Fine Particulate Air Pollution and the "No-Multiple-Versions-of-Treatment" Assumption: Does Particle Composition Matter for Causal Inference?

Epidemiological studies likely need to consider PM2.5 composition even if total outdoor PM2.5 mass concentration is the exposure of interest

Background

Outdoor fine particulate air pollution, <2.5 µm (PM2.5) mass concentrations can be constructed through many different combinations of chemical components that have varying levels of toxicity. This poses a challenge for studies interested in estimating the health effects of total outdoor PM2.5 (i.e., how much PM2.5 mass is present in the air regardless of composition) because we must consider possible confounders of the version of treatment-outcome relationships.

Methods

We evaluated the extent of possible bias in mortality hazard ratios for total outdoor PM2.5 by examining models with and without adjustment for sulfate and nitrate in PM2.5 as examples of potential confounders of version of treatment-outcome relationships. Our study included approximately 3 million Canadians and Cox proportional hazard models were used to estimate hazard ratios for total outdoor PM2.5 adjusting for sulfate and/or nitrate and other relevant covariates.

Results

Hazard ratios for total outdoor PM2.5 and nonaccidental, cardiovascular, and respiratory mortality were overestimated due to the confounding version of treatment-outcome relationships, and associations for lung cancer mortality were underestimated. Sulfate was most strongly associated with nonaccidental, cardiovascular, and respiratory mortality suggesting that regulations targeting this specific component of outdoor PM2.5 may have greater health benefits than interventions targeting total PM2.5.

Conclusions

Studies interested in estimating the health impacts of total outdoor PM2.5 (i.e., how much PM2.5 mass is present in the air) need to consider potential confounders of the version of treatment-outcome relationships. Otherwise, health risk estimates for total PM2.5 will reflect some unknown combination of how much PM2.5 mass is present in the air and the kind of PM2.5 mass that is present.

Using Parametric g-Computation for Time-to-Event Data and Distributed Lag Models to Identify Critical Exposure Windows for Preterm Birth: An Illustrative Example Using PM2.5 in a Retrospective Birth Cohort Based in Eastern Massachusetts (2011-2016)

BACKGROUND

Parametric g-computation is an attractive analytic framework to study the health effects of air pollution. Yet, the ability to explore biologically relevant exposure windows within this framework is underdeveloped.

OBJECTIVES

We outline a novel framework for how to incorporate complex lag-responses using distributed lag models (DLMs) into parametric g-computation analyses for survival data. We call this approach "g-survival-DLM" and illustrate its use examining the association between PM 2.5 during pregnancy and the risk of preterm birth (PTB).

METHODS

We applied the g-survival-DLM approach to estimate the hypothetical static intervention of reducing average PM 2.5 in each gestational week by 20% on the risk of PTB among 9,403 deliveries from Beth Israel Deaconess Medical Center, Boston, Massachusetts, 2011-2016. Daily PM 2.5 was taken from a 1 -km grid model and assigned to address at birth. Models were adjusted for sociodemographics, time trends, nitrogen dioxide, and temperature. To facilitate implementation, we provide a detailed description of the procedure and accompanying R syntax.

RESULTS

There were 762 (8.1%) PTBs in this cohort. The gestational week-specific median PM 2.5 concentration was relatively stable across pregnancy at ∼ 7 μ g / m 3 . We found that our hypothetical intervention strategy changed the cumulative risk of PTB at week 36 (i.e., the end of the preterm period) by - 0.009 (95% confidence interval: - 0.034 , 0.007) in comparison with the scenario had we not intervened, which translates to about 86 fewer PTBs in this cohort. We also observed that the critical exposure window appeared to be weeks 5-20.

DISCUSSION

We demonstrate that our g-survival-DLM approach produces easier-to-interpret, policy-relevant estimates (due to the g-computation); prevents immortal time bias (due to treating PTB as a time-to-event outcome); and allows for the exploration of critical exposure windows (due to the DLMs). In our illustrative example, we found that reducing fine particulate matter [particulate matter (PM) with aerodynamic diameter ≤ 2.5 μ m (PM 2.5 )] during gestational weeks 5-20 could potentially lower the risk of PTB. https://doi.org/10.1289/EHP13891.

Using Parametric g-Computation to Estimate the Effect of Long-Term Exposure to Air Pollution on Mortality Risk and Simulate the Benefits of Hypothetical Policies: The Canadian Community Health Survey Cohort (2005 to 2015)

BACKGROUND

Numerous epidemiological studies have documented the adverse health impact of long-term exposure to fine particulate matter [particulate matter ≤2.5μm in aerodynamic diameter (PM2.5)] on mortality even at relatively low levels. However, methodological challenges remain to consider potential regulatory intervention's complexity and provide actionable evidence on the predicted benefits of interventions. We propose the parametric g-computation as an alternative analytical approach to such challenges.

METHOD

We applied the parametric g-computation to estimate the cumulative risks of nonaccidental death under different hypothetical intervention strategies targeting long-term exposure to PM2.5 in the Canadian Community Health Survey cohort from 2005 to 2015. On both relative and absolute scales, we explored the benefits of hypothetical intervention strategies compared with the natural course that a) set the simulated exposure value at each follow-up year to a threshold value if exposure was above the threshold (8.8 μg/m3, 7.04 μg/m3, 5 μg/m3, and 4 μg/m3), and b) reduced the simulated exposure value by a percentage (5% and 10%) at each follow-up year. We used the 3-y average PM2.5 concentration with 1-y lag at the postal code of respondents' annual mailing addresses as their long-term exposure to PM2.5. We considered baseline and time-varying confounders, including demographics, behavior characteristics, income level, and neighborhood socioeconomic status. We also included the R syntax for reproducibility and replication.

RESULTS

All hypothetical intervention strategies explored led to lower 11-y cumulative mortality risks than the estimated value under the natural course without intervention, with the smallest reduction of 0.20 per 1,000 participants (95% CI: 0.06, 0.34) under the threshold of 8.8 μg/m3, and the largest reduction of 3.40 per 1,000 participants (95% CI: -0.23, 7.03) under the relative reduction of 10% per interval. The reductions in cumulative risk, or numbers of deaths that would have been prevented if the intervention was employed instead of maintaining the status quo, increased over time but flattened toward the end of the follow-up period. Estimates among those ≥65 years of age were greater with a similar pattern. Our estimates were robust to different model specifications.

DISCUSSION

We found evidence that any intervention further reducing the long-term exposure to PM2.5 would reduce the cumulative mortality risk, with greater benefits in the older population, even in a population already exposed to low levels of ambient PM2.5. The parametric g-computation used in this study provides flexibilities in simulating real-world interventions, accommodates time-varying exposure and confounders, and estimates adjusted survival curves with clearer interpretation and more information than a single hazard ratio, making it a valuable analytical alternative in air pollution epidemiological research. https://doi.org/10.1289/EHP11095.