A total of 693 subjects were included in this ongoing study, including 31 subjects who met diagnostic criteria for PTSD and 662 who did not do so. A total of 73 of the subjects met diagnostic criteria for anxiety disorders other than PTSD. The number of subjects at the baseline assessment and at each of the subsequent follow-up assessment points is as follows: baseline: 693; first follow-up: 638; second follow-up: 605; third follow-up: 542; fourth follow-up: 381; fifth follow-up: 266. The 31 subjects with PTSD included 16 male participants (51.6%) and 15 female participants (48.4%), of whom 12 were Caucasian participants (38.7%) and 19 were African American participants(61.3%). The current mean age of the 31 individuals with PTSD was 22.2 +/- 2.5 years. The mean age of onset of PTSD was 15.4 +/- 5.6 years. Of those 31 individuals, 23 (74.2%) met diagnostic criteria for a lifetime SUD, of whom 19 (61.3%) met diagnostic criteria for a CUD, while 14 met diagnostic criteria for a lifetime alcohol use disorder (45.2%).
The mean age of onset of the CUD was 16.7 +/- 2.3 years. Two-thirds (67.7%) of those individuals with PTSD also met diagnostic criteria for a lifetime depressive disorder. Other common diagnoses among that sample included conduct disorder in 10 individuals (32.3%), oppositional defiant disorder in 8 individuals (25.8%), antisocial personality disorder in 5 individuals (16.1%), and attention deficit hyperactivity disorder in 3 individuals (9.7%). Of those 31 individuals, 23 (74.2%) reported having been treated for emotional problems, 5 (16.1%) reported having been treated for a drug problem, and 4 (12.9%) reported having been treated for an alcohol problem. One of the cases of PTSD was military-related, while the others were related to civilian traumas. Persons who met diagnostic criteria for PTSD were significantly more likely to be female (48.4% vs 26.9%, chi-square=6.8, p<0.009) and to be African American (61.3% vs 24.9%, chi-square=20.1, p<0.000) than persons who did not. They were also more likely to have a father with a history of a SUD (74.2% vs. 48.3%, chi-square=7.92, p=0.005) and were significantly more likely to meet diagnostic criteria for several other disorders than person who did not meet criteria for PTSD, including lifetime depressive disorder (chi-square=41.6, p<0.001), lifetime SUD (chi-square 27.2, p<0.001), lifetime CUD (chi-square=26.4, p<0.001), lifetime antisocial personality disorder (chi-square=13.0, p<0.001), and conduct disorder (chi-square=8.0, p<0.005). Those with PTSD were much more likely to have been treated for emotional problems during their lifetime (74.2% vs 38.9%, chi-square=15.3, p<0.001), to be treated for drug problems (16.1% vs 5.8%, chi-square=5.4, p=0.020), or to be treated for alcohol problems (12.9% vs 4.3%, chi-square=5.0, p=0.025).
The 161 subjects who met diagnostic criteria for a CUD included 136 male participants and 25 female participants, including 103 (64.%) Caucasian participants and 58 participants (36.%) of other races. Of those 161 individuals with a CUD, 19 (11.8%) met diagnostic criteria for PTSD. CUD were significantly more common among the individuals with PTSD (n=19, 61.3%) than among those who did not meet diagnostic criteria for PTSD (n=142, 21.5%) (chi-square=26.4, p<0.001). Of the 19 participants who met diagnostic criteria for both CUD and PTSD, 9 had PTSD first and CUD second, 9 had CUD first and PTSD second, and 1 had the onset of both diagnoses at the same age.
Logistic regression demonstrated that the development of CUD was associated with PTSD (Wald=12.7, p=0.000), deviance of peers (Wald=63.4, p=0.000), the TLI of the offspring (Wald=28.8, p=0.000), African American race (Wald=14.2, p=0.000), male gender (Wald=12.0, p=0.001), household SES (Wald=9.2, p=0.002), and paternal history of lifetime SUD (Wald=6.9, 0.009).
Path analysis (mediation analyses, ) demonstrated that PTSD is directly associated with CUD (beta=.19, z=4.40, p<0.001), peer deviance (beta=0.10, z=2.30, p=0.02), but not with TLI (beta=.03, z= .75, p=0.45). Path analysis also showed that higher peer deviance is associated with CUD (beta=.34, z=8.41, p<0.001). In addition, mediational analyses showed that PTSD mediated the association between peer deviance and CUD (beta=.02, z=2.21, p=0.03).
Analyses were conducted to determine whether PTSD significantly contributed to the variance in the development of CUD above and beyond the variance which resulted from all other variables. After adding PTSD, a change in R squared (variance) of 1.4% was noted, raising the percentage of variance that was explained from 25.5% to 26.6%, which was a significant increase (chi-square=6.32, df=1, p=0.012).