Abstract
Context:
Calcium intake during growth is essential for future bone health but varies widely between individuals and populations. The impact on bone of increasing calcium intake is unknown in a population where low calcium intake, stunting, and delayed puberty are common.
Objective:
To determine the effect of prepubertal calcium supplementation on mean age at peak velocity for bone growth and mineral accrual.
Design and Setting:
Prospective follow-up of boys in rural Gambia, West Africa, who had participated in a double-blind, randomized, placebo-controlled trial of calcium supplementation.
Participants:
Eighty boys, initially aged 8.0–11.9 years, were followed up for 12 years.
Interventions:
Subjects received 1 year of calcium carbonate supplementation (1000 mg daily, 5 d/wk).
Main Outcome Measures:
Dual-energy x-ray absorptiometry measurements were carried out for whole body (WB), lumbar spine, and total hip bone mineral content, bone area (BA), and WB lean mass. Super imposition by translation and rotation models was made to assess bone growth.
Results:
Age at peak velocity was consistently earlier in the calcium group compared to the placebo group, for WB bone mineral content (mean, −6.2 [SE, 3.1]; P = .05), WB BA (mean, −7.0 [SE, 3.2] mo; P = .03), lumbar spine and total hip BA. By young adulthood, supplementation did not change the amount of bone accrued (mineral or size) or the rate of bone growth.
Conclusions:
Twelve months of prepubertal calcium carbonate supplementation in boys with a low calcium diet advanced the adolescent growth spurt but had no lasting effect on bone mineral or bone size. There is a need for caution when applying international recommendations to different populations.
Childhood and adolescence are important periods for skeletal growth in length and width and for mineral accumulation. These are important determinants of current and future skeletal health, and hence fracture risk (1, 2). During the pubertal growth spurt, 20–30% of whole body (WB) bone mineral is accrued (3, 4). Although the internal environment (eg, gender, genetics, and ethnicity) plays a major role in the determination of “peak bone mass,” external environmental factors (eg, diet and lifestyle) are also important (1, 4, 5). Adequate dietary calcium is essential for future bone health because calcium is the most abundant mineral in bone. However, calcium intakes vary widely between individuals and populations, and the benefits of increasing calcium intakes are uncertain. There have been inconclusive results from previous intervention trials in children and adolescents aiming to increase dietary calcium intakes to recommended levels. Most trials reported short-lived increases in bone mineral content (BMC) or bone mineral density (BMD) (6–12), and few have shown lasting effects of supplementation (11, 13, 14). A meta-analysis of 19 trials in children and adolescents concluded that calcium supplementation led to a small but sustained increase in forearm BMD, but not in femoral neck or lumbar spine BMD (15).
One reason for the lack of a permanent effect may be that trials were conducted in populations in the developed world in which calcium intakes are deemed adequate by current recommendations. In the developing world, where low habitual calcium intakes, delayed puberty, low BMC, and stunting are common (5), increasing calcium intakes to levels similar to UK or US guidance may therefore have a greater effect. This could lead to sustained increases in BMC/BMD and, possibly, peak bone mass. In our study of rural Gambian prepubertal children with low mean daily calcium intakes (approximately 300 mg/d), the supplement group had 5.5% greater size-adjusted BMC (SA-BMC) at the radius after 12-month supplementation with calcium carbonate (mean total calcium intake, 1056 mg/d). BMC, bone width (BW), and SA-BMC remained higher 1 and 2 years after supplementation had stopped (14, 16). The same cohort has been followed regularly until the end of height growth. Results showed that boys in the calcium group reached peak height velocity earlier and were taller in midadolescence, indicating an earlier pubertal growth spurt, but stopped growing earlier and were eventually shorter than the placebo group (17). The aim of the current study was to determine the timing and pattern of adolescent increases in BMC, bone area (BA), and lean mass, measured by dual-energy x-ray absorptiometry (DXA), and whether there was a sustained supplementation effect on WB, lumbar spine, total hip BMC and BA, and WB lean mass during and after adolescence in boys who had participated in the supplementation trial.
Subjects and Methods
Participants and study design (Supplemental Figure 1)
The original supplementation trial (ISRCTN28836000) has been reported elsewhere (14, 16). Briefly, 80 prepubertal boys aged 8.0–11.9 years at baseline and resident in the rural village of Keneba, The Gambia (latitude between 13°N and 14°N), were recruited to participate in a calcium supplementation trial. Children had year-round abundant UVB sunshine skin exposure and good vitamin D status. Boys were randomized double-blind to calcium (n = 40) or placebo (n = 40) groups in equal blocks of four to minimize possible seasonal effects. The supplement was 1000 mg calcium daily, 5 days per week for 1 year; those in the placebo group received a matching tablet. The mean dietary calcium intake was 338 mg/d SD 142 mg/d. Calcium supplementation increased the calcium intake by 714 mg/d to give a mean total dietary calcium intake of 1056 mg/d in the calcium group. Compliance for each child was 100% (16). The boys had annual anthropometric measurements until 2 years after supplementation. Thereafter, measurements were made every 2 years for over 12 years. DXA measurements became available partway through the first year after the supplement phase had ended. We therefore present DXA data obtained at 1, 2, 4, 8, 10, 12, and 14 years after supplement, referred to as time points Y3, Y4, Y6, Y8, Y10, Y12, and Y14, respectively (Supplemental Figure 1). At Y14, 73% of the calcium group and 63% of the placebo group remained in the study, with mean age 23.5 ± 0.2 years.
Ethical approval for the original study, subsequent follow-up, and cross-calibration of DXA instruments at Y12 was obtained from the Medical Research Council/Gambian Government Joint Ethics Committee. Informed written consent was obtained from the parents/guardians of each child for the trial and for each follow-up. Child assent was given for every measurement. After age 18 years, the participants consented themselves. All participants and investigators involved in data collection were blinded to the supplement allocation throughout the study.
Age calculation and anthropometric measurements
Age of participants was calculated using dates of birth from the demographic data surveillance system. This is based on prospective data collected by the Medical Research Council on the population of Keneba and three surrounding villages for over 60 years (18). Height was measured to the nearest 0.1 cm using a stadiometer; The participant was shoeless, and the head was positioned in the horizontal Frankfort plane. Weight was measured to the nearest 0.1 kg using weighing scales, with participants wearing light clothing. Measurements were obtained by trained operators using standardized protocols and regularly calibrated equipment.
Dual-energy x-ray absorptiometry
DXA measurements were carried out using a Lunar DPX+ scanner (GE Healthcare) for Y3–Y12 scans and analyzed using software version 4.7e (GE Healthcare). The final follow-up measurements at Y14 were made using the GE Lunar Prodigy (Encore 2006, version 10.51.006; GE Healthcare). Cross-calibration of the Y14 measurements to DPX+ equivalents was necessary for inclusion in the final data analysis (19). Most measurements for Y12 were carried out using both the DPX and GE Lunar Prodigy. A linear regression equation was applied to calculate the DPX equivalent for BMC, BA, and lean mass because systematic differences between the two DXA instruments were found for all variables.
Outcome measures were WB, lumbar spine (L1–L4), and total hip BMC (grams), BA (centimeters2), and WB lean mass (grams). All scans were scrutinized for quality of scan acquisition and analysis and for consistency of positioning the region of interest between scans. Gambian boys have low body mass index, little fat mass, and short femoral necks. This can cause inaccuracy and imprecision in measurements of fat mass and femoral neck scans by DXA. For this reason femoral neck and fat mass data could not be analyzed with confidence and are not reported.
Manufacturer's protocols were followed for daily quality assurance and weekly quality control. The range of precision of the two DXA systems assessed from duplicate scans on each machine in Gambian adults was: DPX+ (n = 39), 0.9–4.7% for BA and 1.1–4.9% for BMC; Prodigy (n = 35), 0.9–1.3% for BA and 1.0–1.3% for BMC. Total hip measurements were the least precise. The long-term stability of the DPX and Prodigy were monitored using manufacturer's phantoms and indicated satisfactory stability and no significant drift in scanner performance.
Data analysis
Data were analyzed for outliers and were excluded if significant movement artifact or poor scan quality was found. Figure 1 details the number of scans available at each time point. Descriptive characteristics are presented as means ± SEM at each DXA measurement time point for the calcium and placebo groups separately.
Figure 1.
Mean (±SEM) differences in WB: A, BMC; B, BA; C, SA-BMC; and D, height between the calcium (S) and placebo (P) groups at each time point after correction for baseline height to adjust for possible maturity differences at baseline and, where significant, baseline age. Note the attenuation of BMC differences after appropriate adjustment for BA.
Multiple regression analysis was used (DataDesk 6.3.1; Data Description Inc) to compare the supplement and placebo groups at each time point. Natural logarithms of size variables were taken to normalize data and describe proportional (percentage) differences between calcium and placebo groups. With the dependent variable transformed to natural logarithms, multiplying the regression coefficient by 100 expresses results as sympercent differences, which closely correspond to percentage effects ([difference/mean] * 100) (20). Covariates were baseline age (when significant) and baseline height. Baseline height was included to take into account differences in size, maturity, and age before supplementation. Body size and maturity are closely correlated with BMC and BA, so this was an appropriate way to adjust for baseline bone measurements, which were missing due to DXA being introduced only during Y3. BA was also included in the model to give SA-BMC (21).
The individual growth curves for WB, lumbar spine, and total hip BMC and BA, and WB lean mass were fitted using the super imposition by translation and rotation (SITAR) method, with a dedicated software library written in R (version 3.0.2; http://www.R-project.org) (22). The SITAR method summarizes individual growth curves in terms of a mean curve, fitted as a natural cubic spline, and three sets of subject-specific random effects that transform each subject's curve to closely match the mean curve. The mean age at peak velocity (APV) is obtained by differentiating the mean curve and identifying the age when the velocity peaks. The random effects are termed size, tempo, and velocity, and they have the following properties: size represents an up/down shift of each curve (ie, more or less mineral, larger or smaller bones); tempo is a left/right shift of the curve (later or earlier skeletal growth spurt, corresponding to APV); and velocity is a shrinking/stretching of the age scale that alters the slope (faster or slower bone mineral accrual or bone growth). In summary, the SITAR method assumes that all individual growth curves are essentially the same shape and that changes in size, tempo, and velocity will bring them all into registration. Separate regression models are fitted for each of the three parameters, including fixed effects to go with the random effects. Mean differences between the calcium and placebo groups are estimated by including group as a fixed effect in the three models, to measure the effect of the calcium supplement on respectively the final size or mineral level reached, the APV, and the rate of accrual. The models were initially fitted, including group differences for all three of size, tempo, and velocity, and subsequently fitting only tempo. A natural spline curve with 6 degrees of freedom best fitted the data. Goodness of fit of the SITAR models is judged by the percentage of variance explained, compared to that for the simpler fixed effects model, which is a spline curve fitted to the data treated cross-sectionally.
To describe the pattern of skeletal growth in the population, the supplement and placebo groups were combined and SITAR models fitted. Qualitative comparisons of the APVs were used to describe the pattern of growth with respect to mineralization, bone size, and height growth.
The multiple regression analyses were run constraining the dataset to only those participants with a Y12 measurement to test whether attrition caused bias in the findings. The results were not materially different from those obtained with the full models, and therefore data are not presented. SITAR analysis is obtained by averaging the individual growth curves after they have been aligned. If a few curves are truncated at the upper end, this makes no difference to the results, so long as most curves are complete as they are in this study, and therefore attrition has no impact.
Results
Anthropometric measures at each follow-up from Y3 are presented in Table 1 and DXA measures in Table 2.
Table 1.
Descriptive Characteristics of Gambian Boys at Each DXA Occasion
| Time Point | Group, n | Age, y | Height, cm | Weight, kg | Body Mass Index, kg/m2 |
|---|---|---|---|---|---|
| Y3 | Calcium (40) | 12.5 (0.1) | 142 (0.9) | 30.0 (0.7) | 14.9 (0.2) |
| Placebo (40) | 12.6 (0.1) | 142 (1.3) | 29.7 (0.8) | 14.7 (0.2) | |
| Y4 | Calcium (40) | 13.5 (0.1) | 146 (1.0) | 33.0 (0.8) | 15.3 (0.2) |
| Placebo (40) | 13.6 (0.1) | 146 (1.4) | 32.3 (0.9) | 15.0 (0.1) | |
| Y6 | Calcium (40) | 15.5 (0.1) | 158 (1.1) | 41.3 (1.0) | 16.4 (0.2) |
| Placebo (40) | 15.6 (0.1) | 156 (1.6) | 39.9 (1.3) | 16.2 (0.1) | |
| Y8 | Calcium (39) | 17.5 (0.1) | 167 (0.8) | 49.6 (1.0) | 17.8 (0.2) |
| Placebo (39) | 17.5 (0.1) | 165 (1.4) | 47.2 (1.4) | 17.5 (0.1) | |
| Y10 | Calcium (37) | 19.4 (0.1) | 171 (0.7) | 54.7 (0.9) | 18.7 (0.3) |
| Placebo (39) | 19.5 (0.1) | 172 (1.1) | 54.0 (1.1) | 19.5 (0.1) | |
| Y12 | Calcium (34) | 21.7 (0.2) | 173 (0.8) | 59.1 (1.2) | 19.6 (0.2) |
| Placebo (30) | 21.6 (0.2) | 175 (1.2) | 57.8 (1.2) | 18.9 (0.3) | |
| Y14 | Calcium (29) | 23.5 (0.2) | 174 (0.8) | 58.9 (1.4) | 19.5 (0.4) |
| Placebo (25) | 23.6 (0.2) | 176 (1.4) | 59.8 (1.5) | 19.2 (0.2) |
Data are presented as mean (SEM). Number of participants is indicated for calcium and placebo groups at each time point.
Table 2.
DXA Measurements at Each Time Point in the Calcium and Placebo Groups
| Time Point | Group | Whole Body |
Lumbar Spine (L1–L4) |
Total Hip |
||||
|---|---|---|---|---|---|---|---|---|
| BMC, g | BA, cm2 | Lean Mass, kg | BMC, g | BA, cm2 | BMC, g | BA, cm2 | ||
| Y3 | Calcium | 1230 (41.1) | 1390 (29.3) | 23 (5.7) | 19 (0.7) | 31 (0.8) | 18 (0.8) | 20 (0.5) |
| Placebo | 1150 (75.0) | 1350 (62.8) | 24 (1.3) | 18 (1.4) | 32 (1.6) | 17 (1.2) | 20 (0.7) | |
| Y4 | Calcium | 1440 (36.4) | 1580 (29.5) | 28 (0.7) | 23 (0.9) | 35 (0.8) | 21 (0.6) | 23 (0.5) |
| Placebo | 1370 (47.3) | 1540 (37.5) | 27 (0.8) | 23 (1.0) | 35 (0.9) | 21 (0.8) | 23 (0.5) | |
| Y6 | Calcium | 1710 (50.3) | 1780 (36.3) | 36 (1.0) | 31 (1.4) | 42 (1.0) | 27 (0.8) | 27 (0.6) |
| Placebo | 1630 (64.6) | 1720 (46.5) | 35 (1.2) | 30 (1.6) | 41 (1.2) | 26 (1.1) | 26 (0.6) | |
| Y8 | Calcium | 2180 (51.0) | 2100 (31.3) | 46 (0.9) | 44 (1.5) | 50 (0.9) | 33 (0.7) | 31 (0.4) |
| Placebo | 2020 (74.0) | 2030 (45.7) | 44 (1.2) | 42 (2.0) | 48 (1.2) | 32 (1.1) | 31 (0.5) | |
| Y10 | Calcium | 2430 (53.1) | 2250 (24.6) | 48 (0.7) | 53 (1.4) | 53 (0.7) | 37 (0.9) | 32 (0.3) |
| Placebo | 2350 (72.0) | 2230 (36.7) | 47 (1.0) | 51 (2.0) | 53 (1.1) | 36 (1.1) | 33 (0.4) | |
| Y12 | Calcium | 2780 (66.6) | 2370 (30.6) | 52 (1.0) | 62 (1.7) | 57 (0.8) | 41 (1.1) | 34 (0.3) |
| Placebo | 2710 (80.0) | 2370 (39.2) | 51 (1.1) | 60 (2.0) | 57 (1.1) | 40 (1.4) | 34 (0.6) | |
| Y14 | Calcium | 2880 (72.1) | 2390 (30.5) | 52 (1.1) | 65 (1.6) | 57 (0.8) | 42 (1.2) | 34 (0.3) |
| Placebo | 2850 (85.6) | 2430 (47.4) | 52 (1.3) | 64 (2.3) | 59 (1.1) | 42 (1.5) | 34 (0.5) | |
Data are presented as mean (SEM).
Percentage difference between supplement and placebo groups
Percentage group differences (mean ± SEM) in WB BMC, BA, and SA-BMC and height by time point are shown in Figure 1. The calcium group had higher BMC and BA in midadolescence (Y8, WB BMC 8.3 (2.8)%, P = .004; WB BA 3.8 (1.5)%, P = .01) but not later (Y14, WB BMC 2.1 (3.6)%, P = .6; WB BA, −0.1 (2.0)%, P = .9). SA-BMC was not different between groups at any time point, indicating an effect on skeletal size rather than mineral content per se. As previously reported (17), height was greater in the supplement group at Y6 and lower at Y12 and Y14. Lumbar spine and total hip BMC and lumbar spine BA showed similar patterns, but none were significant. These differences between supplement and placebo groups are consistent and of similar magnitude to those measured by single-photon absorptiometry (SPA) at Y3 where radial BMC (5%) and SA-BMC (5%) were higher (P < .05) but BW was not different (−0.3%) (14). At Y4, BMC differences were sustained (4%; P = .02) and BW was greater in the calcium group (1.2%), meaning the differences in SA-BMC were less (2.5%; P = .06). Together, the DXA and SPA cross-sectional analyses indicate earlier bone growth in the calcium supplement group compared to the placebo group.
SITAR analysis—longitudinal bone growth and mineral accretion
SITAR models were fitted for WB, lumbar spine, and total hip BMC and BA, and WB lean mass (Figure 2, Table 3, and Supplemental Figure 2). All seven models fitted well, explaining more than 90% of the variance of the fixed effects models. There were no significant differences between groups in size or velocity for any of the sites, indicating no overall effect on the amount of bone mineral accrued, appositional growth, or the rate of growth. However, most of the tempo effects approached significance, and for this reason the final models adjusted only for between-group tempo differences. APV was consistently earlier for the supplement group than the placebo group (Table 4), and significantly so for WB BMC and for BA at all three sites.
Figure 2.
SITAR plots of WB BMC, BA, and lean mass in the calcium group (solid lines) and the placebo group (dashed lines). The figure illustrates the distance and velocity curves for each parameter.
Table 3.
Whole Body, Lumbar Spine, Total Hip: Mean Differences in SITAR Parameters Between the Calcium and Placebo Groups
| Site | Measurement | SITAR Parameter | Variance Explained, % | Difference (SE) | P |
|---|---|---|---|---|---|
| Whole body | BMC, g | Size, % | 98 | 1.1 (3.3) | .7 |
| Tempo, mo | −6.3 (3.4) | .06 | |||
| Velocity, % | 1.8 (3.9) | .7 | |||
| BA, cm2 | Size, % | 97 | −0.1 (1.9) | .9 | |
| Tempo, mo | −6.2 (3.1) | .05 | |||
| Velocity, % | −4.0 (4.2) | .3 | |||
| Lean mass, kg | Size, % | 95 | 1.5 (2.5) | .6 | |
| Tempo, mo | −4.9 (3.3) | .1 | |||
| Velocity, % | 3.1 (3.8) | .4 | |||
| Lumbar spine | BMC, g | Size, % | 96 | −0.7 (3.9) | .9 |
| Tempo, mo | −5.6 (3.6) | .1 | |||
| Velocity, % | −1.1 (4.7) | .8 | |||
| BA, cm2 | Size, % | 95 | −2.1 (2.2) | .4 | |
| Tempo, mo | −7.7 (3.2) | .02 | |||
| Velocity, % | 3.9 (5.0) | .4 | |||
| Total hip | BMC, g | Size, % | 94 | 0.0 (1.9) | .9 |
| Tempo, mo | −6.1 (3.7) | .1 | |||
| Velocity, % | 2.6 (6.1) | .7 | |||
| BA, cm2 | Size, % | 94 | −0.8 (1.6) | .6 | |
| Tempo, mo | −8.1 (3.5) | .02 | |||
| Velocity, % | 7.4 (6.6) | .3 |
BMC, BA, and lean mass mean differences in SITAR parameters for WB, lumbar spine and total hip, between the calcium and placebo groups when size, tempo and velocity are allowed to differ.
Table 4.
BMC, BA, and Lean Mass Mean Differences in SITAR Parameters for WB, Lumbar Spine, and Total Hip Between the Calcium and Placebo Groups Constraining Size and Velocity to Be Equal Where Only the Tempo Effect Is Allowed to Differ
| Site | Measurement | Difference (SE), mo | P |
|---|---|---|---|
| Whole body | BMC | −7.0 (3.2) | .03 |
| BA | −6.2 (3.1) | .05 | |
| Lean mass | −3.9 (2.6) | .1 | |
| Lumbar spine | BMC | −5.1 (3.1) | .1 |
| BA | −6.7 (3.0) | .03 | |
| Total hip | BMC | −4.2 (3.1) | .2 |
| BA | −6.0 (3.0) | .05 |
Overall, combining the supplement and placebo groups to describe the pattern of growth in the whole group, the mean APVs for WB BMC, WB BA, and lean mass were, respectively, 16.8, 16.4, and 16.2 years; that for height was 16.1 years (17). The differences in APV relative to BMC were −4.4 (WB BA), −7.0 (lean mass), and −8.2 (height) months, with SE values of 1.5 months.
Discussion
By following the study participants longitudinally through adolescent growth, we have been able to test the effects of childhood supplementation on later growth and bone mineral accrual. SITAR analysis has allowed us to determine the patterns of growth in boys from The Gambia. This study shows that 12 months of calcium carbonate supplementation advances the timing of the adolescent skeletal growth spurt and shortens the prepubertal growth period in Gambian boys accustomed to a low calcium intake. Boys in the calcium group had greater WB BMC and greater WB, lumbar spine, and total hip BA than the placebo group during midadolescence, but not later. There were no differences in the velocity of bone growth or mineral accrual, indicating that the passage through the growth spurt occurred at the same rate in both groups. There was no effect of calcium supplementation on BMC or BA at the end of the follow-up period; ie, the total amount of mineral accrued and skeletal size by the end of the follow-up period were not affected. These data challenge the notion that “more” dietary calcium necessarily translates to “better” bone health. The differences in timing of APV for BA and BMC provide independent evidence that confirms our previous report that the boys in the calcium supplement group had earlier APV for height, were taller in midadolescence, but stopped growing earlier (17).
Most trials in children and adolescents report no sustained effect of calcium supplementation on BMC or BA (6–10). It has been suggested that the lack of sustained effects of calcium salt supplementation indicate a phenomenon known as the bone remodeling transient (7, 16, 23). However, we report temporary increases in BMC and BA for up to 8 years after stopping calcium supplementation, long after bone remodeling rates returned to presupplement levels (14). A 7-year intervention showed faster rates of bone accrual after 4-year supplementation with calcium-citrate-malate but no effect after 7-year supplementation, except in the proximal radius and metacarpals (11). Further analysis indicated these sustained effects were only in taller subjects. It was suggested that this “lasting” effect was due to the calcium requirements of taller children being greater than shorter children (11). In other populations with habitually low dietary calcium intakes, similar to the one in the current study, the supplement effects were not sustained (9, 10).
It has been suggested that a protein-based calcium supplement may result in sustained differences in growth due to shifts in the IGF-1 GH axis (7, 24–26), although this is not consistently reported (9). In children who received a milk-based calcium supplement, persisting positive effects on BMD were reported 3.5 years after supplementation. However, the study lacked long-term follow-up data (13, 27, 28). The results of a trial in Chinese adolescents with low daily calcium intakes (∼400 mg/d) showed no sustained effect of milk supplementation 3 years after supplementation stopped (9). Differences in IGF-1 have been reported in response to calcium salts, suggesting that lasting effects may occur, although most evidence does not support this (29). In the current study, the differences in BMC and BA up to 8 years after supplement reflect differences in the timing of the skeletal growth spurt for height and BA in the calcium group, possibly through effects on IGF-1 production by osteoblasts (29).
This study also clarifies the pattern of bone growth in Gambian boys. In the whole group, APV for height occurred first (16.1 y), then lean mass (16.2 y), WB BA (16.4 y), and WB BMC (16.8 y). This pattern of development is consistent with that in a Canadian cohort and indicates a sequence of height growth accompanied by growth in lean mass, followed by appositional bone growth, and finally mineral consolidation (3, 30, 31). Later APV and lower WB BMC peak velocity in the Gambian boys compared to the Canadian boys (∼250 g/y compared to 407 g/y) indicates a longer growth period in Gambian boys (3). Comparison to published Canadian values suggests differences in skeletal proportions and dimensions in Gambian boys at the end of follow-up. This comparison is limited because the two studies used scanners from different manufacturers, modeled growth in different ways, and had follow-up measurements at different intervals, so more formal comparisons are required to confirm these population differences. The findings, however, are similar to our comparative study of Gambian and UK women, which showed that Gambians had narrower bones and greater volumetric BMD, measured by peripheral quantitative computed tomography, than UK women (32).
There were limitations to the study. There were no baseline DXA measurements to give presupplement data, but analysis of the Y3 DXA data gave similar group differences to the original SPA results (14), and the DXA-measured BA differences paralleled those of height (17). There was some attrition in the study, but over 65% of boys attended the final measurement time point; SITAR is robust to attrition, and there were no material differences in multiple regression analyses when data were restricted to those boys who completed the study. We might expect group differences in limb and trunk proportions due to the earlier growth spurt in the supplemented group because of the different growth rates of the axial and appendicular skeleton before and during puberty (33). This would provide independent evidence of a supplement effect on APV, but we cannot confirm this because sitting height was not measured. Finally, the intervention was for calcium supplementation alone, and correction of other deficits in the Gambian diet together with calcium or protein-based supplements might have more permanent effects on bone in this population, although data from the Chinese trial suggest this may not be the case (9).
In the current study, boys had been supplemented in childhood with 1000 mg elemental calcium for 5 days per week for 1 year, which on average increased calcium intake by 714 mg/d to a mean of 1056 mg/d in the supplemented group (16). This level of daily calcium intake is similar to current UK and US recommendations for 8- to 12-year-old children (UK Reference Nutrient Intake, 550–1000 mg/d [34]; US Recommended Daily Allowance, 1300 mg/d [35]). Our findings show that supplementation of prepubertal boys with calcium to internationally recommended levels was not associated with a sustained increase in BMC or BA by the time they were approaching skeletal maturity as young adult men, and together with other studies from developing populations (9, 10), they suggest a need for caution when applying international recommendations to different populations.
Acknowledgments
We thank the participants in the study, their families, and the staff of the Medical Research Council (MRC) Keneba and MRC Human Nutrition Research who contributed over many years to this intensive study. We particularly thank the late Bakary Dibba whose PhD thesis was based on the initial calcium supplementation study. We also thank Landing Jarjou, Mariama Jammeh, Fatou Manneh, Lamin Jammeh, and Buba Sisa (research team at MRC Keneba) for coordinating the study and performing measurements. At MRC Human Nutrition Research, we thank Gail Goldberg, Sheila Levitt, Jennifer Thompson, and Duangporn Harpanich for data entry, collation, and checking data.
This research is jointly funded by the Medical Research Council (MRC) and the Department for International Development (DFID) under the MRC/DFID Concordat agreement program no. U105960371, U123261351, and MR/J004839/1.
Disclosure Summary: The authors have nothing to disclose.
Footnotes
- APV
- age at peak velocity
- BA
- bone area
- BMC
- bone mineral content
- BMD
- bone mineral density
- BW
- bone width
- DXA
- dual-energy x-ray absorptiometry
- SA-BMC
- size-adjusted BMC
- SITAR
- super imposition by translation and rotation
- SPA
- single-photon absorptiometry
- WB
- whole body.
References
- 1. Heaney RP, Abrams S, Dawson-Hughes B, et al. Peak bone mass. Osteoporos Int. 2000;11(12):985–1009. [DOI] [PubMed] [Google Scholar]
- 2. Matkovic V, Jelic T, Wardlaw GM, et al. Timing of peak bone mass in Caucasian females and its implication for the prevention of osteoporosis. Inference from a cross-sectional model. J Clin Invest. 1994;93(2):799–808. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Bailey DA, Martin AD, McKay HA, Whiting S, Mirwald R. Calcium accretion in girls and boys during puberty: a longitudinal analysis. J Bone Miner Res. 2000;15(11):2245–2250. [DOI] [PubMed] [Google Scholar]
- 4. Matkovic V, Fontana D, Tominac C, Goel P, Chesnut CH., 3rd Factors that influence peak bone mass formation: a study of calcium balance and the inheritance of bone mass in adolescent females. Am J Clin Nutr. 1990;52(5):878–888. [DOI] [PubMed] [Google Scholar]
- 5. Prentice A. Diet, nutrition and the prevention of osteoporosis. Public Health Nutr. 2004;7(1A):227–243. [DOI] [PubMed] [Google Scholar]
- 6. Johnston CC, Jr, Miller JZ, Slemenda CW, et al. Calcium supplementation and increases in bone mineral density in children. N Engl J Med. 1992;327(2):82–87. [DOI] [PubMed] [Google Scholar]
- 7. Lambert HL, Eastell R, Karnik K, Russell JM, Barker ME. Calcium supplementation and bone mineral accretion in adolescent girls: an 18-mo randomized controlled trial with 2-y follow-up. Am J Clin Nutr. 2008;87(2):455–462. [DOI] [PubMed] [Google Scholar]
- 8. Specker B, Binkley T, Fahrenwald N. Increased periosteal circumference remains present 12 months after an exercise intervention in preschool children. Bone. 2004;35(6):1383–1388. [DOI] [PubMed] [Google Scholar]
- 9. Zhu K, Zhang Q, Foo LH, et al. Growth, bone mass, and vitamin D status of Chinese adolescent girls 3 y after withdrawal of milk supplementation. Am J Clin Nutr. 2006;83(3):714–721. [DOI] [PubMed] [Google Scholar]
- 10. Umaretiya PJ, Thacher TD, Fischer PR, Cha SS, Pettifor JM. Bone mineral density in Nigerian children after discontinuation of calcium supplementation. Bone. 2013;55(1):64–68. [DOI] [PubMed] [Google Scholar]
- 11. Matkovic V, Goel PK, Badenhop-Stevens NE, et al. Calcium supplementation and bone mineral density in females from childhood to young adulthood: a randomized controlled trial. Am J Clin Nutr. 2005;81(1):175–188. [DOI] [PubMed] [Google Scholar]
- 12. Zhu K, Du X, Cowell CT, et al. Effects of school milk intervention on cortical bone accretion and indicators relevant to bone metabolism in Chinese girls aged 10–12 y in Beijing. Am J Clin Nutr. 2005;81(5):1168–1175. [DOI] [PubMed] [Google Scholar]
- 13. Bonjour JP, Chevalley T, Ammann P, Slosman D, Rizzoli R. Gain in bone mineral mass in prepubertal girls 3.5 years after discontinuation of calcium supplementation: a follow-up study. Lancet. 2001;358:1208–1212. [DOI] [PubMed] [Google Scholar]
- 14. Dibba B, Prentice A, Ceesay M, et al. Bone mineral contents and plasma osteocalcin concentrations of Gambian children 12 and 24 mo after the withdrawal of a calcium supplement. Am J Clin Nutr. 2002;76(3):681–686. [DOI] [PubMed] [Google Scholar]
- 15. Winzenberg T, Shaw K, Fryer J, Jones G. Effects of calcium supplementation on bone density in healthy children: meta-analysis of randomised controlled trials. BMJ. 2006;333(7572):775. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Dibba B, Prentice A, Ceesay M, Stirling DM, Cole TJ, Poskitt EM. Effect of calcium supplementation on bone mineral accretion in Gambian children accustomed to a low-calcium diet. Am J Clin Nutr. 2000;71(2):544–549. [DOI] [PubMed] [Google Scholar]
- 17. Prentice A, Dibba B, Sawo Y, Cole TJ. The effect of prepubertal calcium carbonate supplementation on the age of peak height velocity in Gambian adolescents. Am J Clin Nutr. 2012;96(5):1042–1050. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Rayco-Solon P, Moore SE, Fulford AJ, Prentice AM. Fifty-year mortality trends in three rural African villages. Trop Med Int Health. 2004;9(11):1151–1160. [DOI] [PubMed] [Google Scholar]
- 19. Crabtree NJ, Shaw NJ, Boivin CM, Oldroyd B, Truscott JG. Pediatric in vivo cross-calibration between the GE Lunar Prodigy and DPX-L bone densitometers. Osteoporos Int. 2005;16(12):2157–2167. [DOI] [PubMed] [Google Scholar]
- 20. Cole TJ. Sympercents: symmetric percentage differences on the 100 log(e) scale simplify the presentation of log transformed data. Stat Med. 2000;19(22):3109–3125. [DOI] [PubMed] [Google Scholar]
- 21. Prentice A, Parsons TJ, Cole TJ. Uncritical use of bone mineral density in absorptiometry may lead to size-related artifacts in the identification of bone mineral determinants. Am J Clin Nutr. 1994;60:837–842. [DOI] [PubMed] [Google Scholar]
- 22. Cole TJ, Donaldson MD, Ben-Shlomo Y. SITAR–a useful instrument for growth curve analysis. Int J Epidemiol. 2010;39(6):1558–1566. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Heaney RP. The bone-remodeling transient: implications for the interpretation of clinical studies of bone mass change. J Bone Miner Res. 1994;9(10):1515–1523. [DOI] [PubMed] [Google Scholar]
- 24. Cadogan J, Eastell R, Jones N, Barker ME. Milk intake and bone mineral acquisition in adolescent girls: randomised, controlled intervention trial. BMJ. 1997;315(7118):1255–1260. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Eastell R, Lambert H. Diet and healthy bones. Calcif Tissue Int. 2002;70(5):400–404. [DOI] [PubMed] [Google Scholar]
- 26. Prentice A, Ginty F, Stear SJ, Jones SC, Laskey MA, Cole TJ. Calcium supplementation increases stature and bone mineral mass of 16- to 18-year-old boys. J Clin Endocrinol Metab. 2005;90:3153–3161. [DOI] [PubMed] [Google Scholar]
- 27. Remer T, Boye KR, Manz F. Long-term increase in bone mass through high calcium intake before puberty. Lancet. 2002;359(9322):2037–2038; author reply 2038. [DOI] [PubMed] [Google Scholar]
- 28. Chevalley T, Bonjour JP, Ferrari S, Hans D, Rizzoli R. Skeletal site selectivity in the effects of calcium supplementation on areal bone mineral density gain: a randomized, double-blind, placebo-controlled trial in prepubertal boys. J Clin Endocrinol Metab. 2005;90(6):3342–3349. [DOI] [PubMed] [Google Scholar]
- 29. Ginty F, Prentice A, Laidlaw A, et al. Calcium carbonate supplementation is associated with higher plasma IGF-1 in 16–18 year-old boys and girls. In: Burckhardt P, Dawson-Hughes B, Heaney RP, eds. Nutritional Aspects of Osteoporosis. 2nd ed San Diego: Elsevier; 2004:45–57. [Google Scholar]
- 30. Baxter-Jones AD, Faulkner RA, Forwood MR, Mirwald RL, Bailey DA. Bone mineral accrual from 8 to 30 years of age: an estimation of peak bone mass. J Bone Miner Res. 2011;26(8):1729–1739. [DOI] [PubMed] [Google Scholar]
- 31. Rauch F, Bailey DA, Baxter-Jones A, Mirwald R, Faulkner R. The ‘muscle-bone unit’ during the pubertal growth spurt. Bone. 2004;34(5):771–775. [DOI] [PubMed] [Google Scholar]
- 32. Laskey MA, de Bono S, Zhu D, et al. Evidence for enhanced characterization of cortical bone using novel pQCT shape software. J Clin Densitom. 2010;13(3):247–255. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Seeman E. Delayed puberty and skeletal growth in mass, size, and density. In: Bonjour J, Tsang R, eds. Nutrition and Bone Development. Philadelphia: Vevey/Lippincott-Raven;1999:245–259. [Google Scholar]
- 34. Nutrition and bone health: with particular reference to calcium and vitamin D Report of the Subgroup on Bone Health (Working Group on the Nutritional Status of the Population) of the Committee on Medical Aspects of Food and Nutrition Policy. London, The Stationary Office: 1998, 1–128. [PubMed] [Google Scholar]
- 35. Institute of Medicine. Dietary Reference Intakes for Calcium and Vitamin D. In: Ross AC, Taylor CL, Yaktine AL, Del Valle HB, eds. Washington, DC: National Academies Press; 2010. [PubMed] [Google Scholar]