Chronic disease exercise videos show low quality on social media

A systematic review and meta-analysis of 17 studies found that exercise videos for people with chronic diseases are widely available on social media but have low-to-moderate educational quality. The analysis included 1,515 videos and assessed them with three principal quality instruments. Pooled scores were below the midpoints of those scales, while video popularity measures did not differ significantly across quality categories.

The researchers searched Web of Science, PubMed, SportDiscus through EBSCO, Embase, Scopus, CNKI and Google Scholar from the databases’ inception through September 2025. Eligible studies assessed the quality of exercise videos for chronic disease management on social media and were available in English or Chinese. Two reviewers independently screened studies and extracted data. The review followed PRISMA guidelines, and the researchers assessed study quality using an eight-item D’Souza framework.

The statistical analysis used random-effects meta-analysis with restricted maximum likelihood in Stata 19. The researchers reported 95% confidence intervals and, for the principal instruments, 95% prediction intervals. Meta-regression was used to compare quality strata by selected video characteristics, engagement metrics and uploader type. The analysis therefore focused on educational quality rather than on whether the videos improved health outcomes or whether viewers learned from them.

On the Global Quality Scale, the pooled mean was 2.96 out of 5, with a 95% confidence interval of 2.63 to 3.29 and a 95% prediction interval of 1.70 to 4.23. The modified DISCERN instrument produced a pooled mean of 2.53 out of 5, with a 95% confidence interval of 2.18 to 2.90 and a 95% prediction interval of 1.24 to 3.84. The JAMA benchmark criteria yielded a pooled mean of 1.49 out of 4, with a 95% confidence interval of 0.65 to 2.33.

Variation among the included studies was substantial. The reported I² values ranged from 93.0% to 98.6%, indicating pronounced heterogeneity in the observed quality results. The wide prediction intervals also show that the quality expected in a new collection of videos could vary considerably. Consequently, the pooled means summarize a broad and uneven evidence base rather than a uniform score for every chronic disease exercise video.

Engagement measures did not distinguish among the quality strata. Video duration, view count, likes, dislikes and comment counts did not differ significantly across the categories. In this dataset, those measures therefore did not provide evidence of higher educational quality, although the analysis did not establish that audience activity is unrelated to quality in every platform or population.

Uploader type showed a different pattern. Videos from individual physician accounts represented 34.2% of the poor-quality stratum but only 3.6% of the good-quality stratum, a difference reported as p = 0.03. Videos from academic or institutional channels were more often classified as good quality, but that comparison had p = 0.068 and is suggestive rather than conclusive. The review identified an association between channel type and assessed quality; it did not show that changing uploader type would cause a video’s quality to improve.

The study’s evidence base was concentrated geographically and by platform. Thirteen of the 17 studies, or 76.5%, were conducted in Turkey, and 16 of the 17, or 94.1%, examined YouTube. Only English- and Chinese-language studies were eligible, and the review did not establish that the pooled results apply equally to other languages, countries, chronic conditions or social media platforms.

Overall, the authors concluded that exercise videos for people with chronic diseases are widely available but educationally suboptimal, while engagement metrics do not indicate quality. Higher-quality content was associated with academic and institutional channels rather than individual physician accounts. The analysis is registered in PROSPERO as CRD420261345615, and the open-access research was published in BMC Public Health on 21 October 2026.

Subject of Research: Educational quality of social media exercise videos for people with chronic diseases

Article Title: Quality assessment of exercise videos for chronic diseases on social media platforms: a systematic review and meta-analysis

Article References: Li, K., Yun, D., Fu, L., Hui, S. S., Liu, J., Li, R., & Hu, R. (2026). Quality assessment of exercise videos for chronic diseases on social media platforms: a systematic review and meta-analysis. BMC Public Health. https://doi.org/10.1186/s12889-026-30063-y

Image Credits: AI Generated

DOI: 10.1186/s12889-026-30063-y

Keywords: chronic disease, social media, exercise videos, health education, systematic review, meta-analysis, YouTube, content quality, online health information, Quality, assessment, exercise