The prevalence and impact of type 2 diabetes in Appalachia is understudied despite the presence of high-risk socioeconomic conditions (e.g., low levels of education and income). Appalachian counties experience greater burdens of poverty, income inequality, unemployment (1), and diabetes compared with non-Appalachian counties (2–5). Rates of comorbid depression have not been documented in this region.

Patients with type 2 diabetes are two times more likely to experience depressive symptoms than their peers without diabetes (6). Depression prevalence in studies using self-report depression inventories was found to be 32.9%, with lower rates (28.5%) observed in studies using diagnostic interview schedules (6,7). Depressive symptoms are associated with worsened blood glucose levels (8), diabetes complications (9), increased functional disability (10), worsened adherence to diabetes regimen (11), higher ambulatory care costs (12), and increased mortality (13).

The current study was conducted to identify rates of self-reported depression and to identify the socioeconomic and medical correlates of depression among type 2 diabetic patients attending family medicine and endocrinology appointments from rural Appalachian counties of southeastern Ohio and West Virginia. It was hypothesized that poverty would increase the risk of comorbid depression in this region.

Participants, recruited from family medicine and endocrinology practices, were diagnosed with type 2 diabetes for ≥1 year and aged ≥18 years, with the ability to provide informed consent. Patients (n = 628) were identified from billing and registry databases. Of these, 288 patients responded to letters of invitation for study participation from their medical provider. Thirty-two participants did not meet inclusion criteria, and 55 declined participation (9.8% refusal rate). A total of 201 (36% response and 78.5% cooperation rates) completed questionnaires and consented to release medical record data.

Participants completed the Beck Depression Inventory (BDI)-II (14), which assesses depressive symptom severity for 2 weeks prior to administration. Demographic questionnaire variables included age, ethnicity, Appalachian heritage, marital status, income, and educational status. Economic resource variables included employment status, total annual income, home ownership, appraisal of financial status and optimism, and presence of health insurance (15).

Medical records were reviewed for the year before study participation. Variables included physician diagnosis of type 2 diabetes, A1C (normal reference range 4.3–5.7% [DCA2000+ Analyzer; Bayer Diagnostics]), total cholesterol, BMI, diabetes complications and risk factors, and prescribed medications.

Analyses were conducted using SAS version 8.0e (16). Logistic regression analyses were conducted to assess predictors of depression status, coded dichotomously as “depressed” (BDI-II score ≥20) or “nondepressed” (≤19). Overall model fit was evaluated using Hosmer-Lemeshow goodness-of-fit statistic (17). Test of discrimination of depression status was evaluated by examining the area under the receiver-operating characteristic curve (ROC) (17). Multiple regression analysis was used to assess correlates of depression severity, using BDI-II score as a continuous outcome measure.

The sample was predominantly female (60%, n = 121), white (94%, n = 188), married (71%, n = 141), completed high school or less (53%, n = 106), and had a mean age of 57 ± 12 years. Forty-five percent (n = 87) reported a total annual income of $20,000 or less, with 43% (n = 87) working outside the home. Eighty percent (n = 161) reported owning their own home.

The mean duration of type 2 diabetes was 9.7 ± 8.2 years, with the majority of patients receiving oral agents (49%, n = 98) or combination treatment (27%, n = 54). Mean A1C value was 7.5 ± 1.6%. Mean number of diabetes complications was 1.1 ± 1.2, with cardiovascular disease (22.1%, n = 44), neuropathy (16.6%, n = 33), and coronary heart disease (13.6%, n = 27) noted most frequently. The mean number of prescribed medications was 7.7 ± 3.9.

The mean BDI-II score was 14.5 ± 12.0. Of participants, ∼58% (n = 116) reported few depressive symptoms (score 0–13), 12% (n = 23) with mild symptoms (14–20), 16% (n = 31) with moderate symptoms (21–28), and 16% (n = 30) with severe symptoms (≥29). When the moderate and severe categories were combined, 31% (n = 61) of participants reported clinically significant levels of depressive symptoms.

Results of the logistic and multiple regression analyses are shown in Table 1. Younger age, unemployment, lack of home ownership, greater number of medications, and higher BMI significantly predicted depression status. Depression severity was associated with younger age, unemployment, and greater number of prescribed medications.

In this study, 31% of Appalachian participants reported comorbid diabetes and depression, a rate comparable with average prevalence estimates in studies using self-report questionnaire assessment methods (5,7).

Models predicting depression status and severity found younger, unemployed participants and those who did not own their own homes at increased risk of depression. These findings are consistent with those observed in urban ethnic minority samples (18,19), suggesting that poverty may represent a common risk factor across ethnic groups for comorbid depression in diabetic individuals. The cross-sectional study design does not permit determination of the casual direction of this relationship.

Contrary to previously published studies (8), we did not observe significant differences in average A1C by depression status. This may be due to the limited variability in A1C observed in the sample as a whole.

Limitations to the current study include the cross-sectional nature of the study design, the use of a clinic patient population, limited recruitment response rate, and the use of self-report questionnaires. Response rates were comparable with those observed in studies of other low-income populations (20–22) and may reflect barriers endemic to the rural research (23). Although rates of depression are higher in self-report questionnaires than psychiatric interviews (6), observed rates were not inflated compared with other national samples.

Findings from this study document a substantial burden of clinically significant depressive symptoms (one in three participants) in this Appalachian region. Poverty increases the risk of depression in diabetes. Further investigation into treatment availability and effectiveness is needed in light of the economic vulnerability of this region.

Table 1—

Regression models to assess socioeconomic and medical correlates of depression

Logistic regression model predicting depression status
PredictorsBSEWaldOR (95% CI)
Younger age −0.04 0.02 5.39* 0.96 (0.93–0.99) 
Unemployment 1.07 0.45 5.75* 2.92 (1.22–6.99) 
Lack of home ownership 1.34 0.49 7.31* 3.81 (1.45–10.03) 
No. of medications 0.12 0.06 4.56* 1.13 (1.01–1.26) 
Duration of diabetes −0.03 0.02 1.15 0.98 (0.93–1.02) 
BMI 0.05 0.03 4.43* 1.10 (1.00–1.11) 
Hosmer-Lemeshow goodness-of-fit test χ2(8, n = 152) = 4.22; P = 0.84  ROC = 0.80 R2 = 0.23 
Logistic regression model predicting depression status
PredictorsBSEWaldOR (95% CI)
Younger age −0.04 0.02 5.39* 0.96 (0.93–0.99) 
Unemployment 1.07 0.45 5.75* 2.92 (1.22–6.99) 
Lack of home ownership 1.34 0.49 7.31* 3.81 (1.45–10.03) 
No. of medications 0.12 0.06 4.56* 1.13 (1.01–1.26) 
Duration of diabetes −0.03 0.02 1.15 0.98 (0.93–1.02) 
BMI 0.05 0.03 4.43* 1.10 (1.00–1.11) 
Hosmer-Lemeshow goodness-of-fit test χ2(8, n = 152) = 4.22; P = 0.84  ROC = 0.80 R2 = 0.23 
Multiple regression model to assess correlates of depression severity
PredictorsBβSE—
Younger age −0.27 −0.27† 0.09 — 
Unemployment 4.30 0.18* 2.01 — 
Income −0.64 −0.09 0.67 — 
Lack of home ownership 2.07 0.07 2.34 — 
Ability to make ends meet −1.79 −0.18 0.99 — 
Exercise −0.83 −0.13 0.49 — 
BMI 0.11 0.08 0.11 — 
No. of complications −1.07 −0.13 0.73 — 
Duration of diabetes −0.05 −0.03 0.13 — 
No. of medications 0.45 0.16* 0.23 — 
Overall model test F = 8.33† R2 = 0.43 Adjusted R2 = 0.43 — 
Multiple regression model to assess correlates of depression severity
PredictorsBβSE—
Younger age −0.27 −0.27† 0.09 — 
Unemployment 4.30 0.18* 2.01 — 
Income −0.64 −0.09 0.67 — 
Lack of home ownership 2.07 0.07 2.34 — 
Ability to make ends meet −1.79 −0.18 0.99 — 
Exercise −0.83 −0.13 0.49 — 
BMI 0.11 0.08 0.11 — 
No. of complications −1.07 −0.13 0.73 — 
Duration of diabetes −0.05 −0.03 0.13 — 
No. of medications 0.45 0.16* 0.23 — 
Overall model test F = 8.33† R2 = 0.43 Adjusted R2 = 0.43 — 
*

P < 0.05,

†

P < 0.0001.

View Large

Partial support for this study was received from the National Institutes of Health (DK071545).

1.
Black DA, Sanders SG: Labor market performance, poverty, and income inequality in Appalachia [article online],
2004
. Available from http://www.arc.gov/images/reports/labormkt/labormkt.pdf. Accessed 18 July 2006
2.
Mokdad AH, Bowman BA, Ford ES, Vinicor F, Marks JS, Koplan JP: The continuing epidemics of obesity and diabetes in the United States.
JAMA
 
286
:
1195
–1200,
2001
3.
Centers for Disease Control:
BFRSS 2005 Data, State of Ohio
[article online],
2005
. Available from http://apps.nccd.cdc.gov/brfss/display.asp?cat=DB&yr=2002&qkey=1364&state=US. Accessed 28 July 2006
4.
Voinovich Center for Leadership and Public Affairs:
Appalachian Rural Health Institute Health Needs Assessment Survey Report of Findings
. Athens, OH, Ohio University Voinovich Center for Leadership and Public Affairs, 2006
5.
Dabney B Gosschalk A: Diabetes in Rural America. In
Rural Healthy People 2010
. College Station, TX, Texas A&M University Health Science Center, School of Rural Public Health, 2006
6.
Anderson RJ, Freedland KE, Clouse RE, Lustman PJ: The prevalence of comorbid depression in adults with diabetes: a meta-analysis.
Diabetes Care
 
24
:
1069
–1078,
2001
7.
Anderson D, Horton C, O'Toole M, Brownson C, Fazzone P, Fisher E: Integrating depression care with diabetes care in real-world settings: lessons from the Robert Wood Johnson Foundation Diabetes Initiative.
Diabetes Spectrum
 
20
:
10
–16,
2007
8.
Lustman PJ, Anderson RJ, de Groot M, Carney RM, Clouse RE: Depression and poor glycemic control: a meta-analytic review of the literature.
Diabetes Care
 
23
:
934
–942,
2000
9.
de Groot M, Anderson RJ, Freedland KE, Clouse RE, Lustman PJ: Association of depression and diabetes complications: a meta-analysis.
Psychosom Med
 
63
:
619
–630,
2001
10.
Egede LE: Diabetes, major depression and functional disability among U.S. adults.
Diabetes Care
 
27
:
421
–428,
2004
11.
Ciechanowski PS, Katon WJ, Russo JE: Depression and diabetes: impact of depressive symptoms on adherence, function, and costs.
Arch Intern Med
 
160
:
3278
–3285,
2000
12.
Egede LE, Zheng D, Simpson K: Comorbid depression is associated with increased health care use and expenditures in individuals with diabetes.
Diabetes Care
 
25
:
464
–470,
2002
13.
Zhang X, Norris SL, Gregg EW, Cheng YJ, Beckles G, Kahn HS: Depressive symptoms and mortality among persons with and without diabetes.
Am J Epidemiol
 
161
:
652
–660,
2005
14.
Beck AT, Steer RA, Brown GK:
BDI-II Beck Depression Inventory Manual
. 2nd ed. San Antonio, TX, Harcourt, Brace & Company,
1996
15.
Auslander WF, Haire-Joshu D, Houston C, Williams JH, Krebill H, Assessing economic assets: relationship to health behaviors and readiness to prevent NIDDM in black women (Abstract).
Diabetes
 
46(Suppl. 1)
:
135A
16.
SAS Institute: Version 8.03e [computer software]. Cary, NC,
2001
17.
Tabachnick PFL:
Using Multivariate Statistics
. 4th ed. New York, Allyn & Bacon, 2001
18.
Jackson-Triche ME, Greer Sullivan J, Wells KB, Rogers W, Camp P, Mazel R: Depression and health-related quality of life in ethnic minorities seeking care in general medical settings.
J Affect Disord
 
58
:
89
–97,
2000
19.
de Groot M, Auslander W, Williams JH, Sherraden M, Haire-Joshu D: Depression and poverty among African-American women at risk for type 2 diabetes.
Ann Behav Med
 
25
:
172
–181,
2003
20.
Blumenthal DS, Sung J, Coates R, Williams J, Liff J: Recruitment and retention of subjects for a longitudinal cancer prevention study in an inner-city black community.
Health Serv Res
 
30
:
197
–205,
1995
21.
Moorman PG, Newman B, Millikan RC, Tse C, Sandler DP: Participation rates in a case-control study: the impact of age, race, and race of interviewer.
Ann Epidemiol
 
9
:
188
–195,
1999
22.
Gilliss CL, Lee KA, Gutierrez Y, Taylor D, Beyene Y, Neuhaus J, Murrell N: Recruitment and retention of healthy minority women into community-based longitudinal research.
J Womens Health Gend Med
 
10
:
77
–85,
2001
23.
DiBartolo MC, McCrone S: Recruitment of rural community-dwelling older adults: barriers, challenges, and strategies.
Aging Ment Health
 
7
:
75
–82,
2003

Published ahead of print at http://care.diabetesjournals.org on 12 March 2007. DOI: 10.2337/dc06-1599.

A table elsewhere in this issue shows conventional and Système International (SI) units and conversion factors for many substances.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C Section 1734 solely to indicate this fact.

DIABETES CARE
2007