Attrition effects in the current study were examined in Mplus (Muthén & Muthén, 1998–2007) by regressing substance use endorsement counts (the sum of the five substance categories examined) on whether individuals returned for later assessments (controlling for the non-independent nature of twin data). The only significant effect indicated that males who did not complete the 17-year-old follow-up had used more substances by age 14 than those who did (r=0.36, z=2.16, p=0.03, Cohen’s d=0.25). Attrition is therefore unlikely to meaningfully impact our findings.
