Abstract

We are, some say, at the threshold of a medical revolution. Current medical practice – which is based on a crude ‘one size fits all’ (or ‘one size fits most’) approach – will be replaced by ‘precision medicine’: an approach where big data and machine learning are harnessed to offer precisely tailored risk predictions, diagnoses, and treatment plans based on an individual’s lifestyle, environment, and genetic make-up. In this talk, I look at the role of sex and gender data categories in the development of precision medicine. I focus specifically on the case of precision medicine research on Alzheimer’s, dementia and related disorders, a well-funded, politically powerful, and socially salient field of biomedicine with a history of contentious debate regarding the role of biological and social factors in disease risk and prediction. I identify an assumption that I call the ‘default predictive value of sex’ and show how this assumption is fuelling calls for the development of sex-specific algorithms and ‘pink and blue’ machine learning models. In doing so, I show how these approaches to precision medicine risk naturalizing gender disparities and ossifying a binary, essentialized conception of sex in diagnostic and predictive tools.