| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Service of Bone Diseases (World Health Organization Collaborating Centre for Osteoporosis Prevention), Department of Rehabilitation and Geriatrics (T.C., J.-P.B., S.F., R.R.) and Service of Nuclear Medicine, Department of Radiology (D.H.), University Hospitals of Geneva, CH-1211 Geneva 14, Switzerland
Address all correspondence and requests for reprints to: Thierry Chevalley, M.D., Service of Bone Diseases, Department of Rehabilitation and Geriatrics, University Hospitals of Geneva, CH-1211 Geneva 14, Switzerland. E-mail: thierry.chevalley{at}hcuge.ch.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: The objective was to assess whether calcium supplementation increases bone mass gain in prepubertal boys in a skeletal site-specific manner.
Methods: In a 12-month double-blind, placebo-controlled trial with 1-yr follow-up, 235 healthy prepubertal boys aged 7.4 ± 0.4 yr (mean ± SD) were randomized to receive two food products providing 850 mg/d calcium (calcium supplement group, n = 116) or an isocaloric placebo (n = 119). Areal bone mineral density (aBMD) was determined by dual-energy x-ray absorptiometry at radius (two sites), hip (two sites), femoral diaphysis (FDia), and L2–L4 vertebrae.
Results: At 12 months, aBMD gain was greater at the FDia and at the mean of the five appendicular skeletal sites in the calcium supplement group in both intention-to-treat analysis [76 ± 32 vs. 64 ± 33 mg/cm2·yr; difference, 12.0 (95% confidence interval, CI, 3.6–20.3), P = 0.006; and 33 ± 16 vs. 28 ± 16 mg/cm2·yr; difference, 5.1 (95% CI, 0.9–9.2); P = 0.018, respectively] and active treatment analysis [81 ± 32 vs. 64 ± 31 mg/cm2·yr; difference, 17.2 (95% CI, 7.9–26.5); n = 174, P < 0.001; and 35 ± 16 vs. 28 ± 14 mg/cm2·yr; difference, 7.5 (95% CI, 2.9–12.2); P = 0.002]. There was no beneficial effect of calcium on lumbar spine. The calcium effect was still detectable by ANOVA repeated measures analysis at the FDia (P = 0.004) and at the mean of the five appendicular skeletal sites (P = 0.002) 1 yr after the end of intervention (active treatment analysis). There was no change in bone size.
Conclusion: In prepubertal boys, calcium-enriched foods increased aBMD at several appendicular skeleton sites, but not at the lumbar spine, and this without any bone size change. This effect was maintained 1 yr after treatment discontinuation.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Healthy prepubertal Caucasian boys were recruited through the Public Health Youth Service of the Geneva region from September 1999 to September 2000. The protocol was approved by the Ethics Committee of the Department of Pediatrics of the University Hospitals of Geneva. Informed consent was obtained from the parents and their children. Families of 4150 boys were invited by mail to fill out a 7-d dietary diary. From this first step, 990 analyzable diaries were received (Fig. 1
). Boys with a spontaneous calcium intake below the 75th percentile were recontacted and invited to participate in the intervention trial because we previously observed a greater effect of calcium supplementation on aBMD gain in prepubertal girls with a spontaneous calcium intake below the median (11). A total of 235 subjects were enrolled after applying the following exclusion criteria: weight to height ratio below the third or above the 97th percentile according to Geneva reference values, presence of physical signs of puberty, chronic disease, gastrointestinal disease with malabsorption, congenital or acquired bone disease, and regular use of medication. They were randomized into two groups receiving either calcium-enriched foods [calcium supplement (Ca-suppl.)] or placebo foods for 12 months. Enrolled subjects were further stratified according to the median of their spontaneous calcium intake, and those with relatively low or high calcium intakes were distributed in equal numbers in both groups. An additional examination took place 1 yr after the end of intervention.
|
Calcium-phosphate salt extracted from milk was used to enrich several food products: chocolate or caramel cakes, biscuits, fruit juices, powdered drinking chocolate, and chocolate bars. On average, the consumption of two calcium-enriched servings per day provided a Ca-suppl. of about 850 mg. Two very similar food products with respect to energy, proteins, and lipids, but without added milk calcium-phosphate salt, were consumed daily by the control (placebo) group. The mean nutrient contents of the milk calcium-phosphate salt foods and placebo were: calcium, 436 and 46 mg; phosphate, 236 and 54 mg; lipids, 6.6 and 6.6 g; carbohydrates, 22.9 and 22.4 g; proteins, 2.3 and 2.2 g; and energy, 160 and 158 kcal, respectively. Over the entire intervention period, each subject received one set of food products every month. Parents were instructed to encourage their sons to consume two food products every day, instead of similar servings usually taken for breakfast or snacks. Compliance was recorded by the parents and verified through regular phone calls and interviews by a trained dietician.
Calcium intake assessment
The spontaneous calcium intakes were assessed by frequency questionnaires at baseline and 6, 12, and 24 months. The mean of the three questionnaires at 0, 6, and 12 months was taken into account for analyzing the relationship between nutritional intakes and anthropometric or bone variables.
Physical activity assessment
Usual physical activity and weight-bearing physical activity were recorded and measured in hours per week (15) or in kilocalories per day after the coding of physical activities in energy costs (16). It included physical education classes, organized sports, recreational activity, and usual walking and cycling.
Measurement of anthropometric and bone variables
At baseline and 12 and 24 months, we measured participants’ weight and standing height using a stapediometer, and body mass index (BMI) was calculated. aBMD was determined by dual-energy x-ray absorptiometry using a Hologic QDR 4500 instrument (Hologic, Waltham, MA). Six skeletal sites were examined: radius [distal metaphysis (RMet) and diaphysis], hip (femoral neck and trochanter), FDia, and L2–L4 vertebrae [lumbar spine (LS)] in anteroposterior view as previously reported (11). The coefficient of variation of repeated measurements at these sites as determined in young healthy adults varied from 1.0–1.6% for BMD and from 0.3–3.0% for bone mineral content (BMC) and projected bone scanned area.
Expression of the results and statistical analysis
Differences in the anthropometric and osteodensitometric gains between the Ca-suppl. and the placebo groups were computed by intention-to-treat analysis (ITT) that included all subjects who entered the study and had measurements repeated at 1 yr. Differences were further calculated by active treatment analysis (AT) that included the subjects who consumed the tested foods until the end of the intervention year. Taking into account the precision of the various anthropometric and osteodensitometric measurements and an expected minimal difference in gains of 20%, a rough estimate of the sample size was 80 subjects in either the Ca-suppl. or placebo groups reaching the 52-wk point. This is for a 5% 
-error (i.e. only 5% chance of finding significance that is not actually present) and a 20% ß-error (i.e. 80% chance of detecting a significant difference if it exists). With an estimate from a similar trial in prepubertal girls that 75% of the subjects will consume the tested foods until the completion of the 1-yr intervention, we decided to enroll at least 100 boys in each group. For both anthropometric and bone variables, gains are expressed in either absolute values (grams per square centimeter per year) or percentage change from baseline as means ± SD. aBMD gains were determined separately for each skeletal site. The average aBMD changes at the five appendicular skeletal sites studied were also calculated. The differences between the Ca-suppl. and placebo groups were evaluated using two-tailed Student’s t test for unpaired values and also expressed as a 95% confidence interval (CI) of the differences. For the differences in gains of osteodensitometric variables, an analysis of covariance was used to control for the influence of gain in both body weight and standing height. A Spearman’s correlation was used for the relationship between FDia aBMD gain and physical activity level. A two-way ANOVA was used to analyze the interaction between calcium supplementation and physical activity on FDia aBMD gain. ANOVA repeated measures analysis was used to report the baseline, yr 1, and 1-yr follow-up aBMD changes in one model. The significance level for two-sided P values was 0.050 for all tests. The data were analyzed using STATA (Stata Corp., College Station, TX) software, version 7.0.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Among the 235 randomized prepubertal boys with age from 6.5–8.5 yr (mean, 7.44 ± 0.39 yr), 232 had their aBMD measured at baseline and after 1 yr (ITT) (Fig. 1
). Changes in all variables are expressed in units per year, although the time interval between the first (preintervention) and second (postintervention) visits was 1.10 ± 0.06 and 1.11 ± 0.06 yr in the placebo and Ca-suppl. groups, respectively. At baseline, there was no difference in the anthropometric characteristics, spontaneous calcium intake, physical activity, or bone densitometry variables among the subjects assigned to consume foods enriched in calcium or not (placebo) (Table 1). From the initial cohort, 174 boys (86 and 88 in the Ca-suppl. and placebo groups, respectively) remained compliant (AT) to the supplements during the whole intervention period (Fig. 1). The other 58 (25%) noncompliant subjects (28 and 30 in the Ca-suppl. and placebo groups, respectively) withdrew from treatment but were remeasured at 1 yr. Among them, median withdrawal time (14.5 wk) from treatment was similar in the Ca-suppl. and placebo groups. In the AT cohort, 97.1 ± 7.2% (n = 86) and 98.9 ± 3.2% (n = 88) of the study foods were consumed throughout the intervention period in the Ca-suppl. and placebo groups, respectively, as estimated by regular interviews of the subjects’ mothers.
|
Anthropometric and bone mineral mass changes
None of the examined subjects displayed any sign of puberty at the end of intervention. The gains in anthropometric and osteodensitometric variables after 12 months of intervention are shown in Table 2. The progression in both standing height and body weight were quite comparable in both groups (Table 2). At LS, no effect of Ca-suppl. on aBMD gain was observed by ITT or AT, whether expressed in absolute values (Table 2) or in percentage change from baseline (Fig. 2). At FDia, a significantly greater aBMD gain, expressed either in milligrams per square centimeter per year or in percentage change from baseline, was observed in the Ca-suppl. group compared with the placebo group by both ITT (P = 0.006 and P = 0.003, respectively) and AT (P < 0.001 and P < 0.001) analyses. These levels of significance were maintained after adjustment for the gain in both body weight and standing height (Table 2 and Fig. 2). At the other four appendicular sites, namely Rmet, radius diaphysis, and femoral neck and trochanter, the aBMD gain was greater, at least in the AT cohort, in the Ca-suppl. group than in the placebo group. However, the differences were not statistically significant (Table 2 and Fig. 2).
|
|
An almost significant relationship was found between aBMD gain at FDia and weight-bearing physical activity (ITT, 
= 0.12, P = 0.068) but not with total physical activity ( = 0.07, P = 0.27), as expressed in kilocalories per day. The calcium response at the FDia level was slightly greater in boys above [Ca-suppl. (n = 63) vs. placebo (n = 53), 80.0 ± 32.6 vs. 67.2 ± 35.9 mg/cm2·yr; difference, 12.8 (0.2–25.3); P = 0.046] than below [Ca-suppl. (n = 65) vs. placebo (n = 51), 71.4 ± 30.0 vs. 61.9 ± 30.6 mg/cm2·yr; difference, 9.4 (–1.9–20.8); P = 0.10] the median of the weight-bearing physical activity estimate. However, a positive interaction between calcium supplementation and physical activity did not reach statistical significance by two-way ANOVA (P = 0.698).
Influence of calcium exposure
At the femoral sites (neck, trochanter, diaphysis), the difference in aBMD gain between Ca-suppl. and placebo groups was approximatively 10% larger by AT compared with ITT (Table 2 and Fig. 2). Such a greater mean calcium effect by AT compared with ITT was not observed at the level of the two radial sites. The magnitude of the calcium effect was further estimated as the difference in gains (expressed in percentage change from baseline) observed between the Ca-suppl. and placebo groups. At the five appendicular skeletal sites, this calcium effect ranged from 8–35% with a mean value of 25% (Table 2). The difference in the mean increment in appendicular aBMD (five appendicular skeletal sites) was significantly greater among the Ca-suppl. compared with the placebo group [ITT: 5.6 ± 2.8 vs. 4.7 ± 2.6%; difference, 0.9% (95% CI, 0.2–1.6); P = 0.014; and AT: 6.0 ± 2.8 vs. 4.7 ± 2.4%; difference, 1.3% (95% CI, 0.5–2.1); P = 0.002] (Fig. 3).
|
None of the differences in bone area and BMC gains between the Ca-suppl. and placebo groups was statistically significant (Table 3). Nevertheless, for FDia and for the mean of the five appendicular sites, there was a much larger gain in BMC than in bone area in favor of the Ca-suppl. group. These differential changes are consistent with the significantly greater aBMD gains in the Ca-suppl. than in the placebo in FDia and in the mean of the five appendicular sites as presented in Table 2 and in Figs. 2 and 3.
|
The effect of calcium-enriched foods on aBMD gain was further analyzed according to the median of the spontaneous calcium intake as previously performed in prepubertal girls (11). The mean calcium intake was 556 ± 11 (n = 116) and 945 ± 18 mg/d (n = 116) when spontaneous calcium intake was below or above the median, respectively. There was no significant difference between groups: 553 ± 128 (placebo, n = 59) and 560 ± 107 mg/d (Ca-suppl., n = 57) vs. 937 ± 186 (placebo, n = 59) and 954 ± 206 mg/d (Ca-suppl., n = 57), respectively. At baseline, neither statural height nor body weight differed significantly between these groups (height, 125.9 ± 5.9 and 124.4 ± 5.4 cm vs. 126.5 ± 5.7 and 125.7 ± 6.4 cm; weight, 25.3 ± 5.7 and 24.5 ± 3.1 kg vs. 25.6 ± 4.6 and 25.5 ± 4.9 kg). Calcium supplementation did not affect lumbar spine aBMD gain in any of the subgroups (data not shown). In the five appendicular skeletal sites, the difference in mean aBMD gain in response to calcium supplementation was not greater in boys with spontaneous calcium intake below the median compared with those above the median (data not shown). In both subgroups, there was no statistically significant difference in baseline aBMD at any skeletal site between the Ca-suppl.- and placebo-assigned subjects (data not shown).
Follow-up 1 yr after end of intervention
A total of 172 boys (Ca-suppl., n = 85; placebo, n = 87) compliant with the trial foods during the first year were reexamined 1 yr after the end of the intervention (Fig. 1). Spontaneous calcium intakes evaluated at the end of the follow-up period were similar to baseline, 787 ± 354 and 749 ± 315 mg/d in the Ca-suppl. and placebo groups, respectively. The maintenance of a greater gain in appendicular aBMD, before and after adjustment for the gain in body weight and standing height, was found in the cohort of boys having all consumed the calcium-enriched foods until the end of the intervention year. Thus, 1 yr after the end of intervention, the following gains in aBMD were recorded: RMet, 25.3 ± 21.4 vs. 16.5 ± 16.5 mg/cm2·yr; difference, 8.8 (95% CI, 3.0–14.6), P = 0.003 and P < 0.001; FDia, 135.1 ± 42.3 vs. 123.7 ± 37.6 mg/cm2·yr; difference, 11.4 (95% CI, –0.7–23.4); P = 0.064 and P < 0.001; and mean of the five appendicular skeletal sites, 57.6 ± 18.4 vs. 51.4 ± 18.1 mg/cm2·yr; difference, 6.1 (95% CI, 0.6–11.7), P = 0.030 and P < 0.001, in the Ca-suppl. and placebo groups, respectively (Table 4). These aBMD gains, expressed in percentage change from baseline at the end of a 12-month calcium intervention and 1 yr after the end of intervention, were statistically significant as reported in one model by ANOVA repeated measures analysis at RMet (P = 0.010), FDia (P = 0.004), and at the mean of the five appendicular skeletal sites (P = 0.002) (Fig. 4). The corresponding P values in the ITT cohort were P = 0.054, 0.080, and 0.073, respectively.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Site selectivity
A positive difference in favor of the calcium-supplemented boys was only observed at appendicular skeletal sites but not at the lumbar spine. The unresponsiveness of the axial site to calcium supplementation was recorded with both ITT and AT. This differential response corroborates previous data obtained in a small cohort of 22 prepubertal twin pairs (3), as well as results recorded in a larger group of 7- to 9-yr-old prepubertal girls enrolled in a trial quite similar to that described in the present report (11). As indicated in a review on the role of calcium on bone health during childhood, most intervention studies that included an axial skeletal site measurement did not find a greater yearly increase in lumbar bone mineral mass accrual in the calcium-supplemented compared with the placebo group (17). This could be due to a greater impact of calcium supplementation on vertebral bone area than on BMC gains at that site. The asynchrony between the greater progression in bone size compared with mineral accumulation is maximal at peak height velocity (18, 19, 20). Thus, in boys, this dissociation is observed around 14 yr of age, i.e. at least 4–5 yr later than the mean age (8.5 yr) of our cohort at the end of the calcium intervention phase. Furthermore, as presented in Table 3, at LS, the increase in bone area compared with BMC gain was not greater in the Ca.-suppl. than in the placebo group. Therefore, it is unlikely that the absence of calcium effect on LS aBMD gain could be ascribed to such a differential tempo in vertebral growth and mineral accumulation within the vertebral body. Thus, the appendicular skeleton appears to be more sensitive than the axial skeleton to the effect of calcium supplementation. Among appendicular regions, the clearest effect was observed in the FDia, a weight-bearing site essentially consisting of compact bone. Nevertheless, aBMD gain could also be positively influenced by calcium supplementation in regions including both cortical and trabecular bone tissues, such as trochanter and radial metaphysis.
At the three femoral sites, the difference in aBMD gain between Ca-suppl. and placebo groups was approximatively 10% higher in the AT compared with the ITT cohort. Such a trend for a dose-response effect was not observed at the level of the two radial sites. It suggests that the response of the forearm might be threshold-dependent rather than dose-dependent.
Spontaneous calcium intake and response to supplementation
Analysis of the results according to the spontaneous calcium intake did not reveal a greater response among boys who had relatively low calcium intake. This observation differs from our previous study in prepubertal girls who underwent the same type of intervention (11). In this previous trial, the positive effect of calcium was significantly larger in girls with a spontaneous calcium intake below (mean = 675 mg/d) than above (mean = 1185 mg/d) the median, which was estimated at 880 mg/d in the AT cohort (11). This analysis suggested that in 7- to 9-yr-old prepubertal girls, 1200 mg/d was very close to the maximum calcium intake for optimal bone mineral mass accrual. The current study in prepubertal boys was not designed to investigate the influence of the spontaneous calcium intake on the effect of supplementation. Indeed, on purpose, we did not enroll boys with a calcium intake above the 75th percentile at baseline. This selection could explain why the difference in the mean response between the Ca-suppl. and the placebo group was not greater in those boys with a spontaneous calcium intake below (mean = 561 mg/d) than above (mean = 968 mg/d) the median, which was estimated at 730 mg/d in the AT cohort. Taking into account this difference in the spontaneous calcium intake between the former study (11) and the present study, it is not possible to establish that the nearly maximal intake for optimal bone mass mineral accrual be set at a higher level, i.e. above 1200 mg/d, in 7- to 9-yr-old prepubertal boys compared with girls in the same age range.
Physical activity and response to calcium supplementation
The difference observed according to the degree of exposure to calcium supplementation between femoral and radial sites might be related to some differential interaction with physical activity at the level of weight-bearing compared with non-weight-bearing skeletal structures (21, 22). The calcium response at the FDia aBMD level was slightly greater in boys above (
gain Ca-suppl. minus placebo, +13 mg/cm2·yr, P = 0.046) than below ( gain Ca-suppl. minus placebo, +9 mg/cm2·yr, P = 0.11), the median of the weight-bearing physical activity estimate. However, a positive interaction between calcium supplementation and physical activity did not reach statistical significance. It is possible that the statistically significant evidence of such a synergism between two environmental factors that exert a positive impact on bone accrual during growth would still require a larger sample size than that of this study.
Bone size and standing height
In this study, the positive effect of milk-extracted calcium salt on aBMD in the appendicular skeleton was not associated with significant increase in bone size or standing height. Therefore, the presence of an association (11, 12, 23) or not (3, 4, 6, 9), as in the present study, of an effect on bone size and standing height cannot be merely ascribed to the nature of the supplemented calcium salt, as previously suggested (11). In girls, the observed calcium effect on bone size and standing height appeared to be transient and possibly related to an acceleration of pubertal maturation (13, 14).
Follow-up after discontinuation of calcium supplementation
The extent to which any benefit on bone mineral mass accrual may be maintained after discontinuation of the intervention remains uncertain. Thus, three follow-up studies have suggested that the effects of calcium on aBMD gain would not be maintained beyond the period of supplementation (5, 24, 25). In contrast, we observed that the benefit was maintained 1 yr (11), 3.5 yr, (13) and even 7.5 yr (14) after the end of intervention in prepubertal girls. As for the effect on bone size, we initially suggested that the contrasting results might be related to the chemical nature of the Ca-suppl.: citrate malate or carbonate salt in the negative studies (5, 24, 25), and phosphate salt as extracted from milk in the positive follow-up study (11). The present study using calcium phosphate salt extracted from milk would favor this hypothesis because the benefit was not lost 1 yr after discontinuation of calcium supplementation, at least in the cohort of boys who kept consuming the tested foods until the end of intervention. However, a recent study does not support this hypothesis because the benefit of bone mineral mass accrual obtained with calcium carbonate supplementation was maintained 1 yr after the end of intervention (26). Therefore, the reason why the benefit of calcium supplementation during childhood may or may not be maintained remains unknown. This issue needs to be further addressed not only because it may shed light on the mechanism by which calcium can affect aBMD increase but also because it is of importance for public health dietary recommendations in the context of early prevention of adult osteoporosis.
In conclusion, this prospective randomized controlled study in prepubertal boys showed that increasing the calcium intake induced a greater aBMD gain in the appendicular but not in the axial skeleton. The calcium effect on aBMD was not associated with a greater gain in bone size or standing height. Finally, this study in prepubertal boys indicates that the benefit on aBMD is not lost 1 yr after the end of intervention. Whether this benefit will be maintained until attainment of peak bone mass remains to be established before considering that increasing calcium consumption in boys during childhood is an efficient measure in the primary prevention of osteoporosis in men.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
First Published Online March 8, 2005
Abbreviations: aBMD, Areal bone mineral density; AT, active treatment analysis; BMC, bone mineral content; BMI, body mass index; Ca-suppl., calcium supplement; CI, confidence interval; FDia, femoral diaphysis; ITT, intention-to-treat analysis; LS, lumbar spine; RMet, distal metaphysis.
Received July 23, 2004.
Accepted February 28, 2005.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  L. Richert, T. Chevalley, D. Manen, J.-P. Bonjour, R. Rizzoli, and S. Ferrari Bone Mass in Prepubertal Boys Is Associated with a Gln223Arg Amino Acid Substitution in the Leptin Receptor J. Clin. Endocrinol. Metab., November 1, 2007; 92(11): 4380 - 4386. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Gugatschka, H. Dobnig, A. Fahrleitner-Pammer, P. Pietschmann, S. Kudlacek, A. Strele, and B. Obermayer-Pietsch Molecularly-defined lactose malabsorption, milk consumption and anthropometric differences in adult males QJM, December 1, 2005; 98(12): 857 - 863. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
