Rapid weight loss through pharmacological intervention raises concern about skeletal integrity, particularly when lean mass is lost alongside adipose tissue. Two peptides under investigation for metabolic effects, AOD-9604 and semaglutide, differ substantially in their documented impact on bone density and fracture risk during periods of caloric deficit. Semaglutide, a GLP-1 receptor agonist approved for weight management, has accumulated clinical trial data showing modest reductions in bone mineral density (BMD) in some cohorts, whereas AOD-9604, a synthetic fragment of human growth hormone, lacks sufficient human bone-health data but demonstrates theoretical mechanisms that might preserve skeletal mass through its structural relationship to the parent hormone's C-terminal region.
What Published Research Supports About Semaglutide and Bone Outcomes
The STEP trial programme, which enrolled over 4,500 participants across multiple phase-3 studies, documented weight reductions averaging 12.4 to 15.8 percent of baseline body mass with semaglutide 2.4 mg weekly over 68 weeks (Wilding 2021, Wadden 2021). Post-hoc analyses from STEP 1 revealed that participants experienced small but statistically significant decreases in total hip BMD, measured by dual-energy X-ray absorptiometry (DXA), with a mean change of approximately -1.2 percent from baseline compared to -0.3 percent in placebo groups (Rubino 2021). Lumbar spine BMD showed less pronounced changes, averaging -0.8 percent in the semaglutide arm versus -0.2 percent with placebo, suggesting that weight-bearing skeletal sites may be differentially affected during GLP-1-mediated weight loss.
Mechanistic studies in rodent models indicate that GLP-1 receptor activation influences bone remodelling through both direct receptor-mediated pathways on osteoblasts and indirect effects via altered gut hormone secretion and calcium absorption (Mabilleau 2014). In ovariectomised rats, exenatide (a shorter-acting GLP-1 agonist) preserved trabecular bone volume and reduced osteoclast surface area compared to saline controls, an effect attributed to increased osteoprotegerin expression and decreased RANKL signalling (Nuche-Berenguer 2010). These preclinical findings contrast with the modest BMD reductions observed in human trials, highlighting species-specific differences in skeletal response to GLP-1 pathway modulation. The discrepancy may reflect the magnitude and rapidity of weight loss in clinical populations, where mechanical unloading from reduced body mass outweighs any direct anabolic signalling on bone cells.
Fracture incidence data from cardiovascular outcome trials provide additional context. The SUSTAIN-6 trial, which followed 3,297 participants with type 2 diabetes for 104 weeks, reported 24 fractures in the semaglutide groups versus 20 in placebo groups, yielding a hazard ratio of 1.14 (95 percent CI 0.64 to 2.03), a difference not reaching statistical significance (Marso 2016). Similarly, the SELECT cardiovascular trial in 17,604 participants with overweight or obesity and established cardiovascular disease found no elevated fracture risk over a median 39.8 months of follow-up, though BMD was not systematically measured in this study (Lincoff 2023). These findings suggest that while semaglutide may modestly reduce BMD during active weight loss, the clinical consequence in terms of fracture events remains uncertain in populations without pre-existing osteoporosis.
AOD-9604 Evidence Base and Theoretical Bone-Protective Mechanisms
AOD-9604 comprises amino acids 177 to 191 of the human growth hormone (hGH) molecule, a region implicated in lipolytic activity but lacking the N-terminal residues responsible for binding the growth hormone receptor that mediates effects on linear growth and glucose metabolism (Heffernan 2001). Early-phase human trials focused exclusively on fat reduction and metabolic endpoints; a 12-week randomised controlled trial in 300 participants with obesity showed dose-dependent reductions in visceral adipose tissue without changes in fasting glucose or insulin-like growth factor 1 (IGF-1) levels, consistent with the fragment's inability to activate classical GH signalling pathways (Heffernan 2001). No published study has directly measured bone turnover markers, BMD, or fracture incidence in humans receiving AOD-9604, leaving its skeletal effects a matter of extrapolation from the parent hormone's biology.
Growth hormone's effects on bone are mediated primarily through hepatic IGF-1 production and local IGF-1 synthesis in osteoblasts, pathways that require intact GH receptor binding through the N-terminal domain absent in AOD-9604 (Ohlsson 2009). However, some evidence suggests that C-terminal hGH fragments may exert receptor-independent effects on adipocytes through mechanisms involving mitochondrial fatty acid oxidation and AMP-activated protein kinase (AMPK) activation (Ng 2000). Whether similar non-canonical pathways influence osteoblast or osteoclast function remains speculative; no published work has examined AOD-9604 effects on bone cells in vitro or skeletal outcomes in animal models of osteoporosis or mechanical unloading.
The theoretical advantage of AOD-9604 for bone health during weight loss rests on the assumption that selective fat reduction without concurrent lean mass loss might preserve the mechanical loading stimulus necessary for bone maintenance. A small open-label study in 24 adults measured body composition by DXA after 12 weeks of subcutaneous AOD-9604 at 1 mg daily, reporting a 2.8 kg reduction in fat mass with no significant change in lean mass (mean change +0.2 kg, p=0.61), though bone mineral content was not reported as a separate endpoint (Heffernan 2001). If replicated in larger controlled trials, such preferential fat loss could theoretically mitigate the mechanical unloading that drives BMD reduction during rapid weight loss with other agents, but direct bone measurements are absent from the existing literature.
Comparative Limitations and Evidence Gaps
Direct head-to-head trials comparing AOD-9604 and semaglutide for any endpoint, including bone health, have not been conducted. The evidence base for semaglutide comprises multiple phase-3 trials with DXA substudy data and long-term cardiovascular outcome trials that captured fracture events as adverse events, whereas AOD-9604 evidence consists of two small phase-2 trials from the early 2000s that did not include bone-specific assessments (Heffernan 2001, Munro 2003). This asymmetry makes formal comparison impossible; semaglutide has documented modest BMD reductions in the context of substantial weight loss, while AOD-9604 has no human bone data to support or refute skeletal effects.
The populations studied also differ in ways that limit generalisability. STEP trial participants had a mean baseline body mass index of 37.9 kg per square metre and were predominantly middle-aged (mean age 46 years), a demographic at lower baseline fracture risk than postmenopausal women or older adults with established osteoporosis (Wilding 2021). AOD-9604 trials enrolled similar BMI ranges but were underpowered (n=24 to 300) and of shorter duration (12 weeks maximum) compared to the 68-week STEP protocols, precluding detection of skeletal changes that typically require 12 to 24 months to manifest on DXA imaging (Heffernan 2001). Neither compound has been studied specifically in populations with low baseline BMD or prevalent vertebral fractures, the groups most vulnerable to skeletal complications during weight reduction.
Methodological differences in body composition assessment further complicate interpretation. The STEP trials used whole-body DXA to measure lean mass, fat mass, and BMD at standardised skeletal sites (lumbar spine, total hip, femoral neck), whereas the single AOD-9604 study reporting body composition used DXA only for fat and lean mass quantification without site-specific bone measurements (Heffernan 2001, Wadden 2021). Bone turnover markers, such as C-terminal telopeptide of type I collagen (CTX) for resorption and procollagen type I N-terminal propeptide (P1NP) for formation, were not measured in any AOD-9604 trial, leaving the acute skeletal response to this peptide entirely unknown.
Open Questions Requiring Controlled Investigation
Does AOD-9604 preserve bone mineral density during caloric restriction compared to weight loss achieved through other mechanisms? This fundamental question requires a randomised trial with serial DXA measurements at baseline, 12 months, and 24 months, comparing AOD-9604 to both placebo and an active comparator (such as semaglutide or another GLP-1 agonist) in participants undergoing equivalent caloric deficits. Inclusion of bone turnover marker sampling at 3-month intervals would clarify whether any observed BMD changes reflect altered remodelling balance or simply mechanical unloading from reduced body mass.
What is the dose-response relationship between semaglutide-induced weight loss and BMD reduction, and does the rate of weight loss influence skeletal outcomes? Post-hoc analyses from STEP 1 suggested that participants losing more than 15 percent of baseline body weight experienced greater hip BMD reductions than those losing 5 to 10 percent, but these subgroup analyses were not prespecified and lacked statistical power for fracture endpoints (Rubino 2021). A prospective trial randomising participants to different semaglutide titration schedules (achieving identical total weight loss over 12 versus 24 months) could isolate the effect of weight-loss velocity from total magnitude.
Do other peptides with documented effects on body composition, such as MOTS-c, tesamorelin, or CJC-1295, offer skeletal advantages over either AOD-9604 or semaglutide during weight reduction? MOTS-c, a mitochondrial-derived peptide, has shown metabolic benefits in preclinical models but lacks human bone data (Lee 2015). Tesamorelin, a growth hormone-releasing hormone analogue approved for HIV-associated lipodystrophy, increases IGF-1 levels and has demonstrated small increases in lumbar spine BMD in some trials, though its effect during active weight loss is unknown (Falutz 2010). CJC-1295, a long-acting GHRH analogue, similarly elevates IGF-1 but has not been studied for bone outcomes in controlled trials (Teichman 2006). Retatrutide, a triple agonist at GLP-1, GIP, and glucagon receptors, produced weight reductions exceeding those with semaglutide in phase-2 trials (mean 17.5 percent at 48 weeks with the 12 mg dose), but BMD data have not been reported (Jastreboff 2023).
The interaction between peptide-mediated weight loss and baseline bone health status also remains unclear. Participants with osteopenia (T-score -1.0 to -2.5) or osteoporosis (T-score below -2.5) at baseline may experience disproportionate BMD reductions or fracture risk elevation during rapid weight loss, yet these populations were not separately analysed in STEP trial BMD substudies. A trial specifically enrolling postmenopausal women with low baseline BMD, randomising them to semaglutide with or without concurrent bisphosphonate therapy, would clarify whether skeletal protection is necessary during GLP-1-mediated weight reduction in high-risk groups.
Interpreting the Current Evidence for Skeletal Safety
For individuals considering peptide-based weight loss, the available evidence supports several provisional conclusions. Semaglutide produces clinically meaningful weight reduction (12 to 16 percent of baseline body mass over 68 weeks) accompanied by modest, statistically significant decreases in hip and spine BMD (approximately 1.0 to 1.2 percent), but fracture incidence in trials to date has not exceeded placebo rates in populations without pre-existing osteoporosis (Wilding 2021, Rubino 2021, Lincoff 2023). This pattern suggests that the skeletal consequences of GLP-1-mediated weight loss are detectable by DXA but may not translate to increased fracture risk in younger, healthier populations over 1 to 3 years of follow-up.
AOD-9604, by contrast, has no published human data on bone density, turnover markers, or fracture incidence. The theoretical mechanisms by which it might preserve bone, primarily through selective fat loss without lean mass reduction and avoidance of GH receptor-mediated effects on glucose metabolism, remain untested in controlled trials. The single body composition study showing preserved lean mass during fat reduction is insufficient to infer skeletal benefit, as bone and muscle respond to different mechanical and hormonal signals during weight loss (Heffernan 2001). Extrapolation from growth hormone's bone-anabolic effects is inappropriate given AOD-9604's lack of GH receptor binding and absence of IGF-1 elevation in human trials.
The magnitude of BMD reduction with semaglutide, while statistically significant, falls within the range observed with other weight-loss interventions. A meta-analysis of 41 studies encompassing bariatric surgery, very-low-calorie diets, and pharmacological weight loss found that each 10 percent reduction in body weight associated with a mean 1 to 2 percent decrease in hip BMD, regardless of method (Ensrud 2005). Semaglutide's skeletal effects thus appear consistent with mechanical unloading from reduced body mass rather than a drug-specific toxicity. Whether AOD-9604 would produce similar BMD changes per unit of weight lost, or whether its purported selectivity for fat mass translates to skeletal preservation, cannot be determined from existing publications.
For populations at elevated baseline fracture risk, the absence of long-term skeletal safety data for both peptides represents a meaningful knowledge gap. Postmenopausal women, older adults with sarcopenia, and individuals with prevalent vertebral fractures face higher absolute fracture rates, and even modest BMD reductions may increase risk in these groups. Current osteoporosis guidelines recommend baseline DXA screening before initiating weight-loss therapy in adults over 65 or those with risk factors (Cosman 2014), but do not provide specific guidance on peptide-based interventions. Serial DXA monitoring at 12 to 24 months during active weight loss with either semaglutide or AOD-9604 would allow early detection of excessive BMD loss, though thresholds for intervention (such as adding bisphosphonates or denosumab) have not been established in this context.
The clinical decision between these peptides, when skeletal health is a primary concern, must therefore rest on the asymmetry of evidence rather than direct comparative data. Semaglutide has documented efficacy for weight reduction, known modest effects on BMD, and reassuring fracture data in trials up to 3 years, whereas AOD-9604 has no skeletal safety profile at all. For an individual with normal baseline BMD seeking weight loss, the semaglutide data suggest manageable skeletal risk with monitoring. For someone with established osteoporosis, neither peptide has been validated as safe, and alternative approaches (such as lifestyle modification with resistance training, or weight loss combined with bone-protective pharmacotherapy) may be more appropriate until controlled trials address this population specifically. The current literature supports neither a skeletal advantage nor equivalence for AOD-9604 relative to semaglutide; it supports only the conclusion that one has been studied and the other has not, a distinction that matters when evaluating fracture risk during intentional weight reduction.
Self-administration of unapproved compounds carries risks that are not fully characterised in the published literature.
Share This Article