Challenges in Health Analytics Due to Lack of Diversity in Genetic Research: Implications and Issues with Published Knowledge

This chapter covers the challenges in health analytics that analysts will encounter due to lack of diversity in genetic research studies. We will cover the need for greater genetic diversity and the differences between race, ethnicity, and genetic ancestry. Challenges with identifying and inferring ancestry from genetics and concordance or lack thereof between genetically determined variables and self-reported. This will include both challenges related to race/ethnicity, but also challenges touched upon earlier in this book such as self-reported chromosomal status (a proxy for genetic sex). The issues with applying genetic knowledge learned from European ancestry populations to individuals of other ancestries will be discussed and the errors in reports that can occur due to patients’ ancestries not matching those of the published research studies. We will also discuss challenges related to ancestry and genetic admixture. On a related note, we will discuss Mendelian disease findings, especially homozygous state in adult humans and how that should theoretically be rare among diseases that affect life expectancy but should be less rare for diseases that do not impact life expectancy. We will describe why caution should be used when obtaining results as some Mendelian diseases are less well studied in large sample sizes, and how care should be taken given the issues with under-representative in genetic research and other challenges. We will describe how to identify problematic prior research in genetics. We will teach readers how to recognize that genetics alone will not explain population-level differences in disease prevalence by race and ethnicity in part because race and ethnicity are social constructs and discrimination, and deferential treatment is a real phenomenon with health consequences. After reading this chapter, you should be able to confidently answer the following questions: