Objective <p>Bariatric surgery effectively reduces obesity-related comorbidities, however fat mass loss is often accompanied by lean mass reduction. This 36-month prospective study (N=55) examined body composition changes following three bariatric procedures compared with a diet-controlled group (Diet). </p> Methods <p>Random intercept mixed-effects models assessed 36-month trajectories of fat and lean mass, comparing surgical groups with Diet, adjusting for age and sex. Associations between body composition changes, measured by dual-energy X-ray absorptiometry and adiponectin (fat-derived) and undercarboxylated osteocalcin (ucOC; bone-derived) hormones were evaluated at 0, 1, 3, 6, 12, 24, and 36 months.</p> Results <p>Roux-en-Y gastric bypass produced the greatest fat mass loss (46.6%), followed by sleeve gastrectomy (30.1%) and laparoscopic adjustable gastric banding (20.5%). Corresponding lean mass losses were 14.2%, 11.7%, and 5.2%. In Diet changes in fat mass (−2.1%) and lean mass (+3.6%) were non-significant. Over 36 months, all surgical groups lost significantly more lean mass than Diet, with the greatest losses after RYGB (15.1%) and LAGB (14.8%). Women experienced 30% greater lean mass loss than men (95% CI: -34%, -26%). Each additional year of age was associated with a 0.2% lean mass decline (95% CI: -0.5%, 0%). After controlling for treatment procedure, age and sex, each twofold increase in adiponectin and ucOC was associated with 5.0% (95% CI: -9.4%, -2.4%) and 3.7% (95% CI: -5.0%, -1.3%) reductions in lean mass, respectively.</p> Conclusions <p>These findings underscore the lean mass loss risk after bariatric surgery, particularly in women and older adults, and highlight the importance of long-term monitoring of body composition, nutrition and bone health.</p> Clinical trial registration <p>The study was registered as ANZCTR: 12613000188730.</p>