Effects of environmental and lifestyle exposures on urinary levels of polycyclic aromatic hydrocarbon metabolites: A cross-sectional study of urban adults in China.

Urinary polycyclic aromatic hydrocarbon (PAH) metabolites, biomarkers of internal PAH exposure, are commonly used to explore the effects of PAH on human health. However, the correlation between environmental PAH exposure and the species or levels of urinary PAH metabolites remains unclear. We collected detailed information on PAH exposure sources, including cigarette smoking, cooking, traffic and diet habits via structured questionnaires, and determined 12 urinary monohydroxylated PAH metabolites (OH-PAHs) among 4092 participants from the Wuhan-Zhuhai cohort. Linear mixed models and generalized linear models were conducted to explore the associations of urinary metabolite levels with single or multiple PAH exposure sources. We also calculated the standardized regression coefficients to further compare the contributions of different sources to urinary OH-PAH levels. Our results showed that increasing levels of urinary 1-, 2-hydroxynaphthalene (1-, 2- OHNa) and 2-hydroxyfluorene (2-OHFlu) were significantly correlated with tobacco smoking (all P < 0.01). The concentrations of 1-, 2- OHNa and 9-hydroxyfluorene (9-OHFlu) were positively correlated with dietary intake (all P < 0.05). Individuals who spent a long time in traffic showed elevated levels of 9-OHFlu and 1-hydroxyphenanthrene (1-OHPh) compared with individuals who spent a short time in traffic (all P < 0.05). Self-cooking was associated only with elevated 1-hydroxypyrene (1-OHP) levels. Moreover, good kitchen ventilation resulted in significantly decreased urinary low-molecular-weight OH-PAH levels. These findings suggested that cigarette smoking, self-cooking, high dietary PAH intake and a long time spent in traffic were associated with increased levels of specific urinary PAH metabolites, and good kitchen ventilation effectively reduced the exposure to low-molecular-weight PAHs in self-cooking participants.