Projecting burden of non-communicable diseases (NCD) attributable to air pollution in the northeast and northwest of Peninsular Malaysia under climate change scenario

Abstract

Air pollution silently claims over 7 million lives annually by triggering cardiovascular (CVD), respiratory diseases (RD) and diabetes mellitus (DM). This study aims to project BoD of NCDs attributable to air pollution in Kelantan and Kedah under SSP5- 8.5 scenario. This study involves daily data on mortality and hospital admissions (HA) for CVD, RD, and DM (2005-2021); air pollution (PM10, PM2.5, SO2, NO2, CO, O3) and meteorological variables (temperature, relative humidity) (2005-2021); and WRF- CMAQ air pollution output (validation: 2014) and projections (SSP5-8.5) for each state. Descriptive analysis was performed to examine temporal trends, while Spearman correlation assessed relationships between air pollution and health outcomes. The distributed lag non-linear model (DLNM) was used to estimate short-term cumulative exposure-response relationships, followed by sensitivity analyses to evaluate model robustness. WRF-CMAQ outputs were validated against observed DOE monitoring data, and health prediction models were validated using both observed and simulated air pollution data for 2014. Future NCD outcomes were projected using negative binomial regression. The best-performing baseline health models were integrated with bias-corrected WRF-CMAQ air pollution projections to estimate attributable fractions and the projected BoD, including years of life lost (YLL), years lived with disability (YLD), and disability-adjusted life years (DALYs). Differences between projected and baseline PM2.5-attributable BoD were subsequently quantified. Between 2005 and 2021, mortality from CVD, RD, and DM, and from HA, increased in Kelantan and Kedah. CVD mortality rose sharply after 2018 in Kelantan (131.5) and Kedah (154.1). RD mortality steadily increased in 2013 (79.6) in Kelantan and nearly 100 in Kedah (2018). DM mortality gradually rose to 15.6 in Kelantan (2021) and 18.1 in Kedah (2020). CVD HA increased consistently; RD HA peaked in 2019 (1652.8) in Kelantan and 1667.7 (Kedah), while DM HA varied over time, higher early on in Kelantan and peaking in Kedah in 2009(46.2). DALYS were highest for CVD, then DM and RD, mostly driven by YLL. All pollutants remained below thresholds, though PM10 and PM2.5 occasionally exceeded WHO AQG limits. Air pollution showed a significant negative association with NCD outcomes in Kelantan and Kedah (p<0.05). Temperature in Kedah showed a positive association with RD mortality (r = 0.047, p <0.05). Among all pollutants, PM25 demonstrated the most consistent cumulative exposure-response relationships across NCD in both states and was therefore selected for future BoD projection. Under the global warming SSP5-8.5 scenario, PM2.5-attributable CVD mortality in Kelantan increased by 368.71 deaths (95% CI: -129 to 854 deaths). PM2.5-attributable RD mortality decreased by 122.02 deaths (95% CI: -307 to 451 deaths). In contrast, PM2.5-attributable RD HA increased substantially with an absolute increase of 9,200.29 admissions (95% CI: 4,351-13,786). PM2.5-attributable DM HA in Kelantan increased modestly with an absolute increase of 14.89 admissions (95% CI: -134 to 124). In Kedah, baseline PM25-attributable NCD outcome estimates were unavailable, precluding calculation of absolute baseline-projection differences. In conclusion, under the SSP5-8.5 scenario, PM2.5 continues to impose a measurable and evolving burden of NCDs in Kelantan and Kedah. Keywords: non-communicable disease, air pollution, climate change scenario.