A novel sarcopenic obesity index served as a clinically meaningful predictor of cardiovascular risk and lent further insight into accompanying genetics and possible therapeutic targets, according to research published July 1 in JACC.

Mihir M. Sanghvi, MBBS, PhD, et al., developed the index using body weight and pectoralis major muscle mass, quantified and derived through a deep learning pipeline of 55,768 cardiovascular magnetic resonance (CMR) examinations from the UK Biobank.

Results showed that a higher index was associated with increased risk of incident heart failure (hazard ratio [HR], 1.31), cardiovascular death (HR, 1.51) and all-cause mortality (HR, 1.37), and showed more consistent associations with adverse outcomes compared with body mass index, waist-to-hip ratio and waist-to-height ratio.

Genome-wide analysis of 52,253 participants identified 16 loci for sarcopenic obesity, including 13 with known signals for cardiovascular diseases and/or cardiac imaging endophenotypes, and transcriptomic profiling using 1,082 human skeletal samples found these loci were specifically modulated during muscle atrophy. Additionally, downstream druggability analysis identified ACVR2B as a high-confidence target.

"These results highlight a potential application of the sarcopenic obesity index as a biomarker for nominating individuals in whom targeted therapies that preserve or augment muscle mass, while reducing adiposity, may be of greatest utility," write Sanghvi, et al. ACVR2B is the target of the human monoclonal antibody agent bimagrumab, which is currently being tested with semaglutide to preserve or increase muscle mass during weight loss.

"[Sarcopenic obesity] is emerging as a systemic cardiometabolic phenotype linking ageing, impaired muscle biology, heart failure vulnerability and frailty," write Pier-Giorgio Masci, MD, PhD; Marili Niglas, PhD; and Jimmy D. Bell, PhD, in an accompanying editorial comment. "By enabling scalable opportunistic quantification of muscle mass from routine CMR, Sanghvi et al., provide proof of concept that deep learning-derived pectoralis muscle mass can serve as a practical surrogate of systemic muscle status in large cardiovascular imaging cohorts."