Longitudinal association between frequency of Internet use and incident disability among community-dwelling older people during the COVID-19 pandemic.

Background: There is limited evidence of a protective effect of Internet use for incident disability (ID) during the COVID-19 pandemic. We investigated the association between frequency of Internet use (FIU) and ID among community-dwelling older people.

Methods: We used longitudinal data from the 2019 and 2022 surveys, including 7,913 residents aged ≥65 without disability at baseline. ID was defined as a new public long-term care insurance certification. FIU at baseline was categorized into daily, weekly, monthly, yearly, and non-users. Changes in FIU before and during the COVID-19 pandemic were categorized into continuing frequent (i.e., daily or weekly), continuing moderate (i.e., monthly or yearly), increase in frequency, from non-users to users, decrease in frequency, from users to non-users, and continuing non-users. Covariates included age, gender, education, perceived economic situation, family structure, body mass index, chronic medical conditions, dietary variety, working status, walking time, and cognitive functioning. Multivariable Poisson regression models were used to estimate adjusted cumulative incidence ratio (aCIR) and 95% confidence interval (CI) for ID.

Results: During the 3-year follow-up, 132 of 4,453 people aged 65-74, 595 of 3,460 people aged ≥75, 287 of 3,660 men, and 440 of 4,253 women developed ID. For FIU at baseline, among people aged ≥75 or men, there was a dose-response relationship between more frequent Internet use at baseline and a lower risk of ID (P-trend was 0.005 in people aged ≥75, and <0.001 in men). Compared to non-users, daily users had a significantly lower risk of ID [aCIR (95% CI) = 0.69 (0.53-0.90) in people aged ≥75, and 0.49 (0.34-0.70) in men]. For changes in FIU, "continuing frequent" and "from non-users to users" had a lower risk of ID than continuing non-users. After stratified analyses, "continuing frequent" remained a significant association in people aged ≥75 or in men, while "from non-users to users" had a significant association in those with daily walking time <30 minutes.

Conclusions: Although FIU may act as a marker of disability, or indicate individual adaptability, our findings suggest that Internet use may be a potential preventive measure against ID in community-dwelling older people when social distancing is required.