Background: The association between daily particulate matter (PM) and mortality is well documented. However, the health effects of sustained exposure to high PM levels over several consecutive days (PM wave) are less understood, and evidence across diverse regions is limited. We evaluated short-term associations between PM waves and mortality across multiple countries and cities.

Methods: We collected data for daily all-cause mortality and PM10 and/or PM2.5 concentrations from 37 countries and 340 cities from 1995 to 2019. PM wavs were defined using different combinations of concentration thresholds (relative or absolute) and duration. City-specific effects of PM waves over lag 0-1 were estimated using two approaches: (1) time-series analysis and (2) a matched design that matched each PM wave day to non-wave days within the same city, year, month, and day of week. The overall effect compared PM wave versus non-wave days, and the added effect further controlled for daily PM. City-specific estimates were pooled using multilevel random-effects meta-analysis.

Results: Preliminary results indicated that PM10 waves were associated with increased mortality. Using the city-specific 90th percentile threshold for ≥3 days, PM10 waves were associated with a 0.96% increase in mortality (95% CI 0.44, 1.47%) for the overall effect and a 0.50% increase (95% CI 0.06, 0.94%) for the added effect in time-series analyses. Using an absolute threshold (PM10 ≥ 75 µg/m3 for ≥3 days) yielded 1.19% (95% CI 0.71, 1.68%) for the overall effect and 0.59% (95% CI 0.24, 0.95%) for the added effect. Results were consistent in the matched design. PM2.5 wave estimates were positive but not statistically significant.

Conclusions: Multi-day high PM10 episodes were associated with elevated mortality beyond the effect of high daily concentrations alone, suggesting additional risk during persistent pollution periods. PM wave indicators may complement daily PM metrics in early warning and response systems.