Changing variance and skewness as leading indicators for detecting ozone exposure-associated lung function decrement


The objective of this study was to develop a novel risk analysis approach to assess ozone exposure as a risk factor for respiratory health. Based on the human exposure experiment, the study first constructed the relationship between lung function decrement and respiratory symptoms scores (ranged 0–1 corresponding to absent to severe symptoms). This study used a toxicodynamic model to estimate different levels of ozone exposure concentration-associated lung function decrement measured as percent forced expiratory volume in 1 s (%FEV$_1$). The relationships between 8-h ozone exposure and %FEV$_1$ decrement were also constructed with a concentration–response model. The recorded time series of environmental monitoring of ozone concentrations in Taiwan were used to analyze the statistical indicators which may have predictability in ozone-induced airway function disorders. A statistical indicator-based probabilistic risk assessment framework was used to predict and assess the ozone-associated respiratory symptoms scores. The results showed that ozone-associated lung function decrement can be detected by using information from statistical indicators. The coefficient of variation and skewness were the common indicators which were highly correlated with %FEV$_1$ decrement in the next 7 days. The model predictability can be further improved by a composite statistical indicator. There was a 50 % risk probability that mean and maximum respiratory symptoms scores would fall within the moderate region, 0.33–0.67, with estimates of 0.36 (95 % confidence interval 0.27–0.45) and 0.50 (0.41–0.59), respectively. We conclude that statistical indicators related to variability and skewness can provide a powerful tool for detecting ozone-induced health effects from empirical data in specific populations.

Stochastic Environmental Research and Risk Assessment