Linear regression analysis was first used to assess the relationship between nicotine metabolism (3OHCOT/COT) and smoking-topography measures (mean puff volume, total puff volume, mean inter-puff interval, mean puff duration, and mean puff velocity) in the total sample. Previous research has shown that ethnicity and sex including contraceptive (HC) use among girls may influence nicotine metabolism among adolescent smokers [18, 19]. Therefore, analyses were then repeated within subsets stratified by sex and ethnicity. Statistical significance was defined as p< .05. SPSS (version 15) was used for all analyses.
