary scale of 28 items was developed, based a priori number of self-report instruments have
on four distress-related domains: emotional burden subscale, physician-related distress sub- been developed, including the ATT39
scale, regimen-related distress subscale, and diabetes-related interpersonal distress. The new (3), Questionnaire on Stress in Patients
instrument was included in a larger battery of questionnaires used in diabetes studies at four with Diabetes-Revised (QSD-R) (4), and
diverse sites: waiting room at a primary care clinic (n ⫽ 200), waiting room at a diabetes specialty Problem Areas in Diabetes scale (PAID)
clinic (n ⫽ 179), a diabetes management study program (n ⫽ 167), and an ongoing diabetes
management program (n ⫽ 158). (5). These measures aim to tap the range
of emotional responses to diabetes and to
RESULTS — Exploratory factor analyses revealed four factors consistent across sites (involv- serve as screening measures for clinical
ing 17 of the 28 items) that matched the critical content domains identified earlier. The corre- and research use. The PAID has been the
lation between the 28-item and 17-item scales was very high (r ⫽ 0.99). The mean correlation most widely used of the measures and has
between the 17-item total score (DDS) and the four subscales was high (r ⫽ 0.82), but the pattern been recently translated into several other
of interscale correlations suggested that the subscales, although not totally independent, tapped languages (6 – 8). PAID scores have been
into relatively different areas of diabetes-related distress. Internal reliability of the DDS and the linked to diabetes self-care behaviors
four subscales was adequate (␣ ⬎ 0.87), and validity coefficients yielded significant linkages with
(5,6) and glycemic control (2,5–9) and
the Center for Epidemiological Studies Depression Scale, meal planning, exercise, and total
cholesterol. Insulin users evidenced the highest mean DDS total scores, whereas diet-controlled are associated with general emotional dis-
subjects displayed the lowest scores (P ⬍ 0.001). tress (5), perceived burden of diabetes
(8), diabetes-related health beliefs (10),
CONCLUSIONS — The DDS has a consistent, generalizable factor structure and good in- diabetes coping (10), and marital adjust-
ternal reliability and validity across four different clinical sites. The new instrument may serve as ment (11,12). The instrument is respon-
a valuable measure of diabetes-related emotional distress for use in research and clinical practice. sive to change (13) and is a useful
measure of several aspects of diabetes-
Diabetes Care 28:626 – 631, 2005 related quality of life (14,15).
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
All of these measures, including the
PAID, have some limitations. Some criti-
From the 1Department of Psychiatry, University of California, San Diego, California; the 2Department of
Family and Community Medicine, University of California, San Francisco, California; the 3Tripler Army
cal areas of interest are covered either too
Medical Center, Honolulu, Hawaii; 4Kaiser Permanente, San Diego, California; the 5Naval Medical Center, briefly or not at all (e.g., in the PAID, only
San Diego, California; and the 6Joslin Diabetes Center, Harvard Medical School, Boston, Massachusetts. one item addresses patients’ feelings
Address correspondence and reprint requests to William H. Polonsky, PhD, CDE, P.O. Box 2148, Del about their health care provider). Anec-
Mar, CA 92014. E-mail: [email protected].
Received for publication 21 September 2004 and accepted in revised form 23 November 2004.
dotal reports suggest that patients may be
Abbreviations: CESD, Center for Epidemiological Studies Depression Scale; DDS, Diabetes Distress confused over the exact meaning of some
Scale; EB, emotional burden subscale; ID, diabetes-related interpersonal distress subscale; JOS, Joslin Dia- items (e.g., in the PAID, “not having clear
betes Center’s Diabetes Outpatient Intensive Treatment program; KP, Kaiser Permanente Diabetes Clinic; and concrete goals for your diabetes care,”
NMC, Naval Medical Center Internal Medicine Clinic; PAID, Problem Areas in Diabetes scale; PD, physician-
related distress subscale; QSD-R, Questionnaire on Stress in Patients with Diabetes-Revised; RD, regimen-
and in the QSD-R, “I suffer from irritabil-
related distress subscale; SDSCA, Summary of Diabetes Self-Care Activities; SMBG, self-monitoring of blood ity”). Finally, there is growing interest in a
glucose; TAMC, Tripler Army Medical Center. brief instrument that can, for both clinical
A table elsewhere in this issue shows conventional and Système International (SI) units and conversion and research purposes, assess (and per-
factors for many substances. haps distinguish among) different types
© 2005 by the American Diabetes Association.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby of diabetes-related emotional distress. Al-
marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. though the ATT39 and QSD-R have es-

626 DIABETES CARE, VOLUME 28, NUMBER 3, MARCH 2005

 Polonsky and Associates

tablished subscales, they are not brief (45 ria: ⱖ18 years old, a diagnosis of type 1 or proached at the start of the program and
and 39 items, respectively). In contrast, 2 diabetes, and no severe visual or cogni- asked to complete a brief questionnaire
the PAID is brief (20 items), but subscales tive limitations. battery that assessed psychological func-
have not been developed. Given these tioning and selected clinical variables. Of
concerns, we decided to develop a new San Diego 158 patients approached, 137 completed
measure, the Diabetes Distress Scale Patients at two San Diego sites, the Kaiser the survey (87% of patients approached).
(DDS), which builds on the strengths of Permanente Diabetes Clinic (KP) and the No information was available regard-
previously developed instruments and Naval Medical Center Internal Medicine ing those who refused participation at the
addresses at least some of their limita- Clinic (NMC), were approached immedi- four sites, so differences between those
tions. This study presents initial data on ately before their medical visit and asked who did and did not complete survey
the factor structure of the DDS based on to complete a battery of self-report mea- forms could not be ascertained.
four independent patient samples and on sures, requiring ⬃15–20 min, which in-
its reliability and validity. cluded instruments assessing psycho-
logical functioning, self-care behaviors, Psychological and clinical measures
RESEARCH DESIGN AND and clinical variables. The study was de- In addition to the DDS, questionnaire bat-
METHODS — P a t i e n t s , d i a b e t e s signed to ascertain the prevalence and se- teries at all sites included the Center for
nurse specialists, dietitians, diabetolo- verity of diabetes-related emotional and Epidemiological Studies Depression Scale
gists, and diabetes-knowledgeable psy- behavioral dysfunction at these two sites, (CESD), a widely used, 20-item question-
chologists from around the country were as a prelude to the development of a com- naire designed to assess the major symp-
asked to review the items previously de- prehensive plan for intervention. Of 233 toms of depression (17). A reliable and
veloped for the PAID, QSD-R, and ATT39 eligible patients approached at KP, 41 re- well-validated instrument, the CESD’s
and to suggest new or similar items for a fused, 2 did not return the survey, and 11 targeted symptoms include depressed
new instrument. From this pool of items, returned incomplete surveys, resulting in mood, changes in appetite and sleep, low
an early draft of the new scale containing 179 completed surveys (76% of the sam- energy, feelings of hopelessness, low self-
50 items was developed and pilot tested ple). At NMC, of 275 eligible patients ap- esteem, and loneliness. Respondents are
with several small groups of patients. proached, 67 refused and 8 returned asked to consider the presence and duration
Feedback from these groups led to dele- incomplete surveys, resulting in 200 of each item/symptom over the past week
tion of items that were vague, difficult for completed surveys (72% of the sample).
and to rate each along a 4-point scale from 0
patients to comprehend, or merely dupli-
(rarely or never) to 3 (most or all of the
cative, resulting in a scale of 28 items. Honolulu
time). Possible scores range from 0 to 60. A
These included seven items from each of Patients at Tripler Army Medical Center
score ⱖ16 is the most common cutoff
four domains central to diabetes-related (TAMC) were invited to join an interven-
point, indicating a “likely depression” (17).
emotional distress, created a priori based tion study examining an intensive group
However, this cannot be equated to a clini-
on focus group discussions: emotional education and skills training experience
burden subscale (EB) (e.g., “feeling over- combined with medical management cal diagnosis of depression.
whelmed by the demands of living with (16). In addition to the eligibility criteria All sites (except for JOS) included the
diabetes”), physician-related distress sub- listed above, patients were also required Summary of Diabetes Self-Care Activities
scale (PD) (e.g., “feeling that my doctor to be in poor glycemic control (most re- (SDSCA), a 12-item, self-report scale that
doesn’t take my concerns seriously cent HbA 1c ⱖ8.5%). Patients in the assesses the frequency of blood glucose
enough”), regimen-related distress sub- TAMC database who met eligibility crite- monitoring, exercise and dietary behav-
scale (RD) (e.g., “feeling that I am not ria were sent a letter describing the study, iors, and medication usage over the pre-
sticking closely enough to a good meal followed several days later by a phone call vious 7 days (18). Adequate reliability
plan”), and diabetes-related interpersonal from the project’s nurse recruiter. Patients and validity have been demonstrated
distress subscale (ID) (e.g., “feeling that were also recruited through mailings to (19). Attention was focused on a subset of
my friends/family don’t appreciate how TAMC physicians. Of the 224 patients the items targeting self-monitoring of
difficult living with diabetes can be”). Fol- contacted who met eligibility require- blood glucose (SMBG) (“how often you
lowing a format similar to those devel- ments, 196 (88%) agreed to join the study checked glucose levels”), exercise (“how
oped for the PAID and QSD-R, patients and to complete a baseline questionnaire often you participated in at least 20 min of
rated the degree to which each item was requiring ⬃30 min, assessing psycholog- physical exercise”), and dietary behavior
currently problematic for them on a ical functioning, self-care behaviors, and (“how often you followed your recom-
6-point Likert scale, from 1 (no problem) clinical variables. Completed surveys mended meal plan”). SMBG response al-
to 6 (serious problem). were obtained from 167 patients (75% of ternatives were “every day,” “most days,”
The new instrument was included as patients approached). “some days,” and “none of the days.” Ex-
part of larger studies of patients with dia- ercise response alternatives were 0 –7
betes at four clinical sites. Human subject Boston days. Dietary response alternatives were
approval was received for each of these Eligible patients enrolled at Joslin Diabe- “always,” “usually,” “sometimes,” “rarely,”
studies at their respective institutions, tes Center’s Diabetes Outpatient Intensive and “never.” Responses to the exercise
and all patients provided informed con- Treatment program (JOS), a 3.5-day item were reverse scored, so that higher
sent. At all sites, patients were deemed group education course integrated with scores on all items reflect better self-
eligible if they met broad inclusion crite- intensive medical management, were ap- management.

DIABETES CARE, VOLUME 28, NUMBER 3, MARCH 2005 627

Psychosocial distress in diabetes

Table 1—Baseline characteristics from general care settings where their di-
abetes was managed. In contrast, TAMC
NMC KP TAMC JOS Total P and JOS subjects came from more highly
specialized programs, which suggested
n 200 179 167 137 683 that these individuals may have been
Male (%) 44.5 57.5 53.9 54.7 52.3 more ill and/or were having more trouble
Age (years) 62.0 59.3 50.9 50.6 56.3 ⬍0.001 managing their diabetes. Therefore, it is
Ethnic background (%)* ⬍0.001 noteworthy that significant site differ-
Non-Hispanic white 54.5 68.2 34.1 52.7 ences occurred (Table 1). Patients from
African American 13.0 9.5 17.4 13.2 TAMC and JOS were significantly
Hispanic 5.0 11.2 4.8 7.0 younger than patients from NMC and KP
Asian/Pacific Islander 22.0 5.6 31.7 19.6 (P ⬍ 0.001). JOS patients reported higher
High school graduate (%) 85.6 88.7 92.2 88.7 depression scores than patients from
Diabetes duration (years) 12.5 15.3 10.5 12.8 ⬍0.001 the other three sites (P ⬍ 0.001). Patients
Medication use (%) ⬍0.05 in the TAMC, in which poor glycemic
Insulin 47.5 58.7 44.9 50.4 control was the critical inclusion crite-
Oral hypoglycemic agents only 42.5 35.8 49.7 42.5 rion, displayed the highest standardized
Diet controlled (%) 10.0 5.6 5.4 7.1 HbA 1c levels and the poorest self-
HbA1c (%) 8.7 7.6 10.4 8.2 8.8 ⬍0.001 reported adherence to meal planning rec-
CESD 15.4 13.9 14.3 20.2 15.7 ⬍0.001 ommendations (P ⬍ 0.001). The
DDS (total score) 31.7 36.1 39.3 50.6 38.5 ⬍0.001 variability of the sites provided a broad
Self-care range of patients on which to assess the
Meal planning† 67.7 64.4 36.7 — 56.9 ⬍0.001 structure, reliability, and validity of the
Exercise‡ 2.6 2.8 3.0 — 2.8 DDS.
SMBG§ 70.7 77.7 69.9 — 72.7
*Percentages do not sum to 100% because smaller ethnic groups (Hispanics and African Americans) were Exploratory factor analyses
excluded from these analyses; †those following recommendations “always” or “usually” during the past 7 The four within-site exploratory factor
days; ‡number of days of exercise (20 min or more) during the past 7 days; §those reporting monitoring analyses of the 28 items yielded between
“every day” or “most days.”
four and eight factors, and in each case the
scree plots suggested four or five viable fac-
Metabolic variables Statistical analyses tors. A review of the analyses for each site
At NMC, TAMC, and JOS, HbA1c was An exploratory factor analysis was per- suggested that four factors were most con-
measured by high-performance liquid formed on the 28-item scale for each site sistent and interpretable, with the remain-
chromatography; the consensus normal separately using principal factor analysis ing factors comprising single items, items
range was 4.0 – 6.0%. At the KP site, with Promax rotation. Cronbach’s ␣ was that accounted for a low percentage of vari-
HbA1c was measured by the Roche/BMC used to assess the internal consistency of ance, or uninterpretable item combina-
method; the normal range was 4.2– 6.7%. the total scale and the subscales, and Pear- tions. In addition, the interitem correlations
At TAMC, all subjects completed HbA1c son correlations compared the DDS total for the first four factors across all four sites
testing within 90 days before survey com- scale and each subscale with the CESD, were highly similar. We therefore com-
pletion. At JOS, all subjects completed SDSCA, and metabolic variables, which bined the sample and ran an exploratory
HbA1c testing on the day of survey admin- were used as validity coefficients. factor analysis, this time extracting four fac-
istration. To make analyses comparable tors only. The pattern matrix for this analy-
across sites, only HbA1c results from RESULTS — The clinical and demo- sis is presented in Table 2.
blood drawn within 90 days of survey graphic characteristics of the four samples A review of item content of each fac-
completion at NMC (n ⫽ 136) and KP are presented in Table 1. Mean age was tor suggested that the factors matched the
(n ⫽ 125) were examined. To compare 56.3 years, and males comprised 52.3% critical content domains proposed earlier:
values directly across sites, HbA1c results of the total sample. The majority (83.3%) factor 1 reflected EB, factor 2 encom-
were standardized such that they re- had type 2 diabetes, and the mean HbA1c passed PD, factor 3 indicated RD, and fac-
flected the percentage above (or below) was 8.8%. In the three samples for which tor 4 reflected ID. To create a brief,
the Diabetes Control and Complications further demographic data were available, concise scale and set of subscales with a
Trial upper limit of the normal range the majority of patients (50.4%) were us- relatively equal number of items, we re-
(6.0%). ing insulin: 42.5% were receiving oral hy- viewed items with good factor loadings
At TAMC and KP, the most recent poglycemic agents only, and 7.1% were and retained those items with relatively
lipid profiles were obtained from clinical managed by diet only. Non-Hispanic high loadings, that displayed unique con-
records. Values were included only if they whites predominated (52.7%), followed tent, or that accurately represented the
had been collected within the past 12 by Asian Americans and Pacific Islanders factor’s content domain. The result was a
months. Lipid profiles were available for (19.6%), African Americans (13.2%), and 17-item scale, with 5 EB items, 5 RD
139 subjects at TAMC and 131 subjects at Hispanics (7.0%). Most subjects (87.7%) items, 4 PD items, and 3 ID items. The
KP. Analyses focused on total cholesterol had graduated from high school. correlation between the original 28-item
only. NMC and KP subjects were recruited scale total and the new 17-item scale total

628 DIABETES CARE, VOLUME 28, NUMBER 3, MARCH 2005

 Polonsky and Associates

Table 2— Rotated pattern matrix for the ex- Internal consistency elevated DDS total scores were associated
ploratory factor analysis of the 28 items Cronbach’s ␣ was computed for the total with being younger and more depressed,
17-item scale and for each subscale for using insulin, poorer self-care, and having
EB PD RD ID each site. Because the results varied little elevated lipid levels.
among the sites, the ␣ values for the com- None of the subscales were signifi-
DDS1 0.678 0.005 0.028 ⫺0.001 bined sample are presented: 17-item scale cantly related to patient sex, ethnicity, ed-
DDS5 0.562 ⫺0.009 0.309 ⫺0.105 total ⫽ 0.93; EB ⫽ 0.88, PD ⫽ 0.88, ucational level, or diabetes duration. The
DDS9 0.804 ⫺0.044 ⫺0.001 0.049 RD ⫽ 0.90, and ID ⫽ 0.88. These ␣ val- EB and RD subscales were linked to
DDS13 0.744 ⫺0.004 0.066 0.056 ues are adequate, especially given the poorer adherence to meal planning (r ⫽
DDS17 0.823 0.012 0.044 ⫺0.036 number of items per scale. 0.21 and r ⫽ 43, respectively) and less
DDS21 0.606 ⫺0.019 0.205 ⫺0.004 exercise (r ⫽ 0.12 and r ⫽ 0.16, respec-
DDS25 0.772 0.041 0.043 0.062 Validity tively). Only RD was related to less fre-
DDS2 ⫺0.044 0.798 0.034 ⫺0.078 Pearson correlation coefficients (or, quent SMBG (r ⫽ 0.19). All four
DDS6 0.063 0.502 0.205 0.033 where appropriate, ␹2 values) were com- subscales were positively associated with
DDS10 ⫺0.090 0.801 0.123 ⫺0.031 puted between the scale total, each of the depressive affect (in all cases, r ⬎ 0.33).
DDS14 0.121 0.482 ⫺0.041 0.253 four subscales, and the CESD, disease Subscale scores were mostly unrelated to
DDS18 0.031 0.842 ⫺0.042 ⫺0.002 management, and metabolic variables for HbA1c but were consistently and posi-
DDS22 0.091 0.755 ⫺0.121 0.090 each site. Similar results occurred across tively linked to total cholesterol (for EB,
DDS26 ⫺0.003 0.833 ⫺0.010 ⫺0.023 all four sites, so the sample was again RD and ID, r ⬎ 0.16).
DDS3 ⫺0.018 ⫺0.045 0.750 0.029 combined and the analyses were com-
DDS7 0.040 0.005 0.610 ⫺0.028 pleted for the sample as a whole (Table 3). CONCLUSIONS — We have devel-
DDS11 ⫺0.038 ⫺0.015 0.581 0.133 The DDS scale total was not signifi- oped a new instrument to assess diabetes-
DDS15 0.178 ⫺0.011 0.714 0.030 cantly related to patient sex, ethnicity, ed- related emotional distress and provided
DDS19 0.292 0.088 0.556 0.007 ucational level, or diabetes duration. Age data regarding its factor structure, inter-
DDS23 0.028 0.059 0.829 ⫺0.046 was negatively correlated with the total nal consistency, and validity. These data
DDS27 0.013 0.025 0.744 0.128 score (r ⫽ ⫺0.29), indicating that indicate that the DDS has a consistent,
DDS4 0.158 0.068 0.146 0.401 younger subjects reported more diabetes- generalizable factor structure and good
DDS8 0.231 0.097 0.015 0.358 related distress than older subjects. Regi- internal reliability and validity across four
DDS12 0.320 0.110 0.074 0.202 men type was associated with different different clinical sites. In contrast to pre-
DDS16 ⫺0.084 ⫺0.007 0.118 0.816 levels of distress. Insulin users reported vious measures, the DDS is more concep-
DDS20 0.508 0.045 ⫺0.128 0.257 the highest DDS total scores (36.9 ⫾ tually driven, drawing items from four
DDS24 0.213 0.079 0.014 0.603 17.1) followed by those taking oral hypo- preestablished domains of diabetes-
DDS28 0.043 ⫺0.067 0.014 0.877 glycemic agents only (35.2 ⫾ 16.2) and, related distress: EB, PD, RD, and ID. Our
Items are listed in the APPENDIX. Note that all items finally, those whose disease was con- findings of a stable factor structure match-
selected for the final 17-item scale are in bold. trolled by diet (26.7 ⫾ 12.1). DDS total ing these four domains is consistent with
scores were positively associated with de- recent results from Snoek et al. (8), who
was very high (r ⫽ 0.99), indicating that pressive symptomatology (CESD; r ⫽ found a relatively similar factor structure
the 17-item version captured most of the 0.56), poorer adherence to meal planning in the PAID, labeling those factors as
variance reflected in the 28-item version recommendations (r ⫽ 0.30), and lower “negative emotions,” “treatment prob-
but with 40% fewer items. levels of exercise (r ⫽ 0.13). The DDS lems,” “food-related problems,” and “lack
The mean correlation between the total was unrelated to glycemic control of social support.”
subscales and the 17-item total score was (r ⫽ 0.01) but was positively associated The DDS has certain potential advan-
0.82. The subscales that most highly cor- with total cholesterol (r ⫽ 0.20). In sum, tages over previous instruments. It is
related with the scale total were EB and
RD (for both, r ⫽ 0.88), whereas ID (r ⫽
0.76) and PD (r ⫽ 0.67) were less Table 3—Zero-order correlations between the DDS and items of interest
strongly associated. The strongest links
were between EB and RD (r ⫽ 0.69), EB DDS Total EB PD RD ID
and ID (r ⫽ 0.61), and RD and ID (r ⫽
0.57). In contrast, the least highly corre- Age ⫺0.29* ⫺0.31* ⫺0.07 ⫺0.29* ⫺0.20*
lated associations were between PD and Years of education 0.04 0.05 0.02 0.05 ⫺0.02
the other three—EB (r ⫽ 0.44), RD (r ⫽ Diabetes duration ⫺0.02 ⫺0.02 0.01 ⫺0.05 0.00
0.45), and ID (r ⫽ 0.42). In total, these CESD 0.56* 0.55* 0.34* 0.42* 0.48*
correlations (in all cases, P ⬍ 0.001) sug- Self-care
gested that the subscales reflected both Meal planning 0.30* 0.21* 0.07 0.43* 0.17*
unique and shared variance: the DDS sub- Exercise 0.13† 0.12† 0.05 0.16* ⫺0.07
scales were not totally independent, but, SMBG 0.08 ⫹0.00 0.00 0.19* 0.04
at the same time, they tapped into rela- HbA1c 0.01 0.02 ⫺0.11† 0.08 0.01
tively different areas of diabetes-related Total cholesterol 0.20* 0.17† 0.03 0.20* 0.22*
distress. *P ⬍ 0.001; †P ⬍ 0.01.

DIABETES CARE, VOLUME 28, NUMBER 3, MARCH 2005 629

Psychosocial distress in diabetes

shorter, and the new subscales allow for 2) Feeling that my doctor doesn’t 25) Feeling overwhelmed by the de-
direct comparison of four different types know enough about diabetes and dia- mands of living with diabetes.
of distress. This may be especially useful betes care. 26) Feeling that I don’t have a doc-
when the instrument is used for planning 3) Feeling that I can’t control my eat- tor who I can see regularly about my
clinical interventions. The DDS also ap- ing. diabetes.
pears applicable to patients from both 4) Feeling that there is no one in my 27) Not feeling motivated to keep
sexes and from at least several major eth- life with whom I can talk really openly up my diabetes self-management.
nic groups. We believe that the items are about my feelings about diabetes. 28) Feeling that friends or family
clearer than in previous instruments, al- 5) Worrying about the future and the don’t give me the emotional support
lowing for less patient confusion. Indeed, possibility that I could develop serious that I would like.
the DDS has a Flesch-Kincaid grade level long-term complications.
of 7.3, suggesting that it should be com- 6) Feeling that I don’t see my doctor
prehensible to a majority of patients. often or long enough. References
There are, however, a number of lim- 7) Feeling that I am not getting 1. Egede LE, Zheng D: Independent factors
itations to this study. First, the number of enough physical exercise. associated with major depressive disorder
patients from each ethnic group was too 8) Feeling that I have to hide my dia- in a national sample of individuals with
small to permit separate factor analyses. betes from others. diabetes. Diabetes Care 26:104 –111,
Although the factor structure was rela- 9) Feeling angry, scared, and/or de- 2003
tively stable across sites, it will be impor- pressed when I think about living with 2. Nichols GA, Brown JB: Unadjusted and
diabetes. adjusted prevalence of diagnosed depres-
tant to verify these findings within diverse sion in type 2 diabetes. Diabetes Care 26:
patient populations. Second, along the 10) Feeling that my doctor doesn’t
744 –749, 2003
same lines, the samples of type 1 and type give me clear enough directions on how 3. Dunn SM, Smartt H, Beeney L, Turtle J:
2 diabetic patients also were too small to to manage my diabetes. Measurement of emotional adjustment in
permit separate analyses. Third, longitu- 11) Feeling that I am not testing my diabetic patients: validity and reliability of
dinal data are not yet available, so little blood sugars frequently enough. ATT39. Diabetes Care 9:480 – 489, 1986
can be said about the DDS’s test-retest re- 12) Feeling that friends or family act 4. Herschbach P, Duran G, Waadt S, Zettler
liability or about the instrument’s ability like “diabetes police” (e.g., nag about eat- A, Amm C, Marten-Mittag B: Psychomet-
ing properly, testing blood sugars, not ric properties of the Questionnaire on
to detect change over time as a function of Stress in Patients with Diabetes–Revised
a clinical intervention. Finally, the mea- trying hard enough).
(QSD-R). Health Psychol 16:171–174,
sures of validity were relatively limited; 13) Feeling “burned out” by the con- 1997
this is not surprising, since none of the stant effort to manage diabetes. 5. Polonsky WH, Anderson BJ, Lohrer PA,
four studies were specifically designed to 14) Feeling that I can’t tell my doctor Welch G, Jacobson AM, Aponte JE,
evaluate the DDS. what is really on my mind. Schwartz CE: Assessment of diabetes-re-
The DDS is a stable, internally consis- 15) Feeling that I am often failing lated distress. Diabetes Care 18:754 –760,
with my diabetes regimen. 1995
tent, conceptually driven measure of dia-
16) Feeling that friends or family 6. Ishii H, Welch GW, Jacobson A, Goto M,
betes-related distress for use in research Okazaki K, Yamamoto T, Tsujii S: The
and clinical practice. Brief and easy to ad- are not supportive enough of my self-
Japanese version of the Problem Area in
minister, it may serve as a valuable tool in care efforts (e.g., planning activities Diabetes Scale: a clinical and research tool
identifying patients experiencing high that conflict with my schedule, encour- for the assessment of emotional function-
levels of distress linked to their diabetes aging me to eat the “wrong” foods). ing among diabetic patients. Diabetes 48:
and pinpointing their specific concerns. 17) Feeling that diabetes controls A319, 1999
All 28 of the preliminary items are listed my life. 7. Lerman-Garber I, Barron-Uribe C, Cal-
18) Feeling that my doctor doesn’t zada-Leon R, Mercado-Atri M, Vidal-
in the APPENDIX; the final 17 items com-
take my concerns seriously enough. Tarnayo R, Quintana S, Hernandez ME,
prising the DDS are in bold. Ruiz-Reyes MdlR, Tamez-Gutierrez LE,
19) Not feeling confident in my day-
Nishimura-Meguro E, Villa AR: Emo-
to-day ability to manage diabetes.
tional dysfunction associated with diabe-
20) Worrying that diabetes limits my tes in Mexican adolescents and young
Acknowledgments — The authors thank Bar-
bara J. Anderson, PhD, and Garry Welch, PhD,
social relationships and friendships. adults with type-1 diabetes. Salud Publica
for their advice, guidance, and support. 21) Feeling that I will end up with Mex 45:13–18, 2003
serious long-term complications, no 8. Snoek FJ, Pouwer F, Welch GW, Polon-
matter what I do. sky WH: Diabetes-related emotional dis-
22) Feeling that my doctor doesn’t re- tress in Dutch and U.S. diabetic patients:
APPENDIX ally understand what its like to have dia- cross-cultural validity of the Problem Ar-
The 28 preliminary items developed for betes. eas in Diabetes scale. Diabetes Care 23:
1305–1309, 2000
the DDS (the final 17 items comprising 23) Feeling that I am not sticking 9. Weinger K, Jacobson AM: Psychosocial
the DDS are in bold) closely enough to a good meal plan. and quality of life correlates of glycemic
1) Feeling that diabetes is taking up 24) Feeling that friends or family control during intensive treatment of
too much of my mental and physical don’t appreciate how difficult living type 1 diabetes. Patient Educ Couns 42:
energy every day. with diabetes can be. 123–131, 2001

630 DIABETES CARE, VOLUME 28, NUMBER 3, MARCH 2005

 Polonsky and Associates

10. Welch GW, Jacobson AM, Polonsky WH: Problem Areas in Diabetes (PAID) ques- 17. Radloff LS: The CES-D scale: a self-report
The Problem Areas in Diabetes Scale: an tionnaire. Diabet Med 20:69 –72, 2003 depression scale for research in the gen-
evaluation of its clinical utility. Diabetes 14. Polonsky WH: Understanding and assess- eral population. Appl Psychol Meas 1:385–
Care 20:760 –766, 1997 ing diabetes-specific quality of life. Diabe- 401, 1977
11. Trief PM, Himes CL, Orendorff R, Wein- tes Spectrum 13:36 – 41, 2000 18. Toobert DJ, Glasgow RE: Assessing diabe-
stock RS: The marital relationship and 15. Polonsky WH: Emotional and quality of tes self-management: The summary of di-
psychosocial adaptation and glycemic life aspects of diabetes management. Curr abetes self-care activities questionnaire.
control of individuals with diabetes. Dia- Diabetes Rep 2:153–159, 2002 In Handbook of Psychology and Diabetes.
betes Care 24:1384 –1389, 2001 16. Polonsky WH, Earles J, Smith S, Pease DJ, Bradley C, Ed. Chur, Switzerland, Har-
12. Trief PM, Wade MJ, Britton KD, Wein- Macmillan M, Christensen R, Taylor T, wood Academic Publishers, 1994, p 351–
stock RS: A prospective analysis of marital Jackson RA: Integrating medical manage- 375
relationship factors and quality of life in ment with diabetes self-management 19. Toobert DJ, Hampson SE, Glasgow RE:
diabetes. Diabetes Care 25:1154 –1158, training: a randomized control trial of the The summary of diabetes self-care activi-
2002 Diabetes Outpatient Intensive Treatment ties measure: results from 7 studies and a
13. Welch GW, Weinger K, Anderson B, (DOIT) program. Diabetes Care 26:3048 – revised scale. Diabetes Care 23:943–950,
Polonsky WH: Responsiveness of the 3053, 2003 2000

DIABETES CARE, VOLUME 28, NUMBER 3, MARCH 2005 631

View publication stats