Do antipsychotics lead to cognitive impairment in dementia? A meta-analysis of randomised placebo-controlled trials

Behavioural and psychological symptoms in dementia (BPSD) are common and often treated with antipsychotics, which are known to have small efficacy and to cause many side effects. One potential side effect might be cognitive decline. We searched MEDLINE, Scopus, CENTRAL and www.ClincalStudyResult.org for randomized, double-blind, placebo-controlled trials using antipsychotics for treating BPSD and evaluated cognitive functioning. The studies identified were summarized in a meta-analysis with the standardized mean difference (SMD, Hedges's g) as the effect size. Meta-regression was additionally performed to identify associated factors. Ten studies provided data on the course of cognitive functioning. The random effects model of the pooled analysis showed a not significant effect (SMD = -0.065, 95 % CI -0.186 to 0.057, I ² = 41 %). Meta-regression revealed a significant correlation between cognitive impairment and treatment duration (R ² = 0.78, p < 0.02) as well as baseline MMSE (R ² = 0.92, p < 0.005). These correlations depend on only two out of ten studies and should interpret cautiously.