Background

Short-term exposure to PM_2.5 has been widely associated with human morbidity and mortality. However, most up-to-date research was conducted at a daily timescale, neglecting the intra-day variations in both exposure and outcome. As an important fraction in PM_2.5, PM₁ has not been investigated about the very acute effects within a few hours.

Methods

Hourly data for size-specific PMs (i.e., PM₁, PM_2.5, and PM₁₀), all-cause emergency department (ED) visits and meteorological factors were collected from Guangzhou, China, 2015–2016. A time-stratified case-crossover design with conditional logistic regression analysis was performed to evaluate the hourly association between size-specific PMs and ED visits, adjusting for hourly mean temperature and relative humidity. Subgroup analyses stratified by age, sex and season were conducted to identify potential effect modifiers.

Results

A total of 292,743 cases of ED visits were included. The effects of size-specific PMs exhibited highly similar lag patterns, wherein estimated odds ratio (OR) experienced a slight rise from lag 0–3 to 4–6 h and subsequently attenuated to null along with the extension of lag periods. In comparison with PM_2.5 and PM₁₀, PM₁ induced slightly larger effects on ED visits. At lag 0–3 h, for instance, ED visits increased by 1.49% (95% confidence interval: 1.18–1.79%), 1.39% (1.12–1.66%) and 1.18% (0.97–1.40%) associated with a 10-μg/m³ rise, respectively, in PM₁, PM_2.5 and PM₁₀. We have detected a significant effect modification by season, with larger PM₁-associated OR during the cold months (1.017, 1.013 to 1.021) compared with the warm months (1.010, 1.005 to 1.015).

Conclusions

Our study provided brand-new evidence regarding the adverse impact of PM₁ exposure on human health within several hours. PM-associated effects were significantly more potent during the cold months. These findings may aid health policy-makers in establishing hourly air quality standards and optimizing the allocation of emergency medical resources.