Proper health care relies on accurate risk prediction models to guide clinical decision making to prevent the development of adverse health outcomes. However, most current clinical risk prediction models focus on clinical risk factors and do not account for variability among individuals that influence disease risk. Precision medicine aims to tailor treatments towards the individual, and polygenic risk scores (PRS) show potential for using an individual’s unique genome to predict the risk for many complex diseases. Despite the successes of using PRSs in research studies, their translation to clinical use remains slow. One major factor limiting their predictive power and clinical utility is that they include the effects of only common variants across the genome and do not account for many other genetic and nongenetic risk factors that also contribute to disease risk. In this dissertation, we first evaluate the utility of PRSs, aiming to enhance their predictive performance by assessing the impact of various factors on PRS effectiveness. We develop optimized PRS models and explore the genetic relationship between cardiometabolic phenotypes and female-specific health conditions. Additionally, we investigate the effects of other risk factors on phenotypes. We test the association of an environmental risk factor, socioeconomic vulnerability, with a wide range of phenotypes and examine how these associations vary across different groups. We also examine the effects of rare variants, which are not included in PRSs, on a few phenotypes and increase the power of rare variants by combining the effects of multiple genes together into rare variant risk scores. Finally, we aim to improve the power of PRS models by incorporating them with rare variants and socioeconomic vulnerability. Our results demonstrate the potential of integrative risk models in improving model performance and support the need for future analyses to combine PRSs with other risk factors, particularly for diverse cohorts.