Osteoporosis is a most common metabolic disease of bone and approximately 40% of postmenstrual women met the criteria of osteoporosis.¹⁾ Osteoporosis is defined as the bone disease with low bone strength and high risk for fracture by The National Institutes of Health Consensus Development Panel on Osteoporosis Prevention, Diagnosis, and Therapy.²⁾ Major risk factors of osteoporosis is low bone mass in adolescence and rapid bone loss by aging and menopause. Therefore, maximizing bone mass in adolescence is important to prevent osteoporosis.³⁾ Measurement of bone mineral density (BMD) is a useful methods to evaluate osteoporosis status.^4,5) In previous studies for adults, body fat mass and lean body mass have been related to BMD. In these studies, lean body mass has been positively associated with BMD regardless of sex and age, while the association between body fat mass and BMD has been inconsistent according to the menopausal status^6,7) and age groups.^8,9,10)

On the other hand, studies in adolescents have shown inconsistent results about gender difference in the relationships between body fat mass and BMD.^11,12) In addition, studies for obese adolescents reported an inverse association between abdominal fat and BMD^13,14) and a recent study about the relationship between regional BMD and body fat percent resulted in age and sex-specific relationships in adolescents of the Korea National Health and Nutrition Examination Survey (KNHANES) in 2009. However, that study did not examine the associations with lean body mass after taken into account of bone mass-free lean body mass.¹⁵⁾ Therefore, it is not clear whether the associations between body fat percent and BMD would be different by regions and independent from lean body mass free from bone mass.

The present study aimed to evaluate the relationships between BMD and body fat percent of whole body, trunk, and extremities in adolescents after adjusting for body weight or bone mass-free lean mass using the data of KNHANES in 2010.

Table 1 shows distribution of study variables in boys and girls. The WBFP, TBFP, and EBFP were higher in girls as compared with boys, while TEFR was higher in boys as compared with girls. The BMD of whole body and total femur in boys were higher as compared with girls, while LSBMD was higher in girls.

Table 2 and Table 3 present the relationships between BMD and body fat percent in boys and girls, respectively. When the relationships were adjusted for body weight and other confounding variables, there were inverse relationships between body fat percent and BMD in both genders. In contrast, the relationships between TEFR and BMD were positive in boys, while there were inverse relationships in girls. On the other hand, in the models adjusted for bone mass-free lean mass of whole body instead of body weight, the relationships of body fat percent and TEFR with BMD were not more significant in girls, while the relationships between TEFR and BMD were still significant in boys.

Osteoporosis, a common metabolic bone disease related with aging and menopause, is an important burden of public health. Therefore, achievement of optimal bone mass in adolescents is critical for prevention of osteoporosis in later life. Bone mass reaches its peak at 18 to 25 years of age and genetic and non-genetic factors such as nutritional status, hormone, and physical activity in adolescence influence on the process of bone mass formation.¹⁹⁾ Among the nutritional factors, the relationships between body fat mass and BMD among adolescents have been inconsistent between studies.^11,12,15)

In the present study, we observed inverse associations between body fat percent of whole body, trunk, and extremities and BMD in boys and girls after adjusting for age, weight, height, serum 25-hydroxy vitamin D concentration, intake of energy and calcium, physical activity, and menstrual status for girls. However, the associations were not more significant when bone mass-free lean mass of whole body was substituted for body weight as a confounder regardless of gender. On the other hand, the associations between TEFR and BMD at whole body and regional BMD were gender-specific: for girls, the relationships were inverse associations in the body weight adjusted models but were not significant in the bone mass-free lean mass of whole body adjusted models; for boys, the associations were positive regardless of adjustment for body weight or bone mass-free lean mass of whole body.

Regional body fat mass, especially abdominal body fat mass, was related with BMD in previous studies. Magnetic resonance imaging-measured abdominal visceral fat had an inverse relationship with spinal and whole BMD of obese or non-obese 30 female adolescents.¹⁴⁾ In the study about 175 obese adolescents, after adjusting for whole and abdominal body fat mass, abdominal fat thickness which measured by ultrasonography and BMD had negative relationships.¹³⁾ The pathophysiology of inverse association between abdominal fat and BMD has been explained by the unfavorable effects of adipokine from abdominal fat on bone growth-related factors or insulin resistance associated with visceral fat.^14,20) As compared with these two studies, we examined the association between trunk fat that covers chest and pelvic fat as well as abdominal fat using a different measurement method and adjustment with different confounding variables in obese and no-obese adolescents. In this study, the strength of associations between body fat percent and BMD were attenuated after adjusting for bone mass-free lean mass of whole body. This finding suggests that the apparent associations between body fat percent and BMD could be due to bone mass contained in body weight.

The underlying mechanisms for gender-specific associations between TEFR and BMD after adjusting for bone mass-free lean mass of whole body is unclear. The differences in sex hormone such as the effects of estrogen on body fat^21,22) and cortical and trabecular bone²³⁾ in girls may play a role in the gender-differences. Although the pathophysiology should be clarified, these findings suggest that increase in fat mass of trunk relative to extremities may favorable for increase in BMD in boys regardless of bone mass-free lean body mass.

There are several limitations need to be considered. First, there may be measurement errors according to subjects' age, height, weight, BMD, and body composition²⁴⁾ although DEXA is a standard method of BMD measurement. Second, uncontrolled residual confounding variables like sexual development and disease status could influence on the relationships. Third, the study design is cross-sectional study and then, a causal relationship between body composition and BMD cannot be assessed. Lastly, these findings could not infer the relationship between body fat and BMD in other regions. In spite of these limitations, our results may be generalized to Korean adolescents since the data represented the adolescent populations in Korea.

In conclusion, after adjusting for bone mass-free lean mass of whole body, body fat percent were not independently related to BMD in adolescents, while ratio of trunk fat mass to extremities fat mass was positively associated with BMD in boys but not in girls.