most severe class, while this is not the case for the moderate class. This relationship between smoking persistence and moderate nicotine withdrawal was maintained even when we expanded our criteria of smoking persistence from any smoking within the last month to any within the last year, as was used by Breslau et al. (1992). Our study may have included additional covariation between nicotine withdrawal smoking persistence compared to Breslau et al. (1992), wherein nicotine withdrawal symptoms were only assessed among those with at least 1 unsuccessful quit attempt. Change in societal views on smoking could in part account for the differential effects of withdrawal on persistence between our samples here and the older cohort described in Madden et al. (1997). Future studies should explore this particular group of adult smokers highlighted here who report specifically high levels of irritability, restlessness and increased appetite during nicotine abstinence, in terms of increased risk for persisting in the smoking habit.
