anderson_depresion_diabetes_comorbida2001.pdf

The Prevalence of Comorbid Depressionin Adults With DiabetesA meta-analysis

RYAN J. ANDERSON, BA1

KENNETH E. FREEDLAND, PHD1RAY E. CLOUSE, MD1,2

PATRICK J. LUSTMAN, PHD1,3

OBJECTIVE To estimate the odds and prevalence of clinically relevant depression in adultswith type 1 or type 2 diabetes. Depression is associated with hyperglycemia and an increased riskfor diabetic complications; relief of depression is associated with improved glycemic control. Amore accurate estimate of depression prevalence than what is currently available is needed togauge the potential impact of depression management in diabetes.

RESEARCH DESIGN AND METHODS MEDLINE and PsycINFO databases andpublished references were used to identify studies that reported the prevalence of depression indiabetes. Prevalence was calculated as an aggregate mean weighted by the combined number ofsubjects in the included studies. We used x2 statistics and odds ratios (ORs) to assess the rate andlikelihood of depression as a function of type of diabetes, sex, subject source, depression assess-ment method, and study design.

RESULTS A total of 42 eligible studies were identified; 20 (48%) included a nondiabeticcomparison group. In the controlled studies, the odds of depression in the diabetic group weretwice that of the nondiabetic comparison group (OR 5 2.0, 95% CI 1.82.2) and did not differby sex, type of diabetes, subject source, or assessment method. The prevalence of comorbiddepression was significantly higher in diabetic women (28%) than in diabetic men (18%), inuncontrolled (30%) than in controlled studies (21%), in clinical (32%) than in community(20%) samples, and when assessed by self-report questionnaires (31%) than by standardizeddiagnostic interviews (11%).

CONCLUSIONS The presence of diabetes doubles the odds of comorbid depression.Prevalence estimates are affected by several clinical and methodological variables that do notaffect the stability of the ORs.

Diabetes Care 24:10691078, 2001

R ecent meta-analyses link depres-sion in diabetes with hyperglycemia(1) and with an increased risk forcomplications of the metabolic disorder(2). There is also evidence from three con-trolled trials to suggest that treatment ofdepression improves glycemic control(35). An accurate estimate of depressionprevalence is needed to help gauge the

potential impact of depression manage-ment in patients with comorbid diabetes.Gavard et al. (6) last reviewed studies ofthe prevalence of depression in diabetesin 1993. Since then, the literature on thissubject has expanded considerably. Inthe present study, we comprehensivelyreviewed the scientific literature to de-termine the odds of clinically significant

depression in those with diabetes ver-sus those without diabetes and to esti-mate the aggregate prevalence. Theseestimates were also studied in relation tothe type of diabetes, sex, source of sub-jects, study design, and method of de-pression assessment.

RESEARCH DESIGN ANDMETHODS

Inclusion/exclusion criteria.MEDLINE and PsycINFO search engineswere used to identify published studiesthat measured the point and/or lifetimeprevalence of depression in adults withdiabetes. The terms depression, depressivedisorder, minor depressive disorder, or dys-thymic disorder were combined with theterms diabetes or diabetes mellitus. Thesearch was limited to studies publishedbefore January 1, 2000. Studies were lim-ited to those that 1) were published oravailable in English, 2) had a sample size$25, and 3) included only adults ($18years of age) diagnosed with type 1 ortype 2 diabetes. Reference lists of pub-lished studies were also examined to ob-tain additional reports.

The review includes all availablestudies that identified clinically relevantdepression, (i.e., depression severeenough to warrant clinical intervention).This definition includes major depressivedisorder as well as minor and subsyndro-mal depression. In patients with diabetesand other chronic medical illnesses, eachof these presentations of depression hasbeen shown to have adverse effects on so-cial and physical functioning and qualityof life that are independent of the effectsof the medical illness (711). Both majorand minor depressions are associatedwith increased medical morbidity andmortality, even after adjustment forhealth status and health behaviors (1214). Similarly, there is evidence that ther-apy to treat these depressive conditions iseffective and associated with improve-ments in mood, functioning, and qualityof life (1517).

c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c

From the Departments of 1Psychiatry and 2Medicine, Washington University School of Medicine; and the3Department of Veterans Affairs Medical Center, St. Louis, Missouri.

Address correspondence and reprint requests to Patrick J. Lustman, PhD, Department of Psychiatry,Washington University School of Medicine, 4940 Childrens Pl., St. Louis, MO 63110. E-mail:[email protected].

Received for publication 7 November 2000 and accepted in revised form 22 February 2001.Abbreviations: BDI, Beck Depression Inventory; OR, odds ratio.A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion

factors for many substances.

E p i d e m i o l o g y / H e a l t h S e r v i c e s / P s y c h o s o c i a l R e s e a r c hO R I G I N A L A R T I C L E

DIABETES CARE, VOLUME 24, NUMBER 6, JUNE 2001 1069

Procedures and statistical analyses.Studies meeting the inclusion criteriawere examined, and study demographics(age, race, sex, and type of diabetes) as wellas depression information (assessmentmethod, prevalence, and mean scale scores)were recorded using a structured formbased on the ones used by the CochraneLibrarys Database of Systematic Reviews,produced by the Cochrane Collaboration(18). Two reviewers independently ab-stracted data from each study. The articleswere divided into subgroups of controlledand uncontrolled studies. For the pur-poses of this review, the term controlleddoes not imply that the condition understudy (depression) was randomly manip-ulated or followed longitudinally. A studywas considered controlled if the preva-lence of depression in diabetic patientswas compared with that of a nondiabeticcomparison group. A study was consid-ered uncontrolled if it did not have a non-diabetic comparison group.

Major depressive disorder was as-sessed with structured or semistructureddiagnostic interviews (e.g., the DiagnosticInterview Schedule [19] or the StructuredClinical Interview for DSM-III-R [20])and diagnosed according to the criteriafor major depressive disorder specified inthe version of the Diagnostic and StatisticalManual of Mental Disorders (21,22) thatwas current at the time of the study. Inthese studies, depression prevalence wasequal to the percentage of patients in thesample who met the criteria for the diag-nosis.

In some studies, depression was as-sessed by self-report symptom scales(e.g., the Beck Depression Inventory[BDI] [23] or the Center for Epidemio-logic StudiesDepression Scale [24]). Inthese studies, depression prevalence wasequal to the percentage of subjects withscale scores above a specified thresholdvalue. The threshold score used to iden-tify a depression case varied somewhatacross studies (e.g., BDI score $10 insome studies [2527] and $13 [28] or$16 [29] in others). This variance is to beexpected, since the threshold score usedto identify depression is to some extentdependent on the importance placed ondepression recognition and the resourcesavailable for this purpose within a partic-ular clinical setting (30,31).

Odds ratios (ORs) were calculatedonly for controlled studies, because thesecalculations derive from comparing the

odds of depression in the diabetic groupwith the odds in the nondiabetic compar-ison group. Potential moderator variableswere analyzed to determine whether itwould be necessary to estimate the oddsof depression separately by these vari-ables. The Breslow-Day test for homoge-neity was used to determine whether thelikelihood of depression in diabetic ver-sus nondiabetic patients differed depend-ing on type of diabetes, sex, source ofsubjects, or depression assessmentmethod. Cochran-Mantel-Haenszel sta-tistics were used to test the significance ofthe aggregate odds of depression in dia-betic versus nondiabetic patients after ad-justing for these variables.

Depression prevalence was calculatedas an aggregate mean, weighted by thenumber of subjects in the study or group-ing of interest (e.g., type 1 vs. type 2 dia-betes). This method factors study samplesize into the calculation of the overallprevalence estimate. We used x2 tests tostatistically compare the prevalence of de-pression in the diabetic and nondiabeticcomparison groups. An overall estimateof depression prevalence adjusted for po-tential moderators (e.g., age, sex, and typeand severity of diabetes) could not be cal-culated with precision because none ofthe studies fully characterized the de-pressed and nondepressed subsets in thisregard.

RESULTS The literature searchidentified 48 studies, 6 of which were ex-cluded for having ,25 subjects or havingpoorly described or inadequate depres-sion or diabetes assessment methods. The42 included studies had a combined sam-ple size of 21,351 subjects. Of these stud-ies, 20 (48%) were controlled (i.e.,included a nondiabetic comparisongroup) and 22 (52%) were uncontrolled(Tables 1 and 2). Of the 20 controlledstudies, 3 (15%) were comprised exclu-sively of patients with type 1 diabetes; 8(40%) included only patients with type 2diabetes; and 9 (45%) included a mixedsample (i.e., one that contained patientshaving either type 1 or type 2 diabetes).None of the mixed-sample controlledstudies reported depression prevalenceseparately by type of diabetes. Of the 22uncontrolled studies, 6 (27%) were lim-ited to patients with type 1 diabetes, 5(23%) to patients with type 2 diabetes,and 11 (50%) to patients with either type1 or type 2 diabetes. The majority (6 of 11

[64%]) of the mixed-sample uncontrolledstudies did not report depression preva-lence separately by type of diabetes. Ofthe 42 total studies, 3 (7%) did not pro-vide enough information to be includedin the OR or point prevalence calcula-tions. Two of these three studies (32,33)reported only the statistical comparisonof the mean depression scale scores of di-abetic versus nondiabetic subjects, andthe other (34) reported only lifetime ratesof depression.

Odds of depression in diabetesTen of the controlled studies reported de-pression estimates separately by type ofdiabetes (type 1, n 5 3; type 2, n 5 7).The odds of depression were significantlyincreased in both type 1 (OR 5 2.9, 95%CI 1.65.5, x2 5 12.8, P 5 0.0003) andtype 2 diabetes (OR 5 2.9, 95% CI 2.33.7, x2 5 84.3, P , 0.0001) over nondi-abetic control subjects. The increasedodds of depression associated with diabe-tes were similar in type 1 versus type 2diabetes (2.9 vs. 2.9, Breslow-Day x2 50.004, P 5 0.95), and the significant ef-fect of diabetes on depression remainedafter controlling for type of diabetes(Cochran-Mantel-Haenszel x2 5 95.5,P , 0.0001).

Seven controlled studies reported theprevalence of depression separately formen and women. The odds of depressionwere significantly elevated in bothwomen (OR 5 1.7, 95% CI 1.42.0, x2 534.0, P , 0.0001) and men (OR 5 1.7,95% CI 1.42.2, x2 5 19.6, P , 0.0001)with diabetes compared with control sub-jects. The increased odds of depressionassociated with diabetes were similar inwomen versus men with diabetes (1.7 vs.1.7, Breslow-Day x2 5 0.08, P 5 0.8),and the significant effect of diabetes ondepression remained after controlling forsex (Cochran-Mantel-Haenszel x2 553.9, P , 0.0001).

The controlled studies were dividedinto subsets of community (n 5 11) andclinical (n 5 7) studies. All of the commu-nity studies identified diabetic and non-diabetic subjects from random samples ofcommunity-dwelling individuals. In theclinical studies, the diabetic and nondia-betic subjects were drawn mostly fromnonrandom samples of patients recruitedfrom health care clinics, patient supportgroups, or physician referrals. The oddsof depression were significantly increasedin people with diabetes versus those with-

Depression in diabetes

1070 DIABETES CARE, VOLUME 24, NUMBER 6, JUNE 2001

Table 1Prevalence of clinically significant depression in adults with diabetes: controlled studies (n 5 20)

Study

Subjects(Diabetic: n)(Control: n)

Sex(% female)

Age(years)

Race(% white)

Depressionassessment

method

Prevalence of depression

Depressionscale scores

Overall(%)

Males(%)

Females(%)

Kokkonen (62) T1: 63e 41.3 20.9 6 1.9 PSE ID $5 14.3 Men: 9.3 6 9.4*Women: 14.2 6 10.2

Medically well: 123a 51.2 21.9 6 1.4 11.4

Songar et al. (25) T1: 60e 68.3 29.7 BDI $14 43.3 15.2Medically well: 30c 60.0 29.2 3.3* 4.1*

Popkin et al. (63) T1: 75e 64.0 31 DIS/DSM-III 10.7 3.7 14.6 1st degree relatives:

34f55.9 36 2.9 0.0 5.3

Amato et al. (64) T2: 197a 68.0 73.9 6 5.9 GDS $21 13.6 11.4 14.7 13.2 6 6.81142a 55.9 74.2 6 6.4 8.7 6.6* 10.6 11.1 6 6.5*

Eaton et al. (50) T2: 148a DIS/DSM 6.1 1600a 5.3

Viinamaki et al. (65) T2: 82a 46.3 66.9 6 0.7 Zung $50 11.0 11.4 10.5 39.4 6 1.3115a 55.7 65.6 6 0.5 6.9 36.9 6 0.8

Leedom et al. (26) T2: 71e 70.4 50 6 2.0 18.3 BDI $10 49.3 12.246e 67.4 46 6 1.7 23.9 21.7 5.9 6 0.7

Palinkas et al. (28) T2: 93a 39.8 72.4 6 8.7 BDI $13 11.5 8.8 13.6 6.51284a 54.5 68.5 6 9.5 4.6\ 2.6 6.2 5.4\

Wing et al. (29) T2: 32e 50.0 52.1 6 7.7 BDI $16 21.8 10.6 6 6.4Spouses: 32e 50.0 50.8 6 8.8 12.5 7.5 6 6.2

Weyerer et al. (66) T2: 55a 72.7 CIS/ICD 8 27.3 Medically well: 122a 54.1 10.6\

Tun et al. (32)# T2: 119e 48.7 63.4 Zung## 39.3Nondiabetic out-

patients: 25e56.0 63.0 34.0

Black (67) T1 & 2: 636a 58.2 0.0 CES-D $16 31.1 22.6 37.9 2196a 58.4 0.0 24.1* 15.9 30.2

Penninx et al. (33)# T1 & 2: 204a 52.9 73.3 6 7.7 CES-D## 10.1 6 9.2Medically well: 719a 42.4 67.2 6 8.6 5.4 6 6.3*

Bourdel-Marchassonet al. (68)

T1 & 2: 237a 50.6 CES-DMen: $17Women: $23

21.3 Men: 9.3 6 9.4*Women: 14.2 6 10.2

2555a 60.7 12.7* Men: 7.2 6 7.6Women: 11.8 6 9.6

Rajala et al. (69) T1 & 2: 62a 40.3 55 Zung $45 19.3 18.9 20.0 480a 58.5 55 11.7 10.1 12.8

Zhang et al. (70) T1 & 2: 209a 58.4 0.0 DIS/DSM-III 1CES-D

3.8 1.8 5.1 8.3

1289a 55.6 0.0 3.6 1.7 5.1 7.8

Wells et al. (71) T1 & 2: 154a\ \ DIS/DSM-III 9.6 Medically well:

1353a\ \ 48.5 4.4

Continued on following page

Anderson and Associates


out diabetes in both the community(OR 5 1.8, 95% CI 1.62.1, x2 5 98.6,P , 0.0001) and the clinical (OR 5 2.1,95% CI 1.52.8, x2 5 19.0, P , 0.0001)studies. The increased odds associatedwith diabetes were similar in the commu-nity and clinical studies (1.8 vs. 2.1,Breslow-Day x2 5 0.37, P 5 0.5), and thesignificant effect of diabetes on depres-sion remained after controlling for subjectsource (Cochran-Mantel-Haenszel x2 5117.6, P , 0.0001).

Seven of the controlled studies usedclinician interviews and psychiatric diag-nostic criteria to determine depression,and the other 11 controlled studies usedthreshold scores on self-report depressionsymptom scales. The pattern of ORs re-lated to method of depression assessmentwas similar to that of type of diabetes, sex,and subject source (Fig. 1). The odds ofdepression were significantly elevated indiabetic patients over control subjectswhether depression was assessed with di-agnostic interviews or with self-reportscales. There was no significant differencein ORs between methods (Fig. 1), and thesignificant effect of diabetes on depres-sion remained after controlling formethod of assessment (Cochran-Mantel-Haenszel x2 5 170.9, P , 0.0001).

Because the incremental increases inthe odds of depression did not differ as afunction of type of diabetes, sex, subjectsource, or method of depression assess-

ment, the findings from all of the con-trolled studies (n 5 18 studies, 17,399subjects) were combined to calculate anaggregate OR. In this combined sample,the odds of depression were twice as highin those with diabetes compared with thecontrol subjects (OR 5 2.0, 95% CI 1.82.2, x2 5 159.8, P , 0.0001).

Prevalence of depression in diabetesPrevalence estimates were determined byaggregating data reported in the con-trolled and uncontrolled studies to capi-talize on the combined subject pool.These rates were determined by sex, sub-ject source, method of depression assess-ment, and type of diabetes. These unad-justed prevalence estimates are displayedin Table 3, and should be viewed cau-tiously, because they do not adjust forother factors (e.g., sex, method of depres-sion assessment, and study design) thatmay affect prevalence.

The aggregate estimate of depressionwas lower in type 1 vs. type 2 diabetes(x2 5 11.5, P 5 0.007). However, therates for type 1 vs. type 2 diabetes werestatistically similar in the studies that de-termined depression by diagnostic inter-view (13.6 vs. 10.9%) or with self-reportscales (29.1 vs. 32.9%) (P . 0.1 for bothcomparisons). As would be predictedfrom the similar ORs for male and femalesubjects, women had an increased preva-lence of depression in comparison to

men, just as they do in the nondiabeticpopulation. The combined prevalencewas significantly higher in women withdiabetes than in men with diabetes (28.2vs. 18.0%, x2 5 42.1, P , 0.0001; OR 51.6, 95% CI 1.41.8). Self-reportbasedestimates were higher than interview-based estimates in both the controlled(26.1 vs. 9.0%, x2 5 100.6, P , 0.0001)and uncontrolled studies (34.9 vs. 14.2%,x2 5 109.2, P , 0.0001) (Fig. 2). Simi-larly, estimates from the uncontrolledstudies were significantly higher than incontrolled studies whether depressionwas assessed with diagnostic interviews(14.2 vs. 9.9%, x2 5 10.3, P 5 0.001) orwith self-report scales (34.9 vs. 26.1%,x2 5 34.2, P , 0.0001). Major depressivedisorder (per diagnostic interviews) waspresent in 11.4% of patients with diabe-tes; elevations in depressive symptoms(per self-report scales) were significantlymore common and were present in 31.0%of patients with diabetes (x2 5 159.8, P ,0.0001).

Of the 42 studies identified by thesearch, 8 (19.0%) determined the lifetimeprevalence of major depression (2 con-trolled and 6 uncontrolled). Lifetimeprevalence is the proportion of study sub-jects who met criteria for the disorder atany point during their life, either before orduring the time of assessment. The life-time prevalence of depression was signif-icantly higher in those with diabetes than

Table 1Continued

Study


Sex(% female)

Age(years)

Race(% white)


method



Overall(%)

Males(%)

Females(%)

Robinson et al. (72) T1 & 2: 60:70e 45.4 51 6 6.6 56.9 PSE/BedfordCol. criteria

8.5

130f 45.4 44 6 10.4 72.3 8.5

Friis et al. (73) T1 & 2: 56b 71.4 57.0 60.7 CES-D $16 60.7 20.4Medically ill: 56b 73.2 53.0 63.2 48.2 14.2

Murrell et al. (74) T1 & 2: 175a 65.7 92.9 CES-D $20 21.7 15.5 25.4 2277a 61.5 16.0 13.4 17.6

Data are means 6 SD, unless otherwise indicated. T1 5 type 1 diabetes, T2 5 type 2 diabetes, n 5 sample size, PSE 5 Present State Examination, DIS/DSM-III 5Diagnostic Interview Schedule/Diagnostic and Statistical Manual-III, GDS 5 Geriatric Depression Scale, Zung 5 Zung Self-Rating Depression Scale, CIS 5 ClinicalInterview Schedule, CES-D 5 Center for Epidemiologic Studies-Depression Scale.Sample randomly selected from a acommunity, bclinic or hospital, or cunspecified setting.Sample not randomly selected from a dcommunity, eclinic or hospital, or funspecified setting.*P # .001 vs. the prevalence of the overall or sex-specific diabetic group; nonsignificant vs. the prevalence of the overall or sex-specific diabetic group; P # .05vs. the prevalence of the overall or sex-specific diabetic group; P # .01 vs. the prevalence of the overall or sex-specific diabetic group; \age- and sex-adjusted;age-adjusted; #these studies reported only mean depression scale scores for diabetic and nondiabetic subjects, and thus were not included in the prevalencecalculations; sex-adjusted; greater (P , .001) in nondiabetic females vs. male counterparts; \\prevalences are any affective disorder including major depression,dysthmia, and mania. Mania represented just 2% of all affective disorders in this study. the number is the percentage of Caucasian subjects in the entire sample.



Table 2Prevalence of clinically significant depression in adults with diabetes: uncontrolled studies (n 5 22)

Study


Sex(% female)

Age(years)

Race(% white)


method



Overall(%)

Males(%)

Females(%)

Berlin et al. (75) T1: 102e 45.1 43 6 13 MINI/DSM-IV 12.7

Cohen et al. (76) T1: 49e 55.1 34.3 6 9.2 SCID/DSM-III-R 14.3*

Mayou et al. (77) T1: 109e 53.2 21.9 6 2.8 PSEID $5

11.0 7.8 13.8

Winocour et al. (78) T1: 130e 36.2 40.3 6 1.1 100.0 ZungMen: .25.7Women: .29.8

14.6 12.0 19.1 13.4 6 0.6

Stone et al. (27) T1: 57e 56.1 43.5 6 15.7 BDI $10 33.3 8.5 6 6.1

Surridge et al. (79) T1: 50e 46.0 38.1 6 14.5 BDI 1 Hamilton 1psychiatric interview

0 0 0 5.9 6 6.3

Marcus et al. (34) T2: 66e 66.6 52.9 6 9.5 IDD-L

Connell et al. (80) T2: 191d 57.6 70.3 6 6.7 86.0 Zung 47.0 49.7 6 8.6

Naliboff and Rosen-thall (81)

T2: 102e 0.0 66.6 BDI $ 17 1 MMPI\ $80 23.5 23.5

Geringer et al. (82) T2: 64e 100.0 63 6 5.3 Zung $50 18.8 18.8 41.5

Biglan et al. (83) T2: 36e SADS/RDC 22.2

Kohen et al. (84) T1: 36e

T2: 64e42.0 T1: 37.7 6 13.3

T2: 62.5 6 12.7 HADS T1: 16.7

T2: 35.9x: 29.0

T1: 4.6 6 4.3T2: 6.3 6 4.3x: 5.7

Peyrot et al. (85) T1: 203e

T2: 431e59.0 60.3 CES-D $16 T1: 42.4

T2: 40.8x:41.3

31.1 48.4

Bailey (86) T1 & 2: 180e 60.0 46 71.1 CES-D $16 33.9 13.5 6 1.1

Haire-Joshu et al. (87) T1: 163e

T2: 23e54.8 42.7 6 14.7 72.6 BDI $10 27.4 7.8

Jalenques et al. (88) T1: 34e

T2: 27e41.0 T1: 49.8

T2: 59.0x: 53.9

Psychiatric interview/DSM-III-R criteria

T1: 20.6T2: 22.2x: 21.3

13.9 32.0

Padgett et al. (89) T1: 33b

T2: 147b51.1 Zung $50 60.5 52.5

Lee et al. (90) T1 & 2: 93e 58.1 27 6 6.7 DSM-III symptomchecklist

5.4#

Von Dras andLichty (91)

T1: 66a

T2: 50a52.6 T1: 33.7 6 10.9

T2: 55.4 6 9.0x: 43.1 6 14.8

Zung 40.0 48.4 6 11.4

Lustman et al. (92) T1: 57b

T2: 57b66.7 T1: 30.8 6 9.7

T2: 49.1 6 13.0x: 40.0 6 15.1

62.3 DIS/DSM-III 14.0

Continued on following page



in control subjects (17.5 vs. 6.8%, respec-tively, x2 5 34.2, P , 0.0001). The ag-gregate estimate of the lifetime prevalenceof major depression based on all eightstudies was 28.5%.

CONCLUSIONS We estimatedthe odds and prevalence of depression indiabetes from 39 studies having a com-bined total of 20,218 subjects. The prin-cipal conclusion of the review is thatdiabetes doubles the odds of depression.The OR of depression is more consistentacross studies than is the prevalence,which varies by sex, study design, subjectsource, and method of depression assess-ment. The overall OR estimate generalizesacross community and clinical settingsdespite differences in prevalence rates be-tween these settings. Both clinicians andepidemiologists can expect individualswith diabetes to be twice as likely to bedepressed than otherwise similar nondia-betic individuals in similar settings (i.e.,individuals selected by similar proce-dures, of the same sex, and assessed withcomparable depression assessment meth-ods). In contrast, the prevalence estimatemust be adjusted for moderators such assex.

Aggregate estimates based on all ofthe eligible studies indicate that major de-pression and elevated depression symp-toms were present, respectively, in 11 and31% of individuals with diabetes. Theodds of depression were significantlyhigher in women than in men with diabe-tes (OR 5 1.8), a pattern that mirrors thefemale preponderance of depression ob-

served in epidemiological surveys of thegeneral population (3537). The findingsare similar to the unadjusted rates re-ported in other medical illnesses (3840)and in an earlier review of the diabetesliterature by Gavard et al. (6) that in-cluded 18 studies. These investigatorsfound that major depression was presentin 14.7% and elevated depression symp-toms in 26% of diabetic patients. Thus, asmany as one in every three individualswith diabetes (at least in those participat-ing in clinical studies) has depression at a

level that impairs functioning and qualityof life (711), adherence to medical treat-ment (4143), and glycemic control (1),and increases the risk of diabetes compli-cations (2).

The prevalence of depression variedsystematically as a function of the methodused to identify depression cases and thestudy design. Furthermore, in both con-trolled and uncontrolled studies, the de-pression rates were approximately twoto three times higher in studies that usedself-report measures versus diagnostic

Figure 1Likelihood of depression by assessment method. The ORs were significantly higher indiabetic patients than in control subjects (p), but they did not differ as a function of method (1.9vs. 2.1, P 5 0.5). n 5 Number of controlled studies used in the calculations. p, Control subjects;f, diabetic patients.

Table 2Continued

Study


Sex(% female)

Age(years)

Race(% white)


method



Overall(%)

Males(%)

Females(%)

Slawson et al. (93) T1 & 2: 25e 20.0 46.6 MMPI-D $70 36.0 35.0 40.0 63.0

Samson et al. (94) T1: 111e

T2: 129e50.8 52.1 6 15.7 SCID/DSM-III-R T1: 16.2

T2: 10.9x: 13.3**

11.9** 13.9**

Data are means 6 SD, unless otherwise indicated. T1 5 type 1 diabetes, T2 5 type 2 diabetes, MINI/DSM-IV 5 Mini International Neuropsychiatric Interview/Diagnostic and Statistical Manual-IV, SCID 5 Structured Clinical Interview for DSM-III-R, PSE 5 Present State Examination, Zung 5 Zung Self-Rating DepressionScale, HADS 5 Hospital Anxiety and Depression Scale, CES-D 5 Center for Epidemiologic Studies-Depression Scale, MMPI-D 5 Minnesota Multiphasic PersonalityInventory-Depression Scale, IDD-L 5 Inventory to Diagnose Depression-Lifetime, SADS 5 Schedule for Affective Disorders and Schizophrenia.Sample randomly selected from a acommunity, bclinic or hospital, or cunspecified setting.Sample not randomly selected from a dcommunity, eclinic or hospital, or funspecified setting.*Includes diagnoses of dysthymia, major depression, and depression NOS; sex adjusted in the threshold scores for clinically significant depression; cutoff scorefor the symptom scale not specified; reported only the lifetime prevalence of depression for diabetic subjects, and was not included in the point prevalencecalculations; \depression scale of the Faschingbauer abbreviated version; diagnosis of depressive neurosis 1 depression NOS; #dysthymia; **diagnoses of majordepression, depression NOS, and dysthymia.



interviews. It is likely that the two ap-proaches identify somewhat different butoverlapping samples of depressed indi-viduals. Diagnostic interviews identifymajor depressive disorder but excludeother clinically relevant presentations.Self-report measures also identify mostcases of major depressive disorder. In pa-tients with diabetes, BDI cutoff scores of$10 and $16 have sensitivities of 0.98and 0.73, respectively, for major depres-sive disorder (31). Self-report measuresmay identify a broader spectrum of de-pression disorders (e.g., dysthymic disor-der, or minor or subsyndromal depres-sion) or symptoms that reflect comorbidpsychiatric illness (e.g., anxiety or sub-stance-abuse disorders) or general dis-tress.

The reason that depression was moreprevalent in the uncontrolled than in thecontrolled studies is unclear. One possi-bility is that the uncontrolled studies in-cluded a higher proportion of individualsrecruited from settings with higher preva-lences of depression (e.g., physician refer-rals or university-based clinics). Althoughsimilar ORs describing the likelihood ofdepression were found between the clin-ical and community studies, the uncon-trolled studies were comprised almostexclusively of clinic-based samples. Ofthe 11 community studies in the con-trolled subset, 8 restricted their samplesto older individuals; this further contrib-utes to lower rates of depression in thecontrolled studies, since depression is lesscommon in older than younger adults(44). There may also be other unmea-

sured differences in clinical or functionalcharacteristics that might account for thedifferences in the controlled and uncon-trolled studies. For example, the associa-tions of depression with hyperglycemiaand with increased risk of complicationsdescribed in recent meta-analytic reviews(1,2) support the hypothesis that the se-verity of diabetes and/or functional im-pairment may increase the risk fordepression.

A difference in the prevalence of de-pression in type 1 vs. type 2 diabetes

could not be established. The ORs fromthe controlled studies were nearly identi-cal between types, and aggregate esti-mates of prevalence using controlled anduncontrolled studies, segregated by de-pression assessment method, also yieldedequivalent depression rates. Many of thestudies, including some with the largestnumbers of patients, did not report thefraction of depressed individuals by typeof diabetes. This omission exemplified amore general failure of many studies tofully characterize the depressed and non-depressed samples. Such information isneeded to assess the effects of other fac-tors (e.g., age, socioeconomic status, andseverity of diabetes) on the prevalence ofdepression. In particular, failure to reportrace or ethnicity is common in the psy-chosocial literature on diabetes (45).

The findings of this review echo theobservation, first made by Willis (46) in1684, that depression is associated withdiabetes. The complex interactions ofphysical, psychological, and genetic fac-tors that contribute to this association re-main uncertain. Depression may occursecondary to the hardships of advancingdiabetes or to diabetes-related abnormal-ities in neurohormonal or neurotransmit-ter function (4749). On the other hand,evidence from prospective studies in theU.S. and Japan indicates that depressiondoubles the risk of incident type 2 diabe-

Figure 2Aggregate prevalence of depression determined from self-report scales or diagnosticinterviews in controlled and uncontrolled studies, and in all studies combined. Estimates based onself-report scales were significantly higher than those based on diagnostic interviews in all threecomparisons. Depression was also higher in uncontrolled studies than in controlled studies thatused the same depression assessment method (diagnostic interviews: 14.2 vs. 9.0%; self-reportscales: 34.9 vs. 26.1%; P , 0.001 for both). p, Diagnostic interview; f, self-report scale. *P ,0.001 between methods.

Table 3Unadjusted prevalence of depression in controlled and uncontrolled studies andsubsets thereof

Grouping ofstudies

Controlled studiesDiabetic subjects:

uncontrolledstudies

Diabetic subjects:controlled 1uncontrolled

studiesNondiabetic

subjectsDiabeticsubjects

All studies 11.4 (18) 20.5 (18)* 29.7 (21) 25.3 (39)Type 1 8.6 (3) 21.7 (3)* 21.2 (10) 21.3 (13)Type 2 6.4 (7) 16.5 (7)* 33.8 (8) 27.0 (15)Male 9.3 (7) 15.0 (7)* 20.7 (8) 18.0 (15)Female 16.3 (7) 24.3 (7)* 33.0 (8) 28.2 (15)Community 12.7 (11) 19.0 (11)* 39.7 (1) 20.1 (12)Clinic 15.1 (7) 26.7 (7)* 32.7 (19) 31.7 (26)Diagnostic interview 5.0 (7) 9.0 (7)* 14.2 (7) 11.4 (14)Self-report 14.4 (11) 26.1 (11)* 34.9 (14) 31.0 (25)

Data are % (n); n indicates number of studies used in the calculation. *The prevalence of depression wasgreater in diabetic subjects compared with nondiabetic control subjects (P , .001); the prevalence ofdepression in diabetic individuals was greater in uncontrolled studies compared with controlled studies (P ,0.05).



tes independent of its association withother risk factors (50,51). In patients withpreexisting diabetes, depression is an in-dependent risk factor for coronary heartdisease, and appears to accelerate the pre-sentation of coronary heart disease (5254). Additional studies are needed toidentify the behavioral and physiologicalmechanisms that account for these findings.

This review has several limitations.First, publication bias (i.e., nonpublica-tion of studies that fail to find the phe-nomenon of interest) may limit thegeneralizability of the findings. However,this possibility is diminished by the factthat depression was not the principal fo-cus of many of the included studies andwas instead only one of a number of mea-sured psychosocial variables. Second, thedepression prevalence estimates may beunstable due to the small sample sizes ofsome studies, the small number of stud-ies, and the fact that many of the sampleswere not population based. Third, themethods used to calculate the overallprevalence of depression in diabetic sub-jects were certainly suboptimal in that wewere unable to perform a multivariateanalysis controlling for all potential mod-erators. The fact that the ORs were uni-formly similar in the bivariate tests wasencouraging, and suggests that the two-fold increased likelihood of depressionassociated with diabetes is the most ro-bust and generalizable finding of this re-view. Additional studies would berequired to determine more precisely theprevalence of major depression and ele-vated depression symptoms in the generalpopulation of individuals with diabetes.These studies should carefully measureand report potential moderators so thatboth adjusted and unadjusted depressionprevalence can be calculated and in-cluded in the findings.

The third U.S. National Health andNutrition Education Examination Surveyfound that 49% of insulin-treated diabe-tic patients and 56% of those on oral hy-poglycemic agents had HbA1c values,8.0%, and very few patients sustainedHbA1c levels ,7.0%, the goal set by theAmerican Diabetes Association (55,56).Depression may oppose efforts to achievenormoglycemia via behavioral (4143)and physiological (57 60) pathways,and, as shown in this review, is clinicallyrelevant in nearly one of every three pa-tients with diabetes. Successful treatmentof depression is associated with improve-

ments in glycemic control (35). Never-theless, two of every three cases ofdepression are left untreated by primarycare physicians (61). Better recognitionand better treatment of depression are im-portant in themselves, but they could alsoimprove medical outcome in a substantialportion of patients with diabetes. Thismeta-analysis helps to define the preva-lence of depression in diabetes using datafrom available studies and firmly estab-lishes the increased risk of comorbidity.

Acknowledgments This work was sup-ported in part by grants from the National In-stitutes of Health (DK-36452, DK-553060,and 5-P60-DK-290579).

The authors are grateful to Samuel J. Lust-man for his support.

References1. Lustman PJ, Anderson RJ, Freedland KE,

de Groot M, Carney RM: Depression andpoor glycemic control: a meta-analytic re-view of the literature. Diabetes Care 23:434442, 2000

2. de Groot M, Anderson RJ, Freedland KE,Clouse RE, Lustman PJ: Association of di-abetes complications and depression intype 1 and type 2 diabetes: a meta-analy-sis (Abstract). Diabetes 49:A63, 2000

3. Lustman PJ, Griffith LS, Clouse RE,Freedland KE, Eisen SA, Rubin EH, Car-ney RM, McGill JB: Effects of nortriptylineon depression and glucose regulation indiabetes: results of a double-blind, place-bo-controlled trial. Psychosom Med 59:241250, 1997

4. Lustman PJ, Griffith LS, Freedland KE,Kissel SS, Clouse RE: Cognitive behaviortherapy for depression in type 2 diabetes:a randomized controlled trial. Ann InternMed 129:613621, 1998

5. Lustman PJ, Freedland KE, Griffith LS,Clouse RE: Fluoxetine for depression indiabetes: a randomized double-blind pla-cebo-controlled trial. Diabetes Care 23:618623, 2000

6. Gavard JA, Lustman PJ, Clouse RE: Prev-alence of depression in adults with diabe-tes: an epidemiological evaluation. Diabe-tes Care 16:11671178, 1993

7. Koenig HG, George LK, Peterson BL,Pieper CF: Depression in medically illhospitalized older adults: prevalence char-acteristics, and course of symptoms ac-cording to six diagnostic schemes. Am JPsychiatry 154:13761383, 1997

8. Sullivan M, LaCroix A, Russo J, Swords E,Sornson M, Katon W: Depression in cor-onary heart disease: what is the appropri-ate diagnostic threshold? Psychosomatics40:286292, 1999

9. Lyness JM, King DA, Cox CA, YoedionoZ, Caine ED: The importance of subsyn-dromal depression in older primary carepatients: prevalence and associated func-tional disability. J Am Geriatr Soc 47:647652, 1999

10. Judd LL, Rapaport MH, Paulus MP,Brown JL: Subsyndromal symptomaticdepression: a new mood disorder? J ClinPsychiatry 55:1828, 1995

11. Jacobson AM, de Groot M, Samson JA:The effects of psychiatric disorders andsymptoms on quality of life in patientswith type I and type II diabetes mellitus.Qual Life Res 6:1120, 1997

12. Penninx BW, Geerlings SW, Deeg DJ, vanEijk JT, van Tilburg W, Beekman AT: Mi-nor and major depression and the risk ofdeath in older persons. Arch Gen Psychia-try 56:889895, 1999

13. Frasure-Smith N, Lesperance F, TalajicM: Depression following myocardial in-farction: impact on 6-month survival.JAMA 270:18191825, 1993

14. Carney RM, Rich MW, Freedland KE,Saini J, teVelde A, Simeone C, Clark K:Major depressive disorder predicts car-diac events in patients with coronary ar-tery disease. Psychosom Med 50:627633,1988

15. Miranda J, Munoz R: Intervention for mi-nor depression in primary care patients.Psychosom Med 56:136141, 1993

16. Eisenberg L: Treating depression and anx-iety in primary care: closing the gap be-tween knowledge and practice. N EnglJ Med 326:10801084, 1992

17. Mynors-Wallis LM, Gath DH, Lloyd-Thomas AR, Tomlinson D: Randomizedcontrolled trial comparing problem solv-ing treatment with amitriptyline and pla-cebo for major depression in primarycare. Br Med J 310:441446, 1995

18. The Cochrane Collaboration: ReviewManager 4.1.1. Nepean, Ontario, Canada,Wintertree Software, 2000

19. Robins LN, Cottler L, Buckholz K, Comp-ton W: The Diagnostic Interview Schedule:Version 4. St. Louis, MO, WashingtonUniversity, 1996

20. Williams JBW, Gibbon M, First MB,Spitzer RL, Davies M, Borus J, Howes MJ,Kane J, Pope HG, Rounsaville B, WittchenHU: The structured clinical interview forDSM-III-R (SCID). II. Multisite test-retestreliability. Arch Gen Psychiatry 49:630636, 1992

21. American Psychiatric Association: Diag-nostic and Statistical Manual of MentalDisorders. Washington, DC, AmericanPsychiatric Association, 1987

22. American Psychiatric Association: Diag-nostic and Statistical Manual of Mental Dis-orders. Washington, DC, American Psy-chiatric Association, 1994



23. Beck AT, Beamesderfer A: Assessmentof depression: the depression inventory.Mod Probl Pharmacopsychiatry 7:151169, 1974

24. Radloff LS: The CES-D Scale: a self-reportdepression scale for research in the gen-eral population. Appl Psychol Measurement1:385401, 1977

25. Songar A, Kocabasoglu N, Balcioglu I, Ka-raca E, Kocabasoglu C, Haciosman M,Somay G: The relationship between dia-betics metabolic control levels and psy-chiatric symptomatology. Integr Psychia-try 9:3440, 1993

26. Leedom L, Meehan WP, Procci W, ZeidlerA: Symptoms of depression in patientswith type II diabetes mellitus. Psychoso-matics 32:280286, 1991

27. Stone JB, Bluhm HP, White MI: Correlatesof depression among long-term insulin-dependent diabetics. Rehabilitation Psy-chology 29:8593, 1984

28. Palinkas LA, Barrett-Connor E, WingardDL: Type 2 diabetes and depressive symp-toms in older adults: a population-basedstudy. Diabet Med 8:532539, 1991

29. Wing RR, Marcus MD, Blair EH, EpsteinLH, Burton LR: Depressive symptomatol-ogy in obese adults with type II diabetes.Diabetes Care 13:170172, 1990

30. Mulrow CD, Williams JW, Gerety MB,Ramirez G, Montiel OM, Kerber C: Case-finding instruments for depression in pri-mary care settings. Arch Intern Med 122:913921, 1995

31. Lustman PJ, Clouse RE, Griffith LS, Car-ney RM, Freedland KE: Screening fordepression in diabetics using the Beck De-pression Inventory. Psychosom Med 59:2431, 1997

32. Tun PA, Perlmuter LC, Russo P, NathanDM: Memory self-assessment and perfor-mance in aged diabetics and non-diabet-ics. Exp Aging Res 13:151157, 1987

33. Penninx B, van Tilburg T, Boeke JP, DeegD, Kriegsman D, van Eijk J: Effects ofsocial support and personal coping re-sources on depressive symptoms: differ-ent for various chronic diseases? HealthPsychol 17:551558, 1998

34. Marcus MD, Wing RR, Guare J, Blair EH,Jawad A: Lifetime prevalence of major de-pression and its effect on treatment out-come in obese type II diabetic patients.Diabetes Care 15:253255, 1992

35. Kessler RC, Nelson CB, McGonagle KA,Lui J, Swartz MS, Blazer DC: Comorbidityof DSM-III-R major depressive disorder inthe general population: results from theU.S. National Comorbidity Survey. Br JPsychiatry 168:1730, 1996

36. Kessler RC, McGonagle KA, Zhao S, Nel-son CB, Hughes M, Eshleman S, WittchenHU, Kendler KS: Lifetime and 12-monthprevalence of DSM-III-R psychiatric dis-orders in the United States: results from

the National Cormorbidity Survey. ArchGen Psychiatry 51:819, 1994

37. Robins LN, Regier DA: Psychiatric Disor-ders in America: The Epidemiologic Catch-ment Area Study. New York, Free Press(McMillan), 1991

38. Carney RM, Freedland KE, Sheline YI,Weiss ES: Depression and coronary heartdisease: a review for cardiologists. ClinCardiol 20:196200, 1997

39. Bottomly A: Depression in cancer pa-tients: a literature review. Eur J CancerCare 7:181191, 1998

40. Frasure-Smith N, Lesperance F: Coronaryartery disease, depression and social sup-port only the beginning. Eur Heart J 21:10431045, 2000

41. DiMatteo MR, Lepper HS, Croghan TW:Depression is a risk factor for noncompli-ance with medical treatment: meta-analy-sis of the effects of anxiety and depressionon patient adherence. Arch Intern Med160:21012107, 2000

42. Littlefield CH, Craven JL, Rodin GM,Daneman D, Murray MA, Rydall AC: Re-lationship of self-efficacy and bingeing toadherence to diabetes regimen among ad-olescents. Diabetes Care 15:9094, 1992

43. McGill JB, Lustman PJ, Griffith LS, Freed-land KE, Gavard JA, Clouse RE: Relation-ship of depression to compliance withself-monitoring of blood glucose (Ab-stract). Diabetes 41:A84, 1992

44. Fava M, Kendler K: Major depressive dis-order. Neuron 28:335341, 2000

45. de Groot M, Lustman PJ: Depressionamong African-Americans with diabetes:a dearth of studies. Diabetes Care 24:407408, 2001

46. Willis T: Diabetes: A Medical Odyssey. NewYork, Tuckahoe, 1971

47. Trulson MH, Himmel C: Effects of insulinand streptozotocin-induced diabetes onbrain norepinephrine metabolism in rats.J Neurochem 44:18731876, 1985

48. MacKenzie R, Trulson M: Effects of insu-lin and streptozotocin-induced diabeteson brain tryptophan and serotonin me-tabolism in rats. J Neurochem 30:205211, 1978

49. Lustman PJ, Skor DA, Carney RM, San-tiago JV, Cryer PE: Stress and diabeticcontrol. Lancet 1:588, 1983

50. Eaton WW, Armenian HA, Gallo J, PrattL, Ford DE: Depression and risk for onsetof type II diabetes: a prospective popula-tion-based study. Diabetes Care 19:10971102, 1996

51. Kawakami N, Takatsuka N, Shimizu H,Ishibashi H: Depressive symptoms andoccurrence of type 2 diabetes among Jap-anese men. Diabetes Care 22:10711076,1999

52. Peyrot M, Rubin RR: Persistence of de-pressive symptoms in diabetic adults. Di-abetes Care 22:448452, 1999

53. Forrest KYZ, Becker DJ, Kuller LH, Wolf-son SK, Orchard TJ: Are predictors of cor-onary heart disease and lower-extremityarterial disease in type 1 diabetes thesame? A prospective study. Atherosclerosis148:159169, 2000

54. Lloyd C, Wilson R, Forrest K: Prior de-pressive symptoms and the onset of coro-nary heart disease (Abstract). Diabetes 46:3A, 1997

55. U.K. Prospective Diabetes Study Group:Intensive blood-glucose control with sul-phonylureas or insulin compared withconventional treatment and risk of com-plications in patients with type 2 diabetes(UKPDS 33). Lancet 352:837853, 1998

56. American Diabetes Association: Standardsof medical care for patients with diabetesmellitus (Position Statement). DiabetesCare 23 (Suppl. 1):S32S42, 2000

57. Levy MI, Davis KL: The neuroendocrinol-ogy of depression. In Schizophrenia andAffective Disorders: Biology and Drug Treat-ment. Rifkin A, Ed. Boston, John Wright/PSG, 1983

58. Ettigi P, Brown G: Psychoneuroendocri-nology of affective disorder: an overview.Am J Psychiatry 134:493501, 1977

59. Young EA, Haskett RF, Murphy-Wein-berg V, Watson SJ, Akil H: Loss of glu-cocorticoid fast feedback in depression.Arch Gen Psychiatry 48:693699, 1991

60. Jacobson L, Sapolsky R: The role of thehippocampus in feedback regulation ofthe hypothalamic-pituitary-adrenocorti-cal axis. Endocr Rev 12:118134, 1991

61. Lustman PJ, Harper GW: Nonpsychiatricphysicians identification and treatmentof depression in patients with diabetes.Compr Psychiatry 28:2227, 1987

62. Kokkonen E-R: Mental health and socialadaptation in young adults with juvenile-onset diabetes. Nordic J Psychiatry 49:175181, 1995

63. Popkin MK, Callies AL, Lentz RD, ColonEA, Sutherland DE: Prevalence of majordepression, simple phobia, and other psy-chiatric disorders in patients with long-standing type I diabetes mellitus. ArchGen Psychiatry 45:6468, 1988

64. Amato L, Paolisso G, Cacciatore F, FerraraN, Canonico S, Rengo F, Varricchio M:Non-insulin dependent diabetes mellitusis associated with a greater prevalence ofdepression in the elderly. Diabete Metab22:314318, 1996

65. Viinamaki H, Niskanen L, Uusitupa M:Mental well-being in people with non-in-sulin-dependent diabetes. Acta PsychiatrScand 92:392397, 1995

66. Weyerer S, Hewer W, Pfeifer-Kurda M,Dilling H: Psychiatric disorders and dia-betes: results from a community study.J Psychosom Res 33:633640, 1989

67. Black SA: Increased health burden associ-ated with comorbid depression in older



diabetic Mexican Americans. DiabetesCare 22:5664, 1999

68. Bourdel-Marchasson I, Dubroca B, Man-ciet G, Decamps A, Emeriau JC, DartiguesJF: Prevalence of diabetes and effect onquality of life in older French living in thecommunity: the PAQUID Epidemiologi-cal Survey. J Am Geriatr Soc 45:295301,1997

69. Rajala U, Keinanen-Kiukaanniemi S, Ki-vela S-L: Non-insulin-dependent diabetesmellitus and depression in a middle-agedFinnish population. Soc Psychiatry Psychi-atr Epidemiol 32:363367, 1997

70. Zhang J, Markides KS, Lee DJ: Health sta-tus of diabetic Mexican Americans: resultsfrom the Hispanic HANES. Ethn Dis 1:273279, 1991

71. Wells KB, Golding JM, Burnam MA: Af-fective, substance use, and anxiety disor-ders in persons with arthritis, diabetes,heart disease, high blood pressure, orchronic lung conditions. Gen Hosp Psychi-atry 11:320327, 1989

72. Robinson N, Fuller H, Edmeades SP: De-pression and diabetes. Diabet Med 5:268274, 1988

73. Friis R, Nanjundappa G: Diabetes, de-pression and employment status. Soc SciMed 23:471475, 1986

74. Murrell SA, Himmelfarb S, Wright K:Prevalence of depression and its correlatesin older adults. Am J Epidemiol 117:173185, 1983

75. Berlin I, Bisserbe JC, Eiber R, Balssa N,Sachon C, Bosquet F, Grimaldi A: Phobicsymptoms, particularly the fear of bloodand injury, are associated with poor gly-cemic control in type 1 diabetic adults.Diabetes Care 20:176178, 1997

76. Cohen ST, Welch G, Jacobson AM, deGroot M, Samson J: The association oflifetime psychiatric illness and increased

retinopathy in patients with type 1 diabe-tes mellitus. Psychosomatics 38:98108,1997

77. Mayou R, Peveler R, Davies B, Mann J,Fairburn C: Psychiatric morbidity in youngadults with insulin-dependent diabetesmellitus. J Psychol Med 21:639645, 1991

78. Winocour PH, Main CJ, Medlicott G,Anderson DC: A psychometric evaluationof adult patients with type 1 (insulin-de-pendent) diabetes mellitus: prevalence ofpsychological dysfunction and relation-ship to demographic variables, metaboliccontrol and complications. Diabetes Res14:171176, 1990

79. Surridge DHC, Williams-Erdahl DL, Law-son JS, Donald MW, Monga TN, Bird CE,Letemendia FJJ: Psychiatric aspects of di-abetes mellitus. Br J Psychiatry 145:269276, 1984

80. Connell CM, Storandt M, Lichty W: Im-pact of health belief and diabetes-specificpsychosocial context variables on self-care behavior, metabolic control and de-pression of older adults with diabetes.Behav Health Aging 1:183196, 1990

81. Naliboff BD, Rosenthall M: Effects of ageon complications in adult onset diabetes.J Am Geriatr Soc 37:838842, 1989

82. Geringer ES, Perlmuter LC, Stern TA,Nathan DM: Depression and diabeticneuropathy: a complex relationship. JGeriatr Psychiatry Neurol 1:1115, 1988

83. Biglan A, Toobert D, Farmer R, Wilson W,Campbell D: Depression among personswith type II diabetes mellitus. Unpub-lished

84. Kohen D, Burgess AP, Catalan J, Lant A:The role of anxiety and depression inquality of life and symptom reporting inpeople with diabetes mellitus. Qual LifeRes 7:197204, 1998

85. Peyrot M, Rubin RR: Levels and risks of

depression and anxiety symptomatologyamong diabetic adults. Diabetes Care 20:585590, 1997

86. Bailey BJ: Mediators of depression inadults with diabetes. Clin Nurs Res 5:2842, 1996

87. Haire-Joshu D, Heady S, Thomas L,Schechtman K, Fisher EB: Depressivesymptomatology and smoking amongpersons with diabetes. Res Nurs Health 17:273282, 1994

88. Jalenques I, Tauveron I, Albuisson E, Lon-jaret D, Thieblot P, Coudert AJ: Preva-lence of anxiety and depressive symptomsin patients with type 1 and 2 diabetes. RevMed Suisse Romande 113:639646, 1993

89. Padgett DK: Sociodemographic and dis-ease-related correlates of depressive mor-bidity among diabetic patients in Zagren,Croatia. J Nerv Ment Dis 181:123129,1993

90. Lee PWH, Lam KSL, Lieh-Mak F, ChungKF, So T: Emotional maladjustment,physical malaise and diabetic control inyoung Chinese patients with diabetes.Psychol Health Med 1:119127, 1996

91. Von Dras DD, Lichty W: Correlates of de-pression in diabetic adults. Behav HealthAging 1:7984, 1990

92. Lustman PJ, Griffith LS, Clouse RE, CryerPE: Psychiatric illness in diabetes: rela-tionship to symptoms and glucose con-trol. J Nerv Ment Dis 174:736742, 1986

93. Slawson PF, Flynn WR, Kollar EJ: Psycho-logical factors associated with the onset ofdiabetes mellitus. JAMA 185:166170,1963

94. Samson JA, de Groot M, Jacobson AM:Comorbid psychiatric diagnoses in menand women with type 1 and type 2 diabe-tes mellitus. Unpublished



anderson_depresion_diabetes_comorbida2001.pdf

Documents

Transcript of anderson_depresion_diabetes_comorbida2001.pdf