Six-cohort study links digital participation to part of socioeconomic health gap

Digital participation was associated with part of the socioeconomic health gap among adults aged 50 and older in a cross-national analysis of six harmonized aging cohorts. Across all six cohorts, higher socioeconomic status was linked to better composite health, but adding reported internet use or another digital-participation measure reduced the fitted association by varying amounts. The study analyzed 449,861 person-waves and found that the pattern was consistent across countries, while the size and apparent components of the association differed substantially among cohorts.

Researchers constructed a composite health index from cognition, physical functioning, depressive symptoms and self-rated health. They related standardized socioeconomic status to this index using complete-case regressions, adjusting for demographic characteristics, available behavioral covariates and survey wave, with standard errors clustered by respondent. The digital analysis used reported internet use in the Health and Retirement Study, English Longitudinal Study of Ageing, Survey of Health, Ageing and Retirement in Europe and China Health and Retirement Longitudinal Study. A broader communication measure was examined separately in the Mexican Health and Aging Study, while the Korean Longitudinal Study of Aging contributed comparisons of health gradients. Descriptive mediation, cross-cohort decompositions and 1,000 respondent-level bootstrap resamples were used to examine how participation might account for differences in the socioeconomic-health association.

On cohort-standardized scales, a one-standard-deviation increase in socioeconomic status was associated with better health ranging from 0.185 standard deviations in the Korean cohort to 0.411 standard deviations in the US cohort. Adding internet use reduced the fitted socioeconomic-health coefficient by 11.3% to 24.2% in the four cohorts with internet-specific measures. The reduction was smaller, at 6.3%, in the Mexican cohort when its broader communication indicator was included. These results do not mean that digital use alone caused the observed differences. Instead, the analysis shows that reported digital participation is socially patterned and is associated with a measurable portion of the relationship between socioeconomic position and later-life health.

The cross-cohort comparisons suggested that differences in the socioeconomic patterning of digital participation generally made the larger contribution. Differences between participants and non-participants after accounting for access and internet use also varied by cohort and by the measure of participation used. In the English cohort, an activity measure produced an additional 5.6-percentage-point reduction in the fitted socioeconomic-health coefficient, with a 95% confidence interval of 4.4 to 6.7 points. By contrast, a device-count measure in the Chinese cohort changed the coefficient by only 0.2 points, with a 95% confidence interval running from minus 0.3 to 0.6 points. The contrast indicates that the way digital participation is defined can materially affect conclusions about how strongly it is connected with the health gradient.

Adjacent-wave analyses and an additional cohort comparison did not establish a simple pattern of uniformly better future health among digital participants. Digital participants had higher adjusted health scores at the following wave in some analyses, but the survey-weighted participant–non-participant difference in the Mexican cohort was small and uncertain: 0.021 standard deviations, with a 95% confidence interval of minus 0.024 to 0.067. Education-stratified results also differed by cohort, showing that the relationship cannot be reduced to a universal digital divide. The study was observational, used available survey measures and varied in its indicators of digital participation, so it cannot determine whether expanding digital access would narrow socioeconomic inequalities in health. Its findings instead identify a specific question for direct evaluation: whether affordable and accessible digital support can help disadvantaged older adults sustain participation, and whether any resulting health effects are distributed equitably.

Subject of Research: Association between digital participation and socioeconomic inequalities in later-life health across six aging cohorts

Article Title: Digital participation and socioeconomic inequalities in later-life health: a cross-national analysis of six harmonized aging cohorts

Article References: Liu, L., & Zhang, Q. (2026). Digital participation and socioeconomic inequalities in later-life health: a cross-national analysis of six harmonized aging cohorts. BMC Public Health. https://doi.org/10.1186/s12889-026-30004-9

Image Credits: AI Generated

DOI: 10.1186/s12889-026-30004-9

Keywords: digital participation, socioeconomic inequality, later-life health, aging cohorts, internet use, health gradient, cross-national analysis, digital divide, older adults, composite health index, Digital, participation