Key result
Machine learning using address-level SDOH vastly outperforms ADI in predicting diabetes prevalence.
Why the study?
Area indices assessing SDOH collapse dimensions and flatten address-level variation, complicating the study of SDOH pathways related to diabetes outcomes.
Cross-Sectional (n=2,369,365)
Effect estimate: Adjusted R2 0.948
Machine learning models utilizing discrete address-level SDOH features significantly outperform traditional area deprivation indices in predicting tract-level diabetes prevalence.
No takes yet. Share an insight, caveat, or question.
May refine tract-level diabetes prediction over ADI; leaves open clinical utility and prospective validation.
HUANG et al. (2026) conducted a cross-sectional in Diabetes prevalence (n=2,369,365). Address-level discrete SDOH measures vs. Area deprivation index (ADI) was evaluated on Tract-level diabetes prevalence (Adjusted R2 0.948). A machine learning model using address-level discrete social determinants of health predicted diabetes prevalence better than the area deprivation index (testing adjusted R2 0.948 vs 0.381).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: