3778 Effects of metformin on biomarkers in older people with sarcopenia: analysis from the MET-PREVENT randomised controlled trial

Abstract Background Chronic inflammation and metabolic dysfunction are posited to contribute to sarcopenia and physical frailty; both are targets for metformin therapy. We investigated correlations between physical performance measures and inflammatory and metabolic biomarkers in a group of older people with sarcopenia and frailty/prefrailty and investigated the effect of metformin treatment on this biomarker panel. Methods We analysed samples collected at baseline and follow-up (4 months) from the randomised controlled MET-PREVENT trial. MET-PREVENT recruited participants aged 65 and over with probable sarcopenia (EWGSOP2 guidelines) and 4 m walk speed <0.8 m/s. Participants received 500 mg metformin 3x/day or matching placebo for 4 months. Blood sampling and physical performance measures (handgrip strength, 4 m walk speed, six-minute walk distance, 5x sit to stand) were conducted at baseline and 4 months. Biomarkers were measured using ELISA and Luminex platforms for insulin, CRP, adiponectin, leptin, MCP-1, IL-1b, IL-6, IL-8, and TNF-a. Baseline correlations and correlations of changes between baseline and follow-up, were analysed using Spearman’s test; median change between baseline and follow-up in the metformin and placebo groups was compared using Mann–Whitney tests. Results Seventy-two participants were studied, mean age 80 years. Higher baseline IL-6 concentrations correlated with lower six-minute walk distance, (r = −0.40, p = 0.008); other correlations were non-significant. Increased IL-1B and decreased adiponectin between baseline and 4 months were correlated with worsening grip strength (r = −0.25, p = 0.04; r = 0.29, p = 0.016); increased TNF-a and decreased MCP-1 were correlated with worsening 5x sit-to-stand time (r = 0.38, p = 0.02; r = −0.32, p = 0.04). Metformin treatment reduced circulating insulin concentrations more than placebo (median change between baseline and 4 months: -178 pg/ml [IQR -462, 180] vs 147 pg/ml [IQR -89, 353]; between-group p = 0.04) but did not significantly change other biomarkers compared with placebo. Conclusions In this trial population, only weak and inconsistent correlations were seen between physical performance and inflammatory/metabolic biomarkers, and metformin did not beneficially affect most biomarkers measured.