Concentrations of the Sunscreen Agent Benzophenone-3 in Residents of the United States: National Health and Nutrition Examination Survey 2003-2004

Antonia M. Calafat; Lee-Yang Wong; Xiaoyun Ye; John A. Reidy; Larry L. Needham

Disclosures

Environ Health Perspect. 2008;116(7):893-897. 

In This Article

Results

We detected BP-3 in 96.8% of the 2,517 samples at concentrations ranging from 0.4 to 21,700 μg/L; the geometric mean and 95th percentile concentrations were 22.9 μg/L (22.2 μg/g creatinine) and 1,040 μg/L (1,070 μg/g creatinine), respectively ( Table 1 ).

The final model for adults included sex, race/ethnicity, age, age squared (p = 0.038), creatinine concentration (log scale), and the interaction terms creatinine*sex (p < 0.001) and age*race/ethnicity (p = 0.04) ( Table 2 ). Females had significantly higher BP-3 concentrations (p ≤ 0.04) than did males, regardless of creatinine level [see Supplemental Tables S1 and S2 (http://www.ehponline.org/members/2008/11269/suppl.pdf). Although BP-3 concentrations increased linearly as log creatinine increased for both sexes (p < 0.001), the increase was more pronounced for males than for females (β for males, 1.12; for females, 0.65). Also, as age increased, BP-3 LSGM concentrations showed a significant quadratic trend for Mexican Americans (p = 0.016) and a significant linear positive trend for non-Hispanic blacks (p = 0.022) but no significant linear or quadratic trend for non-Hispanic whites (Figure 1). LSGM concentrations of BP-3 for non-Hispanic whites were significantly higher than for non-Hispanic blacks, regardless of age (p ≤ 0.01), and significantly higher than for Mexican Americans only for 20- to 29-year-olds (p = 0.01). LSGM concentrations of BP-3 were significantly higher for Mexican Americans than for non-Hispanic blacks only for 30- to 39-year-olds (p = 0.01) [see Supplemental Tables S1 and S2 (http://www.ehponline.org/members/2008/11269/suppl.pdf)].

LSGM concentrations of BP-3 (in micrograms per liter) by age and race/ethnicity: (A) children and adolescents and (B) adults. Error bars indicate 95% CIs.

The final model for children and adolescents included sex (p < 0.001), race/ethnicity, age, creatinine concentration (log scale) (p < 0.001), and a race/ethnicity*age (p = 0.01) interaction term ( Table 2 ). LSGM concentrations of BP-3 increased as log creatinine increased (β = 0.77, p < 0.001). LSGM BP-3 concentrations for girls [30.2 μg/L; 95% confidence interval (CI), 21.4-42.6 μg/L] were significantly higher (p < 0.001) than for boys (16.1 μg/L; 95% CI, 13.2-19.8 μg/L). BP-3 concentrations also decreased linearly as age increased (p = 0.0005) for non-Hispanic whites but not for Mexican Americans and non-Hispanic blacks [Figure 1; see also Supplemental Table S3 (http://www.ehponline.org/members/2008/11269/suppl.pdf)]. LSGM concentrations of BP-3 for non-Hispanic whites were significantly higher than LSGM concentrations for non-Hispanic blacks, regardless of age, and for Mexican Americans only at younger ages [p < 0.001 at 8.5 years, p < 0.01 at 12 years; Supplemental Table S4 (http://www.ehponline.org/members/2008/11269/suppl.pdf)]. LSGM BP-3 concentrations were significantly higher for Mexican Americans than for non-Hispanic blacks only for older children (p = 0.01, at 12 and at 15.6 years, p = 0.03 at 17.4 years) [Supplemental Table S4 (http://www.ehponline.org/members/2008/11269/suppl.pdf)].

For participants with urinary concentrations above the 95th percentile of BP-3, sex (p < 0.001) and race/ethnicity (p = 0.03), but not age, were significantly associated univariately. In the final multiple logistics regression, sex (p < 0.001) and race/ethnicity (p = 0.03) were significant [Supplemental Table S5 (http://www.ehponline.org/members/2008/11269/suppl.pdf)]. Females were 3.5 times more likely than males to be above the 95th percentile [adjusted odds ratio (OR) = 3.5; 95% CI, 1.9-6.5]. Non-Hispanic whites were 6.8 times more likely to have BP-3 concentrations above the 95th percentile (adjusted OR = 6.8; 95% CI, 2.9-16.2) than were non-Hispanic blacks, and Mexican Americans were four times more likely to be above the 95th percentile (adjusted OR = 4.04; 95% CI, 1.1-15.5) than were non-Hispanic blacks. We found no significant difference between non-Hispanic whites and Mexican Americans.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.

processing....