Answers are grounded in this paper and cite their sources.
Clinicians spend countless hours counseling obese patients about weight loss, and the scientific literature is teeming with studies comparing the effectiveness of diets for attaining and maintaining weight loss in the setting of obesity. Among bone and mineral researchers, there is increasing interest in understanding the skeletal effects of both obesity and the weight loss undertaken to treat it.1, 2 However, the treatment of obesity is not the only potential indication for caloric restriction (CR). For years, some individuals—many already at normal weight—have pursued CR with the desire to slow the biological rate of aging and to increase longevity.3 This potential was explored in the Comprehensive Assessment of Long-term Effects of Reduced Intake of Energy (CALERIE) phase 2 trial, a 2-year randomized controlled trial (RCT) that tested the feasibility, safety, and efficacy of CR in nonobese young and middle-aged adults. Effects on the primary outcomes of resting metabolic rate and core body temperature have been published.4 In this issue of the Journal of Bone and Mineral Research, Villareal and colleagues5 report the effects of the CR intervention on bone metabolism and BMD. CALERIE phase 2 enrolled adults 20 to 50 years old with BMI 22 to 28 kg/m2, randomizing 218 participants to CR or ad libitum control diets. The CR intervention was designed to produce weight loss during the first year followed by weight maintenance during the second year. The CR group achieved an average 11.7% CR over the 2 years, resulting in a mean weight loss from baseline of 8.4 kg at 12 months (an 11.5% decrease) and 7.5 kg at 24 months (a 10% decrease). Weight change in the control group was negligible. Villareal and colleagues5 show that spine and proximal femur BMD declined significantly in the CR group, with a ∼2% decline by 2 years. Markers of bone resorption increased at 6 and 12 months in the CR group, and the bone formation marker bone-specific alkaline phosphatase decreased at 12 and 24 months. Should we be concerned that this decline in bone mass will mean a higher risk of fracture in the future? Or, need we not be concerned because the decline is simply the skeleton's appropriate adaptation to a new, lower body weight? Before attempting to answer these questions, it may be useful to review what we have learned from other human studies about the skeletal effects of weight loss. Most of the published studies addressing weight loss and bone health have examined older adults. Large epidemiologic studies have shown clearly that weight loss is associated with higher fracture risk in older adults,6-14 although such studies are generally limited by the possibility that the weight loss was in part the unintentional result of illness that also negatively impacted bone. This potential for confounding prompted the questioning of older women and men in two large prospective cohort studies about weight loss intention.15, 16 Even voluntary weight loss in the setting of overweight or obesity predicted incident bone loss and fracture at the hip in older women15 and incident hip bone loss in men.16 A number of RCTs, most of 6 to 12 months' duration, have examined the effects of weight loss interventions on BMD in overweight or obese older adults17-23 In general, the weight loss groups have experienced decreases in body weight of 9% to 10% and decreases in BMD of about 2%. Loss of bone mass in these trials is more consistently shown at the hip than spine. Decreases in BMD are substantiated by changes in bone turnover markers, making it unlikely that the trials' observations are simply DXA artifacts in the setting of changing soft tissue mass.24 What is the mechanism for loss of bone mass during weight loss? The process is probably multifactorial. A probable contributor is the mechanical unloading of the skeleton with weight loss. Decreased mechanical stress is known to impact both cortical and trabecular bone25, 26; the osteocyte-secreted protein sclerostin is now recognized as a load-responsible inhibitor of bone formation and may orchestrate such effects.27 Muscle is the primary source of anabolic mechanical stimuli for bone tissue,28, 29 so the loss of lean mass with weight loss may be detrimental to the skeleton. Fat not only loads the skeleton but also secretes adipokines. Adiponectin, which is inversely associated with fat mass, stimulates osteoclast differentiation and action30 and is negatively associated with BMD,31 and leptin's association with bone is more complex, but it may also be an independent determinant of BMD.32, 33 Thus, when adiponectin rises and leptin falls with weight loss, bone mass could decline. A reduction in adiposity also decreases circulating estrogens,34 which are critical for bone integrity in men and women. Regarding nutrition, dietary restriction may result in inadequate intake of calcium or vitamin D, and further, intestinal calcium absorption capacity has been shown to decline with weight loss.35, 36 Finally, because caloric restriction in young mice increases bone marrow fat while decreasing bone mass,37 and because greater marrow fat is associated with lower BMD in humans,38 some have hypothesized a causal role for marrow fat in weight loss–induced BMD decline. However, this causal role has not been demonstrated. In contrast to the RCTs of weight loss in overweight or obese older adults, RCTs in younger adults have not convincingly shown decreases in bone mass with weight loss.39-42 Although one lifestyle intervention trial found a small decline in bone loss at the hip with a 4.5% loss of body weight,42 other trials have not found BMD decreases despite 7% to 10% weight loss.39-41 Why might this be the case? Possibly, a greater amount of muscle mass or the distinct sex hormone milieu in younger adults could mitigate negative skeletal effects of weight loss. The younger adult skeleton might be less vulnerable to weight loss. In addition, younger adults may be more likely than older adults to change other lifestyle factors such as exercise during dietary restriction and weight loss. An exception to the relative preservation of BMD in younger adults is after bariatric surgery. Bariatric surgery patients, who are predominantly premenopausal women, routinely lose ∼30% of preoperative weight, and dramatic skeletal changes are well documented in prospective cohort studies, particularly after Roux-en-Y gastric bypass surgery. Reported changes include BMD decreases of 5% to 10% at the hip by DXA, decreases of 5% to 7% at the spine by QCT, and alterations in trabecular and cortical microarchitecture associated with decreased bone strength.43-46 Bariatric surgery not only causes more extreme mechanical unloading of the skeleton and more pronounced changes in adipokines than nonsurgical weight loss, but also, particularly for Roux-en-Y gastric bypass, induces substantial changes in gut-secreted hormones such as peptide YY and leads to malabsorption of nutrients including calcium.45, 47 Thus, although it is highly important to understand and to address these skeletal changes as potential long-term complications of bariatric surgery, there may be limited generalizability to moderate nonsurgical weight loss. CALERIE phase 2 now joins the landscape of RCTs, which until thus far has shown decreases in BMD with weight loss in overweight and obese older, but not younger, adults. In design, the trial is unique in its enrollment of normal weight participants (mean BMI 25 kg/m2). The trial is noteworthy in the young age of its participants (mean 38 years), in its relatively long duration (24 months), and in its CR strategy designed to produce weight loss during the first year followed by weight maintenance during the second year. As summarized above, BMD by DXA at the spine and proximal femur declined significantly in the CR group, with a ∼2% decline by 2 years, and bone turnover markers indicated an increase in resorption with no change or even a decrease in formation. All participants received a daily multivitamin with 1000 IU vitamin D and a calcium supplement of 1000 mg, and intakes of protein and micronutrients were similar between the two groups or greater in the CR group in follow-up, making it unlikely that dietary composition specifics were responsible for the change in bone mass. Interestingly, self-reported physical activity was lower in the CR group in follow-up, perhaps because these nonobese participants cut back on exercise in order not to lose too much weight, or because they had low energy levels due to the CR. Decreased physical activity may have fueled the negative skeletal effects. In multiple regression analysis, Villareal and colleagues5 found that greater declines in fat-free mass by DXA were associated with greater declines in hip BMD in the CR group. What are the implications of the CALERIE phase 2 results? The question of greatest direct interest to this study is whether or not younger, nonobese adults who lose weight by CR are at higher risk for skeletal complications in the future. Unfortunately, no RCT has been powered to determine the effects of CR-induced weight loss in younger adulthood on fracture. We are left, then, to speculate: Is the decline in bone mass with weight loss simply an appropriate physiological adaptation to a new, lower body weight? Or should we be concerned that it is a harbinger of excessive skeletal fragility? Villareal and colleagues5 argue that it is the former. They used their baseline data to generate equations to predict BMD from body weight, age, and sex, and then they compared actual 24-month BMD measured by DXA with expected BMD derived from the prediction equations. They did not find statistically significant differences between actual and expected values and thus conclude that the bone loss was proportional to the amount of weight lost. One arguing that weight loss–induced skeletal changes are innocuous could also point to the absence of higher fracture incidence in the Look AHEAD trial, which evaluated the effects of a weight loss intervention in over 5000 overweight and obese adults with type 2 diabetes.48 Alternatively, one could argue that the decline in bone mass in this and other weight loss trials is a cause for concern, and that the question is really how big a concern. Lower BMI is associated with lower BMD and higher fragility fracture risk,49 and so any meaningful decline in BMI and decline in BMD (even if proportional) presumably moves a person along the curve to a higher risk of fracture. Based on what we know about that curve, the BMI-related difference in fracture risk may be quite small at overweight or obese BMI values,49 and so our level of concern may be low. (In fact, in frank obesity, other factors may complicate the situation and increase fracture risk, particularly site-specific fracture risk.1) However, it is well established that underweight is a substantial risk factor for fracture, and for CALERIE phase 2 participants who were normal weight when they embarked on CR (eg, those who started with a BMI of 22 kg/m2), a 10% decrease in body weight may have meant verging on underweight. Particularly if they maintain that new, lower weight in the future as they age, then there might be reason to worry. One could also argue that the decline in bone mass is a cause for concern because we cannot rule out that it overshoots the new body weight and results in excessive skeletal fragility. In the Study of Osteoporotic Fractures, older women with intentional weight loss had a two-fold higher risk of hip fracture than those who had been stable at the same final weight,15 although outcomes in the older adults may have limited generalizability to younger adults. Bariatric surgery appears to increase fracture risk compared to the age- and sex-matched general population,50 although outcomes after extreme surgical weight loss may have limited generalizability to moderate nonsurgical weight loss. With CALERIE phase 2, the 2% BMD decrease is distinctly unexpected for the age range; we would expect BMD to remain roughly stable in these participants until approximately menopause for women and a generally similar age in men. Further, a potentially concerning finding Villareal and colleagues5 present in Fig. 1 of their report is that even though the weight loss in CALERIE phase 2 happened in the first year, with weight maintenance during the second year, BMD declined progressively over the 2 years. This pattern has been reported in postmenopausal women51 and in Roux-en-Y gastric bypass patients.52 If CALERIE phase 2 participants had been followed beyond 24 months, would the investigators have observed a continued decline in BMD? Future studies of the skeletal effects of weight loss should address this question by monitoring participants even longer after weight stabilization. Finally, a disquieting observation from prior studies is that when bone mass is lost during weight loss, it does not seem to be regained in full when weight is regained.51, 53 This mirrors the observation that the BMD decline during spaceflight and the microarchitectural deterioration sustained during non–weight-bearing after orthopedic surgery do not recover in full after return to normal weight-bearing, at least not during the periods of time covered by human studies of reloading.26, 54 If BMD change during caloric restriction is simply a proportional adaptation to a new weight, then one would hope it would increase proportionally to weight regain. Can the decline in bone mass during weight loss be prevented? This question is relevant regardless of the extent to which one is alarmed by the ∼2% decrease in BMD in the RCTs of weight loss in overweight and obese older adults and in CALERIE phase 2. Several strategies have been shown to attenuate—although not prevent—the loss of bone mass associated with caloric restriction in older adults. The first of these is exercise, and particularly weight-bearing and muscle-loading exercise, which mitigates the increase in bone resorption and decline in BMD when added to caloric restriction.55 Exercise may have direct positive effects on bone and also result in the maintenance of more lean mass during the loss of fat mass. When dietary strategies have been considered, both higher calcium intake18 and, separately, higher protein intake56 have attenuated declines in BMD during caloric restriction. (In the bariatric surgery population, a multipronged intervention of exercise, calcium, vitamin D, and protein supplementation was recently shown to attenuate postoperative declines in BMD.46) Each of these measures is feasible and inexpensive and thus likely warrants consideration, in the absence of contraindications, even while we await data about future fracture risk. In summary, we now have evidence that weight loss by CR decreases bone mass in overweight and obese older adults and also in nonobese (including normal weight) younger adults. The implications for fracture risk are uncertain and quite likely differ by age, other risk factors for fracture, and initial and final weight or BMI. It remains to be determined how strongly clinicians should counsel patients embarking on CR about a negative impact on skeletal health and recommend measures to attenuate a negative impact. We have a responsibility to our patients and to the general population to understand this issue. ALS is supported by the U.S. Department of Veterans Affairs, Veterans Health Administration, Clinical Science Research and Development Service, Career Development Award-2 (5 IK2 CX000549); her work is also supported by an American Society for Bone and Mineral Research Junior Faculty Osteoporosis Clinical Research Award. The author states that she has no conflicts of interest.
No takes yet. Share an insight, caveat, or question.
Anne L. Schafer (2015) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: