Development of a Bayesian Joint Logistic Model to Better Study the Association between Haplotypes and Disease

In 2011, there will be an estimated 1,596,670 new cancer cases and 571,950 cancer-related deaths in the US. With the ever-increasing applications of cancer genetics in epidemiology, there is great potential to identify genetic risk factors that would help identify individuals with increased genetic susceptibility to cancer, which could be used to develop interventions or targeted therapies that could hopefully reduce cancer risk and mortality. In this dissertation, I propose to develop a new statistical method to evaluate the role of haplotypes in cancer susceptibility and development. This model will be flexible enough to handle not only haplotypes of any size, but also a variety of covariates. I will then apply this method to three cancer-related data sets (Hodgkin Disease, Glioma, and Lung Cancer). I hypothesize that there is substantial improvement in the estimation of association between haplotypes and disease, with the use of a Bayesian mathematical method to infer haplotypes that uses prior information from known genetics sources. Analysis based on haplotypes using information from publically available genetic sources generally show increased odds ratios and smaller p-values in both the Hodgkin, Glioma, and Lung data sets. For instance, the Bayesian Joint Logistic Model (BJLM) inferred haplotype TC had a substantially higher estimated effect size (OR=12.16, 95% CI = 2.47-90.1 vs. 9.24, 95% CI = 1.81-47.2) and more significant p-value (0.00044 vs. 0.008) for Hodgkin Disease compared to a traditional logistic regression approach. Also, the effect sizes of haplotypes modeled with recessive genetic effects were higher (and had more significant p-values) when analyzed with the BJLM. Full genetic models with haplotype information developed with the BJLM resulted in significantly higher discriminatory power and a significantly higher Net Reclassification Index compared to those developed with haplo.stats for lung cancer. Future analysis for this work could be to incorporate the 1000 Genomes project, which offers a larger selection of SNPs can be incorporated into the information from known genetic sources as well. Other future analysis include testing non-binary outcomes, like the levels of biomarkers that are present in lung cancer (NNK), and extending this analysis to full GWAS studies.

ECONOMIC ANALYSIS OF THE DEMAND FOR PRESCRIPTION MEDICINE IN THE UNITED STATES

The purpose of this study was to understand the role of principle economic, sociodemographic and health status factors in determining the likelihood and volume of prescription drug use. Econometric demand regression models were developed for this purpose. Ten explanatory variables were examined: family income, coinsurance rate, age, sex, race, household head education level, size of family, health status, number of medical visits, and type of provider seen during medical visits. The economic factors (family income and coinsurance) were given special emphasis in this study.^ The National Medical Care Utilization and Expenditure Survey (NMCUES) was the data source. The sample represented the civilian, noninstitutionalized residents of the United States in 1980. The sample method used in the survey was a stratified four-stage, area probability design. The sample was comprised of 6,600 households (17,123 individuals). The weighted sample provided the population estimates used in the analysis. Five repeated interviews were conducted with each household. The household survey provided detailed information on the United States health status, pattern of health care utilization, charges for services received, and methods of payments for 1980.^ The study provided evidence that economic factors influenced the use of prescription drugs, but the use was not highly responsive to family income and coinsurance for the levels examined. The elasticities for family income ranged from -.0002 to -.013 and coinsurance ranged from -.174 to -.108. Income has a greater influence on the likelihood of prescription drug use, and coinsurance rates had an impact on the amount spent on prescription drugs. The coinsurance effect was not examined for the likelihood of drug use due to limitations in the measurement of coinsurance. Health status appeared to overwhelm any effects which may be attributed to family income or coinsurance. The likelihood of prescription drug use was highly dependent on visits to medical providers. The volume of prescription drug use was highly dependent on the health status, age, and whether or not the individual saw a general practitioner. ^

ASSOCIATIONS AND INTERRELATIONSHIPS OF SOCIOECONOMIC STATUS, BEHAVIORAL FACTORS AND BIOLOGICAL FACTORS WITH BIRTH WEIGHT

Although the influences of socioeconomic, behavioral and biological factors on birth weight have been extensively studied, most studies have been limited to clinical populations. This study examines such relationships in a national probability sample, the National Health and Nutrition Examination Survey of 1971-1974. The study sample consisted of 2161 white children and 812 black children, aged 1 to 5 years. Analyses were performed on a subsample consisting of 753 white and 138 black children whose mothers were also selected into the survey. Detailed analyses examined interrelationships among socio-economic, behavioral and biological factors by means of multiple regression and partial correlation procedures in the white population. These analyses were not carried out among blacks because of an observed clustering bias introduced in the black subsample that hampered generalization to the US population.^ The results among the whites indicated that the biological factors of maternal height, maternal weight, maternal size (weight/height('2)), maternal age and sex of child were independently related to birth weight and were also interrelated with socioeconomic factors such as family income, education of the mother and education of the head of the household. The joint effect was significantly associated with birth weight.^ Mothers' dietary practices represented the behavioral factors. Selected nutrients from the mothers' 24-hour dietary recall were used to develop indices of dietary quality. Dietary quality was significantly interrelated with socioeconomic status, biological factors and birth weight.^ The findings of this study suggest that smaller, younger mothers of lower socioeconomic status and female children were significantly associated with lower birth weight. The findings also suggest that dietary quality is a mediating factor among socioeconomic status and biological factors in that mothers with more financial and educational resources have better dietary practices. Such mothers may also practice other health behaviors that would prevent having a low birthweight baby. This dissertation contributes primarily to the further conceptualization and empirical testing of the interrelationships among socioeconomic, behavioral and biological factors with respect to birth weight. ^