Mortality risks from a spectrum of causes associated with sand and dust storms in China

Sand and Dust Storms (SDS) are an emerging public health concern, contributing up to 2 billion tonnes of dust annually and degrading air quality across more than 150 countries. Climate and land-use changes have intensified SDS frequency and intensity. The 2021 WHO air quality guidelines aimed to address SDS, but insufficient and unspecific epidemiological evidence has hampered inclusion. Prior studies and meta-analyses have indicated overall increases in circulatory and respiratory mortality associated with SDS, but evidence on specific sub-causes and other disease categories remains limited, often derived from single-city studies with small samples. Inhalable particulate matter (PM10), including PM2.5 and PM2.5–10, is a principal component of SDS and implicated in adverse health effects. Some evidence suggests that heavy PM2.5 pollution events have added effects on mortality beyond daily concentrations, but whether SDS events—characterized by high PM2.5–10 and PM2.5—have added effects on mortality remains unclear. This nationwide multicenter time-series study in China set out to (1) investigate the overall short-term effects of SDS on mortality from a spectrum of causes to identify SDS-sensitive outcomes, and (2) assess added short-term effects of SDS events on mortality beyond particulate matter concentrations.

Existing literature shows SDS are linked to increased respiratory and circulatory mortality. A recent meta-analysis reported pooled increases of 2.33% for circulatory and 3.99% for respiratory mortality in the days following SDS. However, most studies were single-location with limited power and rarely examined cause-specific subcategories (e.g., ischemic vs hemorrhagic stroke; COPD). Studies have connected PM10 elevations during SDS with cardiovascular mortality increases and suggested mechanisms including airway irritation and inflammation, microbial and chemical constituents on dust particles, and systemic oxidative stress. Evidence on SDS impacts on genitourinary, nervous, and digestive diseases is sparse, with some morbidity studies and experimental work indicating potential adverse effects. These gaps motivate a large-scale national assessment of SDS impacts across multiple mortality causes and evaluation of added effects beyond conventional PM pollution.

Study design: A nationwide multicenter two-stage time-series analysis was conducted using daily data from 214 Chinese counties selected to represent primary SDS transmission routes and heterogeneity in SDS exposure. The study period focused on the SDS season each year (1 February–31 May) from 2013 to 2018. Data: Daily county-level mortality counts were obtained from the China CDC Disease Surveillance Points System. Eighteen mortality outcomes were analyzed: seven broad categories (all-cause, non-accidental, circulatory, respiratory, digestive, nervous, genitourinary) and 11 specific causes (including ischemic heart disease and subtypes; ischemic and hemorrhagic stroke and subtypes; hypertensive heart disease; chronic lower respiratory disease; COPD). Exposure and covariates: County-specific PM10 and PM2.5 daily averages were computed from hourly data from China's National Air Pollution Monitoring System; PM2.5–10 was calculated as PM10 minus PM2.5. Meteorological variables (temperature, relative humidity) came from ERA5-Land at county central points, aggregated to daily values. Official sand-dust weather records were collected from the China Meteorological Administration. SDS event definition: A county-day was classified as an SDS event if (1) an official sand-dust weather record existed for that day, (2) PM10 > 50 µg/m3, and (3) PM2.5/PM10 < 0.4. Alternative definitions used PM2.5/PM10 thresholds of 0.35 and 0.45, and a definition without the PM2.5/PM10 criterion, for sensitivity analyses. Statistical analysis: First-stage county-specific models used generalized linear models with quasi-Poisson distribution: log E(Yt) = Intercept + β Zt + ns(Time, df=2 per SDS period) + ns(temperature, df=3) + ns(relative humidity, df=3) + day-of-week indicator. Zt categorized days as SDS event (1), non-SDS with PM2.5 pollution (2; daily PM2.5 ≥ 75 µg/m3), or clean (0). Effects were estimated as excess risk (ER) comparing SDS days to clean days; risks for non-SDS PM2.5 pollution days were also estimated. Second-stage random-effects meta-analysis pooled county-specific estimates: ER = (exp(βmeta) − 1) × 100%. Additional analyses: Added effects of SDS were examined by additionally adjusting for PM2.5–10 and PM2.5 in the GLMs. Lag analyses evaluated single-day lags (lag 1–3). Stratified analyses by sex (male, female) and age (<75, ≥75 years) were performed for all-cause, non-accidental, circulatory, and respiratory mortality, with Z tests for subgroup differences. Sensitivity analyses: Robustness was assessed by (a) varying SDS definitions, (b) changing spline degrees of freedom for time (to 3) and meteorology (to 4 or 5), (c) using 21-day moving average temperature and 7-day moving average humidity, (d) refitting models using whole-year data, and (e) restricting to counties with daily death counts exceeding one during SDS periods.

- Study scope: 2,024 SDS events identified across 214 Chinese counties during SDS seasons (Feb 1–May 31) from 2013–2018; 1,495,724 total deaths in the SDS periods, with 19,082 deaths on SDS event days. - Spatial pattern: SDS events were more frequent in northern counties near dust sources than in southern regions. - Main pooled effects at lag 0 (excess risk, ER, %; 95% CI): - Broad categories: Respiratory diseases 8.90 (4.96, 12.98); Circulatory diseases 2.12 (-0.39, 4.70); Genitourinary 2.57 (-1.56, 6.87); Nervous 1.33 (-1.62, 4.37); Digestive 0.24 (-2.14, 2.68); All-cause 1.65 (-0.26, 3.59); Non-accidental 1.33 (-0.63, 3.33). - Specific causes: Ischemic stroke 7.49 (3.12, 12.05); Intracerebral hemorrhagic stroke 5.40 (1.25, 9.73); Hypertensive heart disease 4.05 (0.41, 7.83); Hemorrhagic stroke 3.67 (-0.23, 7.72); Myocardial infarction 3.45 (0.34, 6.66); Acute myocardial infarction 3.37 (0.28, 6.55); Acute ischemic heart disease 3.33 (0.07, 6.70); Ischemic heart disease 1.21 (-2.19, 4.72); Chronic ischemic heart disease -0.20 (-3.23, 2.92); Chronic lower respiratory disease 12.51 (6.31, 19.08); COPD 11.55 (5.55, 17.89). - Temporal pattern: Strongest and most significant associations were at lag 0; similar but slightly attenuated effects at lag 1; negligible or no associations at lags 2–3. - Added effects: Compared with non-SDS PM2.5 pollution days, SDS days showed more enhanced risks for several outcomes. After adjusting for PM2.5–10 and PM2.5, significant added effects of SDS persisted for ischemic stroke, chronic lower respiratory disease, and COPD. - Subgroup analyses: Respiratory mortality risks were more pronounced in males and older adults (≥75 years), whereas all-cause, non-accidental, and circulatory mortality risks were slightly higher in females and younger (<75) adults; differences were not statistically significant. - Robustness: Results were consistent across alternative SDS definitions (with slightly higher estimates under more lenient criteria), varying spline specifications, meteorological adjustments, study periods, and county subsets.

This study addressed key gaps by providing a comprehensive national assessment of short-term SDS effects on a wide spectrum of mortality causes. The findings confirm substantial respiratory mortality increases during SDS and identify specific circulatory and respiratory sub-causes at elevated risk, notably ischemic stroke, intracerebral hemorrhagic stroke, hypertensive heart disease, chronic lower respiratory disease, and COPD. Mechanistically, inhalation of dust particles can damage airway and alveolar tissues, induce inflammation and oxidative stress, and carry microbial or chemical constituents that exacerbate toxicity. Systemic inflammatory responses and autonomic effects may contribute to thrombotic and hypertensive events, explaining the observed associations with stroke and hypertensive heart disease. Importantly, SDS events exerted added mortality effects beyond conventional PM pollution events and beyond PM2.5 and PM2.5–10 concentrations, suggesting that SDS-specific particle composition, surface chemistry, and bioactive constituents may increase bioreactivity and health risks. The identification of immediate (same-day) impacts underscores the need for rapid public health responses during SDS episodes.

Short-term exposure to SDS events is associated with increased mortality across multiple causes, with the largest effects for respiratory diseases and significant elevations for specific circulatory and respiratory sub-causes. SDS events also have added effects beyond PM mass concentrations for ischemic stroke, chronic lower respiratory disease, and COPD. These results support the development of SDS-specific air quality guidelines and standards, proactive allocation of healthcare resources before and during SDS events, and public education on protective measures. Future research should refine SDS exposure definitions, expand geographic coverage, and disentangle contributions from natural versus anthropogenic particulate sources and constituents to better target interventions.

- Exposure misclassification is possible due to the lack of a universal SDS definition; however, the study combined official records with PM10 levels and PM2.5/PM10 ratios and tested alternative definitions with consistent results. - Mortality data accessibility limited inclusion of more comprehensive or national coverage, though this remains the largest study of its kind in China to date. - The study could not distinguish health effects attributable to natural versus anthropogenic PM sources during SDS; future work should address source-specific contributions.