Share this article on:

Physical Activity and Bone Mineral Density in Adolescents with Vitamin D Deficiency


Medicine & Science in Sports & Exercise: April 2010 - Volume 42 - Issue 4 - p 646-650
doi: 10.1249/MSS.0b013e3181bb813b
Clinical Sciences

Introduction: Studies have shown that physical activity (PA) is superior to many other environmental factors in determining bone mineral density (BMD), but none has examined the independent relationship between PA and vitamin D status.

Purpose: The aim of this study was to assess the relationship among amount of PA, vitamin D (25(OH)D), and BMD.

Methods A total of 166 female ballet dancers and sedentary adolescents were divided by tertiles of serum levels of 25(OH)D (<11.3, 11.3-14.9, and ≥15 ng·mL−1). Diet, PA, and menstruation were assessed by questionnaires; BMD was measured in three sites by dual-energy x-ray absorptiometry.

Results: Across 25(OH)D tertiles, there were no differences in mean participant age, weight, height, PA, calcium and energy intake, BMD, or parathyroid hormone. PA was positively associated with BMD in participants with vitamin D deficiency. Multivariable regression analysis, controlling for age, body mass index, parathyroid hormone, and bone turnover markers, showed that total body, femoral neck, and lumbar spine BMD were all positively related to PA, with regression coefficients increasing as vitamin D levels dropped across tertiles.

Conclusions: PA is positively related to BMD in vitamin D-deficient female adolescents and with increasing magnitude as serum vitamin D levels drop. These findings suggest that PA may counteract the detrimental effect of marked vitamin D deficiency on bone mass.

1"Hadassah Optimal" Sport Medicine Center, Department of Orthopedic Surgery, Hadassah-Hebrew University Medical Center, Jerusalem, ISRAEL; 2Exercise, Nutrition, and Lifestyle Clinic, Safra Children's Hospital, Sheba Medical Center, Tel Hashomer, ISRAEL; 3Department of Epidemiology and Preventive Medicine, Sackler School of Medicine, Tel Aviv University, Tel Aviv, ISRAEL; 4Department of Clinical Nutrition, Rambam Health Care Campus, Haifa, ISRAEL, and 5Bone and Mineral Metabolism Unit, Rambam Health Care Campus, Haifa, ISRAEL

Address for correspondence: Naama W. Constantini, M.D., FACSM, "Hadassah Optimal" Sport Medicine Center, Department of Orthopedic Surgery, Hadassah-Hebrew University Medical Center, Jerusalem, Israel; E-mail:

Submitted for publication April 2009.

Accepted for publication August 2009.

The prevalence of osteoporosis and its associated morbidity, mortality, and health costs is continuingly increasing, although simple methods of treatment are available (20). Attaining high peak bone mass in young adulthood is an important preventive measure of osteoporosis. Adolescence is a critical period of bone mass accrual, and the maximal rate of calcium accumulation in females occurs at the age range of 10.4-14.6 yr (2). Although most peak bone mass is determined genetically, several biological and environmental factors can affect bone accretion as well. In healthy individuals, these are particularly sex/gender, estrogen, smoking, calcium and vitamin D intakes, race/ethnicity, sun exposure, and physical activity (PA) (16). PA can increase bone mass in children and adolescents, specifically weight-bearing exercises with jumping and resistance training components (10,19). Several studies have suggested that PA is superior to other major environmental factors in determining peak bone mass (12-14,18,21,23). In addition, there seems to be a positive interaction between PA and dietary calcium in enhancing bone accretion. Whereas vitamin D is another key factor in bone metabolism, its relationship with PA in enhancing bone mass has not been directly examined. The purpose of this study was to assess the relationship between PA and bone mineral density (BMD) in adolescent females, about 25-hydroxyvitamin D (25(OH)D) status.

Back to Top | Article Outline


The study population consisted of 54 Caucasian female adolescent ballet dancers recruited by addressing ballet dance schools, asking those who danced ≥15 h·wk−1 to participate in the study, and 112 Caucasian female adolescents from the general population, who were found to consume a diet containing less than 800 mg of calcium in a previous study (17), with both groups from the city of Haifa, Israel. Combining the two groups was possible because many areas of similarity were found, namely biological age (years from menarche), height, weight, body mass index (BMI), energy, and calcium intake (all P > 0.1). This grouping allowed for a very wide spectrum of PA amount (1-30 h·wk−1) in a population homogenous for biological age and body mass. These are two highly important factors that determine BMD, and such a wide range of PA is very difficult to achieve with other forms of sampling. All participants and parents signed informed consent forms, and the study was approved by the Institutional Review Board of Rambam Medical Center, Haifa, Israel. We excluded one dancer who reported using oral contraceptives, nine dancers who refused the blood tests or BMD measurements, four dancers who did not complete all baseline measurements, and five who did not have 25(OH)D data available for analysis because of technical reasons.

The amount of PA performed was assessed by self-reported activity records (4). A semiquantitative food frequency questionnaire that included all commercial dairy products and all "typical" teenage foods (fast foods, snacks), developed for the previous study on BMD and calcium intake (17), was used. A computer program was used to calculate the nutritional and energy values using food tables from the Israeli Ministry of Health. In addition, data regarding menstruation, such as age at menarche and the number of menses in the passing year, were also recorded. Menstrual irregularity was defined as having seven or less menses per year, which corresponds to consistently having long cycles of more than 45 d-the upper limit of normal in this age group (1). Weight and height were measured by standard methods, and BMI was calculated.

The main outcome of the study was BMD, which was measured by dual-energy x-ray absorptiometry (Lunar DPX software version 3.6, Madison, WI) at the level of L2-L4 vertebra (lumber BMD (LBMD)), femoral neck (FBMD), and total body (TBMD). The coefficient of variation (CV) of the BMD measurements at these sites was 1%-1.6%. Blood samples for 25(OH)D, parathyroid hormone (PTH), bone alkaline phosphatase (BAP), and osteocalcin (OC) were collected in the morning after an overnight fast. 25(OH)D was measured by RIA (DiaSorin, Stillwater, MN; intra-assay CV at the range of 10.8%-11.7%; interassay CV at this range 8.2%-9.4%). Intact PTH was measured by IRMA (Nichols Allegro, San Juan, CA; intra-assay CV at this range 3.4%; interassay CV at this range 5.6%). BAP was measured by IRMA (Tandem-R Ostase; Hybritech, Inc/Beckman Coulter, San Diego, CA; intra-assay CV at this range 4.2%; interassay CV at this range 7.2%). OC was measured by RIA (Cis Bioindustries, France; intra-assay CV at this range 3.0%-3.7%; interassay CV at this range 5.5%-6.6%). The second urine of the morning was collected on the same day of blood sampling for urinary deoxypyridinoline (DPD) levels (Metra Biosystems, Mountain View, CA; intra-assay CV at this range 8.4%; interassay CV at this range 4.8%) and creatinine.

Vitamin D insufficiency was defined as a 25(OH)D serum level below 30 ng·mL−1 (3,8), whereas vitamin D deficiency was defined when the level was below 15 ng·mL−1 (15).

Back to Top | Article Outline

Statistical analysis.

Normal distribution was evaluated with visual inspection of histograms and determination of skewness and kurtosis, and all variables except serum PTH, DPD, PA, and BMI were considered normally distributed. After logarithmic transformation, serum PTH, DPD, PA, and BMD assumed normal distribution to enhance their compliance with the assumptions of regression models. We divided subjects into tertiles according to 25(OH)D levels and carried out analyses as follows. ANOVA and Tukey's post hoc analysis were used for evaluating the difference among the three 25(OH)D groups. The relationships of PA levels and BMD values at each of the regions studied were analyzed using a stepwise multivariable regression analysis, controlling for necessary covariates including age, OC, BAP, BMI, and PTH. The analyses were conducted separately for every 25(OH)D tertile. Residuals of final models were normally distributed. Results are expressed as mean ± SD. A P ≤ 0.05 was considered statistically significant. Given an adjusted R 2 of 0.23, the study had a power of 82% to detect a significant (P < 0.05) association in a multivariable linear regression.

Back to Top | Article Outline


Table 1 presents demographic, clinical, and laboratory data of study participants, by 25(OH)D tertiles. All subjects were vitamin D-insufficient because the highest value encountered was 29 ng·mL−1. Ninety-seven subjects (64%) were vitamin D-deficient. Mean amount of PA did not differ between the groups (post hoc analysis P > 0.2). Only urinary DPD reached a statistically significant difference between the groups, whereas BAP had a borderline statistical significance. The levels of these bone turnover markers were higher across decreasing tertiles.



The types of activities performed by dancers were ballet classes, mostly including bar exercises and jumps/dance routines. Sedentary subjects reported participation in school gym classes, which typically include various types of mild to moderate PA and very few afternoon activities such as dance lessons, swimming, or walking. For the whole cohort, PA was positively associated with BMD. Each daily hour of PA was associated with an increase of 0.03, 0.04, and 0.06 g·cm−2 in TBMD, LBMD, and FBMD, respectively. Table 2 presents the relationship between the amount of PA and BMD at the various sites in each 25(OH)D tertile. In the low tertiles, which represent vitamin D deficiency, higher BMD was found in participants with higher levels of PA (P = 0.02 for TBMD, P = 0.16 for LBMD, and P = 0.01 for FBMD).



Variables significantly associated with TBMD, LBMD, and FBMD are given in Table 3. Among participants in tertile 1, PA was significantly (P < 0.001) associated with BMD at all sites, yet in tertile 3, no such association was found. Multivariable regression analysis of PA with BMD as dependent variables by 25(OH)D tertiles is shown in Figure 1. TBMD was positively related to PA, with increasing magnitude as 25(OH)D levels dropped across tertiles. Regarding LBMD and FBMD, there was a positive association with PA among subjects in the lowest 25(OH)D tertile. FBMD was also associated with PA among subjects in the second 25(OH)D tertile. No association was found between LBMD and FBMD and PA among subjects in the highest tertile, who were not vitamin D-deficient. The figure demonstrates the increasing effect of PA on BMD across decreasing 25(OH)D levels.





Back to Top | Article Outline


PA and vitamin D are two major factors that determine bone mass, but we are unaware of previous studies that directly addressed the relationship between the three. We used data from adolescent females performing a wide range of PA and discovered that PA was increasingly related to BMD as serum 25(OH)D levels drop. This effect was maximal among subjects with marked 25(OH)D deficiency in all sites-LBMD, FBMD, and TBMD. Given the normal BMD in this tertile, despite marked vitamin D deficiency but in the face of a relatively high amount of PA (10 h·wk−1), we suggest that PA may counteract the known deleterious effects of 25(OH)D deficiency on bone health.

The major limitation of our study is its cross-sectional, observational nature. This prevents cause-and-effect interpretations, but multivariable analyses allowed elucidating independent relationships. Because this is the first study that has addressed and identified the interrelations among PA, vitamin D, and BMD, new insights emerge, which set the stage for controlled intervention trials in vitamin D-deficient populations. It should be noted that the majority of participants, namely the sedentary adolescents, were a priori known to consume a diet low in calcium. Hence, this does not represent the general population but rather uses this special situation to clarify the effect of PA in the presence of vitamin D deficiency. In addition, because no participant had sufficient 25(OH)D levels, our findings are limited to female adolescents with low vitamin D levels.

A high peak bone mass attained in young adulthood is an important determinant of BMD in later life (16), and many environmental factors act on the genetic background influencing bone accrual. Regarding vitamin D, randomized supplementation studies in female adolescents have shown increases in BMD even without additional calcium supplementation (7,22). This adds to findings from observational studies, underscoring the importance of vitamin D in promoting bone mass (5,11). At older age, calcium and vitamin D supplementation assists in decreasing bone loss (20). Another major determinant of BMD is estrogen, and adolescents with menstrual irregularities could have low BMD, which might persist into adulthood (6,24). Because the time from menarche and the proportion of participants with menstrual irregularities did not differ between tertiles, this suggests that estrogen and menstruation do not explain our findings. The importance of PA in increasing BMD has been demonstrated in numerous studies, as summarized in the position statement from the American College of Sports Medicine (10). The equal proportion of dancers in each tertile assisted in maintaining a similar amount and a high level of weight-bearing PA (7.8-11.0 h·wk−1), which, we think, counteracted the deleterious effects of low vitamin D on BMD. To our knowledge, this is the first study to address the interaction between serum 25(OH)D levels and PA in affecting BMD.

It is notable that despite being vitamin D-deficient or insufficient, participants had normal BMD values across all 25(OH)D tertiles, when compared with available reference curves (9): TBMD and LBMD are around the 50th percentile for age, gender, and race, and FBMD is around the 90th percentile. An explanation may be the relatively high mean amount of PA in each tertile (10 h·wk−1), underscoring its beneficial effects. The higher levels of BAP and DPD in the lower 25(OH)D tertiles can reflect increased bone turnover brought about by vitamin D deficiency-but the net result is a normal BMD in all sites. It should be noted that although urinary DPD levels were higher in the lowest tertile of 25(OH)D, reflecting increased bone resorption, multivariable analyses did not identify an independent relationship with BMD.

In summary, we have demonstrated that PA of weight-bearing type of at least 10 h·wk−1 is an independent factor related to BMD in the presence of very low serum 25(OH)D levels. The association between PA and BMD grew stronger with decreasing 25(OH)D levels. These findings suggest that PA may act as a protective factor in preserving bone mass in the face of vitamin D deficiency and reinforce its pivotal role in bone health. Additional intervention studies in vitamin D-deficient subjects or animal models, similar to those combining calcium supplementation and PA, may assist in understanding the mechanism and prerequisites for the activity-induced bone accrual.

Naama W. Constantini and Gal Dubnov-Raz are joint first authors. This study was supported by grants from the Chief Scientist of The Israel Ministry of Health and the Rambam Medical Center Research Foundation.

Prof Ish-Shalom received grant support from Aventis (2004), Eli-Lilly (1996, 2000, 2004, 2005-present), Beaufour Ipsen Group (2001), Merck Sharp & Dome (1997, 2000, 2002), Novartis (2003), NPS (2003), Medison Pharma (2001), Sunlight Ultrasound Technologies (1999), Neopharm Pharmaceutical (1998), Smith Kline Beecham (1997), and Transpharma (2007-present) and consulting and lecture fees from Aventis, Chemical and Technical Supplies, Eli-Lilly, Merck Sharp & Dome, Meditec, Transpharma, and NPS Allelix Corp. All other authors have nothing to declare.

All authors report no conflicts of interest. The results of the present study do not constitute endorsement by American College of Sports Medicine.

Back to Top | Article Outline


1. Adams Hillard PJ. Menstruation in young girls: a clinical perspective. Obstet Gynecol. 2002;99:655-62.
2. 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;15:2245-50.
3. Bischoff-Ferrari HA, Giovannucci E, Willett WC, Dietrich T, Dawson-Hughes B. Estimation of optimal serum concentrations of 25-hydroxyvitamin D for multiple health outcomes. Am J Clin Nutr. 2006;84:18-28.
4. Bouchard C, Tremblay A, Leblanc C, Lortie G, Savard R, Thériault G. A method to assess energy expenditure in children and adults. Am J Clin Nutr. 1983;37:461-7.
5. Cheng S, Tylavsky F, Kröger H, et al. Association of low 25-hydroxyvitamin D concentrations with elevated parathyroid hormone concentrations and low cortical bone density in early pubertal and prepubertal Finnish girls. Am J Clin Nutr. 2003;78:485-92.
6. Csermely T, Halvax L, Schmidt E, et al. Occurrence of osteopenia among adolescent girls with oligo/amenorrhea. Gynecol Endocrinol. 2002;16:99-105.
7. El-Hajj Fuleihan G, Nabulsi M, Tamim H, et al. Effect of vitamin D replacement on musculoskeletal parameters in school children: a randomized controlled trial. J Clin Endocrinol Metab. 2006;91:405-12.
8. Holick MF, Chen TC. Vitamin D deficiency: a worldwide problem with health consequences. Am J Clin Nutr. 2008;87:1080S-6S.
9. Kalkwarf HJ, Zemel BS, Gilsanz V, et al. The bone mineral density in childhood study: bone mineral content and density according to age, sex, and race. J Clin Endocrinol Metab. 2007;92:2087-99.
10. Kohrt WM, Bloomfield SA, Little KD, Nelson ME, Yingling VR, American College of Sports Medicine Position Stand: physical activity and bone health. Med Sci Sports Exerc. 2004;36(11):1985-96.
11. Lehtonen-Veromaa MK, Mottonen TT, Nuotio IO, Irjala KM, Leino AE, Viikari JS. Vitamin D and attainment of peak bone mass among peripubertal Finnish girls: a 3-y prospective study. Am J Clin Nutr. 2002;76:1446-53.
12. Lloyd T, Beck TJ, Lin HM, et al. Modifiable determinants of bone status in young women. Bone. 2002;30:416-21.
13. Lloyd T, Chinchilli VM, Johnson-Rollings N, Kieselhorst K, Eggli DF, Marcus R. Adult female bone density reflects teenage sports-exercise patterns but not teenage calcium intake. Pediatrics. 2000;106:40-4.
14. Lloyd T, Petit MA, Lin HM, Beck TJ. Lifestyle factors and the development of bone mass and bone strength in young women. J Pediatr. 2004;144:776-82.
15. Misra M, Pacaud D, Petryk A, Collett-Solberg PF, Kappy M; Drug and Therapeutics Committee of the Lawson Wilkins Pediatric Endocrine Society. Vitamin D deficiency in children and its management: review of current knowledge and recommendations. Pediatrics. 2008;122:398-417.
16. Ondrak KS, Morgan DW. Physical activity, calcium intake and bone health in children and adolescents. Sports Med. 2007;37:587-600.
17. Rozen GS, Rennert G, Rennert HS, Diab G, Daud D, Ish-Shalom S. Calcium intake and bone mass development among Israeli adolescent girls. J Am Coll Nutr. 2001;20:219-24.
18. Specker B, Vukovich M. Evidence for an interaction between exercise and nutrition for improved bone health during growth. Med Sport Sci. 2007;51:50-63.
19. Strong WB, Malina RM, Blimkie CJ, et al. Evidence based physical activity for school-age youth. J Pediatr. 2005;146:732-7.
20. Tang BM, Eslick GD, Nowson C, Smith C, Bensoussan A. Use of calcium or calcium in combination with vitamin D supplementation to prevent fractures and bone loss in people aged 50 years and older: a meta-analysis. Lancet. 2007;370:657-66.
21. Välimäki MJ, Kärkkäinen M, Lamberg-Allardt C, et al. Exercise, smoking, and calcium intake during adolescence and early adulthood as determinants of peak bone mass. Cardiovascular Risk in Young Finns Study Group. BMJ. 1994;309:230-5.
22. Viljakainen HT, Natri AM, Karkkainen M, et al. 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. 2006;21:836-44.
23. Welten DC, Kemper HC, Post GB, et al. Weight-bearing activity during youth is a more important factor for peak bone mass than calcium intake. J Bone Miner Res. 1994;9:1089-96.
24. Wiksten-Almströmer M, Hirschberg AL, Hagenfeldt K. Reduced bone mineral density in adult women diagnosed with menstrual disorders during adolescence. Acta Obstet Gynecol Scand. 2009;88:543-9.


©2010The American College of Sports Medicine