Skip to main content

Advertisement

SpringerLink
Log in

A Simulation-Based Assessment of Alternative Explanations for Apparent Confounding in “PM Decomposition” Studies

  • Published:
Environmental Modeling & Assessment Aims and scope Submit manuscript

Abstract

Determining whether there is a causal link between chronic exposure to PM2.5 (atmospheric particulate matter with a diameter of 2.5 µm or less) and mortality is a fundamental challenge in determining appropriate air quality standards for PM2.5. While numerous epidemiological studies have found a statistical association between PM2.5 and mortality, unmeasured confounding in this relationship remains a concern. Several recent studies have examined the possibility of unmeasured confounding of the long-term association between PM2.5 exposure and mortality by decomposing PM2.5 into two orthogonal components — a “global” component that measures the national trend in pollution, and a “local” component that measures the local trend in pollution after controlling for the national trend. Generally, these “PM decomposition” studies find that while PM2.5 and mortality are trending downward over time at the national level, areas with steeper declines in PM2.5 do not have correspondingly steep declines in mortality. This finding is consistent with the long-term relationship between PM2.5 and mortality being confounded by some other, unmeasured, long-term trends. However, alternative explanations for these findings have been proposed under which there is still a causal link between PM2.5 and mortality. This study conducts simulation-based tests of four of these proposed alternative explanations — the omission of spatial variation in PM2.5 and mortality, confounding at the local level, measurement error, and an association between PM2.5 and mortality at a different time scale than those tested in the PM decomposition studies. We find that none of these alternative explanations can reproduce the results in the PM decomposition studies while simultaneously allowing for a causal link between PM2.5 and mortality.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14
Fig. 15
Fig. 16
Fig. 17
Fig. 18

Similar content being viewed by others

Data and code availability

The simulated datasets and the code to analyze them are available from the corresponding author on reasonable request.

Notes

  1. This would have been a futile task because along with such demographic features, one would also need to have the actual survival outcomes of those populations, in which data are not publicly available. Our simulations instead assume no demographic heterogeneity from location to location other than the simulated effect on mortality risk of the assumed C-R relationship, as described further below.

  2. Since the equation defining the C-R function is on a log-linear scale, we employ log hazard ratios to characterize PM2.5 mortality risk as inputs to our simulation.

  3. Each of the 10 simulated datasets is equivalent to the single dataset that would be available to an epidemiological study based on observed data. Thus, we do not consider multiple testing bias in our simulations (see Bender and Lange [32]).

  4. When the hazard ratio is quite small, as is the case in the PM2.5 literature, the logarithm of the hazard ratio is approximately equal to the hazard ratio minus 1, or the percent increase (i.e., 0.01 if the relative risk is 1.01).

  5. Measurement errors with such a standard deviation imply that the true PM2.5 exposures may differ from the estimated levels with a 95% range of about 4 µg/m3. This is very large given that the average PM2.5 level for 2000–2006 across all the data used by Greven et al. [8] is only about 12.9 µg/m3 (we did truncate measurement errors when they fell below 0, but the main point is that standard deviations above 2 µg/m3 are very large, and perhaps unrealistically large).

  6. In the simulations which involve no systemic bias and an untruncated error term, bifurcation occurs at large standard errors, and the local coefficient becomes totally attenuated and not significantly greater than 0. These simulations, however, allow for implausible magnitudes of error with outcomes that include negative PM2.5 values and month-to-month changes (measured in terms of the annual average of daily PM2.5 values) that would require the most recent month to have average PM2.5 concentrations to be greater than 40 µg/m3 compared to the PM2.5 concentrations from 13 months prior. Thus, the results from these simulations are not presented in this paper.

References

  1. Dockery, D. W., Pope, C. A., Xu, X., Spengler, J. D., Ware, J. H., Fay, M. E., Ferris, B. G., Jr., & Speizer, F. E. (1993). An association between air pollution and mortality in six US cities. New England Journal of Medicine., 329(24), 1753–1759.

    Article  CAS  Google Scholar 

  2. Eftim, S., Samet, J., Janes, H., McDermott, A., & Dominici, F. (2008). Fine particulate matter and mortality: A comparison of the six cities and American Cancer Society cohorts with a Medicare cohort. Epidemiology, 19(2), 209–216.

    Article  Google Scholar 

  3. Laden, F., Schwartz, J., Speizer, F., & Dockery, D. (2006). Reduction in fine particulate air pollution and mortality: Extended follow-up of the Harvard six cities study. American Journal of Respiratory and Critical Care Medicine, 173(6), 667–672.

    Article  CAS  Google Scholar 

  4. Lepeule, J., Laden, F., Dockery, D., & Schwartz, J. (2012). Chronic exposure to fine particles and mortality: An extended follow-up of the Harvard six cities study from 1974 to 2009. Environmental Health Perspectives, 120(7), 965–970.

    Article  Google Scholar 

  5. Pope, C. A. III, Thun, M. J., Namboodiri, M. M., Dockery, D. W., Evans, J. S., Speizer, F. E., & Heath, C. W. (1995). Particulate air pollution as a predictor of mortality in a prospective study of U.S. adults. American Journal of Respiratory and Critical Care Medicine, 151(3), 669–674.

  6. Pope, C. A., Ezzati, M., & Dockery, D. W. (2009). Fine-particulate air pollution and life expectancy in the United States. New England Journal of Medicine, 360, 376–386.

    Article  CAS  Google Scholar 

  7. Qian, DI., Wang, Y., Zanobetti, A., Wang, Y., Koutrakis, P., Choirat, C., Dominici, F., & Schwartz, J. D. (2017). Air pollution and mortality in the Medicare population. New England Journal of Medicine, 376, 2513–2522.

    Article  Google Scholar 

  8. Greven, S., Dominici, F., & Zeger, S. L. (2011). An approach to the estimation of chronic air pollution effects using spatio‐temporal information. Journal of the American Statistical Association. J Am Stat Assoc. 106(494), 396‐406.

  9. Janes, H., Dominici, F., & Zeger, S. L. (2007). Partitioning evidence of association between air pollution and mortality (author’s response). Epidemiology, 18(4), 427–428.

    Article  Google Scholar 

  10. Janes, H., Dominici, F., & Zeger, S. L. (2007). Trends in air pollution and mortality: An approach to the assessment of unmeasured confounding. Epidemiology, 18(4), 416–423.

    Article  Google Scholar 

  11. Moolgavkar, S. (1994). Air pollution and mortality. New England Journal of Medicine, 330(17), 1237–1238.

    Article  CAS  Google Scholar 

  12. Vedal, S. (1997). Ambient particles and health: Lines that divide. Journal of the Air & Waste Management Association, 47(5), 551–581.

    Article  CAS  Google Scholar 

  13. Gamble, J. (1998). PM2.5 and mortality in long-term prospective cohort studies: Cause-effect or statistical associations? Environmental Health Perspectives, 106(9), 535–549.

  14. Wang, B., Eum, K., Kazemiparkouhi, F., Cheng, Li., Manijourides, J., Pavlu, V., & Suh, H. (2020). The impact of long-term PM2.5 exposure on specific causes of death: exposure-response curves and effect modification among 53 million U.S. Medicare beneficiaries. Environmental Health, 19(20).

  15. Schwartz, J., & Dockery, D. (1992). Particulate air pollution and daily mortality in Steubenville. Ohio. American Journal of Epidemiology, 135(1), 12–19.

    Article  CAS  Google Scholar 

  16. Spix, C., Heinrich, J., Dockery, D., Schwartz, J., Völksch, G., Schwinkowski, K., Cöllen, C., & Wichmann, H. (1993). Air pollution and daily mortality in Erfurt, East Germany, 1980–1989. Environmental Health Perspectives, 101(6), 518–526.

    Article  CAS  Google Scholar 

  17. Kelsall, J., Samet, J., Zeger, S. L., & Xu, J. (1997). Air pollution and mortality in Philadelphia, 1974–1988. American Journal of Epidemiology, 146(9), 750–762.

    Article  CAS  Google Scholar 

  18. Lumley, T., & Sheppard, L. (2000). Assessing seasonal confounding and model selection bias in air pollution epidemiology using positive and negative control analyses. Environmetrics, 11(6), 705–717.

    Article  CAS  Google Scholar 

  19. Moolgavkar, S. H. (2005). A review and critique of the EPA’s rationale for a fine particle standard. Regulatory Toxicology and Pharmacology, 42(1), 123–144.

    Article  CAS  Google Scholar 

  20. Eum, K., Suh, H., Pun, V. C., & Manjourides, J. (2018). Impact of long-term temporal trends in fine particulate matter (PM2.5) on association of annual PM2.5 exposure and mortality: an analysis of over 20 million Medicare beneficiaries. Environmental Epidemiology, 2(2):e009.

  21. Pun, V. C., Kazemiparkouhi, F., Manijourides, J., & Suh, H. (2017). Long-term PM2.5 exposure and respiratory, cancer, and cardiovascular mortality in older US adults. American Journal of Epidemiology, 186(8), 961–969.

  22. Dominici, F., Zeger, S. L., Janes, H., & Greven, S. (2012). Reply to: Additional information to EPA for these two studies. Memo dated November 28, 2012.

  23. Pope, C. A., & Burnett, R. T. (2007). Confounding in air pollution epidemiology: The broader context. Epidemiology, 18(4), 424–426.

    Article  Google Scholar 

  24. Cox, L. A. (2017). Socioeconomic and air pollution correlates of adult asthma, heart attack, and stroke risks in the United States, 2010–2013. Environmental Research, 155, 92–107.

    Article  CAS  Google Scholar 

  25. Cox, L. A. (2021). Socioeconomic correlates of air pollution and heart disease. In Quantitative risk analysis of air pollution health effects (pp. 357–372). International Series in Operations Research and Management Science, Springer.

  26. Moolgavkar, S. H., Chang, E. T., Watson, H. N., & Lau, E. C. (2018). An assessment of the Cox proportional hazards regression model for epidemiologic studies. Risk Analysis, 38(4), 777–794.

    Article  Google Scholar 

  27. Butland, B. K., Armstrong, B., Atkinson, R. W., Wilkinson, P., Heal, M. R., Doherty, R. M., & Vieno, M. (2013). Measurement error in time-series analysis: A simulation study comparing modelled and monitored data. BMC Medical Research Methodology, 13(136), 1–12.

    Google Scholar 

  28. Cox, L. A. (2018). Effects of exposure estimation errors on estimated exposure-response relations for PM2.5. Environmental Research, 164, 636–646.

    Article  CAS  Google Scholar 

  29. Cox, L. A. (2021). How do exposure estimation errors affect estimated exposure-response relations? In Quantitative Risk Analysis of Air Pollution Health Effects (pp. 449–474). International Series in Operations Research and Management Science, Springer.

  30. Dionisio, K. L., Chang, H. H., & Baxter, L. K. (2016). A simulation study to quantify the impacts of exposure measurement error on air pollution health risk estimates in copollutant time-series models. Environmental Health, 15(114), 1–10.

    Google Scholar 

  31. Rhomberg, L. R., Chandalia, J. K., Long, C. M., & Goodman, J. E. (2011). Measurement error in environmental epidemiology and the shape of exposure-response curves. Critical Reviews in Toxicology, 41(8), 651–671.

    Article  Google Scholar 

  32. Bender, R., & Lange, S. (2001). Adjusting for multiple testing – When and how? Journal of Clinical Epidemiology, 54(4), 343–349.

    Article  CAS  Google Scholar 

  33. Zeger, S. L., Thomas, D., Dominici, F., Samet, J. M., Schwartz, J., Dockery, D., & Cohen, A. (2000). Exposure measurement error in time-series studies of air pollution: Concepts and consequences. Environmental Health Perspectives, 108(5), 419–426.

    Article  CAS  Google Scholar 

  34. U.S. E.P.A. (2004). Advisory Council on Clean Air Compliance Analysis response to agency request on cessation lag, EPA-COUNCIL-LTR-05–001, December.

  35. Bell, M., Dominici, F., Ebisu, K., Zeger, S., & Samet, J. (2007). Spatial and temporal variation in PM2.5 chemical composition in the United States for health effects studies. Environmental Health Perspectives, 115(7), 989–995.

Download references

Funding

This work was funded by the Texas Commission on Environmental Quality (TCEQ).

Author information

Authors and Affiliations

  1. The Brattle Group, Washington, D.C., USA

    Wonjun Chang

  2. NERA Economic Consulting, Washington, D.C., USA

    Garrett Glasgow, Bharat Ramkrishnan & Anne E. Smith

Authors
  1. Wonjun Chang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Garrett Glasgow
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Bharat Ramkrishnan
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Anne E. Smith
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Anne E. Smith developed the simulation approach. Wonjun Chang, Garrett Glasgow, and Bharat Ramkrishnan contributed to the development of the simulation code. Wonjun Chang, Garrett Glasgow, Bharat Ramkrishnan, and Anne E. Smith contributed to the writing of the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Garrett Glasgow.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Chang, W., Glasgow, G., Ramkrishnan, B. et al. A Simulation-Based Assessment of Alternative Explanations for Apparent Confounding in “PM Decomposition” Studies. Environ Model Assess 27, 665–692 (2022). https://doi.org/10.1007/s10666-022-09829-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10666-022-09829-2

Keywords

Search

Navigation