Anorexia nervosa (AN) is a severe psychiatric disorder characterized by chronic energy restriction, profound weight loss, and increased mortality and morbidity. Individuals with AN have an increased risk of metabolic disease and premature death. However, the physiological mechanisms underlying these outcomes remain incompletely understood. Skeletal muscle is the largest metabolic organ in the body and plays a central role in metabolic health, physical function, and survival. Clinical evidence indicates that individuals with AN experience a reduction in muscle mass and strength, some of which persist despite weight restoration. Human studies are limited by ethical constraints, variability in recovery duration, and an inability to assess underlying mechanisms directly. Consequently, the extent to which muscle size, strength, and protein turnover recover after AN and the molecular factors contributing to persistent impairments remain unclear.
The aim of this study was to assess skeletal muscle size, strength, fiber morphology, protein synthesis, and molecular regulators of catabolic and anabolic signaling during simulated AN and across short- and long-term recovery using a refined rodent model. The authors hypothesized that simulated AN would induce severe impairments in muscle size and strength that would not fully resolve with weight restoration, accompanied by sustained alterations in protein synthesis and degradation signaling.
Female Sprague Dawley rats underwent 30 days of simulated AN induced by 50 to 6% caloric restriction, a duration estimated to approximate many years of disease in humans. Rats were assigned to no recovery (AN-0R), short-term recovery (5 or 15 days; AN-5R, AN-15R), or long-term recovery (30 days; AN-30R) with ad libitum feeding after restriction. Age-matched health controls were involved in both short and long-term cohorts. Muscle strength was evaluated longitudinally by using rear paw grip strength and electrically stimulated dorsiflexion and plantarflexion. Muscle size, density, and fat area were measured in vivo by using peripheral quantitative computed tomography (pQCT).
Ex vivo analyses involved measurements of muscle mass, fiber cross-sectional area (CSA), myosin heavy chain (MyHC) distribution, 24-hour protein synthesis by deuterium oxide labeling, and gastrocnemius mRNA expression of anabolic, myogenic, inflammatory, proteolytic, and autophagy markers. Data were analyzed separately for short- and long-term recovery by using one-way ANOVA with Dunnett’s post-hoc tests or independent t-tests with baseline bodyweight or baseline functional values involved as covariates where appropriate. Statistical significance was set at p < 0.05.
Simulated AN resulted in marked losses in body weight (30%), muscle mass (22-45% depending on muscle group), and grip strength (20%) compared with controls (all p < 0.05). Short-term recovery partially restored bodyweight and muscle mass, though 5 days of recovery were insufficient to normalize muscle mass in muscles (p < 0.001 vs. control). After 15 days of recovery, muscle mass and pQCT-derived muscle area exceeded control values (p ≤ 0.002), which indicates rapid tissue regeneration. Despite long-term recovery, AN-30R rats exhibited 4% lower body weight (p = 0.019), 9-13% lower muscle mass across several muscles (p = 0.005-0.028), and 14% lower grip strength (p < 0.001), demonstrating incomplete functional recovery despite near-normal muscle size.
Electrically stimulated muscle testing showed continuous impairments in intrinsic muscle function. Maximal plantarflexion force was 21% lower in AN-0R rats (p < 0.001) and remained significantly reduced after 5 and 15 days of recovery (p ≤ 0.015). Plantarflexion force remained 6% lower in AN-30R rats (p = 0.021), although normalization of contraction timing properties occurred after long-term recovery. Dorsiflexion strength showed a nonsignificant 16% reduction after long-term recovery (p = 0.073). Collectively, these findings indicate sustained deficits in muscle quality.
Muscle fiber analyses demonstrated a 29% reduction in fiber CSA during acute AN (p < 0.001), which is caused by reductions in MyHC I and hybrid fibers, while MyHC II fibers were unaffected. Fiber CSA normalized with recovery, and no differences in fiber size or distribution persisted after 30 days. Whole-muscle protein synthesis rates did not differ significantly between groups (p > 0.10). However, key molecular regulators of anabolic signaling were changed. Igf1 mRNA expression was elevated during short-term recovery (p ≤ 0.015) but reduced after long-term recovery (~33%, p = 0.011). Inhibitors of protein synthesis (Redd1 and Deptor) were significantly elevated in AN-30R rats (p ≤ 0.017). Inflammatory markers were unchanged, although myostatin expression was elevated during acute AN (p = 0.008). Autophagy-related markers were suppressed during AN and short-term recovery (p < 0.001) but normalized following long-term recovery.
In summary, simulated AN caused severe impairment in muscle mass and strength, with muscle size recovering more rapidly and completely than muscle function. Persistent reductions in muscle strength and sustained alterations in anabolic signaling after weight restoration suggest long-lasting deficits in muscle quality. These findings highlight that weight restoration alone may be insufficient to fully restore musculoskeletal health following AN and underscore the need for targeted interventions aimed at improving muscle function during recovery.
Reference: Rosa-Caldwell ME, Breithaupt L, Kaiser UB, Muhyudin R, Rutkove SB. Changes in muscle strength and moderators of protein turnover in a rodent model of anorexia nervosa and recovery. J Nutr Physiol. 2025;4:100010. doi:10.1016/j.jnphys.2025.100010
