Bone Mineral in Prepubertal Children: Gender and Ethnicity†
Presented in part as Horlick M, Thornton J, Wang J, Fedun B, Levine LS, Pierson RN. 1999 Determinants of bone mineral in prepubertal children: Gender, ethnicity, age, weight, and height. Pediatr Res 44:91A (abstract).
Previous reports of gender and ethnic differences in bone mineral in prepubertal children have been inconsistent due to different methodologies, the problematic nature of bone density by dual-energy X-ray absorptiometry (DXA) calculated as the ratio of bone mineral mass to projected bone area (BA), and the generally small study populations. The aim of this study was to test the hypothesis that gender and ethnic differences in bone mineral by DXA are present in prepubertal children. The subjects were 336 healthy Asian, black, and white prepubertal children (172 females and 164 males). Total body bone mineral content (TBBMC) was adjusted for total body BA (TBBA), age, height, and weight. Adjusted mean TBBMC was greater in males than in females (p = 0.01). The gender difference was independent of ethnicity. Adjusted mean TBBMC was different for black compared with nonblack children (p = 0.001). The ethnic difference was a function of TBBA and weight. This study in a multiethnic population of prepubertal children shows (1) a gender difference in TBBMC and (2) an ethnic difference in TBBMC.
OSTEOPOROSIS IS a major health problem, which is expected to increase in prevalence with improved longevity in the next century.(1) Fracture risks vary between countries, between women and men, and between ethnic groups. The greatest fracture rate has been identified in white postmenopausal women.(2, 3) Overall rates are highest in whites, intermediate in Asians, lowest in blacks, and are related to peak bone mass and bone mass decline.(4, 5) Studies indicate that half of reference young adult bone mass is achieved before puberty. Prevention of osteoporosis by optimizing bone mass accretion during childhood and adolescence may be more effective than treatment later in life and may be most effective before the onset of puberty.(6, 7) Although much of bone mass is determined genetically, exercise and nutrition have been shown to influence bone mineral in childhood.(8–11) Because fracture rates vary by gender and ethnicity, establishing the timing of onset and characteristics of gender and ethnic differences in bone mass would help evaluate individuals and groups thought to be at particular risk.
Gender and ethnic differences in bone mass have been noted in young children, but reports are inconsistent.(12–23) Investigators' use of different methods, different sites, and different measurements complicates comparison of these studies. For example, although most recent reports used dual-energy X-ray absorptiometry (DXA), some investigators analyzed bone mineral density while others analyzed bone mineral content (BMC). Bone mineral density by DXA is not “true” bone density but rather a calculated value—BMC in grams divided by bone area (BA) in square centimeters (g/cm2) —“areal” bone density. Measurement of BMC in grams by DXA correlates well with total body calcium.(24) However, BA by DXA, expressed as anterior-posterior projection of bone in square centimeters, is a two-dimensional measurement of a three-dimensional structure.(25) Therefore, this calculated areal bone density is recognized as being misleading, especially when bone size varies.(26–28) Greater areal bone density means more mineral for a given BA, which could be explained as readily by thicker bones—the third dimension not measured—as by greater true density. Volumetric bone density is seldom used in children because of high radiation exposure.
Animal carcass analysis showing correlation of BMC results with ash weight has been used to validate the applicability of DXA scanning to pediatric subjects.(29–31) In vitro neutron activation measurement of total body calcium has been used to validate measurement of BMC by DXA in adults, but this is not suitable for pediatric subjects.(24) Because of availability, speed, and low radiation exposure, use of DXA scans for clinical evaluation of children in whom there is concern about bone mineral status is increasing.(32)
The aim of this study was to test for gender and ethnic differences in bone mineral by DXA in prepubertal children, before the effects of gonadal steroids and other hormones of puberty are present. Total body BMC (TBBMC) by DXA was adjusted for recognized covariates—total body BA (TBBA), height, weight, and age.(25, 27, 28) To our knowledge this is the first study of a population of exclusively prepubertal children of Asian, black, and white backgrounds.
MATERIALS AND METHODS
The subjects were 336 Asian, black, and white children (172 females and 164 males) from the New York metropolitan area. Volunteers were recruited by local newspaper notices, announcements at schools and after-school activity centers, and by word of mouth. Consent was obtained from each volunteer's parent or guardian, and assent was obtained from each volunteer as well. Ethnicity was established by consistent Asian, black, or white background of both parents and all four grandparents by questionnaire. Subjects of mixed ethnic background were not included. There were no height or weight restrictions to entry into the study. A medical history from the parent or guardian and a physical examination at the time of the bone mineral evaluation confirmed normal health status. All subjects were prepubertal by the criteria of Tanner.(33) The Institutional Review Board of St. Luke's-Roosevelt Hospital Center approved the study.
Body weight was measured to the nearest 0.1 kg on a balance beam scale (Weight Tronix, New York, NY, U.S.A.) and height was measured to the nearest 0.1 cm using a wall-mounted stadiometer (Holtain, Crosswell, Wales, U.K.).
TBBMC in grams and TBBA in centimeters squared were measured by DXA (Lunar DPX, pediatric software version 3.8 g; Lunar Corp., Madison, WI, U.S.A.). Scan mode was chosen according to the weight guidelines provided by the manufacturer. The CV for repeated TBBMC measurements in human subjects is approximately 1.5%.(34) Repeated studies of a phantom in our unit revealed a CV of 0.6% for BMC and of 0.8% for BA.
Descriptive statistics were calculated for each variable and were reported as the mean and SD. A two-factor analysis of covariance was used to analyze TBBMC. The two factors were gender and ethnicity. The covariates were TBBA, age, weight, and height. The mean TBBMC values presented were all adjusted to the study population's mean TBBA, mean age, mean weight, and mean height.
All statistical calculations were performed using the Statistical Analysis System (SAS) statistical software package (version 7.00) for personal computers. The level of significance for all statistical tests was 0.05.
Subject characteristics and results of the DXA bone measurements are presented in Table 1. Subjects were 135 Asian children, 79 black children, and 122 white children (172 females and 164 males). The mean age for the entire study group was 8.4 (±1.5), with a range in age of 6–11 years. All were prepubertal by Tanner criteria.
Table Table 1. Characteristics and Bone Measurement Results for Prepubertal Children (Mean [SD])8
The results of analysis of covariance for gender and ethnic differences are presented in Table 2. There were no interactions between gender and ethnicity or between gender and any of the covariates. The difference in adjusted mean TBBMC between males and females was a constant and was independent of ethnicity. The adjusted mean TBBMC was significantly higher for males than for females (difference = 14 g; p = 0.01).
Table Table 2. Adjusted Mean TBBMC in Grams by Gender and Ethnicity (Mean [Standard Error of the Mean])8
There were interactions between ethnicity and two of the covariates, TBBA and weight. The effect of ethnicity on TBBMC was significant. The adjusted mean TBBMC was significantly greater for black children than for nonblack children (difference = 31 g; p = 0.001). Because there were significant interactions between ethnicity and two of the covariates, the difference in TBBMC between black and nonblack children was a function of TBBA and weight. The difference in TBBMC (black minus nonblack) is given by the following equation:
For most children in this study population the difference calculated using TBBA and weight is positive (i.e., black children have higher values for TBBMC than nonblack children).
This study shows gender and ethnic differences in bone mineral by DXA in prepubertal children. We suggest that methodological differences and small subject populations in previous studies of prepubertal children have resulted in inconsistent findings. The problematic nature of areal bone density by DXA has contributed to this inconsistency.(25–28, 32)
The finding of greater TBBMC in males compared with females is consistent with previous studies of total body composition in prepubertal children—males have relatively more lean and less fat mass than females and bone is a component of the lean body.(12, 17, 35) The suggestion that there is greater TBBMC in prepubertal males than in females after adjustment for TBBA, height, weight, and age is less expected.
Although bones of male fetuses were found to be longer and heavier than female bones in cadaveric analyses, subsequent studies of gender differences during childhood have resulted in inconsistent or contradictory findings.(36) In several reports using TBBMC or areal bone density by DXA in subjects ranging in age from 1 to 20 years, gender differences were either not observed or were only seen after the onset of puberty.(21, 37–44)
Regional studies of prepubertal children have suggested gender differences. Tanner reported greater bone width of the humerus by X-ray in males compared with females as young as 3 years old, and two studies of young children reported greater BMC of the radius by single-photon absorptiometry (SPA) in males compared with females after controlling for differences in body size.(12–14) In contrast, Nelson et al. found a gender difference (females greater than males) in DXA lumbar BMC in children ages 8–10 years but the pubertal status of subjects in that study was not established.(21) Using DXA in prepubertal Australian children, Jones and Dwyer reported higher size-adjusted BMD at the hip in males but higher size-adjusted BMD at the spine in females.(16) By quantitative computerized tomography (QCT) of prepubertal white males and females in California, Gilsanz et al. reported an independent effect of gender on size of the vertebrae (males greater than females) but not on the size of the femoral shaft.(15) These studies suggest the presence of gender differences in regional bone mass, and discrepancies in specific findings by site are most likely related to the different features of bone measured by regional X-ray, SPA, DXA, and QCT and to the small populations studied.
The current study of 336 prepubertal Asian, black, and white children shows a small but significant prepubertal gender difference in adjusted TBBMC by DXA.
Bone mineral differences between black and nonblack children also are consistent with previous body composition studies showing ethnic distinctions in lean body mass, including bone mineral, in prepubertal black children compared with Asian and white children or in black children and adolescents compared with Mexican-Americans and whites.(17, 19, 20, 21)
Although cadaveric analyses of black and white fetal skeletons suggested ethnic differences in bone length and weight, findings of regional bone mineral studies have not been consistent.(36) Several reports from the United States, measuring BMC or BMD by SPA and DXA, have noted a skeletal advantage of black children, although not all included pubertal assessment or analyzed prepubertal children separately.(18–21, 45, 46) At least one study in the United States found no difference in lumbar BMD between black and white children and adolescents.(47) Using QCT, Gilsanz et al. suggested that the difference in spine BMD between American black and white females emerges during puberty, and M-C Wang et al., using bone mineral apparent density (calculated from regional DXA using assumptions about the geometry of vertebrae), agreed.(22, 46) Gilsanz et al. reported a significant prepubertal black-white difference in femoral length by QCT, but only in boys not girls, and not in other axial or appendicular bone QCT images.(23) Prentice et al. reported greater radial BMD by SPA of white British infants compared with black Gambian infants after correction for weight.(14) Additionally, South African black children were found to have marginally but not significantly greater radial bone mass compared with white children after correction for height.(48) Social, nutritional, and genetic contrasts between study populations, as well as differences in technique, may contribute to the discrepancies in findings. Our data show an ethnic difference (black compared with Asian and white) in prepubertal TBBMC. This black/nonblack difference is a function of TBBA and weight.
It has been reported that Asians have lower bone mineral density than whites, but several studies, including two from this center comparing Asian and white adult men and women, show that this may be an artifact of smaller bone size.(46, 49–52) In this study there was no difference in adjusted TBBMC between Asian and white prepubertal children of the same gender.
This study shows subtle but significant gender and ethnic differences in TBBMC by DXA in prepubertal children. The findings suggest that the problematic nature of areal bone density by DXA has contributed to the previous inconsistencies in results, and that use of TBBA as a covariate, rather than as a component of calculated areal BMD, makes more meaningful interpretation possible.(26–28)
Although the immediate clinical implication of the gender difference of 14 g (1.3% of mean TBBMC) and the ethnic (black/nonblack) difference of 31 g (2.8% of mean TBBMC) remains to be determined, the findings indicate that interpretation of pediatric DXA bone mineral measurements should be adjusted for covariates (including TBBA) and that reference values should be gender and ethnic specific. Although there are differences in analytic algorithms between manufacturers and there are no uniform standards for pediatric subjects as there are for adults, the principle of interpretation of TBBMC using TBBA is applicable to all scanners.(32, 53)
This study substantiates that the processes that determine fracture risk in adulthood begin well before puberty, and that interventions to enhance bone mass accrual or bone strength, “the bottom line,” should start early in life.(54)
Supported by The National Institute of Health grant DK 37352.