Optimizing Air Pollution Warning Systems: A Global Assessment of PM2.5-Mortality Nonlinearity and Alert Thresholds

Short-term exposure to ambient fine particulate matter (PM_2.5) is a well-recognized driver of cardiovascular morbidity and mortality. However, current air pollution alert systems are often suboptimal in protecting public health, largely because they do not fully account for the complex, nonlinear exposure-response relationship between PM_2.5 levels and cardiovascular outcomes. To address this limitation, this study establishes a globally representative nonlinear exposure-response function and determines an optimal public health alert threshold that effectively balances health benefits with the reduction of societal disruption.

We performed a systematic review and meta-analysis covering 100 epidemiological studies, yielding 123 effect estimates published up to May 2025. To estimate the nonlinear curve, we applied a novel three-stage meta-regression model that integrates spline functions with structural causal modeling theory. Furthermore, by utilizing global gridded datasets regarding PM_2.5 concentrations, population distribution, and baseline mortality from 2000 to 2023, we quantified the cardiovascular mortality burden and employed a ROC-like analysis to identify the optimal alert value.

Our meta-analysis indicates a pooled risk ratio of 1.009 (95% CI: 1.0074–1.011) for cardiovascular mortality per 10 μg/m³ increment in short-term PM_2.5. The derived exposure-response curve reveals a distinct supralinear shape: marginal risks are elevated at lower concentrations, plateau at moderate levels (~75-150 μg/m³), and surge sharply again beyond 150 μg/m³. In 2023, pollution episodes exceeding the WHO first-stage interim target (75 μg/m³) were associated with an estimated 59,399 (95% CI: 38,126–82,413) attributable cardiovascular deaths globally. The analysis identifies 136 μg/m³ (95% CI: 129–148) as the optimal alert threshold. Implementing warnings at this specific level could potentially prevent 73.2% (95% CI: 71.8%–76.6%) of attributable deaths while impacting only 32% of at-risk person-days.

In conclusion, a significant nonlinear relationship governs short-term PM2.5 exposure and cardiovascular mortality. The optimal alert value identified in this study provides critical evidence for designing more scientific, efficient, and health-oriented air pollution warning systems, thereby maximizing public health protection while minimizing social costs.