Advertisement

A Machine-Learning Algorithm for Estimating and Ranking the Impact of Environmental Risk Factors in Exploratory Epidemiological Studies

  • Jessica G. Young
  • Alan E. HubbardEmail author
  • Brenda Eskenazi
  • Nicholas P. Jewell
Chapter
  • 39 Downloads

Abstract

Epidemiological research, such as the identification of disease risks attributable to environmental chemical exposures, is often hampered by small population effects, large measurement error, and limited a priori knowledge regarding the complex relationships between the many chemicals under study. However, even an ideal study design does not preclude the possibility of reported false positive exposure effects due to inappropriate statistical methodology. Three issues often overlooked include (1) definition of a meaningful measure of association; (2) use of model estimation strategies (such as machine-learning) that acknowledge that the true data-generating model is unknown; (3) accounting for multiple testing. In this paper, we propose an algorithm designed to address each of these limitations in turn by combining recent advances in the causal inference and multiple-testing literature along with modifications to traditional nonparametric inference methods.

Keywords

Machine-learning Epidemiology Multiple testing Causal inference 

References

  1. 1.
    Chen, J., van der Laan, M. J., Smith, M. T., & Hubbard, A. E. (2007). A comparison of methods to control type I errors in microarray studies. Statistical Applications in Genetics and Molecular Biology, 6, Article 28.Google Scholar
  2. 2.
    Chevrier, J., Eskenazi, B., Holland, N., Bradman, A., & Barr, D. B. (2008). Effect of exposure to polychlorinated biphenyls and organochlorine pesticides on thyroid function during pregnancy. American Journal of Epidemiology, 68, 298–310.CrossRefGoogle Scholar
  3. 3.
    Dudoit, S., van der Laan, M. J., & Pollard, K. S. (2004). Multiple testing, part I. Single-step procedures for control of general type I error rates. Statistical Applications in Genetics and Molecular Biology, 3, Article 11.Google Scholar
  4. 4.
    Eskenazi, B., Marks, A. R., Brandman, A., Fenster, L., Johnson, C., Barr, D. B., et al. (2006). In utero exposure to dichlorodiphenyltrichloroethane (DDT) and dichlorodiphenyldichloroethylene (DDE) and neurodevelopment among young Mexican American children. Pediatrics, 118, 233–41.CrossRefGoogle Scholar
  5. 5.
    Fenster, L., Eskenazi, B., Anderson, M., Bradman, A., Harley, K., Hernandez, H., Hubbard, A., Barr, D.B., (2005). Association of in utero organochlorine pesticide exposure and fetal growth and length of gestation in an agricultural population. Environmental health perspectives, 114(4), pp. 597–602.CrossRefGoogle Scholar
  6. 6.
    Hubbard, A. E., & van der Laan, M. L. (2008). Population intervention models. Biometrika, 95, 35–47.MathSciNetCrossRefGoogle Scholar
  7. 7.
    Ioannidis, J. P. A. (2005). Why most published research findings are false. PLoS Medicine, 2, e124.CrossRefGoogle Scholar
  8. 8.
    Little, R. J., & Rubin, D. B. (2002). Statistical analysis with missing data (2nd ed.). New York: Wiley.CrossRefGoogle Scholar
  9. 9.
    Patel, C. J., Bhattacharya, J., & Butte, A. J. (2010). An environment-wide association study (EWAS) on type 2 diabetes mellitus. PLoS One, 5, e10746.CrossRefGoogle Scholar
  10. 10.
    Pollard, K. S., & van der Laan, M. J. (2003). Resampling-based multiple testing: asymptotic control of type I error and applications to gene expression data. In Division of biostatistics, Technical Report No. 121, University of California, Berkeley.Google Scholar
  11. 11.
    Robins, J. M. (1998). Marginal structural models. In 1997 Proceedings of the American Statistical Association, Section on Bayesian Statistical Science (pp. 1–10). Alexandria: American Statistical Association.Google Scholar
  12. 12.
    Rosenbaum, P. R. (1984). Conditional permutation tests and the propensity score in observational studies. Journal of the American Statistical Association, 79, 565–574.MathSciNetCrossRefGoogle Scholar
  13. 13.
    Rothman, K. J. (1990). No adjustments are needed for multiple comparisons. Epidemiology, 1, 43–6.CrossRefGoogle Scholar
  14. 14.
    Rubin, D. B. (1986). Statistics and causal inference: Comment: Which ‘ifs’ have causal answers. Journal of the American Statistical Association, 81, 961–962.Google Scholar
  15. 15.
    Sinisi, S. E., & van der Laan, M. J. (2004). Loss-based cross-validated Deletion/Substitution/Addition algorithms in estimation. In Division of biostatistics, Technical Report No. 143, University of California, Berkeley.Google Scholar
  16. 16.
    Stitelman, O. M., Hubbard, A. E., & Jewell, N. P. (2010). The impact of coarsening the explanatory variable of interest in making causal inferences: Implicit assumptions behind dichotomizing variables. U.C. Berkeley Division of Biostatistics Working Paper Series. Working Paper 264.Google Scholar
  17. 17.
    Succop, P. A., Clark, S., Chen, M., & Galke, W. (2004). Imputation of data values that are less than a detection limit. Journal of Occupational and Environmental Hygiene, 1, 436–441.CrossRefGoogle Scholar
  18. 18.
    United States Census Bureau (2000). Poverty thresholds 2000, current population Survey. www.census.gov/hhes/poverty/poverty00/pv00thrs.html
  19. 19.
    van der Laan, M. J., & Gruber, S. (2010). Collaborative double robust targeted maximum likelihood estimation. The International Journal of Biostatistics, 6, Article 17.Google Scholar
  20. 20.
    van der Laan, M. J., & Hubbard, A. E. (2006). Quantile-function based null distribution in resampling based multiple testing. Statistical Applications in Genetics and Molecular Biology, 5, Article 14.Google Scholar
  21. 21.
    van der Laan, M. J., Hubbard, A. E., & Jewell, N. (2010). Learning from data: semiparametric models versus faith-based inference. Epidemiology, 21, 479–81.CrossRefGoogle Scholar
  22. 22.
    van der Laan, M. J., & Petersen, M. (2007). Causal effect models for realistic individualized treatment and intention to treat rules. The International Journal of Biostatistics, 3, Article 3.Google Scholar
  23. 23.
    van der Laan, M. J., Polley, E. C., & Hubbard, A. E. (2007). Super Learner. U.C. Berkeley Division of Biostatistics Working Paper Series. Working Paper 222.Google Scholar
  24. 24.
    van der Laan, M. J., & Robins, J. M. (2003). Unified methods for censored longitudinal data and causality. New York: Springer.CrossRefGoogle Scholar
  25. 25.
    van der Laan, M. J., & Rubin, D. B. (2006). Targeted maximum likelihood learning. The International Journal of Biostatistics, 2, Article 11.Google Scholar
  26. 26.
    Wang, Y., Petersen, M. L., Bangsberg, D., & van der Laan, M. J. (2006). Diagnosing bias in the inverse probability of treatment weighted estimator resulting from violation of experimental treatment assignment. In Division of biostatistics, Technical Report No. 211, University of California, Berkeley.Google Scholar
  27. 27.
    Young, J., Hubbard, A. E., Eskenazi, B., & Jewell, N. P. (2009). A machine-learning algorithm for estimating and ranking the impact of environmental risk factors in exploratory epidemiological studies. In Division of biostatistics, Technical Report No. 250, University of California, Berkeley.Google Scholar
  28. 28.
    Young, J. G., Logan, R. W., Robins, J. M., & Hernán, M. A. (2019). Inverse probability weighted estimation of risk under representative interventions in observational studies. Journal of the American Statistical Association, 114, 938–947.MathSciNetCrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2020

Authors and Affiliations

  • Jessica G. Young
    • 1
  • Alan E. Hubbard
    • 2
    Email author
  • Brenda Eskenazi
    • 3
  • Nicholas P. Jewell
    • 2
    • 4
  1. 1.Department of Population Medicine, Harvard Medical School & Harvard Pilgrim Health Care InstituteHarvard UniversityBostonUSA
  2. 2.Division of BiostatisticsUniversity of California at BerkeleyBerkeleyUSA
  3. 3.Division of Environmental Health Sciences, School of Public HealthUniversity of California at BerkeleyBerkeleyUSA
  4. 4.London School of Hygiene & Tropical MedicineLondonUK

Personalised recommendations