Furthermore, based on the sensitivity and specificity analyses, a lower cutoff might more accurately classify smokers and nonsmokers if participants exhale quickly selleck Carfilzomib versus slowly. If exhalation speed is not monitored, a more conservative CO cutoff value of 3 ppm may be warranted. Other studies have successfully used cutoffs within this range (e.g., Cropsey et al., 2008). Importantly, in the current study, increasing the cutoff from 3 to 4 ppm during the fast condition resulted in an increase in the false negative rate from 6% to 29% (Table 1). If higher cutoff values are used, minimum exhalation speeds should be considered. In the current study, the median for the fast condition was 4 s (interquartile range = 3�C5 s) and the median for the slow condition was 14 s (interquartile range = 10�C18 s).

Only 3% of exhalations during the fast condition were ��10 s; thus, a minimum exhalation speed of 10 s would be reasonable when a higher CO cutoff is used. Maximum speeds are not recommended because if individuals with larger lung capacities need longer to empty their lungs, ending too early would underestimate their CO levels. The current study used the piCO+ monitor. To determine the generality of the current findings, we conducted a pilot test, using the same procedure with two different CO monitors (Micro+ and COmpact, Bedfont Scientific USA). Two moderate smokers were counterbalanced to which monitor and condition they tested first. For both participants, their CO was consistently higher during both slow conditions than during both fast conditions (Micro+ mean fast = 21.

3 �� 3.0 and slow = 30.5 �� 3.7 ppm; COmpact mean fast = LEVEL 4.8 �� 0.5; slow = LEVEL 6 �� 0.0) regardless of the monitor and condition order. Finally, there may be individual subject variables that affect CO but were not systematically investigated in the current study. For example, lung volume is positively associated with CO output (r = .64; Terheggen-Lagro, Bink, Vreman, & van der Ent, 2003). Participants with greater lung volume would be expected to take longer to empty their lungs. Lung functioning is also negatively related to smoking severity (Beck, Doyle, & Schachter, 1981; Gold et al., 1996). Lactose intolerance has also been associated with CO output (McNeill, Owen, Belcher, Sutherland, & Fleming, 1990). Finally, factors that affect the half-life of CO (e.g.

, physical activity), as well as time since last cigarette, can influence CO outcomes (Benowitz et al., 2002). Fortunately, the finding in the current study, that speed of exhalation affects CO output, could not be accounted for by differences in lung volume, lung function, lactose intolerance, or CO half-life because all Brefeldin_A participants were exposed to both conditions. Funding This work was supported by the National Institutes of Drug Abuse at the National Institutes of Health (R01DA019580). Declaration of Interests None declared.