The authors state that they have no conflicts of interest.
A Positive Dose–Response Effect of Vitamin D Supplementation on Site-Specific Bone Mineral Augmentation in Adolescent Girls: A Double-Blinded Randomized Placebo-Controlled 1-Year Intervention†
Article first published online: 6 MAR 2006
Copyright © 2006 ASBMR
Journal of Bone and Mineral Research
Volume 21, Issue 6, pages 836–844, June 2006
How to Cite
Viljakainen, H. T., Natri, A.-M., Kärkkäinen, M., Huttunen, M. M., Palssa, A., Jakobsen, J., Cashman, K. D., Mølgaard, C. and Lamberg-Allardt, C. (2006), A Positive Dose–Response Effect of Vitamin D Supplementation on Site-Specific Bone Mineral Augmentation in Adolescent Girls: A Double-Blinded Randomized Placebo-Controlled 1-Year Intervention. J Bone Miner Res, 21: 836–844. doi: 10.1359/jbmr.060302
- Issue published online: 4 DEC 2009
- Article first published online: 6 MAR 2006
- Manuscript Accepted: 3 MAR 2006
- Manuscript Revised: 15 FEB 2006
- Manuscript Received: 3 OCT 2005
- vitamin D3 supplementation;
- adequate calcium intake;
- adolescent girls;
- bone mineral augmentation
The effect of vitamin D supplementation on bone mineral augmentation in 212 adolescent girls with adequate calcium intake was studied in a randomized placebo-controlled setting. Bone mineral augmentation determined by DXA increased with supplementation both in the femur and the lumbar vertebrae in a dose-responsive manner. Supplementation decreased the urinary excretion of resorption markers, but had no impact on formation markers.
Introduction: Adequate vitamin D intake protects the elderly against osteoporosis, but there exists no indisputable evidence that vitamin D supplementation would benefit bone mineral augmentation. The aim of this 1-year study was to determine in a randomized double-blinded trial the effect of 5 and 10 μg vitamin D3 supplementation on bone mineral augmentation in adolescent girls with adequate dietary calcium intake.
Materials and Methods: Altogether, 228 girls (mean age, 11.4 ± 0.4 years) participated. Their BMC was measured by DXA from the femur and lumbar spine. Serum 25-hydroxyvitamin D [S-25(OH)D], intact PTH (S-iPTH), osteocalcin (S-OC), and urinary pyridinoline (U-Pyr) and deoxypyridinoline (U-Dpyr) were measured. Statistical analysis was performed both with the intention-to-treat (IT) and compliance-based (CB) method.
Results: In the CB analysis, vitamin D supplementation increased femoral BMC augmentation by 14.3% with 5 μg and by 17.2% with 10 μg compared with the placebo group (ANCOVA, p = 0.012). A dose–response effect was observed in the vertebrae (ANCOVA, p = 0.039), although only with the highest dose. The mean concentration of S-25(OH)D increased (p < 0.001) in the 5-μg group by 5.7 ± 15.7 nM and in the 10-μg group by 12.4 ± 13.7 nM, whereas it decreased by 6.7 ± 11.3 nM in the placebo group. Supplementation had no effect on S-iPTH or S-OC, but it decreased U-DPyr (p = 0.042).
Conclusions: Bone mineral augmentation in the femur was 14.3% and 17.2% higher in the groups receiving 5 and 10 μg of vitamin D, respectively, compared with the placebo group, but only 10 μg increased lumbar spine BMC augmentation significantly. Vitamin D supplementation decreased the concentration of bone resorption markers, but had no impact on bone formation markers, thus explaining increased bone mineral augmentation. However, the positive effects were noted with the CB method but not with IT.
During puberty, bone mass increases dramatically, reaching 80–90% of peak bone mass in late adolescence.(1,2) Although peak bone mass is strongly determined by genetic factors,(3) it is believed that environmental factors such as physical activity and nutrition can modulate bone mass accretion during growth.(4,5) Low bone mass is a risk factor for the development of osteoporosis. The main goal in its prevention is to optimize bone acquisition during growth. In addition, low bone mass in adolescent girls is related to increased risk of distal forearm fractures, which commonly occur in girls at their peak height velocity.(6,7)
Vitamin D insufficiency occurs commonly in many parts of the world,(8–10) because synthesis of the vitamin in the skin is lowered during winter and food sources of vitamin D are scarce. Rickets is seen in children with severe vitamin D deficiency, but previous studies have shown that even mild vitamin D insufficiency can have detrimental effects on bone mineral acquisition(11–13) and bone remodeling(12–14) in adolescence.
No international consensus on the optimal serum 25-hydroxyvitamin D [S-25(OH)D] concentration has been reached. It has been proposed that concentrations >80 nM would be optimal,(15) but a dissenting opinion claims that concentrations >50 nM show replete vitamin D status.(16) Furthermore, the optimal S-25(OH)D concentration for children and adolescents has not been discussed.
There is a complex interaction between calcium and vitamin D in bone metabolism.(16) Many previous intervention trials in growing children and adolescents have focused on the importance of calcium on bone mineral accrual.(17,18) In Finland, the dietary intake of calcium is typically adequate among adolescents because of high consumption of nonvitamined milk products. This provides an opportunity to study the effect of vitamin D3 supplementation on bone mineral augmentation, serum 25(OH)D, and biochemical markers of bone turnover in adolescent girls in a randomized, controlled, double-blinded setting.
MATERIALS AND METHODS
Subjects and subject selection
A total of 228 adolescent girls, mean age 11.4 ± 0.4 years, were studied in the capital region of Helsinki (60° N) in southern Finland. Recruitment was conducted in primary schools. The subjects included were healthy, used no medicines known to affect calcium metabolism, and were of white origin. Ethical approval was obtained from the Ethical Committee of Helsinki and Uusimaa Hospital District. The subjects and their parents gave informed written consent in agreement with the Helsinki Declaration before entering the study.
The subjects first presented between September 2001 and March 2002 and were later examined every sixth month (Fig. 1). The study lasted 1 year. During each visit, height, weight, and pubertal development were measured. Background data on diet and physical activity as well as serum and second void urinary samples were collected. BMD was measured twice during the study.
A stratified randomization process was performed three times for equal size blocks. The stratification factor was pubertal development (Tanner stage). In each Tanner stage, an equal number of subjects was randomly assigned into three treatment groups receiving 0 (placebo), 5, or 10 μg of vitamin D3 as supplement (Scanpharm). The randomization was done by a person not involved in the project. The vitamin D3 content of the tablets was confirmed by analysis performed at the Danish Institute for Food and Veterinary Research (Søborg, Denmark). The subjects were instructed to take one tablet per day for 28 days during each month of the 1-year study. The blister packs were returned to the researchers during each visit, and compliance was confirmed by pill counts. The study was double-blinded.
The fasting serum samples were collected between 8:00 a.m. and 10:00 a.m. and stored at −20°C before analysis. S-25(OH)D concentration [S-25(OH)D2 and S-25(OH)D3] in the samples was measured with high-performance liquid chromatography (HPLC) analysis at the Danish Institute for Food and Veterinary Research.(10) The CVs for the intra- and interassay were 4.3% and 6.3%, respectively.
Serum intact PTH (S-iPTH) was measured with a commercial two-site immunoenzymometric assay (IEMA; OCTEIA; IDS, Boldon, UK), with intra- and interassay CVs of 3.0% and 5.4%, respectively. Serum and urinary concentrations of calcium, phosphate, and creatinine (S-Ca, S-Pi, U-Ca, and U-Crea, respectively) were analyzed using an automated KoneLab spectrophotometer (Thermo Clinical Labsystems, Espoo, Finland) at the Department of Applied Chemistry and Microbiology, University of Helsinki, following routine methods. The intra- and interassay CVs for these analyses were <7.5%, except for U-Ca and U-Crea, which were <15%. The U-Ca and U-Pi concentrations were expressed as mmol/mmol Crea.
The biochemical markers of bone turnover and creatinine were analyzed at the Department of Food and Nutritional Sciences and Biosciences Institute, University College Cork. S-OC was measured in duplicate samples using an ELISA (N-MID; Osteometer Biotech, Osteopark, Denmark). The intra-assay CV was 11%. U-Pyr and U-Dpyr were analyzed in duplicate, using an automated analysis system (Gilson ASPEC; Gilson, Villiers-le-Bel, France) as described elsewhere.(19) The intra-assay CVs for Pyr and Dpyr were 5% and 3%, respectively. Interassay variation was avoided by analyzing all samples from an individual in the same run. The mean U-Pyr and U-DPyr were expressed as nmol/mmol creatinine. Creatinine was determined in fasting second void urine samples using a diagnostic kit (Metra Creatinine Assay Kit; catalogue no. 8009; Quidel). The intra-assay CV was 1.6%.
The bone area (BA), BMC, and areal BMD (aBMD) were measured at the beginning and end of the study with DXA model A (QDR 4500; Hologic, Waltham, MA, USA) from the lumbar spine L2–L4 vertebrae and left femur region. The femur region includes the femoral neck, trochanter, Ward's triangle, and interarea that is analyzed as a single area of interest. Calibration of the measurement was performed using a spine phantom; the interassay CV for the phantom was 0.31%. Intra-assay CVs were determined with duplicate measurements of 10 subjects. CVs for BMD in the left femur and lumbar vertebrae were 0.67% and 1.39%, respectively.
The site-specific bone mineral augmentation was determined after changes in BMC, as suggested for use in growth studies.(20) The change in BMC is confounded by the change in BA. For example, their correlation was strong (r = 0.8, p < 0.001); therefore, the change in BMC was adjusted with the change in BA.(11,21–23) In addition, BMC values were adjusted for change in weight and Tanner stage.
Tanner stage, nutrient intakes, and other background data
The pubertal stage of the subject was assessed ad modum Tanner.(24) A self-assessment protocol concerning the evaluation of breast development and timing of menarche was completed during the interview.
The subjects completed, together with their parents, a form containing questions on their medical history, use of vitamin D and calcium supplements, and time spent outdoors, on sunny holidays, and in physical activity. The dietary vitamin D and calcium intakes were evaluated using a food frequency questionnaire (FFQ), which is a validated semiquantitative questionnaire covering >70 foods.(12) The nutrient contents of the foods were calculated using the Finnish National Food Composition Database, Fineli, version 2001, which is maintained by the National Public Health Institute of Finland, Nutrition Unit. The physical activity of the subjects consisted of school trips, guided activity in free time, and free time activity on their own.(12) The activity was calculated in minutes per day. All forms were checked by the researchers, and if needed, additional information was requested.
Statistical analyses were performed with SPSS version 12.0 for Windows (SPSS, Chicago, IL, USA). The normal distribution of variables was tested with a Kolmogorov-Smirnov test; if normality was not present, logarithmic transformations were made.
Repeated-measures ANOVA was used to detect the effect of supplementation on serum 25(OH)D and iPTH. The posthoc tests were performed with Tukey's honestly significant different (HSD) and Dunnett's test. Bone data and biochemical markers of bone turnover were tested both with ANOVA and analysis of covariance (ANCOVA) to show the effect of confounding factors. Comparison of groups in ANCOVA was performed with contrasts, and for each analysis, covariates are shown in footnotes.
The change in the biochemical markers of bone turnover was tested for a determinant of the bone mineral augmentation in the femur and the lumbar vertebrae in a linear regression model.
Both the intention-to-treat (IT) and compliance-based (CB) approach was applied in the statistical analyses. In IT analysis, all subjects were included, whereas in CB analysis, those exceeding 80% compliance were included; thus, the number of subjects decreased to 64, 58, and 71 in the placebo, 5-μg group, and 10-μg group, respectively.
Results were considered statistically significant at p < 0.05. p values between 0.05 and 0.1 are considered as trends. In tables, SD is presented in brackets after the mean value. In figures, SE is shown.
Characteristics at baseline and after 1 year
Of the 228 girls studied at baseline, 16 (7.0%) dropped out because of reasons not related to the study protocol (Fig. 1). The baseline characteristics of the 212 girls that completed the study are shown in Table 1, and the change in them after 1 year is shown in Table 2.
Effect of vitamin D supplementation on serum 25(OH)D and iPTH
Both analytic methods (IT and CB) showed a dose-response of vitamin D3 supplementation on S-25(OH)D (Table 2). A significant main effect was shown both with supplementation (p = 0.002) and initial 25(OH)D concentration (p = 0.006), but their interaction was not significant (Fig. 2A).
For further analyses of S-25(OH)D concentration, we divided the subjects into two distinct cohorts, depending on their recruitment time. In autumn (September–October), 102 girls initiated the intervention, whose results are seen in Fig. 2B. In winter (February–March), 78 girls initiated the intervention, and their results are presented in Fig. 2C. These cohorts represented the seasonwise extremes. The girls recruited in November (N = 32) were thus excluded from these specific analyses regarding S-25(OH)D concentration.
The overall response in S-iPTH to supplementation was negligible (Fig. 3).
Effect of pubertal development on BMC
Both the BMC and the BA of the femur and the lumbar vertebrae increased significantly (p < 0.001) in the placebo group throughout the study. The change in BMC correlated strongly with the change in BA (r = 0.784 and r = 0.844; p < 0.001), weight gain (r = 0.288 and r = 0.284; p < 0.001), and change in pubertal development (r =0.194 and r = 0.282; p < 0.006) in the femur and lumbar vertebrae, respectively.
Effect of vitamin D supplementation on BMC and BMD
The BMD in the femur and vertebrae increased slightly more (p < 0.1) in the vitamin D–supplemented groups compared with the placebo group both in the femur and lumbar vertebra when tested with ANOVA (Table 2). On the other hand, the changes in BMC and BA did not differ among the study groups at either site. No covariates were used in the ANOVA, and the possibility of confounding factor exists.
To correct the effect of confounding factors, the BMC was analyzed with ANCOVA, in which change in BA, change in weight, and change in Tanner stage were used as covariates. The femoral BMC augmentation did not differ among groups in ANCOVA when all subjects were included (p = 0.256; N = 210). After requesting 80% compliance, a significant difference among groups was seen (p = 0.024; N = 176). Bone mineral augmentation in the femur was 14.3% and 17.2% higher in the groups receiving 5 and 10 μg of vitamin D, respectively, compared with the placebo group (Fig. 4A). In the lumbar vertebrae, the BMC augmentation did not differ among study groups when all subjects were included (p = 0.144; N = 208). When taking the compliance into account, the difference reach significance (p = 0.039; N = 185), and the BMC augmentation was 12.5% higher in group receiving 10 μg of vitamin D than in the placebo (Fig. 4B).
Because BMC augmentation depends on pubertal stage, we further tested whether the effect of supplementation differed between initially early puberty (Tanner stage 1 or 2) and mid puberty (Tanner stages 3–5) subgroups. We observed that the response to supplementation was similar between subgroups in the femur, but it differed in the lumbar vertebrae (Table 3). Among mid-puberty girls, the BMC augmentation in the lumbar vertebrae was dose-dependent and significant (ANCOVA, p = 0.012), whereas it was not among early puberty girls.
Biochemical markers of bone turnover
The concentrations of both bone resorption markers, U-Pyr and U-DPyr, decreased in the placebo group (p < 0.03) during the study, but the formation marker, S-OC, did not change (p = 0.312).
Vitamin D supplementation further decreased U-Dpyr concentration (p = 0.042; Table 2). On the other hand, supplementation had no effect on S-OC concentrations (ANCOVA, p = 0.401).
Together, the change in S-OC and U-DPyr and U-Pyr explained 13% and 12% of the BMC augmentation in the total femur and lumbar spine, respectively.
Effect of puberty and supplementation on serum and urinary calcium and phosphate concentrations
The U-Ca/Cr ratio decreased during the study (p < 0.001), but no change was seen in U-Pi/Cr excretion. S-Ca increased during the study (p < 0.001) as did S-Pi concentration (p = 0.05). Vitamin D supplementation had no effect on these variables.
This is the first controlled study in which the effect of two doses of vitamin D3 supplementation on bone mineral augmentation in 11- to 12-year-old girls with adequate calcium intake was studied. Bone mineral augmentation in the femur was 14.3% and 17.2% higher in the groups receiving 5 and 10 μg of vitamin D, respectively, compared with the placebo group. In the lumbar vertebra, the BMC augmentation was 12.5% higher with the 10-μg dose than in the placebo. Vitamin D supplementation decreased the concentration of bone resorption markers but had no impact on bone formation markers.
The change in BMC was strongly associated with the change in BA, weight, and puberty development both in the femur and the lumbar vertebrae, as has been reported.(25–27) Although BA and BMC contribute equally to bone strength,(27) the focus in this study was on the bone material properties and its augmentation, which are confounded with changes in BA. Confounding factors can obscure the effect of vitamin D on bone mineral augmentation if they are not controlled in the analysis.
The IT approach is favored in medical sciences.(28) Recently, this topic has been highlighted because it is generally thought that the IT analyzed results could lead to different conclusions than those with efficacy analyzed.(29) In this study, the mean compliance was 89%.(8) The subjects with poor compliance weakened the effect of the supplementation on S-25(OH)D and BMC augmentation, because they had not taken their pills for >2.4 months, which justifies emphasizing the CB results. In addition, the groups were not distorted in basic characteristics when CB analysis was used.
Vitamin D supplementation influenced S-25(OH)D and S-iPTH
Vitamin D supplementation increased the S-25(OH)D concentration in a dose-responsive manner. With 10 μg daily, a stable vitamin D status was maintained throughout the year in the autumn cohort. Guillemant et al.(30) reported that three oral doses of 2.5 mg of vitamin D3 at 2-month intervals for 13- to 17-year-old adolescent boys held the postsummer 25(OH)D concentration stable during the 6 months of the study. Conversely, Lehtonen-Veromaa et al.(11) failed to maintain the postsummer 25(OH)D concentration with 10 μg/day, possibly because they gave vitamin D2, which has lower bioavailability than vitamin D3(31) and increases S-25(OH)D concentration less than D3.(31,32)
An S-25(OH)D concentration of 80 nM for adults has been proposed to be optimal.(15,16) Our study indicates that positive effects on BMC augmentation in adolescent girls are seen when mean S-25(OH)D >50 nM is reached. It should be borne in mind that the analytical methods of assessing S-25(OH)D concentration differ. We used HPLC, which in the DEQAS comparison slightly underestimates the S-25(OH)D in relation to other methods. However, higher concentrations could indeed have a more pronounced effect on BMC augmentation. We agree with Lips(16) that dietary intake of calcium should be taken into account when adequate vitamin D status is defined.
Effect of vitamin D supplementation on BMC augmentation
Vitamin D supplementation increased BMC augmentation 14.3% and 17.2% higher in groups receiving either 5 or 10 μg vitamin D, respectively, than in the placebo group. Similar results were shown in the lumbar vertebrae, albeit only the highest dose increased the BMC augmentation significantly in the whole group. Our results agree with previous studies in which vitamin D promotes bone mineral accretion in adolescent girls.(11,34) In addition, our results implicate that the effect of the threshold nutrients on bone mineral augmentation during growth are vaster than during other phases of life.(27) It has been shown that the long bones grow faster before puberty, whereas during puberty, the spine growth accelerates.(35) This could be the reason why mid-puberty girls responded more efficiently to vitamin D supplementation in the lumbar vertebrae than early puberty girls.
Vitamin D intervention studies have been done mainly in the elderly.(36–39) Intervention trials in which vitamin D supplementation was beneficial to BMD or decreased fracture risk are few,(36,39) whereas others detected no change.(40,41) In many studies where positive effects were seen, calcium supplementation was always included as well.(37,39) We believe that the effect of vitamin D on bone is mediated through calcium balance as those vitamin D–deficient persons with adequate intake of calcium will benefit from vitamin D supplementation.
Effect of vitamin D supplementation on biochemical markers of bone turnover
Supplementation had no effect on the bone formation marker, osteocalcin, which had a slightly decreased concentration during the study as was expected to occur during normal puberty.(14,35) Decreased bone for The U-Pyr concentration, which reflects bone resorption, decreased more in the vitamin D–supplemented groups that in the placebo group, but the effect was not dose-dependent. This implies that supplemental vitamin D has antiresorptive properties that spare bone mineral. We assume that these functions are caused by improved calcium balance(42) because it has been shown that calcium decreases bone resorption.(43) Our results conflict with those of Schou et al.,(44) who reported that vitamin D supplementation had no effect on bone turnover markers in 20 children 6–14 years of age. However, the number of subjects was small, and the results were not controlled by individual variation (e.g., age, pubertal development) that could have obscured the effect. Adequate vitamin D status was associated with decreased bone resorption markers in other studies,(11,14) and in vitamin D deficiency, all biochemical markers of bone turnover are known to be elevated.(45)
Our study indicates that daily vitamin D3 supplementation improves bone mineral augmentation, because 14–17% more mineral was laid down in the femur. The highest dose (10 μg) increased mineral augmentation in the lumbar vertebra by 12.5% in the whole group, but mid-pubertal girls responded even more efficiently. In this study, the positive effects were noted among complaint subjects, which implies that at this age group, the use of supplementation is not as feasible public health strategy as fortifying specific foods. Without follow-up and peer support, the compliance probably would have been even lower. The main focus of this study was to define which vitamin D dose has an effect on bone. Appropriate tools to increase vitamin D intake among adolescents should be further studied.
We conclude that the current vitamin D recommendation for adolescent girls, at least in the northern latitudes, is too low to ensure sufficient vitamin D status during winter. Intake of vitamin D at rates of 10–15 μg/day aids to maintain stable S-25(OH)D concentrations during winter. Vitamin D induced BMC augmentation by decreasing bone resorption, but not affecting bone formation, which was reflected by the biochemical markers of bone turnover. Optimizing bone mineral gain in adolescence is crucial to the prevention of osteoporosis later in life. Increasing vitamin D intake to 10–15 μg/day aids in attaining this goal.
HV was involved in the recruitment of subjects, organizing the sampling, data collection, measuring BMD by DXA, statistical analysis, and writing the manuscript. AN and MH participated in the sampling and data collection and helped with bone mineral measurements. AP was involved in the recruitment of the subjects and laboratory analysis. JJ was responsible for the analysis of S-25(OH)D. KC was responsible for the laboratory analysis of the bone markers. CM was responsible for and involved in the study design and counseling in practical things throughout the study. MK and CL-A were responsible for and involved in the study design, writing of the study protocol, recruitment of the subjects, and writing of the manuscript.
- 61989 Epidemiology of fractures of the distal end of the radius in children as associated with growth. J Bone Joint Surg Am 71A: 1225–1231., , , ,
- 292005 To D or not to D. Commentaries. BoneKEy-Osteovision 2: 28–31.
- 342003 Bone mass and density response to a 12-month trial of calcium and vitamin D supplement in preadolescent girls. J Musculoskelet Neuronal Interact 3: 63–70., , ,