Document Type


Date of Degree

Spring 2016

Access Restrictions

Access restricted until 07/03/2020

Degree Name

PhD (Doctor of Philosophy)

Degree In


First Advisor

Burns, Trudy L

First Committee Member

Burns, Trudy L

Second Committee Member

Janz, Kathleen F

Third Committee Member

Levy, Steven M

Fourth Committee Member

Smith, Brian J

Fifth Committee Member

Ryckman, Kelli K


The foundation for osteoporosis risk is established during the time periods of childhood, adolescence, and young adulthood, periods of development when bone mass is being accrued rapidly. The relative quantity of bone mass accrued is influenced by both lifestyle and genetic factors. The purpose of this dissertation project was to discover single nucleotide polymorphisms (SNPs) associated with: (1) The rate of hip bone accrual (measured as bone mineral content or BMC) during the adolescent growth spurt, and (2) Total hip bone mass measured as BMC around the age of 19 when the amount of bone accrued is approximately at its peak. Additionally, SNP × longitudinal lifestyle factor (calcium intake per day, vitamin D intake per day, and minutes of moderate to vigorous physical activity (MVPA) per day) multiplicative interaction effects were assessed. Each cohort member’s vector of longitudinal physical activity measurements was summarized as belonging to one of a set of specific trajectory groups using finite mixture modeling. The same was then done for calcium intake and vitamin D intake. The source of the data utilized was the Iowa Bone Development Study (IBDS), which includes genetic and longitudinal bone measurement information.

To discover SNPs, a genome-wide association study (GWAS) design was utilized. Females and males were analyzed separately and together. The association between SNPs and the rate of hip bone accrual during the adolescent growth spurt was assessed using linear mixed models controlling for body size, and the association between SNPs and peak hip bone mass was assessed using an ordinary linear regression model, also controlling for body size. Approximately 500,000 SNPs were tested in each GWA analysis; significance was assessed at a familywise error rate of 0.05, the individual test cutoff of which was determined by using SimpleM, a modified Šídák correction.

No statistically significant SNPs were detected at the 0.05 familywise error rate threshold established by SimpleM (p < 1.76×10-7); however genes near suggestive SNPs (24 total) were assessed for biological relevance. Of most biological relevance were two suggestive SNPs (rs2051756 and rs2866908, p-values of 1.25×10-6 and 4.28×10-6, respectively) that were detected in an intron of the DKK2 gene through the GWA analysis exploring peak bone mass in females. The DKK2 gene is part of the Wnt signaling pathway and is associated with embryonic development; additionally, it is expressed more highly in osteoarthritic osteoblasts than in normal osteoblasts. No statistically significant results were found from the SNP × lifestyle factor multiplicative interaction effect tests. The potential importance of the DKK2 gene to peak hip bone mass accrual in females should be studied further in order to understand the pathophysiology of this suggested novel association identified during a discovery GWA analysis.


Bone, Genetics, Longitudinal Data, Peak Bone Mass


xxii, 245 pages


Includes bibliographical references (pages 237-245).


This thesis has been optimized for improved web viewing. If you require the original version, contact the University Archives at the University of Iowa:


Copyright © 2016 Camden Phillip Bay

Included in

Epidemiology Commons