mHealth-based lifestyle intervention for hypertension in community-dwelling adults: a cluster-randomized controlled trial with exploratory multi-omics analyses

Evidence for mobile health (mHealth) interventions in mixed community populations with and without hypertension remains limited. We evaluated the effectiveness of a socioecological model-guided, smart device-based, self-management-oriented lifestyle (3SLIFE) trial and examined whether baseline gut microbiota and fecal metabolites could modify and predict intervention response. This is a prespecified secondary analysis of the 3SLIFE trial, a two-arm, cluster randomized controlled trial conducted from April 2023 to July 2025 in 20 Chinese communities/villages. The communities/villages were randomly allocated in a 1:1 ratio to the intervention (190 participants) or control (193 participants) groups, with a 6-month intervention program (combined lifestyle education, interactive courses, personalized feedback, and family and community support delivered through social media) and follow-ups continuing to 24 months. The main outcome was the between-group difference in hypertension prevalence change (blood pressure [BP] ≥ 140/90 mmHg), evaluated using a mixed-effects model. Hypertension-related outcomes and plasma metabolite levels were also evaluated. Modifying effects and predictive performance of gut microbiota and fecal metabolites on intervention response were further investigated. This trial is registered with chictr.org.cn, ChiCTR2300070575. Compared to the control group, the intervention group showed the reductions in hypertension risk at 6 months (OR = 0.28 [95% CI 0.10, 0.77]) and 24 months (OR = 0.20 [0.07, 0.60]), as well as in diastolic BP (-2.68 mmHg [-4.52, -0.83]) and mean arterial pressure (-2.68 mmHg [-4.80, -0.56]) at 24 months, baPWV (-64.20 cm/s [-114.11, -14.29]) at 12 months, and 10-year ASCVD risk score at 3 months (-0.20 [-0.37, -0.04]) and 6 months (-0.24 [-0.40, -0.07]). The intervention group achieved higher BP control rates among non-medicated participants with baseline BP ≥ 130/80 mmHg at 6, 12 and 24 months. The 14 gut microbial species and 40 fecal metabolites may modify intervention effects and show the highest predictive accuracy for intervention response at 24 months (AUC: 0.85). Coprococcus comes and Bacteroides eggerthii were identified as potential species linking fecal metabolites to altered plasma metabolites in the multi-omics network. Our findings support the potential sustained effectiveness of mHealth-based lifestyle strategies for hypertension management in community-dwelling populations, and provide preliminary evidence that microbiota-metabolite profiling may facilitate precision interventions, which however warrant confirmation in more, larger-sample trials.