Associations of internet use and specific patterns with cognitive function in older adults: a systematic review and meta-analysis

Background Global aging is accompanied by cognitive decline. While the “technological reserve” hypothesis suggests internet use may mitigate this, prior meta-analyses have not clarified the specific effects of different usage patterns. This study aims to systematically evaluate associations between distinct dimensions of internet use and cognitive function in older adults. Methods Conducted in accordance with PRISMA 2020 guidelines, we searched PubMed, Embase, CINAHL, PsycINFO, and Web of Science from inception to March 15, 2026. Two Reviewers independently extracted data. Pooled Odds Ratios (ORs) or Hazard Ratios (HRs) served as primary effect measures; Correlation coefficients ( r ) or standardized beta coefficients (β) were converted into unified OR-scale estimates for meta-analysis, acknowledging inherent limitations introduced by combining mixed effect measures from original studies. Results Thirty-one studies comprising 158,648 participants (cross-sectional and longitudinal observational designs) were included. The random-effects meta-analysis of 25 studies yielded a pooled OR of 1.49 (95% CI: 1.31–1.70), indicating general internet use was significantly associated with better cognitive function ( z = 5.97, P < 0.001 ). Substantial heterogeneity was observed ( I 2 = 99.51%). A 95% prediction interval (1.15–1.93) was further calculated to quantify uncertainty. Subgroup analyses showed significant positive associations for social ( OR = 1.55), entertainment ( OR = 1.46), and instrumental use ( OR = 1.38). However, meta-regression revealed no statistically significant differences among these categories ( F (2, 24) = 0.36, P = 0.63). Sensitivity analysis excluding outliers yielded a similar pooled point estimate, although substantial heterogeneity persisted. (pooled OR = 1.46, 95% CI: 1.29–1.67). Conclusions Internet use was positively associated with better cognitive function among older adults. This positive association was observed across social, entertainment, and instrumental internet use patterns. Given high between study heterogeneity and the observational nature of available evidence, further longitudinal and interventional research is needed to clarify whether diverse digital engagement may support healthy cognitive aging. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/recorddashboard , identifier: CRD420261359423.