Prevalence of depression in myocardial infarction: A PRISMA-compliant meta-analysis.

BACKGROUND: Depression is common in the aftermath of myocardial infarction (MI) and may not only lead to impaired long-term quality of life, but also cause increased mortality among patients with MI. The reported prevalence of depression among patients with MI varied considerably across studies, for which a pooled prevalence was obtained in the only 1 meta-analysis conducted in March 2004. Subsequently, numerous relevant studies have been published, indicating the need for an update on the pooled prevalence. Therefore, this study was aimed at updating the pooled prevalence of depression among patients with MI.

METHODS: A comprehensive literature search in 3 electronic databases, PubMed, Embase, and PsycINFO, was performed in April 2018. The heterogeneity across studies was examined by the Cochran's Q test and quantified by the I statistic. If significant heterogeneity was observed, meta-regression analyses and subgroup analyses were performed to identify the source of heterogeneity. Publication bias was assessed by a funnel plot and verified by the Egger's and Begg's tests.

RESULTS: Nineteen eligible studies conducted in 10 countries were included, which consisted of 12,315 patients with MI, among whom 3818 were identified with depression. High heterogeneity was observed across the eligible studies (I = 98.4%), with the reported prevalence of depression ranging from 9.17% to 65.88%. The pooled prevalence of depression among patients with MI was 28.70% (95% CI: 22.39-35.46%) by a random effects model. Subgroup analyses showed that the pooled prevalence differed significantly by region, tool used to identify depression, study quality, sex, race, anterior MI, and diabetes status (P < .05). Meta-regression analyses did not identify any moderators of heterogeneity, and the heterogeneity was high within most subgroups. Nonetheless, for unmarried subjects, the heterogeneity was low (I = 19.5). The Egger's test and the Begg's test indicated no evidence of publication bias (P > .05).

CONCLUSIONS: Given the high pooled prevalence of depression found in this study and the association between depression and adverse health outcomes among patients with MI, more psychological resources including early assessment and effective treatment of depression should be allocated to patients with MI.