Date of Award


Document Type

Campus Access Dissertation


Epidemiology and Biostatistics



First Advisor

Jihong Liu


Introduction: Childhood and adolescence represent critical times for bone growth and the accretion of bone mass. Children who fail to maximize their bone mass during this period may face an increased risk of osteoporosis, a skeletal disorder that is marked by low bone mass, bone deterioration and increased fracture risk, in later life. Diet represents one potentially modifiable avenue through which bone mass can be optimized during childhood. To date, calcium and vitamin D have received a great deal of attention in the field of bone research. However, emerging evidence suggests that other aspects of diet such breastfeeding, fruit and vegetable intake and protein consumption may also have important consequences on bone health. This dissertation sought to examine the effects of these three factors on bone mass in British children at 10 and 16 years of age.

Methods: Children from the Avon Longitudinal Study of Parents and Children (ALSPAC), a geographically based birth cohort study of children from the former county of Avon in England, served as our study population. In our analyses, we included white children from ALSPAC who were singleton births with available data on bone outcomes, diet and relevant covariates. Total body less head bone mineral density (BMD) was measured by dual energy x-ray absorptiometry (DXA) during clinic visits when the children were 10 and 16 years of age. Infant feeding information was gathered using questionnaires that were mailed to parents when their infants were 6 months of age and was used to examine breastfeeding initiation and duration, as well as feeding type (breast milk only, breast milk and formula, formula only) at 5 months of age.

Fruit and vegetable intake as well as protein consumption were assessed using 3 day diet diaries when the children were 11 years of age. Total fruit and vegetable intake, intake of specific fruits and vegetables and intake of micronutrients (vitamin C, folate, magnesium, potassium and calcium) found in high amounts in fruits and vegetables were examined. To study the effect of protein, total protein intake from all sources as well as protein intake from specific sources (non-dairy animal foods, dairy foods, plant-based foods and other foods) was examined.

Analysis of covariance was used to adjust for a range of covariates including age at the time of the DXA scan, pregnancy, social and behavioral factors, and body size (lean mass, fat mass, and height). Interactions between sex and each aspect of diet were examined and results were presented separately for boys and girls when significant interactions were identified.


Breastfeeding and BMD: At 10 years of age, ever breastfed boys had lower bone mineral density (BMD) [â -5.25, p=0.02] than never breastfed boys. A similar, but partially attenuated relationship was observed at 16 years of age [â -4.06, p=0.40]. At age 10, each additional month of breastfeeding in boys was associated with a 0.68 g/cm2 [p=0.02] decrease in BMD [age 16: â -0.52, p=0.46]. Lastly, compared to boys fed only formula at 5 months, boys fed only breast milk [â -7.10, p=0.01] had lower BMD at age 10. By age 16, similar, but partially attenuated results were again observed with boys fed only breast milk [â -6.90, p=0.28] having lower BMD than those fed only formula. In girls, no differences in BMD were detected according to the breastfeeding variables.

Fruits and vegetables and BMD: At 10 years of age, children in the highest tertile of green leafy vegetable consumption had 4.07 g/cm2 [p<0.01] higher BMD than children in the lowest [p for trend <0.01]. By 16 years of age, evidence of an association between green leafy vegetable consumption and BMD was no longer detected. With the exception of calcium intake, which was positively associated with BMD [p=0.40], no other specific vegetables, fruits or micronutrients were found to be associated with BMD.

Protein and BMD: The relationship between protein and BMD differed by calcium intake. In children who consumed less than 850 milligrams of calcium per day, increasing grams of total protein [â 0.28, p<.01] and non-dairy animal protein [â 0.22, p<.01], were positively associated with BMD at 10 years of age. Similar associations continued to be observed in BMD at 16 years of age. No associations between protein and BMD were identified in children who consumed more than 850 milligrams of calcium per day.

Conclusion: Potentially important relationships between breastfeeding, fruit and vegetable consumption and protein intake and bone density in children were identified in this dissertation. The deficits in bone density that were observed in breastfed boys were unexpected and may have resulted from the lower vitamin D content of breast milk compared to formula. Green leafy vegetables were identified as having a positive effect on BMD. If this finding is confirmed in future studies, efforts to increase green leafy vegetable consumption in children may be warranted. Finally, total protein intake and animal protein intake were found to have a positive effect on bone density, but only when calcium intake was low. Future research aimed at understanding the mechanism behind this association is needed, but it may be due to the effect of protein on improving intestinal absorption of calcium.