You are here

Identifying the shape of the association between long-term exposure to low levels of ambient air pollution and the risk of mortality: An extension of the Canadian Census Health and Environment Cohort using innovative data linkage and exposure methodology

Principal Investigator: 

The University of British Columbia, Canada

This study will investigate mortality effects of low levels of air pollution in Canada using Canadian Census data from about 6 million people. The shape of the exposure-response function will be characterized using newly developed flexible non-linear exposure-response functions. See also this Program Summary of HEI's research program on low levels of air pollution. 

Funded under
In review

Poster Abstract for HEI Annual Conference 2022

The shape of the PM2.5-mortality exposure-response relationship in the MAPLE, ELAPSE and Medicare cohorts. What have we learned from the studies of exposure to low levels?

Michael Brauer1,2 Francesca Dominici3, Bert Brunekreef4, on behalf of the MAPLE, Harvard (Medicare) and ELAPSE and teams

1.         The University of British Columbia, Vancouver, British Columbia, Canada

2.         Institute for Health Metrics and Evaluation, University of Washington, Seattle, Washington, USA

3.         Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA

4.         Institute for Risk Assessment Sciences, Utrecht University, Utrecht, Netherlands

Background. With widespread evidence of the mortality risk associated with long-term exposure to PM2.5 there is urgency to reduce ambient concentrations. Evaluating the benefits of reductions in concentrations depends upon understanding both the magnitude of the PM2.5 mortality risk and the shape of this relationship. We present results on the shape of the relationship between PM2.5 and all-cause mortality based on harmonized exposure assessment and statistical modeling approaches applied to three large projects in Europe and North America.

Methods. The study populations include all Medicare enrollees from 2000 to 2016 (N = 73 million), census-based Canadian cohorts (“stacked” 1991, 1996, 2001 CanCHEC, N =7.1 million), and six European administrative cohorts followed up from 2000 to 2016 (N = 26 million). Satellite-based estimates of annual average PM2.5 concentrations were linked to participants’ residential addresses. Cox proportional hazard models incorporating PM2.5 as time-varying exposure (the Medicare cohort applied an equivalent Poisson additive model) estimated all-cause mortality risk for those ages 65+ with harmonized covariates (age, sex, calendar years/year of enrollment; individual and area level socioeconomic status and ethnicity, regions). Exposure-response relationships were evaluated by 1) restricting analyses to person-years with concentrations below 5, 7.5, 10 and 12 µg/m3 2) applying the extended Shape-Constrained Health Impact Function (eSCHIF) which relates risk to concentration in a form suitable for quantitative benefits analysis. 

Results. In the Medicare cohort, positive associations between PM2.5 exposure and all-cause mortality were observed in all restricted cohort analyses. Using the eSCHIF the shape was slightly supralinear, with an increase at the lowest concentrations, a plateau between 4-7 µg/m3 and another increase at higher concentrations. In MAPLE, positive associations were observed in the full cohort and when restricting to person-years <12 µg/m3.  Using the eSCHIF there was a steep supralinear relationship from 2 – 5 µg/m3, a flat relationship to ~8 µg/m3 and a near linear increase at higher concentrations. Within ELAPSE there were consistent positive associations only in the full Swiss cohort, in analyses excluding person-years > 7.5 µg/m3 and >5 µg/m3 in the Danish and Norwegian cohorts, respectively. eSCHIF shapes indicated supralinear relationships in the Danish cohort, sub-linear relationships in the Norwegian and Rome cohorts and flat and somewhat protective relationships in the Swiss cohort.

Conclusions. There was evidence of associations between PM2.5 and all-cause mortality even after restricting analyses to person-years exposed to < 5 µg/m3, although results differed between cohorts.