Liability for Cannabis Use Disorder in Adolescent Girls

            The prevalence of substance use disorder (SUD) diagnosis is higher in males than females.  However, the shorter interval from first drug exposure to SUD diagnosis in females is consistent with higher liability to SUD compared to males despite lower prevalence. At the Center for Education and Drug Abuse Research (CEDAR) we have developed an index of transmissible liability (TLI) for substance use among offspring of fathers with and without SUD.  Our research has been conducted primarily on males due to insufficient number of female subjects in the CEDAR sample.  A large sample is required for reliable estimation of item parameters using item response theory.  The aim of this study was to derive a female version transmissible liability index (F-TLI) by harmonizing data from CEDAR’s female sample (N=228) with two other longitudinal NIDA-funded projects at the University of Pittsburgh: 1) “Female Adolescent Drug Abuse: Biobehavioral Development” (FAS, N=655) and “Teen Tobacco Use in a Birth Cohort” (MHP, N=142).

A pool of 298 items were selected from the CEDAR sample, and were correlated with SUD status of both parents, then submitted to EFA and CFA.  Any item with a factor loading of less than .40 was eliminated from further analyses.  Unidimensionality of 83 items was confirmed using CFA (chi-square=4153.07, df=3320, RMSEA<.033, CFI=.97, TLI=.97).  Only 12 of these items overlapped with the items of the boys TLI.  Similar items in the MHP and FAS samples were selected for data harmonization.  The three groups scored significantly different on the scale, with the FAS sample scoring higher than the MHP or CEDAR groups (M scores .23675, -.20332, and -.78725, respectively, F=136.82, p<.000).  However, the F-TLI scores for each group predicted future cannabis use disorder (CUD) in the CEDAR group (OR=2.98, CI= 1.81-4.89, p<.001), the MHP group (OR=8.80, CI= 2.44-31.74, p<.001) and showed a trend for the FAS group (OR=1.41, CI= .98-2.01, p=.06). 

While the instrument is still in development, these analyses provide support for a very different set of factors that contribute to liability for substance use, particularly cannabis use, in girls relative to boys.  The items of both TLI are indicative of behavioral undercontrol, however, the girls’ items were more reflective of deviant behaviors such as fighting and defiance whereas boys TLI items reflect more impulsivity.  These differences imply a need for gender-specific interventions.