Background: There is conflicting evidence on the hypothesis that the risk of depression after stroke is influenced by the location of the brain lesion. We undertook a systematic review to examine the hypotheses that depression is more commonly associated with left-hemisphere strokes than with right-hemisphere strokes and with lesions of the left anterior brain than with other regions.
Methods: We did a computer-aided search of MEDLINE, BIDS ISI, and PsychLit databases supplemented by hand searches of key journals. We included all reports on the association of depression after stroke with the location of the brain lesion. Studies were systematically and independently examined by two investigators. Fixed-effects and random-effects meta-analyses were done.
Findings: 143 reports were identified by the search strategy. 48 were eligible for inclusion. Not all reports included original data. Only two reports of original data supported the hypotheses and seven clearly did not. The pooled (random-effects) relative risk of depression after a left-hemisphere stroke, compared with a right-hemisphere stroke, was 0.95 (95% CI 0.83-1.10). For depression after a left anterior lesion compared with all other brain areas the pooled (random-effects) relative risk was 1-17 (0.87-1.62). Restriction of the analyses to reports from high-quality studies or major depressive disorder did not substantially affect the findings. Nor were they affected by stratification of the time between stroke and the assessment of depression. Multiple publications from the same samples of patients were apparent.
Interpretation: This systematic review offered no support for the hypothesis that the risk of depression after stroke is affected by the location of the brain lesion.