Frequency and risk factors associated with depression in elderly visiting Primary Health Care (PHC) settings: Findings from the Cretan Aging Cohort

Background: Depression is a frequent and serious disease often undiagnosed in elderly. We aimed to examine (1) the prevalence/correlates associated with depression and (2) the magnitude of depression underdiagnosis/related factors, in a large community-dwelling elderly population in Crete/Greece visiting Primary Health Care (PHC) settings. Methods: A sub-sample of 2428 non-demented individuals with Mini Mental State Examination (MMSE)>19 were recruited from a population-based cohort of 3140 elderly (>60 years). In phase I, all participants were assessed with a structured questionnaire including demographics, life-style, sleep, physical health and cognitive function. Diagnosis of depression was based on history/treatment of depression. In phase II, in a sub-sample of 293 participants, depression was diagnosed after a thorough extensive neuropsychiatric/neuropsychological evaluation. Results: In phase I, 10.8% reported a diagnosis/treatment of depression. In multivariate analysis, female gender, lack of physical activity, caregiver dependence and use of benzodiazepines were associated with depression. In phase II, depression was prevalent in 28.7% of our sample, whereas about 61% of those diagnosed with depression were undetected in phase I. Factors associated with non-detection of depression were sleep complaints/subjective short sleep, low MMSE and medical comorbidities. Two screening questions related to mood increased possibility to detect undiagnosed depression up to 90%. Limitations: Due to the cross-sectional design, causality between correlated factors cannot be examined. Conclusion: Prevalence of depression is high and largely underdiagnosed in elderly in Crete/Greece. Sleep and memory complaints appear to be surrogate markers of depression, whereas screening mood in PHC settings may significantly increase detection of depression in elders.