Social Determinants May Improve Disease Risk Prediction Beyond Genetics Alone

Social, behavioral, and environmental determinants of health improved disease risk prediction beyond genetics alone in a new analysis of All of Us Research Program data published in The American Journal of Human Genetics.

The study, led by investigators at the Icahn School of Medicine at Mount Sinai, evaluated whether non-genetic factors could be integrated with polygenic scores to improve risk prediction for 6 common complex diseases: asthma, chronic kidney disease, coronary heart disease, hypercholesterolemia, breast cancer, and prostate cancer.

Researchers applied multiple correspondence analysis (MCA) to more than 100 survey-based and community-level variables related to environmental conditions, health behaviors, health care access, and social well-being. The approach generated low-dimensional “embeddings” designed to summarize non-genetic risk patterns without requiring investigators to preselect disease-specific risk factors.

The analysis used All of Us data from 413 457 participants. After restricting the cohort to individuals with usable whole-genome sequencing data, linked electronic health records, and other required information, the final modeling cohort included 171 614 participants.¹

Including MCA-derived social, behavioral, and environmental factors in models that already included demographics and polygenic scores consistently improved disease risk prediction. Improvements in area under the receiver operating characteristic curve ranged from 0.007 to 0.027 across the 6 diseases.¹

For 4 of the 6 diseases studied, gains in predictive performance from the MCA embeddings exceeded gains attributable to polygenic scores. The embeddings also recovered known contributors to disease risk, such as economic status and smoking, while identifying less commonly examined factors, including loneliness and spirituality.¹

“Genes are an important part of the equation, but they do not determine destiny,” Samira Asgari, PhD, assistant professor of genetics and genomic sciences at the Icahn School of Medicine at Mount Sinai and senior corresponding author of the study, said in a press release.²

The findings may be particularly relevant as health systems increasingly seek to integrate genetic risk scores, electronic health record data, and social determinants of health into more comprehensive risk prediction tools. For primary care clinicians, the study reinforces the importance of evaluating disease risk in the context of patients’ lived circumstances, not genetic predisposition alone.¹

The investigators also examined whether social and environmental context modified genetic risk. They found little evidence of broad interaction between polygenic risk and MCA-derived non-genetic factors. Adding interaction terms produced minimal improvement in model performance, with ΔROC-AUC of 0.001 or less. Genetic association results also remained highly stable after the MCA embeddings were added to models, suggesting that inherited risk and non-genetic context largely contributed additively in the diseases studied.¹

The authors cautioned that the study was not designed to prove causality or identify single risk factors that directly cause disease. Many survey responses were collected at a single point in time, limiting the ability to determine whether specific social or behavioral factors preceded disease onset. Survey nonresponse also was incomplete and nonrandom, which may introduce selection bias.¹

Still, investigators concluded that the results support incorporating social, behavioral, and environmental factors into clinical models of disease risk. Future studies are expected to evaluate how these determinants can be combined with additional biological data and to explore mechanisms linking social experiences with disease development.¹

References

1. Biji A, Ferar K, Pejaver V, Kenny EE, Liu B, Asgari S. Integrating social determinants of health and genetic risk in disease risk models. Am J Hum Genet. Published online June 22, 2026. doi:10.1016/j.ajhg.2026.05.014
2. The Mount Sinai Hospital/Icahn School of Medicine at Mount Sinai. Study finds social determinants of health can match or exceed genetic risk in predicting common diseases. Published June 22, 2026.