BackgroundMetro system has become popular in urban areas. However, short-term effects of size-fractionated particulate matter (PM) on cardiac autonomic function in metro system remain unexplored. ObjectivesTo explore the contribution of ambient PM to in-cabin PM and investigate the short-term effects of exposure to size-fractionated PM and black carbon (BC) in metro system on cardiac autonomic function in young healthy adults. MethodsThirty nine young healthy adults were asked to travel in metro system during 9:00–13:00 on a weekends between March and May 2017. We performed continuous ambulatory electrocardiogram monitoring for each of them, and measured real-time size-fractionated PM, BC, nitrogen dioxide, nitric oxide, carbon dioxide, ozone, noise, temperature and relative humidity in metro cabin. We also collected the data of ambient PM2.5 (aerodynamic diameter < 2.5 µm) concentrations in Beijing. Linear regression model was used to estimate the infiltration factor of ambient PM2.5 to assess the relationship between metro cabin PM and ambient PM. Mixed-effects model was used to estimate the associations between changes in HRV parameters and PM0.5 (aerodynamic diameter < 0.5 µm), PM0.5–2.5 (aerodynamic diameter between 0.5 µm and 2.5 µm), PM2.5–10 (aerodynamic diameter between 2.5 µm and 10 µm), and BC, respectively. ResultsWe found that size-fractionated PM in metro systems were significantly associated with HRV parameters. Per IQR (interquartile range) increase in PM0.5 (1.6*107/m3) in 1-h moving average concentration was associated with a 13.96% (95% CI: − 18.99%, − 8.61%) decrease in SDNN (standard deviation of normal-to-normal intervals). Similar inverse associations were found between size-fractionated PM exposure and LF (low frequency power), HF (high frequency power), respectively, and smaller particles had greater effects on HRV parameters at shorter lag time. Sex of participants modified the adverse associations between size-fractionated PM and HRV. An IQR of 1-h PM0.5 increasing was associated with a decrease of 6.05% (95% CI: − 22.87%, − 14.44%) in males and a 34.87% (95% CI: − 49.59%, − 15.85%) in females in LF (P for interaction = 0.026). The infiltration factor of ambient PM2.5 was 0.39 (95% CI: 0.33, 0.45). It is estimated that PM2.5 originated from ambient air may account for 20.2% of the PM measured in metro cabin. Per IQR increase in BC (5.5 μg/m3) in 5-min, 1-h, and 2-h moving averages, a primary tracer for ambient PM from combustion source, was associated with decreases of 0.84% (95% CI: − 1.20%, − 0.47%), 2.22% (95% CI: − 3.20%, − 1.22%), and 4.44% (95% CI: − 6.28%, − 2.56%) in SDNN, respectively. ConclusionsShort-term exposure to PM may disturb metro commuter's cardiac autonomic function, and the potential effects depend on the size of PM and the sex of commuters. Ambient PM from combustion source may have adverse effects on the cardiac autonomic function of passengers in cabin.