July 2010
Reviewer: Jeff C. Huffman, MD
Depressive disorder and incident diabetes mellitus: the effect of characteristics of depression
Campayo A, de Jonge P, Roy JF, et al
Am J Psychiatry 2010; 167(5):580-588
Background: Depression and diabetes mellitus (DM) are common and their prevalence is rising. Several studies have suggested that depression is a risk factor for the development of DM, yet most studies have been methodologically limited. A recent study found that major depression is associated with incident DM, but many cases of clinical depression in the community (especially in older adults) may not meet criteria for major depression. A study of ‘real-world’ community dwelling persons could help to clarify the connection between depressive disorders and incident DM.
Methods: This was a secondary analysis of a project in Spain that evaluated the incidence and risk factors (including medical illness) for depression and dementia. In this project, carefully trained lay interviewers performed 25-90 minute structured interviews at subjects’ homes at baseline, 2.5 years, and 5 years (interviewers were blinded to prior interviews’ results); research psychiatrists met with subjects for whom there was diagnostic unclarity. Depression was assessed using the Geriatric Mental State Schedule, with ‘AGECAT’ criteria used to assess severity of cases; DM was assessed through a series of self-report questions. Data on multiple potential confounding variables was also collected to assess the independent connection between depression and DM.
Results: A total of 4803 subjects age 55 or over enrolled in the study; 3521 had no missing data at baseline and had no DM at baseline and were included in this analysis. Overall, 379 (10.8%) were diagnosed with a depressive disorder at baseline, and 163 subjects (25 of whom had depression at the baseline assessment) developed DM during the 5 year follow-up period. On multivariable logistic regression, depression was associated with the development of DM, independent of multiple demographic, medical, and medication-related factors (hazard ratio 1.65 (95% CI= 1.02-2.66]; p=.04). On further analysis, patients with several characteristics of depression—nonsevere depression, depression persisting at the 2.5 year follow-up, and untreated depression—were all associated with incident DM.
Discussion: This interesting data from a large cohort of patients supports the growing literature that depression is an independent risk factor for multiple serious medical illnesses. In this case, depression, even if relatively mild, was associated with developing DM in older adults, especially if it went untreated and continued to persist. This underscores a potential need to address even milder forms of depression, both to improve quality of life and to reduce medical risk, in older adults.
The study has some methodological limitations as well as perhaps some clinical ones. An important methodologic limitation was the use of self-report to diagnose DM—without lab results or medical records required. In addition, though there were a large number of subjects in the overall study, there were only 25 depressed subjects who developed DM, and then when depression subtypes began to filter down these 25 subjects even further, it did lead to forming conclusions about the depression/DM connection based on small numbers of patients. Further work may well be required, especially about the findings regarding the characteristics of depression. Furthermore, clinically, studies examining the effects of milder depression often put clinicians in a bind—we do not have clear, evidence-based treatments for depression that is milder than major depression (aside from dysthymia), and therefore it is not clear that we should be treating patients with milder depression, given that we do not have evidence-based treatments to use.
At the same time, this is important work that further links depression to negative medical outcomes, and provides one more reason for clinicians to lean toward treating patients whose depression falls into a grey area of treatment versus no treatment, as untreated and mild depression was independently associated with subsequent development of DM in this prospective study.