= .237) and mean puff duration (t = −2.064, p = .048, adjusted R2 = .101). The regression scatter plots are shown in Figure 2. Among boys, metabolism accounted for 19.2% of the variability in mean puff volume; metabolism and mean puff duration and number of puffs were inversely related, with metabolism accounting for 10.1% of the variance in mean puff duration and 23.7% of the variance in number of puffs. No other relationships reached significance among sex-stratified or ethnicity-stratified samples and topography measures. Metabolism did not significantly predict CO levels in any sample (data not shown). Table 2 shows unstandardized regression coefficients (B), standard errors, and standardized regression coefficients (β) for topography analyses.
