Prediction of the three-year course of recurrent depression in primary care patients: different risk factors for different outcomes

Background
The objectives of this study are: (1) identification of predictors for the three-year course of recurrent depression in the rarely studied, but relevant sample of primary care patients, and (2) investigation whether different outcome indicators, time to recurrence, proportion depression-free time and mean severity of depressive symptoms during follow-up, are associated with different risk factors.

Methods
Depression course was established by assessing 110 patients three-monthly with the Composite International Diagnostic Interview and the BDI, during a three-year period. Eight (groups of) predictors, assessed at baseline, were examined: socio-demographics, parental depression, history and severity of depression, anxiety, coping potential, social dysfunctioning and physical functioning.

Results
Time to recurrence was predicted by number of previous episodes (OR = 1.91). Both proportion depressive disorder-free time and mean depression severity during follow-up were predicted by: severity of depression (B = − .19 and .21 respectively), anxiety (B = − .32 and .33), social dysfunctioning (B = − .21 and .22) and physical functioning (B = .24 and − .39). Mean severity was additionally predicted by: educational level (B = − .21), duration of the longest prior episode (B = .32), and coping potential (B = − .40). Coping potential and number of previous episodes were marginally significant predictors for all three outcomes.

Limitations
Although substantial, sample size was restricted.

Conclusion
Different outcome variables are predicted by different risk factors. Restriction to one outcome may lead to missing important determinants of the depression course. Number of prior episodes and coping potential seem to warrant special attention from the GP.