Comparisons were made between relatives of 78 control probands without ODD and CD (Controls, N=265), relatives of 10 control probands with ODD and without CD (ODD, N=37), relatives of 19 ADHD probands without ODD and CD (ADHD, N=71), relatives of 38 ADHD probands with ODD and without CD (ADHD+ODD, N=130), and relatives of 50 ADHD probands with ODD and CD (ADHD+ODD+CD, N=170). 11% of control probands (10/88) had ODD. Of ADHD probands, 82% had ODD (88/107) and 47% had conduct disorder (50/107). The ADHD and ADHD+ODD+CD groups had a significantly lower mean SES (i.e., higher Hollingshead’s score) compared to the Controls (). Therefore, pairwise comparisons between Controls and the ADHD and ADHD+ODD+CD groups controlled for SES. No differences were found in the age of probands or relatives.
Disruptive Behavior Disorders in Relatives
All three ADHD groups had significantly higher rates of ADHD (ADHD+ODD+CD=27.5%, OR=1.6 [1.3,1.9], p<0.001; ADHD+ODD=17.2%, OR=1.5 [1.2,2.0], p<0.001; ADHD=16.9%, OR=2.0 [1.2,3.1], p=0.004) compared to the Control group (5.4%, ), and the ADHD+ODD+CD group had a significantly higher rate of ADHD compared to the ODD group (8.1%, OR=1.6 [1.2,2.3], p=0.004). Similarly, rates of ODD were significantly higher in all three ODD groups (ADHD+ODD+CD=26.8%, OR=1.5 [1.3,1.8], p<0.001; ADHD+ODD=11.6%, OR=1.3 [1.0,1.7], p=0.04; ODD=16.2%, OR=3.2 [1.1,8.9], p=0.03) compared to the Control group (5.8%, ), and the ADHD+ODD+CD group had significantly higher rates of ODD compared to the ADHD+ODD group (11.6%, OR=2.8 [1.5,5.3], p=0.002). The ADHD+ODD+CD group also had significantly higher rates of CD or antisocial personality disorder (27.6%, ) compared to Controls and the two other ADHD groups (ADHD+ODD=8.5%, OR=4.1 [1.7,9.8], p=0.001; ADHD=12.7%, OR=1.6 [1.1,2.4], p=0.02; Controls=6.4%, OR=1.4 [1.2,1.6], p<0.001). Because ADHD alone in the proband did not increase the risk for ODD in the relatives, and ODD alone in the proband did not increase the risk for ADHD in the relatives, we can rule out the hypothesis of variable expressivity. The hypotheses of independent transmission and family subtype are still viable because all of the ADHD groups had elevated rates of ADHD and all of the ODD groups had elevated rates of ODD.
Rates of disorders in first-degree relatives
Comorbidity, or lack thereof, of ADHD and ODD in relatives of probands with ADHD+ODD will determine which hypothesis of familial transmission outside the context of CD is the best fit. In the ADHD+ODD group, ODD was significantly more common in relatives with ADHD than without ADHD (40.9% versus 4.7%, OR=14.0 [4.1,47.2], p<0.001), signaling co-segregation of the two disorders in this group of families. Therefore, we can rule out the hypothesis of independent transmission. There was no evidence for nonrandom mating in the ADHD+ODD group (both p=1.00 for ADHD mother with ODD father and ODD mother with ADHD father using Fisher’s exact test). Because the comorbidity in the ADHD+ODD group was not due to the assortative mating of parents with ADHD and ODD, we can accept family subtype as the best fitting hypothesis.
Other Disorders in Relatives
The ADHD+ODD+CD group had significantly higher rates of alcohol or drug dependence (, 31.5% versus ODD=6.2%, OR=1.9 [1.2,3.0], p=0.005; Controls=15.0%, OR=1.2 [1.0,1.3], p=0.05), major depressive disorder (, 27.4% versus ADHD+ODD=8.5%, OR=4.1 [2.0,8.3], p<0.001; ADHD=10.0%, OR=1.8 [1.1,3.1], p=0.03; ODD=2.8%, OR=2.4 [1.3,4.3], p=0.005; Controls=9.1%, OR=1.3 [1.1,1.5], p<0.001), bipolar disorder (, 11.8% versus ADHD+ODD=3.8%, OR=3.4 [1.0,10.8], p=0.04; ADHD=1.4%, OR=3.1 [1,1,8.3], p=0.03; Controls=3.4%, OR=1.3 [1.0,1.6], p=0.02), multiple (≥2) anxiety disorders (, 30.6% versus Controls=14.7%, OR=1.2 [1.1,1.4], p=0.002). In addition, the ADHD+ODD group had a significantly higher rate of alcohol or drug dependence compared to the ODD group (21.5% versus 6.2%%, OR=2.0 [1.0,4.1], p=0.05) but did not differ significantly from the Controls.