JMIR Publications

ABSTRACT

Background:

eHealth literacy is widely assumed to drive how people seek health information online—yet this assumption rests on a body of evidence that has never been quantitatively synthesized across population groups or examined against the algorithmically mediated environments in which today's users actually navigate health content. Contemporary measurement of eHealth literacy relies heavily on the eHealth Literacy Scale (eHEALS), a tool calibrated to Web 1.0-era search behaviors. Whether eHEALS retains predictive validity for online health information seeking behavior (OHIS) across generationally distinct cohorts—particularly digital natives who acquire health knowledge through curated feeds and short-video platforms rather than deliberate search—remains an open and consequential question.

Objective:

This study aims to quantify the strength and heterogeneity of the association between eHealth literacy and OHIS, and to identify boundary conditions across generation, morbidity status, and information source credibility.

Methods:

Following PRISMA guidelines, we searched PubMed, Embase, Web of Science Core Collection, PsycINFO, and Library, Information Science & Technology Abstracts for studies published through April 28, 2025. Eligible studies enrolled individual-level participants, assessed eHealth literacy with validated instruments, and measured active OHIS. Two independent reviewers extracted data and appraised study quality using the modified Newcastle-Ottawa Scale. Pearson r values were transformed to Fisher's z and pooled under a random-effects model; moderator analyses were performed for the three prespecified subgroups.

Results:

Of 8,090 nonduplicate records, 30 studies entered the qualitative synthesis and 18 (19 independent effect sizes) the meta-analysis. The overall pooled correlation was r = 0.30 (95% CI 0.18–0.41; P < .001), indicating a small-to-moderate association. Subgroup analyses revealed a strikingly uneven pattern: among non-Gen Z participants the correlation was r = 0.39, whereas in Gen Z it was near zero (r = 0.08)—suggesting that eHEALS-measured literacy is largely disconnected from how this cohort seeks health information. The association was substantially stronger among patients than nonpatients (r = 0.56 vs. r = 0.23) and for professional versus nonprofessional sources (r = 0.38 vs. r = 0.26). No significant publication bias was detected (Egger's test, P = .38).

Conclusions:

The near-zero eHealth literacy–OHIS association in Gen Z is the study's most consequential finding: it indicates that eHEALS has limited predictive validity for a generation that navigates health content through algorithmically curated feeds, short-video platforms, and AI-assisted interfaces rather than deliberate keyword search. Interpreted through a Motivation-Ability-Opportunity lens, perceived ability no longer constrains seeking behavior in digital natives—motivational activation and platform affordances do. These findings challenge the field to move beyond self-report confidence measures toward platform-sensitive, performance-based instruments, and call for intervention designs that pair literacy skills with motivational and environmental cues rather than treating literacy as a standalone determinant of health information behavior.