These represent a DSM-III-R (American Psychiatric Association, 1987) diagnosis of MD, and a broadened DSM-III-R diagnosis of GAD, where the minimum duration is reduced from 6 months to 1 month (as in Hettema et al., 2006, and described in Kendler, Neale, Kessler, Heath, & Eaves, 1992), which was the duration criterion in DSM-III. DSM-III-R criteria were used at the onset of data collection for the add to favorites VATSPSUD and continued to be used after the publication of the DSM-IV (American Psychiatric Association, 1994) in order to maintain consistency across waves of data collection. Nicotine Dependence ND was determined using the Fagerstr?m Test for Nicotine Dependence (FTND) (Heatherton, Kozlowski, Frecker, & Fagerstrom, 1991), with possible scores of 0�C10.
Importantly, items in the FTND do not address negative affect: otherwise, the independent variable (FTND score) and the outcome variable (anxious/depressive symptoms) would be confounded. Only individuals who reported being or having ever been a regular smoker were administered the FTND. Neuroticism Neuroticism (N) was scored using the short form of the Eysenck’s Personality Questionnaire (EPQ) (Eysenck, Eysenck, & Barrett, 1985). An individual’s N-score was the total number of items they endorsed out of 12. Female twin pairs who were part of the female-twin only sample were assessed for N at three timepoints; those who were part of the sample that also included male twins were measured at two timepoints. Cronbach’s alpha was .84 for women and .82 for men. The mean N-score across timepoints was taken for each individual.
Statistical Analyses Preliminary analyses involving intraindividual regressions were conducted on the complete sample using PROC GENMOD in SAS 9.1 (Cary, NC) in order to adjust significance levels for twinship and model the covariance structure of hierarchical data. Twin-based regressions were the primary method by which we addressed our research questions. In these regressions, co-twins�� phenotypes are used as predictors, enabling us to test whether genetic factors are influential. For twin-based analyses, data on nicotine withdrawal-induced anxiety or depression were only available for one twin in a pair in some cases. In order to maximize statistical power, regression analyses were performed on all subjects so that Twin 1��s history of ND or MD/GAD was used to predict Twin 2��s symptoms of anxiety/depression during nicotine withdrawal and vice versa.
To account for the fact that information about each twin subject is used both as a predictor and dependent variable in our regression analyses, SEs were corrected using PROC GENMOD. The following regression models were tested: (1) (2) (3) (4) In each regression, Entinostat age, sex, and zygosity are covariates. Withdrawal-induced depression and anxiety were modeled as ordinal variables, and the ��descending�� option was used.
Related posts: