Prevalence and correlates of undiagnosed depression among U.S. adults with diabetes: The Behavioral Risk Factor Surveillance System, 2006

Abstract

Aims

Many people with depression may be undiagnosed and thus untreated. We sought to assess the prevalence and correlates of undiagnosed depression among adults with diabetes.

Methods

Data of U.S. adults from the Behavioral Risk Factor Surveillance System in 2006 were analyzed. Cox proportional hazard regression analysis was used to estimate prevalence ratios (PRs) and 95% confidence intervals (CIs) of correlates for undiagnosed depression.

Results

The unadjusted and age-adjusted prevalences of undiagnosed depression were 8.7% and 9.2%. About 45% of diabetes patients with depression were undiagnosed. After adjustments for all correlates, female gender (PR, 1.4; 95% CI: 1.1–1.8), poor or fair health (PR, 2.8; 95% CI: 2.1–3.6), lack of social and emotional support (PR, 2.5; 95% CI: 1.8–3.3), life dissatisfaction (PR, 3.5; 95% CI: 2.2–5.5), use of special equipment (PR, 1.4; 95% CI: 1.1–1.8), no leisure-time physical activity (PR, 1.5; 95% CI: 1.2–1.9), and comorbid cardiovascular disease (PR, 1.5; 95% CI: 1.2–1.9) were associated with undiagnosed depression.

Conclusions

Undiagnosed depression among people with diabetes was common. Because depression is associated with increased risk of diabetes-related complications, early detection of depression is needed in clinical settings.

Introduction

In 2007, approximately 23.5 million U.S. adults aged 20 years or older (10.7%) had diabetes, with 90% to 95% of all diagnosed cases being type 2 diabetes [1]. Among people with diabetes, depression has been associated with poor general health [2], functional limitations or disabilities [2], [3], and increased health care use and expenditure [4]. In addition, depression is associated with poor glycemic control [5] and increased risk for diabetes complications [6], [7]. Furthermore, emerging evidence shows that depression is associated with increased risk for death from coronary heart disease and from all causes [8].

Estimates of the prevalence of depression in the United States have varied widely, ranging from 3.8% to 27.3% in studies based on structured patient interviews, and from 11.5% to 60.7% in studies based on self-administered questionnaires [9]. In a recent report based on 2006 data from Behavioral Risk Factor Surveillance System (BRFSS), we estimated the prevalence of both major and minor depression among people with diabetes to be 8.3% [10]. Depression, however, is often undiagnosed or unrecognized in clinical settings. For example, results from the World Health Organization (WHO) Collaborative Study on Psychological Problems in General Health Care in 14 countries have showed that 42% to 48% of depressed people were not recognized as being “psychiatric cases” by primary care physicians [11]. And a recent study showed that less than 25% of hospital patients with acute coronary syndrome and moderate to severe depressive symptoms had documentation indicating their symptoms were recognized before they were discharged from the hospital [12]. According to the Pathway Study [13], about 49% of patients with major depression and diabetes were unrecognized by the primary health care system. However, little is known about the prevalence of undiagnosed depression among people with diabetes in the general population.

Previous studies have shown that female sex, younger age, low education level, low family income, perceived poor physical health, smoking, body mass index (BMI) ≥30 kg/m², a glycosylated hemoglobin A1C concentration ≥8.0%, and ≥2 diabetes complications were each significantly associated with increased risk for major depression [14], [15]; that a high level of supportive relationships was associated with a decreased risk for depressive symptoms (defined as a score of 16 or higher on the Center for Epidemiological Studies-Depression [CES-D] scale) among rural African American adults with type 2 diabetes [16]; and that cardiovascular disease was associated with an increased risk for diagnosed depression (ascertained by International Classification of Diseases, Ninth Revision [ICD-9] standards or the use of antidepressant medication) [17].

It is unknown whether these factors have an association with undiagnosed depression similar to their association with diagnosed depression. Assessment of correlates for undiagnosed depression may have etiological implications because people who have depression that has never been diagnosed or recognized by physicians may have characteristics different from those of people whose depression has been diagnosed [17]. Therefore, in this study we sought to (a) assess the prevalence of undiagnosed depression among U.S. adults with diabetes, (b) identify the possible correlates for undiagnosed depression, and (c) examine the possible effect modification of social and emotional support on the association between patients’ sense of “life satisfaction” and their risk for undiagnosed depression.