Background. Previous longitudinal studies of the association between depression and cognitive dysfunction have had relatively short follow-up periods. This report presents a long-term study of the association between baseline syndromal depression and cognitive outcome measured 9 to 12 years later.

Methods. Self-CARE (D) depression, cognitive function and pre-morbid intelligence were recorded on 1083 subjects on entry to the Medical Research Council trial of treatment of hypertension in older adults in 1983–5. In 1994–5, we aimed to re-interview all survivors to assess cognitive function using the MMSE. We used multivariate analysis to explore whether baseline depression predicted cognitive outcome after this long follow-up period.

Results. Baseline depression was crudely associated with poorer cognitive outcome at time 2. However, this long-term prospective association was no longer apparent after adjusting for baseline cognitive performance, which was associated with baseline depression and robustly predicted cognitive outcome at time 2. We found that gender modified the association between depression and poorer cognitive outcome, so that the association was statistically significant only among men.

Conclusion. Propensity for depression and failing cognition may have common determinants that still need to be established by future neurobiological investigations in conjunction with further long-term prospective epidemiological research.