The main objective of this study was to evaluate the effects of race/ethnicity, genetic admixture, and body fat on BMC among a sample of peripubertal children. There have been inconsistencies in results from studies investigating the influence of body fat on BMC among various populations. Some studies indicate that overweight children and adolescents have higher BMC compared with their normal-weight peers whereas others conclude that overweight is associated with lower BMC [19
]. It is possible that the inconsistencies can be attributed to the population studied, the method used to assess body habitus (i.e., BMI versus DXA), or the confounding effect that body size, fat distribution, and body composition have on the evaluation of adiposity and bone mass [17
]. In this study we show, in a multiethnic cohort of children, that percent fat assessed by DXA and ancestral genetic background (but not race/ethnicity) both contribute to BMC. Further, the relationship between admixture and BMC is fairly consistent in both sexes and across the maturation process. However, the impact of fat accumulation on BMC differs according to genetic admixture.
The degree of ancestry has been shown to be of clinical importance among adult women. Hill et al. [8
] demonstrated that Tobagonian women with greater West African ancestry had 10–18 and 29% higher bone density compared to non-Hispanic black and white women in the United States, respectively. Others have demonstrated an inverse relationship between bone mass and European ancestry among women [9
]. Although previous pediatric studies have examined the relationship between fat mass and BMC [17
], the specific contributions of genetic markers and body fat to BMC have not been reported. Our study showed that ancestral genetic background was associated with BMC. Interestingly, the relationships were maintained, in general even after stratifying by sex or Tanner stage, suggesting a significant influence of genetic admixture irrespective of sexual dimorphism or body composition changes resulting from reproductive maturation.
The effect of ancestral genetic background may be mediated by fat distribution [17
], insulin dynamics [35
], adipokines [36
], and/or hormones [37
], all of which influence BMC. The racial/ethnic differences and interactions between fat deposit and BMC observed by Afghani and Goran underscore not only the complexity of body composition relationships in terms of physiology but also the importance of including genetic factors in the analysis [17
]. The results of our study demonstrate the differential impact of genetic factors and adiposity on BMC that may not be identified using racial/ethnic category as the independent variable. The extent to which inherent physiological factors may also contribute to BMC is an avenue worth exploring.
Behavioral factors such as physical activity and dietary calcium and vitamin D have been consistently reported to impart an effect on BMC [38
]; however, we did not detect a contribution of either in this sample. In our sample, racial/ethnic differences in these independent variables were shown; however, when included in regression models, there was no relationship between the physical activity or diet with BMC. The lack of association suggests that, in our sample, racial/ethnic differences in BMC were not accounted for by disparate physical activity and/or diet levels. It is likely that the relatively small amount of physical activity, particularly in terms of moderate to vigorous activity, was not sufficient to explain the variance in this sample of healthy children. In addition, it is plausible that in this cohort the dietary calcium and vitamin D intakes were adequate to prevent deficiency but not high enough to produce a contribution in the presence of genetic and other environmental factors (high adiposity). Optimizing calcium and vitamin D intake has been shown to increase bone mineralization in children and adolescents, primarily in children with the lowest intake [38
]. Nevertheless, throughout the pubertal transition, the interaction of genetic, biological, physiological, and environmental factors all play a role in the determination of not only bone mass accrual but also bone loss throughout life. Significant evidence indicates the full genetic potential is only attained if nutrition, physical activity, and other lifestyle factors are optimized. Our results suggest that efforts to improve dietary quality and engagement in physical activity in this population should be addressed.
The strengths of this study were the use of ancestral genetic admixture in addition to racial/ethnic classification to separate the biological and nonbiological aspects of race and robust measures of fat and lean mass in a multiethnic population with a wide range of body habitus. To the best of our knowledge, this was the first study to investigate the independent roles of genetic admixture and adiposity in the pediatric skeletal system. There are, however, limitations in using DXA as it may imprecisely estimate BMC, particularly in overweight individuals [41
]. However, because our participants were early pubertal children with BMIs in the average range for children and well below the average range for adults, as well as the type of DXA used, we believe that the observed BMC findings in our study were accurate. We do not rule out the possibility that other measures [e.g., serum calcium, parathyroid hormone, vitamin D, leptin, insulin-like growth factor (IGF)] levels may also mediate the effect of ancestry on BMC; however, for reasons of limitations in amount of sera collected and funding, we do not have these measures available. In addition, the cross-sectional nature of the study design prevented the inference of long-term relationships over the pubertal transition; longitudinal data will be required to determine the long-term contribution of individual adiposity and genetic makeup on ultimate bone health. Further, the relatively small sample included only participants from a discrete geographic area, limiting generalizablity. Nevertheless, although the interaction terms were not significant, it is not likely because of power, because power calculations indicated that we needed a sample of 55 to obtain a significant effect, and we had this sample size. Additional analyses examining sex-related differences across Tanner stages were evaluated. This analysis demonstrated a sex-related difference at Tanner 1 with an effect in girls but not in boys, whereas the effect reverses as puberty progresses. However, the separation of samples according to this stratification might bring bias to our results, preventing an accurate conclusion, partly because of the variation in reproductive maturation across groups (i.e., variance in admixture estimates would be limited by stratification). The evaluation of the role of genes in the etiology of racial/ethnic differences in physiological outcomes requires the understanding that genetic similarity cannot be inferred simply based on racial categories. When accounting for the genetic heterogeneity of the participants in our study, the genetic and biological factors influencing fat mass accumulation in African Americans in fact may be the same aspects decreasing susceptibility to osteoporosis in the adult population among this group. Conversely, excess body fat may worsen the influence of AMINADM on BMC. These results indicate population differences in body composition are at least in part the consequence of ancestral genetic background, suggesting the utility of inclusion of admixture for identifying susceptibility for osteoporosis and obesity. Future research on the evaluation of genes influencing bone mass accrual in children is warranted.