arXiv subjects

David H. Rehkopf

Publications and source records attributed to David H. Rehkopf.

5 recordsLinked to original sources

arXiv2026

Geospatial Foundation Models Capture Health-Relevant Dimensions of Place Beyond Conventional Social Risk Indices

Area-based social risk indices summarize residents' socioeconomic conditions but incompletely capture physical features of place that may affect health. We evaluated whether numerical representations of physical place produced by four geospatial foundation model families from 2022 satellite data explained residual variance in tract-level associations between the Area Deprivation Index, Social Deprivation Index, and Social Vulnerability Index with health outcomes. We used LightGBM to predict variables from the American Community Survey and 40 chronic disease and health-behavior outcomes from CDC PLACES across 82,646 census tracts in the contiguous United States, evaluating performance across 10 held-out states. Among survey variables, models were moderately predictive of some variables including housing type (R-squared up to 0.54) but weak for disability, unemployment, and income disparity. For health outcomes, models explained up to 54% of variance left unexplained by social risk indices, with the largest gains for annual checkups, arthritis, and high blood pressure. Mean total variance explained by geospatial foundation models across the 40 health-related outcomes increased from 0.31 in the smallest tract-size decile to 0.39 in the largest. Geospatial foundation models capture health-relevant features of place not represented by conventional social risk indices and may usefully augment them in epidemiological analyses.

stat.AP↗

arXiv2024

Towards Causal Interpretation of Sexual Orientation in Regression Analysis: Applications and Challenges

This study presents an approach to analyze health disparities in Sexual and Gender Minority (SGM) populations, with a focus on the role of social support levels as an example to allow causal interpretations of regression models. We advocate for precisely defining the exposure variable and incorporating mediators into analyses, to address the limitations of comparing counterfactual outcomes solely between SGM and heterosexual populations. We define sexual orientation into domains (attraction, behavior, and identity), and emphasize a consideration of these elements either separately or together, depending on the research question. We also introduce social support measured before and after the disclosure of sexual orientation to facilitate inference. We illustrate this approach by examining the association between SGM status and depression diagnosis with data from the 2020 and 2021 National Health Interview Survey. We find a direct effect of SGM status on depression (OR: 3.07, 95% CI: 2.64 - 3.58) and no indirect effect through social support (OR: 1.07, 95% CI: 0.87-1.31). Our research emphasizes the necessity of the comprehensive measurement of sexual orientation and a focus on intervenable variables like social support in order to empower SGM communities and address SGM related health inequalities.

stat.AP↗

arXiv2023

Potential for allocative harm in an environmental justice data tool

Neighborhood-level screening algorithms are increasingly being deployed to inform policy decisions. We evaluate one such algorithm, CalEnviroScreen - designed to promote environmental justice and used to guide hundreds of millions of dollars in public funding annually - assessing its potential for allocative harm. We observe the model to be sensitive to subjective model decisions, with 16% of tracts potentially changing designation, as well as financially consequential, estimating the effect of its positive designations as a 104% (62-145%) increase in funding, equivalent to \$2.08 billion (\$1.56-2.41 billion) over four years. We also observe allocative tradeoffs and susceptibility to manipulation, raising ethical concerns. We recommend incorporating sensitivity analyses to mitigate allocative harm and accountability mechanisms to prevent misuse.

stat.AP↗

arXiv2021

Machine learning in the social and health sciences

The uptake of machine learning (ML) approaches in the social and health sciences has been rather slow, and research using ML for social and health research questions remains fragmented. This may be due to the separate development of research in the computational/data versus social and health sciences as well as a lack of accessible overviews and adequate training in ML techniques for non data science researchers. This paper provides a meta-mapping of research questions in the social and health sciences to appropriate ML approaches, by incorporating the necessary requirements to statistical analysis in these disciplines. We map the established classification into description, prediction, and causal inference to common research goals, such as estimating prevalence of adverse health or social outcomes, predicting the risk of an event, and identifying risk factors or causes of adverse outcomes. This meta-mapping aims at overcoming disciplinary barriers and starting a fluid dialogue between researchers from the social and health sciences and methodologically trained researchers. Such mapping may also help to fully exploit the benefits of ML while considering domain-specific aspects relevant to the social and health sciences, and hopefully contribute to the acceleration of the uptake of ML applications to advance both basic and applied social and health sciences research.

cs.LG↗

arXiv2020

A data-driven prospective study of incident dementia among older adults in the United States

We conducted a prospective analysis of incident dementia and its association with 65 sociodemographic, early-life, economic, health and behavioral, social, and genetic risk factors in a sample of 7,908 adults over the age of 50 from the nationally representative US-based Health and Retirement Study. We used traditional survival analysis methods (Fine-Gray models) and a data-driven approach (random survival forests for competing risks) which allowed us to account for the competing risk of death with up to 14 years of follow-up. Overall, the top five predictors across all groups were lower education, loneliness, lower wealth and income, and lower self-reported health. However, we observed variation in the leading predictors of dementia across racial/ethnic and gender groups. Our ranked lists may be useful for guiding future observational and quasi-experimental research that investigates understudied domains of risk and emphasizes life course economic and health conditions as well as disparities therein.

stat.AP↗