The Association Between eHealth Literacy and Health Behaviors During and Since the COVID-19 Pandemic: Systematic Review and Meta-Analysis.

Background: Since COVID-19, health information seeking, service navigation, and routine care have become increasingly digitally mediated. It remains unclear whether the association between eHealth literacy and health behaviors is consistent across behavioral domains, populations, and analytic frameworks. Objective: This systematic review and meta-analysis synthesized COVID-19 and post-COVID-19 evidence on the association between eHealth literacy and health behaviors and examined variation across study contexts. Methods: We conducted a PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020-compliant, PROSPERO-registered review (CRD420251009048). PubMed, Embase, Web of Science Core Collection, CINAHL Ultimate, and Scopus were searched from January 1, 2020, through March 27, 2026. Eligible observational studies assessed eHealth literacy and reported analyzable associations with health behavior outcomes collected in 2020 or later. Health behaviors were classified as health decision-making, health-promoting, or health management behaviors. Correlation coefficients, grouped odds ratios (ORs), and continuous ORs were synthesized separately using random-effects models with Knapp-Hartung adjustment. Certainty was assessed using the Grading of Recommendations Assessment, Development, and Evaluation framework. Results: In total, 19 studies were included: 10 contributed correlation coefficients, 6 grouped ORs, and 3 continuous ORs. In the correlation-based synthesis, higher eHealth literacy was associated with more favorable health behaviors (pooled r=0.43, 95% CI 0.36-0.51; 95% prediction interval 0.22-0.61), with substantial heterogeneity (I2=80.50%). In the grouped OR synthesis, higher eHealth literacy was also associated with more favorable health behaviors (pooled OR 2.12, 95% CI 1.47-3.05; 95% prediction interval 1.11-4.06), with moderate heterogeneity (I2=46.33%). In the continuous OR synthesis, all 3 studies showed positive associations, but the pooled effect was not statistically significant (pooled OR 1.07, 95% CI 0.89-1.30; 95% prediction interval 0.84-1.37), with very high heterogeneity (I2=97.98%). Subgroup analyses showed a significant difference only by geriatric status in the grouped OR synthesis. Certainty was low for the grouped OR synthesis and very low for the other 2 syntheses. Conclusions: Higher eHealth literacy was generally associated with more favorable health behaviors in the correlation-based and grouped OR syntheses, whereas evidence from the continuous OR synthesis was inconclusive. Given the predominantly cross-sectional evidence base, heterogeneity, risk-of-bias concerns, and low to very low certainty, the association should be interpreted as contextual and associative rather than causal or uniform. This review is innovative in synthesizing COVID-19 and post-COVID-19 evidence, applying a functional classification of health behaviors, and analyzing distinct effect measures separately. Unlike previous reviews that summarized the association more broadly, it avoids a single mixed pooled effect and provides a cautious, context-specific interpretation. In practice, interventions should pair eHealth literacy improvement with trustworthy digital services, clinician support, and behavior-specific conditions that help translate digital information into sustained health-related action.