EPA-0229 - Association between depression and mortality depends on methodology used

Introduction Previous research demonstrates various associations between depression, cardiovascular disease (CVD) incidence and mortality. Differences between studies may occur as a result of different methodologies. Objectives This work investigated the impact of using two different methods to measure depression and two different methods of analysis to establish relationships. Aims The work investigated the association between depression, CVD incidence (CVDI) and mortality from coronary heart disease (MCHD), smoking related conditions (MSRC), and all causes (MALL), in a major population study using depression measured from a validated scale and a depression measure derived by factor analysis, and analyses based on continuous data and grouped data. Methods Data from the PRIME Study (N=9,798 men) on depression and ten year CVD incidence and mortality were analysed using Cox proportional hazards models. Results Using continuous data, no relationships with CVDI were found, but both measures of depression resulted in the emergence of positive associations between depression and mortality (MCHD, MSRC, MALL). Using grouped data, no associations with CVDI or MCVD were found, and associations between the measure derived from factor analysis and MSRC and MALL were also lost. Positive associations were only found between depression measured using validated items, MSRC and MALL. Conclusions These data demonstrate a possible association between depression and mortality but detecting this association is dependent on the methodology used. Different findings based on methodology present clear problems for the determination of relationships. The differences here suggest the preferential use of validated scales and suggest against over-reduction via factor analysis and grouping.