OBJECTIVESExposure misclassification is a major obstacle to obtain accurate dose-response relationships. In order to solve this problem, the impact of hair treatment on total mercury in hair was assessed in Japanese women.

METHODSA cross-sectional study was carried out among 327 women at age 24-49 years to determine hair mercury levels and estimate daily mercury intakes from seafood by using a food frequency questionnaire.

RESULTSHair mercury levels in the women and daily mercury intake ranged from 0.11 to 6.86 (median 1.63) μg/g and from 0.77 to 144.9 (median 15.0) μg/day, respectively. The hair mercury was positively correlated with the daily mercury intake (p<0.001). When the women were divided into two subgroups based on artificial hair-waving, hair coloring/dyeing, residence (non-fishing and fishing areas), and working status, a significant difference in the hair mercury level was observed between the women with and without artificial hair-waving only (p<0.001). The multiple regression analysis showed that the log-transformed hair mercury level was significantly related to the log-transformed daily mercury intake (standardized regression coefficient βs=0.307) and artificial hair-waving (βs=-0.276); but not to hair coloring/dyeing, residence, working status or age. Permanent hair treatment was estimated to reduce total mercury in hair by approximately 30%, after adjusting for daily mercury intake and other possible factors.

CONCLUSIONSThese findings suggest that hair mercury is not the best biomarker of methylmercury exposure when a study population includes women with artificial hair-waving.