A randomised, prospective 12-month study in women with obesity combined proteomics, metabolomics, advanced imaging, and dynamic meal testing across multiple adipose depots, liver, and blood before and after low-calorie diet plus bariatric surgery (sleeve gastrectomy or Roux-en-Y gastric bypass). Key findings: surgery normalised non-invasive liver phenotype markers consistent with MASLD reversal, preferentially reduced android over gynoid fat, and drove proteomic remodelling in abdominal but not gluteal adipose tissue. Candidate molecules—all-trans-retinol dehydrogenase, lumican, and androgenic steroids—emerged as putative regulators of metabolic improvement. Outcomes were statistically indistinguishable between the two surgical procedures.
The depot-specificity finding is the most actionable insight here. For years, clinicians have known visceral/android fat drives cardiometabolic risk more than subcutaneous gluteal fat; this study now offers molecular granularity explaining why surgically reducing android fat translates to metabolic benefit. The retinol dehydrogenase and lumican signals are intriguing—both have been linked to adipose inflammation and fibrosis in preclinical work—but identifying them in human tissue biopsies elevates their translational relevance considerably. Limitations include an all-female cohort limiting generalisability, the absence of a non-surgical weight-loss control arm, and modest sample sizes typical of invasive biopsy studies. The equivalence of SG and RYGB outcomes challenges assumptions about procedure-specific hormonal mechanisms. Overall, this is a methodologically rigorous, confirmatory-plus-discovery study that incrementally deepens mechanistic understanding without yet changing clinical practice.