Ambient VOCs pollution and emergency hospitalizations in Hong Kong

Abstract

Numerous studies have examined the association of air pollution with cause-specific cardiorespiratory diseases in the general population. Previous studies have mainly focused on the criteria air pollutants, such as particulate matters, nitrogen dioxide, carbon monoxide, sulfur dioxide, and ozone. However, questions remain for other coexisting gaseous entities, especially ambient volatile organic compounds (VOCs). Epidemiological studies have linked indoor VOCs to allergic asthma, loss of lung function, and heart rhythm disorders in particular occupational places, but not in general. One explanation is that daily monitoring for ambient VOCs is not available in many places. The other is that more attention has been paid to the capacity of VOCs to generate ozone or secondary organic aerosols than the independent health effects of ambient VOCs. The current thesis aims to estimate the short-term cardiorespiratory effects of ambient VOCs in the Hong Kong population. VOC speciation data from April 2011 to December 2014 were obtained from the Environmental Protection Department of Hong Kong, and emergency cause-specific cardiorespiratory hospitalizations during the study period were provided by the Hospital Authority of Hong Kong. A generalized additive model (GAM) with the quasi Poisson framework was utilized to assess associations of VOC chemicals with emergency hospitalizations for cause-specific cardiorespiratory diseases while adjusting for time-varying confounders, including secular and seasonal trends, temperature, relative humidity, day-of-week, and holidays. VOC mixture emitted from the same source may have similar physicochemical composition and toxicity. So, I further conducted source apportionment of ambient VOC speciation by the Positive Matrix Factorization model to identify the relevant source profile of ambient VOCs and to measure their potential contributions. Then I used the GAM integrated with a distributed-lag linear model to estimate associations of the emergency hospitalizations for cardiorespiratory diseases with source-apportioned VOCs. Stratification analyses by season, age, and sex were conducted to detect season variation and susceptible populations, and sensitivity analyses were conducted to test the robustness of the findings. I found evidence of positive links of multiple VOC chemicals with emergency hospitalizations for congestive heart failure (CHF) and chronic obstructive pulmonary disease (COPD). I also observed a potential negative association of ambient isoprene with emergency COPD hospitalizations. Six VOC sources were apportioned during the study period, including gasoline emissions, liquefied petroleum gas (LPG) usage, aged VOCs, architectural paints, household products, and biogenic emissions. I found that daily increase in emergency COPD hospitalizations was positively linked to ambient VOCs from gasoline emissions, architectural paints, and household products, and was negatively related to biogenic VOCs. Daily increases in ambient VOCs from architectural paints and household products were positively associated with emergency CHF hospitalizations in cold seasons. My work suggests that ambient VOCs could trigger an acute exacerbation of cardiorespiratory diseases. Relevant air quality guidelines for ambient VOCs should be issued to protect the general population. Further studies on personal exposure assessment and the biological mechanisms of VOCs and health would be helpful.