| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Medicine, University of Melbourne, Royal Melbourne Hospital (M.A.C., L.M.P., M.F., J.D.W.), Parkville, Victoria 3050, Australia; and School of Health Sciences, Deakin University (C.A.N., C.M.), Burwood 3125, Victoria, Australia
Address all correspondence and requests for reprints to: Dr. John D. Wark, Department of Medicine, Royal Melbourne Hospital, Parkville, Victoria 3050, Australia. E-mail: jdwark{at}unimelb.edu.au.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The most rapid period of skeletal development occurs over several years in childhood and adolescence, accounting for 40–50% of the total accrual of skeletal mass (4, 5, 6). This period may provide the best opportunity to maximize peak bone mass. Calcium is the major mineral in bone, and increasing dietary calcium intake has been proposed as an effective way of increasing peak bone mass. A significant increase in calcium requirements has been demonstrated during periods of most rapid growth and skeletal consolidation (particularly during infancy and adolescence, followed by childhood and young adulthood) (7). A number of studies have assessed the effect on bone of increased dietary calcium (as supplements) in children between the ages of 7 and 14 yr (8, 9, 10, 11, 12). These studies have all demonstrated increases in bone density with increased calcium intake. This positive effect on bone density, ranging from 1.6–5.1% gain compared with controls, has been seen when total calcium intake is increased to 1200–1600 mg/d.
The interaction between calcium supplementation and pubertal status is controversial, and the mechanism of the effects of augmented calcium intake during bone development is not well understood. In a study conducted by Johnston et al. (9), there was no effect of increased calcium intake on bone density in female twins who were either postmenarcheal or who passed through puberty during the study, whereas those twins who were premenarcheal showed significant BMD increases in the lumbar spine and distal radius. In our previously reported study in twin girls (8) with a mean age of 14 yr, calcium supplementation increased both hip (1.3%) and spine (1.5%) bone density significantly compared with co-twin controls over 6 months, despite 74% of our subjects having achieved menarche. The bone density difference was maintained during calcium supplementation to 18 months. A recent review (13) found the bone density results of several calcium supplementation trials in younger and older children to be inconsistent and unclear, with some trials observing significant effects at different pubertal ages and at some bone sites, but not others.
We now report results from a randomized, co-twin, placebo-controlled, single-blind intervention study, measuring the effect of increased calcium intake on BMD measurements in 51 pairs of premenarcheal female twins with a mean age of 10.3 ± 0.2 yr. Our co-twin study design using female twins and determining within-pair differences in bone mineral measurements allows for greater control of environmental factors compared with most studies in the literature, in which unrelated individuals were studied. A major additional advantage of the co-twin study design is that it controls for all or, an the average, half of the additive genetic effects on (change in) bone mineral measurements at a stage of life when genetic variance in bone mineral measures has been shown to surge strongly (14). For these reasons, our co-twin design confers a substantial advantage in statistical power compared with studies in unrelated individuals (15, 16).
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
One hundred and forty-eight young female twin pairs, aged 8–17 yr, enrolled with the Australian Twin Registry were approached between October 1995 and March 2000 to participate in one of two studies: this calcium supplementation trial in premenarcheal girls or a longitudinal bone study. Sixty-four pairs of premenarcheal female twins (including one triplet set), aged 8–13 yr, agreed to participate in the clinical trial.
The study was approved by both the Clinical Research and Ethics Committee of Royal Melbourne Hospital and the Australian Twin Registry. For each twin pair, written informed consent was obtained from each individual participant and from at least one parent or legal guardian.
Study design
This co-twin randomized clinical trial of the effect of calcium supplementation was conducted for a period of 24 months. One member in each twin pair was randomly assigned to receive 1200 mg calcium as calcium carbonate in tablet form (Caltrate, 600 mg), and the other twin was given a matched placebo in a single-blind manner. For the triple set, one was randomly assigned to receive placebo and second and third members received calcium, contributing two pairs to the analysis. The placebo tablet was similar in appearance, taste, and composition, but contained no active ingredient. All tablets were supplied by Whitehall Pty. Ltd. (Sydney, Australia). Twins were instructed to take two tablets nightly at approximately the same time. Tablet counts were performed at the end of each 3-month period to assess compliance.
Subjects visited the study center at baseline and 6, 12, 18, and 24 months, when the following information was collected. Subjects completed questionnaires to assess their medical history, use of medication, and physical activity (8, 14, 17). Dietary calcium intake was measured using a 4-d food record [assessed using dietary analysis program Diet 3; Xyris Software (Australia) Pty. Ltd., Highgate Hill, Australia]. The food record was completed over 3 weekdays and 1 weekend day and was recorded in household measurements with the option of using scales if preferred. Dietary calcium was also assessed by short food frequency questionnaire (8, 14, 17). Physical activity was determined over the previous 6 months by questionnaire (17). Hours of sport per week and hours of walking per week were each recorded in one of the following categories: 0–1, 2–3, 4–7, or more than 7 h/wk. Height and weight were measured to the nearest 0.1 cm and 0.1 kg, respectively, at each visit. Subjects also were visited at their homes at 3, 9, 15, and 21 months, where they completed the short food frequency and short physical activity questionnaires.
Areal BMD (aBMD) was measured by dual energy x-ray absorptiometry (DXA) using a QDR 1000W densitometer (Hologic, Inc., Waltham, MA). The lumbar spine (LS; L2–L4), total hip (TH), femoral neck (FN), and total forearm (FA) BMD and total body bone mineral content (TB BMC) were measured at baseline and 6, 12, 18, and 24 months. Fat mass and lean mass were determined by DXA using Hologic software (version 6.10).
Statistical methods
Descriptive statistics are reported for all variables (unless stated). Paired t tests (two-tailed) and regression analysis of the within-pair differences in bone mineral measures between twins were performed using the statistical package SPSS (version 10.0). The within-pair absolute differences in gain in aBMD, BMC, and bone area were calculated as the (calcium twin at x months – calcium twin at 0 months) – (placebo twin at x months – placebo twin at 0 months). A t test was used to test the gain in bone mineral measures from baseline at each time point in the calcium and placebo groups, respectively. The within-pair percentage difference was calculated by dividing the absolute gain (described above) by the baseline value for the respective placebo group and multiplied by 100.
Densitometry results were expressed as the within-pair percent difference with the 95% confidence intervals. To assess the influence of calcium supplementation on bone mineral status independent of bone and body size, we adjusted for age, height, and weight using linear regression and calculated the adjusted within-pair percentage differences. Baseline anthropometric measurements were plotted for the calcium- and placebo-supplemented twins in a box plot (version 10.0, SPSS, Inc., Chicago, IL). An extreme outlier was considered as a value more than 3 box lengths from the upper (75th percentile) or lower (25th percentile) edge of the box, whereas an outlier was a value between 1.5 and 3 box lengths from the upper or lower edge of the box.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The baseline characteristics of the 51 pairs of participants are presented in Table 1
. There were no differences between the calcium and placebo groups for any baseline measure (by unpaired t test). Similarly, there was no within-pair difference in the baseline measures (by paired t test). All twin pairs were premenarcheal at baseline (51 pairs). There was no baseline difference between calcium and placebo twins in calcium intake by 4-d food record (Table 1) or short food frequency questionnaire (data not shown). There was no change in calcium intake over time in either the placebo- or calcium-treated groups. There was no within-pair difference in calcium intake from dietary sources at any time point (data not shown).
|
Tablet compliance did not differ between the calcium-supplemented and placebo-supplemented groups for any interval. Overall, the placebo group took 76.6% of their tablets, and the calcium group took 76.2%. The calcium group compliance ranged from 75.3–79.1%, and the placebo group compliance ranged from 75.1–79.9%. There was no significant change in compliance over time. The mean daily additional intake of calcium from the supplement (as determined by average compliance) was 914 mg/d. The calcium-supplemented group received an average of 1631 mg calcium/d (diet plus supplement) during the first 12 months of intervention compared with the placebo group, who received an average of 718 mg/d calcium.
There were increases in weight, height, and bone mineral measurements at each interval in both the calcium- and placebo-supplemented groups. During the 24-month study there was no significant difference between the placebo- and calcium-supplemented groups with respect to gains in height and weight at any time point.
During the first 6 months, the calcium-supplemented group showed significantly greater gains in aBMD at the TH (0.012 ± 0.004 g/cm2, or 1.6%; P < 0.01) and at the FN (0.011 ± 0.005 g/cm2, or 1.2%; P < 0.05; n = 51; Tables 2 and 3). From baseline to 12 months, significant within-pair differences in change in aBMD were detected at the TH (0.011 ± 0.006 g/cm2, 1.4%; P < 0.05) and LS (0.012 ± 0.005 g/cm2, 1.6%; P < 0.05; n = 48 pairs; Tables 2 and 3). At 18 months there was a marginal within-pair difference in change from baseline in TH aBMD (0.013 ± 0.007, 1.6%; P = 0.055), but no difference in LS and FN aBMD (n = 42 pairs; Tables 2 and 3). At 6, 12, and 18 months, there were no within-pair differences in gain in FA BMD or TB BMC. There was no significant within-pair difference in change in aBMD or TB BMC at any site measured after 24 months of calcium intervention (n = 24 pairs; Tables 2 and 3). At no interval was there a within-pair difference in bone area at any site as measured by DXA.
|
|
and Fig. 1). At 12 months, calcium-supplemented twins had a greater gain in adjusted aBMD at the TH (0.013 ± 0.005 g/cm2, 1.6%; P < 0.05), LS (0.012 ± 0.004 g/cm2, 1.6%; P < 0.01) and TB BMC (26.058 ± 8.579 g, 2.5%; P < 0.01), but not at the FN (Table 3 and Fig. 1). At 18 months, there was a significant difference in the within-pair change in adjusted TH aBMD (0.017 ± 0.007, 2.4%; P < 0.01) and adjusted TB BMC (50.429 ± 16.105, 4.6%; P < 0.01; Table 3 and Fig. 1). After 24 months of calcium intervention, there was a significant difference in adjusted TB BMC (41.019 ± 17.751, 3.7%; P < 0.05; Table 3 and Fig. 1). There was no within-pair difference in adjusted measures of bone mineral at the FA at any time point (data not shown).
|
All individual participants were premenarcheal at baseline. After 6 months, three of 102 individuals had reached menarche [one concordant pair and one discordant pair (became concordant by 12 months)]. At the 12 month point, five of 96 individuals had reached menarche [two concordant pairs and one discordant pair (concordant by 18 months)]. At 18 months, 17 of 84 individuals were postmenarcheal [seven concordant pairs and three discordant pairs (two pairs become concordant by 24 months)]. At the completion of the study, 13 of 48 individuals were postmenarcheal (five concordant pairs and three discordant pairs). There was no difference in the gain in measures of bone mineral between those who reached menarche and those that remained premenarcheal at any time point. Similarly, there were no within-pair differences in gain in bone mineral measures in those pairs concordant for being postmenarcheal compared with those that remained concordant for premenarcheal status.
There was no baseline difference in age, height, weight, lean mass, fat mass, and measures of bone mineral between those pairs that completed 24 months of intervention compared with those that ceased intervention at any earlier time point (data not shown). The results for mono- and dizygotic twins were similar (data not shown).
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Some calcium intervention studies have demonstrated significant differences in outcomes between pre- and postpubertal subjects (9, 11). The effects we report here were similar for TH, a 1.9% greater increase with calcium at 6 months rising to 2.4% at 18 months, compared with our previous study with a 1.3% effect after 12 months (8). In contrast, the response at the LS was not as marked and tended to be less at 18 months in the present study, but remained with a 1.6% greater increase at 18 months in our previous study. A key difference between our two trials was that at baseline and 18 months, 0% and 20% of participants in the current study were postmenarcheal compared with 74% and 86% in our previous study, respectively. Johnston et al. (9) found a significant effect on aBMD with increased calcium intake (average, 719 mg calcium/d as citrate malate) in 22 prepubertal monozygotic twin pairs. Significant increases of 5.1% at the distal radius (cortical bone) and 2.8% at the lumbar spine (more trabecular bone) were observed, but there was no effect of calcium in the 23 pairs of monozygotic twins who were postpubertal or who passed through puberty during the study period. In contrast, we found no effect of supplementation at the total forearm in our prepubertal twin pairs, but an increase in aBMD at the LS and TH after 12 months of intervention. In the study by Johnston et al. (9), the sample size (n = 22 prepubertal twin pairs) was smaller than ours, and males were included. These factors may help to explain the differences observed between the studies. Molgaard et al. (18) and Bailey et al. (19) indicated that a significant relationship of bone accretion with pubertal stage occurs in boys and girls, with the peak annual accretion occurring earlier in girls (13 yr) than boys (14 yr). Tanner stage (20) has been found to be a significant determinant of BMD in girls, but not in boys. Therefore, the underlying rate of change in bone mineral measures may well differ between girls and boys of the same age, and responsiveness to interventions also may differ.
Other calcium intervention studies involving premenarcheal females have found increases in bone density (11, 12). Bonjour and co-workers (12) found that supplementation with 804 mg calcium/d added to foods in prepubertal females for 48 wk resulted in a 1.6–2.4% increase in bone density compared with controls at various sites. Bonjour et al. (12) a1so reported that gains in BMD were more pronounced in girls on a low calcium intake before the study.
Lloyd et al. (11) found similar results as in our present study with a substantially smaller level of calcium supplementation (360 vs. 934 mg/d). Approximately 32% of their female population had reached menarche compared with 5% of our female twins at the end of the first 12 months. Lloyd et al. (11) reported that among subjects with above-median Tanner score, the calcium-supplemented girls had higher rates of bone acquisition than the placebo group. In our previous study in which 74% had reached menarche, we found increases in BMD of 1.3% and 1.5% at the TH and LS, respectively, with 6 months of calcium supplementation in the entire group.
There is continuing uncertainty about the level of dietary calcium required to ensure adequate availability of calcium for optimal accrual of bone mass. Dietary calcium has a fundamental role in the development of growing bones. Calcium must be absorbed in sufficient quantities to adequately provide for periods of rapid growth (such as infancy and adolescence) and to offset daily excretory losses that would otherwise deplete previously obtained skeletal reserves (21). There is a threshold requirement for dietary calcium, which is determined predominantly by the skeletal needs. This means that the skeletal response will occur when calcium intake is increased from deficiency levels to a threshold zone. Increasing calcium intake when the level of dietary calcium already exceeds the threshold will most likely not produce additional gains in bone mass (22). Heaney et al. (21) estimated calcium threshold levels for optimally nourished children and adolescents to be approximately 1400–1600 mg/d. Our calcium-treated group achieved significant gains is aBMD compared with the placebo groups with an overall calcium intake that approximated the upper limit of the threshold estimated by Heaney et al. (21).
There were some interesting contrasts between our two studies. The current twin population was substantially younger and smaller than our previous study group (8). Although we provided our twins with 1200 mg calcium as calcium carbonate tablets, the previous study used 1000-mg effervescent calcium tablets (containing calcium carbonate and calcium lactate gluconate). The average total calcium intake (dietary and tablet) was very similar (both increased to >1600 mg/d). The different types of calcium salts may have influenced intestinal absorption of calcium, but any differences would be expected to be small. Another explanation for the observed differences between our two trials may be that during periods of rapid growth, calcium absorption is up-regulated to meet the body’s needs (7). Mean height velocity has been shown to peak at approximately 12 yr of age in females (7). The older twin population, with a mean age of 14.0 yr (baseline), would have probably achieved their maximum height velocity 2 yr before study commencement. Our current twin population, with a mean age of 10.3 yr (baseline), was in an accelerated phase of growth and development, with a greater ability to retain calcium.
Our study of female twins used a classic twin approach by determining within-pair differences in BMD, accounting for many genetic and environmental confounders (23), in contrast to most studies in the literature in which unrelated individuals were used. The co-twin design also has greater statistical power than a comparable study of unrelated individuals with a similar sample size, in part due to the matching of subjects for age, genetic and environmental factors, skeletal maturity, and anthropometric factors (14). A potential problem with the co-twin intervention design is the possibility of cross-over, particularly when the co-twins cohabit, as our young twins did. However, each cross-over would be predicted to reduce rather than exaggerate any treatment effect. Our current study has demonstrated 1.5–2.0% increases in aBMD in young premenarcheal females. An increase in hip bone density of 0.05 g/cm2 (4.6% of the average young adult female value) has been projected to represent a 50% reduction in osteoporotic fracture risk in later life (24). If our increase in bone density were to persist postsupplementation (through to skeletal maturity), the results may translate into a significant reduction in osteoporotic fracture risk in later life. We acknowledge the possible effects of multiple t tests, but as there were consistent trends in the data and our previous study (8), these cannot be explained by random nominal statistical significance.
In conclusion, supplementation with 1200 mg calcium/d was associated with increases in bone mineral measures in premenarcheal females. All subjects are being followed postintervention to determine the persistence of the effect of calcium supplementation after the adolescent phase of growth. If these gains remain after cessation of supplementation, these effects may increase peak bone mass and reduce the future risk of developing osteoporosis and related fractures.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: aBMD, Areal BMD; BMD, bone mineral density; DXA, dual energy x-ray absorptiometry; FA, forearm; FN, femoral neck; LS, lumbar spine; TB BMC, total body bone mineral content; TH, total hip.
Received November 14, 2003.
Accepted July 14, 2004.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  T. Lehtimaki, J. Hemminki, R. Rontu, V. Mikkila, L. Rasanen, M. Laaksonen, N. Hutri-Kahonen, M. Kahonen, J. Viikari, and O. Raitakari The Effects of Adult-Type Hypolactasia on Body Height Growth and Dietary Calcium Intake From Childhood Into Young Adulthood: A 21-Year Follow-up Study--The Cardiovascular Risk in Young Finns Study Pediatrics, October 1, 2006; 118(4): 1553 - 1559. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Zhu, Q. Zhang, L. H. Foo, A. Trube, G. Ma, X. Hu, X. Du, C. T Cowell, D. R Fraser, and H. Greenfield Growth, bone mass, and vitamin D status of Chinese adolescent girls 3 y after withdrawal of milk supplementation Am. J. Clinical Nutrition, March 1, 2006; 83(3): 714 - 721. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-P. Bonjour Is Peripuberty the Most Opportune Time to Increase Calcium Intake in Healthy Girls? IBMS BoneKEy, March 1, 2005; 2(3): 6 - 11. [Full Text] [PDF]
|
|
|
|
|
|
E. Seeman and G. J. Strewler Clinical and Basic Research Papers - September and October 2004 Selections IBMS BoneKEy, November 1, 2004; 1(11): 1 - 5. [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 |
