Participants attended a study Vismodegib medulloblastoma session at the Qingdao Centers for Disease Control (QCDC), conducted by QCDC staff members trained in the study protocol. After a description of the study procedures and confidentiality, parental consent, and twin assent to participate, twins completed surveys that assessed depressive symptoms, smoking behavior, and other characteristics not reported here. The study was approved by Institutional Review Boards at the University of Southern California and the QCDC. Measures Zygosity Zygosity was determined by simultaneous detection of multiple short tandem repeat loci in blood samples (Lv, Zhan, & Qin, 2003). In the Chinese National Twin Registry, the probability of correctly identifying monozygosity based on these markers was �� .996.

Surveys As part of a multistage translation process, bilingual researchers, public health experts, study staff, and QCDC staff members translated, verified, and evaluated each question on the survey for local idioms and reading level (Unger et al., 2011). The Center for Epidemiological Studies Depression Scale The CESD is a 20-item self-report measure of past-week depressive symptoms with suitable psychometric properties in prior studies of Chinese adolescents (Chou, 1999, 2000; Radloff, 1977; Yang, Soong, Kuo, Chang, & Chen, 2004). Each item lists a particular symptom for which respondents indicated how often they felt that way in the past week: rarely or none of the time (0�C1 days, 0 point), some or a little of the time (1�C2 days, 1 point), occasionally or a moderate amount of the time (3�C4 days, 2 points), or most or all of the time (5�C7 days, 3 points).

Although the factor structure is not entirely consistent across all studies, a meta-analysis of CESD factor analyses in 28 diverse samples (N = 22,340) found a clear four-factor solution that distinguishes PA, NA, SF, and IP (Shafer, 2006). An exploratory factor analysis using principal axis factoring with a promax rotation in our data found two primary factors accounting for 34% and 13% of the variance, respectively, with one factor exhibiting strong loadings from PA items and another with strong loadings from items representing NA, SF, and IP dimensions. Nevertheless, as in previous work (Leventhal et al., 2008; Pettit et al.

, 2008), we utilized the four-dimension approach and computed subscale scores for each of the four dimensions by computing its respective items�� average score (these continuous scores were then later categorized for twin modeling, see ��Data Analysis�� section Anacetrapib below). This approach was adopted for several reasons. First, several prior studies examining relations between depressive symptom dimensions on the CESD and smoking variables have utilized the four-subscale approach and have found that all four of the different subscales exhibit different patterns of associations with smoking characteristics (Leventhal et al., 2008; Mickens et al., 2011; Pomerleau et al., 2003).