The Role of Air Nicotine in Explaining Racial Differences in Cotinine Among Tobacco-Exposed Children

Stephen E. Wilson, MD, MSc; Robert S. Kahn, MD, MPH; Jane Khoury, PhD; Bruce P. Lanphear, MD, MPH


CHEST. 2007;131(3):856-862. 

In This Article


Sample Characteristics

The characteristics of this study population are outlined in Table 1 . African-American children were slightly older than white children, and resided in households with a smaller volume. African-American families used the HEPA air cleaners less frequently than white families during both the first 6 months and second 6 months of the study. There were no racial differences in the proportion of girls in the study, the season of enrollment, or the distribution of active filters. Also, there were no differences by race in the reported levels of ETS exposure in settings outside of the home.

Cotinine and Air Nicotine

African-American children had significantly higher levels of cotinine in both hair and serum compared to white children ( Table 1 ). Racial differences in cotinine biomarkers persisted > 12 months (Figures 1 and 2) However, there were no significant racial differences in air nicotine levels at the 6-month (2.95 µg vs 2.55 µg, p = 0.62) and 12-month (3.27 µg vs 2.17 µg, p = 0.17) study visits.

Figure 1.

Geometric mean levels of serum cotinine over 12 months by race. The error bars represent the 95% confidence intervals around the mean.

Figure 2.

Geometric mean levels of hair cotinine over 12 months by race. The error bars represent 95% confidence intervals around the mean.

Repeated-Measures Analysis

African-American race was independently associated with serum cotinine over time ( Table 2 ). In the bivariate model, African-American race was significantly associated with serum cotinine. The addition of various covariates attenuated the relationship between race and cotinine, reducing the race estimate from 0.435 (p = 0.015) to 0.280 (p = 0.040), but African-American race and cotinine remained statistically significant. Transformation of the race coefficient in the adjusted model generated a value of 1.32, indicating the ratio of serum cotinine means between African-American children and white children. Other factors independently associated with serum cotinine included air nicotine, filter run time, home ETS exposure, and total home volume. Notably, age, gender, and season were not associated with serum cotinine levels.

Racial differences in hair cotinine were even more striking ( Table 3 ). Adding potential confounders to the initial model increased the race estimate from 1.36 (p < 0.0001) to 1.40 (p < 0.0001). African-American race and air nicotine were independently associated with hair cotinine in the fully adjusted model. Transformation of the race coefficient provided a value of 3.9, indicating a fourfold difference in hair cotinine levels between African Americans and whites over the 12-month study period. Next, we tested for an interaction by introducing a race/air nicotine product term into both models. The p values for these product terms exceeded 0.80, indicating the absence of effect modification.


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.
Post as: