The Effect of Dietary Protein on Bone Health During Energy Restriction and Exercise Training in Overweight and Obese Adults

Christian Stephen Wright, Purdue University


With nearly two-thirds of Americans classified as overweight or obese, quality of life has dramatically decreased in the wake of increasing metabolic disorders. Diet- and exercise-induced energy restriction are commonly used for the treatment of obesity and its associated metabolic disorders, particularly in conjunction with a higher protein weight loss diet (HPWL, >1.2 g protein/kg/day). However, body weight is closely related to bone mass and WL is shown to decrease bone mineral density (BMD) and increase the risk of skeletal fracture. Moreover, though beneficial for WL and certain metabolic outcomes, controversy exists surrounding the proposed detrimental effects of dietary protein on bone. The proposed protein-induced bone loss effect has been termed the acid-ash hypothesis, where the consumption of excess dietary protein increases acid production and acidifies the environment surrounding the bone leading to demineralization, a negative calcium balance, and the exacerbation of WL-induced bone loss. In contrast, others have cited the beneficial effects of a HPWL on bone health, showing increases in calcium absorption and circulating insulin-like growth factor 1 (IGF-1) concentrations. Conflicting outcomes are therefore reported, showing either an attenuation, neutral, or exacerbation of bone loss with a HPWL diet. Growing evidence submits that the conflicting outcomes reported are a result of differences in predominant source of dietary protein. Different sources of dietary protein vary in their amino acid, micronutrient, and bioactive compositions which may differentially influence bone health. Such effects include greater increases in IGF-1 concentrations with meat versus soy consumption, lower bone turnover rates with a high dairy HP diet versus a mixed HP diet, and greater increases in cortical volumetric BMD and femur calcium content with a HP soy diet versus a HP milk diet. In addition, data from our lab suggest that both age and hormonal status influence these protein source effects on bone mass as an omnivore, HPWL diet vs. a lacto-vegetarian, normal protein (NP) WL diet promoted a decrease in BMD in only postmenopausal women. Due to the conflicting outcomes in the literature and the potential influence of protein source on these effects, the purposes of research conducted in Studies 1-3 (Chapter 2-4) were to determine the effects of dietary protein on bone quantity and morphology in overweight and obese adults (humans and rodents) during energy restriction. Given the potential anabolic effects of exercise-induced mechanical loading and whey protein (WP) supplementation on BMD, Study 1 (Chapter 2) retrospectively investigated the effects of 1) WP supplementation (0g, 10g, 20g, and 30g consumed twice daily with breakfast and lunch), 2) total protein intake (TPro; dietary intake + supplementation) and 3) change in TPro (post-pre) on BMD and bone mineral content (total body, total femur, femoral neck, and lumbar spine) following a 36-week exercise (EX) intervention (resistance 2 d/wk and aerobic 1 d/wk) in obese adults. We hypothesized that WP supplementation would increase measurements of bone quantity in a dose-dependent manner, and that both higher TPro and greater positive changes in TPro would increase measurements of bone quantity. However, contrary to our hypothesis, neither increased WP supplementation, higher TPro, nor greater changes in TPro were associated with changes in BMD over time. Though contrary to our hypothesis, these results contribute additional evidence to further discrediting the acid-ash hypothesis and the detrimental effects of dietary protein on bone. Given the variability in the literature concerning the effects of a HPWL diet on bone health and the lack of any comprehensive, systematic, and analytical evaluation of the current literature, the purpose of Study 2 (Chapter 3) was to utilize a systematic review and meta-analysis approach to investigate the effect of a HPWL diet on bone quantity in healthy adults. We hypothesized that a HPWL diet (≥25% of energy from protein or ≥1.0 g/kg/day) would attenuated WL-induced decreases in bone quantity more so than a NPWL diet. Though the HP diets did show a statistical attenuation of total body and lumbar spine BMD following WL, the amount of attenuation was within or near the margin of error for DXA (0.01 g/cm2). Thus we were unable to conclude if said attenuation was legitimate or clinically relevant. Therefore, contrary to our hypothesis, the current literature supports neither a positive nor a negative effect of a HPWL diet on bone quantity in adults. The importance and future implications of Study 2 cannot be understated. Not only is Study 2 the first and only systematic review and meta-analysis of its topic, but by using the pinnacle of evidence-based research it also further clarifies and solidifies the effects of HPWL diets on bone while simultaneously addressing areas of concern and proposing future research endeavors. Aware of the potential influence of protein source in HPWL diets, Study 3 (Chapter 4) was a highly-controlled, mechanistic-driven WL intervention which investigated the effects of 1) protein source (beef, milk, soy, control), and 2) protein quantity (high vs. normal protein) on bone morphology (bone quantity and microarchitecture) and body composition in a rat model of postmenopausal obesity. Based upon preliminary data from our lab and previously conducted clinical interventions, we hypothesized that the HP, high-beef diet would exacerbate WL-induced bone loss, while the HP, high-milk diet would attenuate WL-induced bone loss more so than other energy restricted diets. In agreement with our preliminary data, a HP, high-beef WL diet increased trabecular separation and created more plate-like trabecular following WL than any other energy restricted diet. Post-intervention trabecular bone volume fraction was also lower with the HP, high-beef diet in comparison to the HP, high-soy diet. However, contrary to our hypothesis, the HP, high-milk diet did not influence changes in bone morphology following WL. Study 3 brings clarity to the disparity in the literature concerning the effects of HPWL diets on bone health. Results show that protein source and not simply protein quantity differentially influences bone health during WL. Such findings not only generate additional questions and promote further investigations into the effects of HPWL diets on bone health, but also suggest that protein source specific recommendations should be made in order to prevent increases in WL-induced bone loss in postmenopausal women. Collectively, Studies 1-3 represent a multidisciplinary and translational approach for investigating the effects of dietary protein on bone in overweight and obese adults during energy restriction, utilizing primary clinical research, analytical techniques, and basic animal research to systematically and sequentially address the research topic.




Campbell, Purdue University.

Subject Area


Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server