Cigarette use during the study was assessed weekly using the timeline followback method (Sobell & Sobell, 1992); self-reports of abstinence selleck inhibitor were verified by expired CO ��8 parts per million (Jarvis, Tunsall-Pedoe, Feyerabend, Vesey, & Saloojee, 1987). Dropouts were considered nonabstainers. The main predictor was ADHD diagnosis by subtype (i.e., ADHD-inattentive, ADHD-hyperactive/impulsive, and ADHD-combined) following the adult ADHD Clinical Diagnostic Scale (ACDS) version 1.2 (Adler & Cohen, 2004). The ability of the ACDS to establish ADHD diagnosis and subtype following DSM-IV criteria was demonstrated in community-based studies (Kessler et al., 2006, 2010). The ACDS raters had a minimum of M.A. degree and were trained and certified in the use of the ACDS (per Adler et al., 2005).
Other covariates selected because of prior evidence of their potential effects on smoking abstinence were total ADHD symptoms at baseline, nicotine dependence level and other smoking history (number of cigarettes smoked daily at baseline and age began smoking), past psychiatric history (major depressive disorder, any anxiety disorder, alcohol abuse/dependence, and drug abuse/dependence), and demographic characteristics (age, gender, marital status, education, and employment status). For measuring ADHD symptom level, we used the ADHD-Rating Scale (DuPaul, Power, Anastapolous, & Reid, 1998). The presence of comorbid psychiatric diagnoses was assessed with the Composite International Diagnosis Interview (CIDI) version 2.1 (http://www.crufad.unsw.edu.au/cidi/cidi.htm).
Nicotine dependence level was measured with the Fagerstr?m Test for Nicotine Dependence (FTND) using 7 as the cut score to distinguish participants according to low to medium (<7) or high (��7) dependence level. This cut score produced maximum agreement with the presence of DSM-III-R�Cdiagnosed nicotine dependence compared with cuts of 5, 6, or 8 (Moolchan et al., 2002). Evidence exist indicating the reliability and validity of the instruments used to measure the covariates, that is, ADHD-RS (D?pfner et al., 2006; R?sler, Retz, & Stieglitz, 2010), CIDI (Andrews & Peters, 1998), and FTND (de Meneses-Gaya, Zuardi, Loureiro, & de Souza Crippa, 2009). Statistical Methods Differences in demographic characteristics, smoking and psychiatric history, and ADHD symptom ratings were analyzed by the ��2 test for categorical variables and the t test for continuous variables.
Statistical significance was set at p Brefeldin_A �� .05. Logistic regression modeling was applied to analyze prolonged smoking abstinence as a function of the independent variables. Clinical sites were entered as a fixed effect. Adjusted odds ratios (AORs) and 95% CI associated with main effects and interaction terms were assessed. SAS PROC LOGISTIC (SAS 9.2, Cary, N.C.) was used to conduct the analyses.