Do nonpharmacological interventions prevent cognitive decline? a systematic review and meta-analysis

At present, prevention is particularly important when there is no effective treatment for cognitive decline. Since the adverse events of pharmacological interventions counterbalance the benefits, nonpharmacological approaches seem desirable to prevent cognitive decline. To our knowledge, no meta-analyses have been published on nonpharmacological interventions preventing cognitive decline. To investigate whether nonpharmacological interventions play a role in preventing cognitive decline among older people, we searched related trials up to March 31, 2019, in MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials (CENTRAL), ClinicalTrials and the Cochrane library databases. Randomized controlled trials (RCTs) were included if they enrolled participants older than 60 years of age who had a risk of cognitive decline, and the interventions were nonpharmacological. Two reviewers independently extracted data and assessed study quality. The Grading of Recommendations Assessment Development and Evaluation (GRADE) approach was used to rate the quality of evidence. Heterogeneity was quantified with I2. Subgroup analysis and meta-regression were used to research the sources of heterogeneity. Influence analyses were used to detect and remove extreme effect sizes (outliers) in our meta-analysis. Publication bias was assessed with funnel plots and Egger test. Primary outcomes were the incidence of mild cognitive impairment (MCI) or dementia and Alzheimer’s Disease Assessment Scale–Cognitive Subscale (ADAS-Cog) scores. Second outcomes were activities of daily living (ADL) and Mini-Mental State Examination (MMSE) scores. A total of 22 studies with 13,264 participants were identified for analysis. In terms of prevention, nonpharmacological interventions appeared to be more effective than control conditions, as assessed by the incidence of MCI or dementia (RR, 0.73; CI, 0.55–0.96; moderate-certainty evidence), while the results of ADAS-Cog suggested no significant differences between two groups (MD, −0.69; CI, −1.52–0.14; very low-certainty evidence). Second outcomes revealed a significant improvement from nonpharmacological interventions versus control in terms of the change in ADL (MD, 0.73; CI, 0.65–0.80) and MMSE scores (posttreatment scores: MD, 0.25; CI, 0.02–0.47; difference scores: MD, 0.59, CI, 0.29–0.88). Univariable meta-regression showed association between lower case of MCI or dementia and two subgroup factors (P = 0.013 for sample size; P = 0.037 for area). Multiple meta-regression suggested that these four subgroup factors were not associated with decreased incidence of MCI (P > 0.05 for interaction). The Naive RR estimate was calculated as 0.73. When the three studies that detected by outlier and influence analysis were left out, the Robust RR was 0.66. In conclusion, nonpharmacological therapy could have an indicative role in reducing the case of MCI or dementia. However, given the heterogeneity of the included RCTs, more preregistered trials are needed that explicitly examine the association between nonpharmacological therapy and cognitive decline prevention, and consider relevant moderators.