%0 Journal Article %J SSM Popul Health %D 2018 %T Machine learning approaches to the social determinants of health in the Health and Retirement study. %A Seligman, Benjamin %A Tuljapurkar, Shripad %A David Rehkopf %K Biomarkers %K Computer science %K Machine learning %K Neural network %K Social Factors %K Social Support %X

Background: Social and economic factors are important predictors of health and of recognized importance for health systems. However, machine learning, used elsewhere in the biomedical literature, has not been extensively applied to study relationships between society and health. We investigate how machine learning may add to our understanding of social determinants of health using data from the Health and Retirement Study.

Methods: A linear regression of age and gender, and a parsimonious theory-based regression additionally incorporating income, wealth, and education, were used to predict systolic blood pressure, body mass index, waist circumference, and telomere length. Prediction, fit, and interpretability were compared across four machine learning methods: linear regression, penalized regressions, random forests, and neural networks.

Results: All models had poor out-of-sample prediction. Most machine learning models performed similarly to the simpler models. However, neural networks greatly outperformed the three other methods. Neural networks also had good fit to the data (between 0.4-0.6, versus <0.3 for all others). Across machine learning models, nine variables were frequently selected or highly weighted as predictors: dental visits, current smoking, self-rated health, serial-seven subtractions, probability of receiving an inheritance, probability of leaving an inheritance of at least $10,000, number of children ever born, African-American race, and gender.

Discussion: Some of the machine learning methods do not improve prediction or fit beyond simpler models, however, neural networks performed well. The predictors identified across models suggest underlying social factors that are important predictors of biological indicators of chronic disease, and that the non-linear and interactive relationships between variables fundamental to the neural network approach may be important to consider.

%B SSM Popul Health %V 4 %P 95-99 %G eng %1 http://www.ncbi.nlm.nih.gov/pubmed/29349278?dopt=Abstract %R 10.1016/j.ssmph.2017.11.008 %0 Journal Article %J PLOS ONE %D 2018 %T Poverty dynamics, poverty thresholds and mortality: An age-stage Markovian model %A Bernstein, Shayna Fae %A David Rehkopf %A Tuljapurkar, Shripad %A Horvitz, Carol C. %E Komarova, Natalia L. %K Life Expectancy %K Mortality %K Poverty %K Socioeconomic factors %X Recent studies have examined the risk of poverty throughout the life course, but few have considered how transitioning in and out of poverty shape the dynamic heterogeneity and mortality disparities of a cohort at each age. Here we use state-by-age modeling to capture individual heterogeneity in crossing one of three different poverty thresholds (defined as 1×, 2× or 3× the “official” poverty threshold) at each age. We examine age-specific state structure, the remaining life expectancy, its variance, and cohort simulations for those above and below each threshold. Survival and transitioning probabilities are statistically estimated by regression analyses of data from the Health and Retirement Survey RAND data-set, and the National Longitudinal Survey of Youth. Using the results of these regression analyses, we parameterize discrete state, discrete age matrix models. We found that individuals above all three thresholds have higher annual survival than those in poverty, especially for mid-ages to about age 80. The advantage is greatest when we classify individuals based on 1× the “official” poverty threshold. The greatest discrepancy in average remaining life expectancy and its variance between those above and in poverty occurs at mid-ages for all three thresholds. And fewer individuals are in poverty between ages 40-60 for all three thresholds. Our findings are consistent with results based on other data sets, but also suggest that dynamic heterogeneity in poverty and the transience of the poverty state is associated with income-related mortality disparities (less transience, especially of those above poverty, more disparities). This paper applies the approach of age-by-stage matrix models to human demography and individual poverty dynamics. In so doing we extend the literature on individual poverty dynamics across the life course. %B PLOS ONE %V 13 %P e0195734 %8 Apr-05-2019 %G eng %U http://dx.doi.org/10.1371/journal.pone.0195734 %N 5 %! PLoS ONE %R 10.1371/journal.pone.0195734 %0 Journal Article %J EBioMedicine %D 2016 %T Racial and Socioeconomic Variation in Genetic Markers of Telomere Length: A Cross-Sectional Study of U.S. Older Adults. %A Hamad, Rita %A Tuljapurkar, Shripad %A David Rehkopf %K Age Factors %K Aged %K Aged, 80 and over %K Alleles %K Cross-Sectional Studies %K Ethnic Groups %K Female %K Gene Frequency %K Genetic Markers %K Genome-Wide Association Study %K Geriatric Assessment %K Humans %K Male %K Middle Aged %K Polymorphism, Single Nucleotide %K Population Surveillance %K Socioeconomic factors %K Telomere Homeostasis %K United States %X

BACKGROUND: Shorter telomere length (TL) has been associated with stress and adverse socioeconomic conditions, yet U.S. blacks have longer TL than whites. The role of genetic versus environmental factors in explaining TL by race and socioeconomic position (SEP) remains unclear.

METHODS: We used data from the U.S. Health and Retirement Study (N=11,934) to test the hypothesis that there are differences in TL-associated SNPs by race and SEP. We constructed a TL polygenic risk score (PRS) and examined its association with race/ethnicity, educational attainment, assets, gender, and age.

RESULTS: U.S. blacks were more likely to have a lower PRS for TL, as were older individuals and men. Racial differences in TL were statistically accounted for when controlling for population structure using genetic principal components. The GWAS-derived SNPs for TL, however, may not have consistent associations with TL across different racial/ethnic groups.

CONCLUSIONS: This study showed that associations of race/ethnicity with TL differed when accounting for population stratification. The role of race/ethnicity for TL remains uncertain, however, as the genetic determinants of TL may differ by race/ethnicity. Future GWAS samples should include racially diverse participants to allow for better characterization of the determinants of TL in human populations.

%B EBioMedicine %V 11 %P 296-301 %G eng %R 10.1016/j.ebiom.2016.08.015 %0 Book Section %B Allocating Public and Private Resources across Generations %D 2006 %T Consequences of Educational Change for the Burden of Chronic Health Problems in the Population %A Mark D Hayward %A Eileen M. Crimmins %A Zhang, Zhenmei %E Gauthier, Anne H. %E Chu, Cyrus %E Tuljapurkar, Shripad %K Educational attainment %K Educational Change %K Functional Limitation %K Functional Problem %K Life Table %X Changes in the public and individual burden of chronic health problems have significant implications for the allocation of public and private resources across generations. Preston (1984) noted almost two decades ago that population ageing in the United States was accompanied by the rapid expansion of public programs benefiting the health of elderswhile public programs benefiting children’s education contracted. Health care is the principal public service provided to the elderly while education is the counterpart for children. Within a historical time period, political choices about the funding of age-targeted service programs have an urgency that oftentimes sweeps aside the fact that investments in children’s well-being pay substantial dividends decades later when children become the elders of a population. In large part, this reflects a lack of attention both by policy makers and by demographers of these long-run associations. Here, we provide new insights into the longrun consequences of investments in children for the burden of chronic health problems by conducting a thought experiment in which we simulate how sweeping historical changes in a population’s educational achievement potentially alters active life expectancy and the prevalence of functioning problems in the population. %B Allocating Public and Private Resources across Generations %I Oxford University Press %C Oxford, UK %P 227-242 %@ 978-1-4020-4480-9 %G eng %L pubs_2005_Hayward_etal.pdf %4 Education/Health Status/Life Expectancy %$ 10502 %+ National Institute on Aging and the National Institute on Child Health and Human Development; Revision of Penn State Population Research Insitute Working Paper 02-02 %R https://doi.org/10.1007/978-1-4020-4481-6_9