Longitudinal epidemiology of major depression as assessed by the Brief Patient Health Questionnaire (PHQ-9)

Abstract

Background Most diagnostic interviews depend on recall of past symptoms and may be vulnerable to recall bias. The objective of this study was to describe the epidemiology of major depression using an approach that is less reliant on recall using an assessment scale, the 9-item Brief Patient Health Questionnaire (PHQ-9). The pattern observed is relevant both to understanding major depression epidemiology and to assessing a possible role for the PHQ-9 as a screening instrument.Method Random digit dialing was used to select a sample of 3304 community residents. Each respondent was assessed with a baseline interview followed by a series of 6 subsequent follow-up interviews 2 weeks apart.Results Prevalence was between 2.5% and 3.3% during each interview, consistent with prior reports. The incidence of new episodes was surprisingly high, and many of the episodes were brief. Similarly, high rates of recovery (according to PHQ-9 scoring) were seen early in follow-up but declined subsequently. Episodes of major depression detected by the PHQ-9 tended to be preceded and followed by elevated levels of depressive symptoms.Conclusions The long-standing episodes of major depression typical of clinical practice appear to represent a minority of episodes occurring in the community. These results suggest that, in general population screening applications, the PHQ-9 will identify many respondents having brief and perhaps self-limited episodes. Although some episodes are characterized by large increases in symptoms, many respondents appear to move above and below the diagnostic threshold as a result of small changes in their symptom levels. Efforts to develop more effective approaches to screening may benefit from severity-based decision rules and serial measurement strategies.