Date of Award

May 2016

Degree Type


Degree Name

Master of Science (MS)


Nutrition Science and Dietetics


Margaret A. Voss


bone, gymnast, micronutrient, pre-menarcheal

Subject Categories

Medicine and Health Sciences


Objective: The purpose of this study was to explore the relationships between dietary intake of vitamin K, vitamin A, vitamin D, fiber, and bone mineral density in a sample of pre-menarcheal female gymnasts and non-gymnasts in the Syracuse-area. The goal was to observe whether higher intakes of these vitamins were associated with higher bone mineral density as measured by dual-energy X-ray absorptiometry (DXA), and to potentially examine the impact of activity level on these relationships.

Methods: A cross-sectional design was used to examine a subset of data from a larger on-going longitudinal study on bone growth and physical activity in young females. The original study was designed to compare activity, and bone density data between gymnast and non-gymnasts. The data used in this analysis were collected from January 2009 to September 2012 from 52 pre-menarcheal female gymnasts and non-gymnasts between the ages of 9-11. Participants were asked to complete Harvard’s Youth/Adolescent Questionnaire biannually to quantify nutrient intake. Annual bone density measurements were taken of the lumbar spine and total body using dual-energy X-ray absorptiometry (DXA). Participants also were asked to self-report physical activity semi-annually during an interview session in which they described the type of activity and hours per week of participation. Tanner breast and pubic stages were self-assessed by subjects at each interview session. Standing height, sitting height, and weight measures were collected at the interview sessions as well.

Participants: The participants were 52 pre-menarcheal female gymnasts and non-gymnasts between the ages of 9-11. The selection criteria for this analysis was limited to participants who self-reported as Tanner Stage I (Tanner breast I and Tanner pubic I) at the time of the annual DXA scan. This criterion minimizes the potentially confounding effects of estrogen on bone density and mineral concentration. Non-Caucasian subjects were not included in this analysis as the original cohort of subjects were predominantly Caucasian and the sample size for the current study was too small to account for racial variation. A factor analysis was performed to decrease the dimensionality of the full data set. A principle component analysis was used to observe groupings of dietary variables and to better understand over all dietary patterns. Pearson correlations, and simple linear regressions were used to identify associations between specific micronutrients and minerals and bone mineral density. Multiple regression analyses and ANOVA (general linear model, GLM) were used to further examine the relationships relationship between bone mineral density and the intake of vitamin A, vitamin D, biologically active vitamin K, and the ratio of vitamin A:D intake while adjusting for physical activity, weight, and age of participants. Statistical significance for all tests was set at the α=0.05 level.

Results: Inverse relationships were found between bone mineral density and dietary intake of vitamin A (p=0.018), beta-carotene (p=0.008), fiber (p=0.029), and carotene (p=0.007). Bone mineral density increased with a combination of increased vitamin K1 intake and increased body mass (p=0.0006) and vitamin A and increased body mass (p=0.0001). A complex relationship appears to exist between bone mineral density and the combination of vitamin A and vitamin D intakes. The lowest bone mineral densities were observed in the highest quartile of vitamin A intake; this quartile was well above the tolerable upper limit for the age range of the focal population age range. The lowest bone densities were also associated with the highest levels of vitamin D, which was an unexpected result. However, this quartile of vitamin D intake was also above daily recommend intake levels for the focal population. Although not statistically significant, bone mineral density was lowest when vitamin A was either very high (4th quartile of vitamin A intake, mean ± stdev = 17,317.34 ±4358.86 IU) or very low (1st quartile of vitamin A intake, mean ± stdev = 5243.01 ±1230.68 IU) with respect to vitamin D intake. Intermediate levels of the ratio of the vitamin A intake with reference to vitamin D intake produced the highest bone densities.

Conclusions: Higher intakes of fiber, vitamin A, carotene, and beta-carotene appear to be associated with lower measures of bone mineral density in this study population. Increased difference in the intake of vitamin A with respect to vitamin D appears to be associated with lower measures of bone mineral density, even with intake levels of vitamin D above the EAR and RDA suggested for this age group. This relationship may result from excess preformed vitamin A ingestion leading to increased retinoic acid in the body. Excess retinoic acid has the potential to impair vitamin D’s ability to carry out its own genomic functions. These relationships warrant further exploration in order to improve the understanding of the genomic impact of vitamin A’s impact on bone metabolism in the preadolescent life stages, as well as to aid in the prevention of negative effects on bone mineral density during this time period.


Open Access