<p>Combined summary datasheet for SoMe posts.</p>

Introduction Metabolic dysfunction associated with steatotic liver disease (MASLD)/non-alcoholic fatty liver disease (NAFLD) represents a significant public health concern. Social media (SoMe) increasingly influences health perceptions in lower-middle-income countries, with one-third of Sri Lanka’s population using SoMe for health information. Assessing MASLD content quality on SoMe is therefore important. Aims & methods This cross-sectional study assessed accuracy, completeness, and quality of MASLD content across Facebook, YouTube, TikTok, Instagram, and X in Sinhala, English, and Tamil from Sri Lanka (January 2005-December 2024). Board-certified gastroenterologists independently reviewed posts using standardised scales for accuracy (0–3), completeness (0–5), and global quality score (GQS) (0–5). Posts were categorised by source profile and content type, with user interactions analysed. Results Analysis included 289 posts: 158 (54.7%) YouTube, 101 (34.9%) Facebook, 14 (4.8%) TikTok, 11 (3.8%) X, 5 (1.7%) Instagram. Languages: 214 (74.0%) Sinhala, 54 (18.7%) Tamil, 21 (7.3%) English. Content sources: undisclosed identity (36.0%), non-healthcare persons (26.0%), healthcare professionals (22.1%), alternative healthcare professionals (14.2%), healthcare institutions (1.7%). Health promotion (61.9%) was the predominant content type. Mean accuracy was 1.78/3 (59.3%), with healthcare professionals scoring highest (2.35/3, 78.5%) versus others (51.0–55.1%; p < 0.001). Completeness averaged 2.1/5 (42%), with English content scoring higher than Sinhala and Tamil. GQS averaged 2.4/5 (48.4%). 82% of posts were classified as “Rotten” (<60% score for each metric). Facebook and YouTube showed significantly higher completeness and GQS (p < 0.05). User engagement metrics showed no correlation with content quality. Conclusion Most SoMe content originated from non-healthcare sources. Healthcare professionals delivered the most accurate content. Facebook and YouTube showed relatively higher content quality scores, though comparisons are limited by the small number of posts from other platforms. Overall quality remained suboptimal across platforms, with 82% failing adequate standards. User engagement didn’t correlate with quality. These findings highlight the need for improved quality control and health literacy initiatives for MASLD information on SoMe platforms.