Using Distributed Lag Non-Linear Models to Estimate Exposure Lag-Response Associations between Long-Term Air Pollution Exposure and Incidence of Cardiovascular Disease.

Abstract

Long-term air pollution exposure increases the risk for cardiovascular disease, but little is known about the temporal relationships between exposure and health outcomes. This study aims to estimate the exposure-lag response between air pollution exposure and risk for ischemic heart disease (IHD) and stroke incidence by applying distributed lag non-linear models (DLNMs). Annual mean concentrations of particles with aerodynamic diameter less than 2.5 µm (PM2.5) and black carbon (BC) were estimated for participants in five Swedish cohorts using dispersion models. Simultaneous estimates of exposure lags 1–10 years using DLNMs were compared with separate year specific (single lag) estimates and estimates for lag 1–5- and 6–10-years using moving average exposure. The DLNM estimated no exposure lag-response between PM2.5 total, BC, and IHD. However, for PM2.5 from local sources, a 20% risk increase per 1 µg/m3 for 1-year lag was estimated. A risk increase for stroke was suggested in relation to lags 2–4-year PM2.5 and BC, and also lags 8–9-years BC. No associations were shown in single lag models. Increased risk estimates for stroke in relation to lag 1–5- and 6–10-years BC moving averages were observed. Estimates generally supported a greater contribution to increased risk from exposure windows closer in time to incident IHD and incident stroke.