A Systematic Literature Review and Meta-analysis of Real-World Interventions for Cognitive Ageing in Healthy Older Adults.

Activities running in community-based-settings offer a method of delivering multimodal interventions to older adults beyond cognitive training programmes. This systematic review and meta-analysis investigated the impact of randomised controlled trials (RCTs) of 'real-world' interventions on the cognitive abilities of healthy older adults. Database searches were performed between October 2016 and September 2018. Forty-three RCTs were eligible for inclusion with 2,826 intervention participants and 2,234 controls. Interventions to enhance cognitive ability consisted of participation in activities that were physical (25 studies), cognitive (9 studies), or mixed (i.e., physical and cognitive; 7 studies), and two studies used other interventions that included older adults assisting schoolchildren and engagement via social network sites. Meta-analysis revealed that Trail Making Test (TMT) A, p =  0.05, M = 0.43, 95% CI [-0.00, 0.86], digit symbol substitution, p =  0.05, M = 0.30, 95% CI [0.00, 0.59], and verbal fluency, p =  0.04, M = 0.31, 95% CI [0.02, 0.61], improved after specific types of interventions versus the control groups (which were either active, wait-list or passive controls). When comparing physical activity interventions against all control groups, TMT A, p =  0.04, M = 0.25, 95% CI [0.01, 0.48], and digit span forward, p =  0.05, M = 0.91, 95% CI [-0.00, 1.82], significantly improved. Results remained non-significant for all outcomes when comparing cognitive activity interventions against all control groups. Results therefore suggest that healthy older adults are more likely to see cognitive improvements when involved in physical activity interventions. In addition, TMT A was the only measure that consistently showed significant improvements following physical activity interventions. Visuospatial abilities (as measured by TMT A) may be more susceptible to improvement following physical activity-based interventions, and TMT A may be a useful tool for detecting differences in that domain.