shows all patients in the cohort.
Table 1 Characteristics* of children with ASD, aged 5–18y by parent-reported lifetime history of psychiatric comorbidity (n = 4343).
Participants were more likely to have a single psychiatric comorbidity (26.9%) than two (14.4%), three (6.3%), or four (1.5%) comorbidities (data not shown). Of those with at least one comorbidity, 45.2% had two or more.
The most common comorbid diagnosis was AD/HD or ADD (38.1%), followed by anxiety disorders (26.2%), depression (11.0%), and bipolar disorder (5.2%), as seen in . Overall, only 23/4343 (0.5%) reported a diagnosis of schizophrenia, and thus this disorder was not included in the analysis.
Table 2 Characteristics of children with ASD by presence of ≥1 psychiatric comorbidity, aged 5–18y, significantly different from peers without particular comorbidity (P < .01 by chi-square).
shows significant independent variables (P < .01) on bivariate analysis by child/family characteristic for each reported psychiatric comorbidity.
There was no difference in proportion of overall comorbidity by gender on bivariate analysis. However, older age, diagnosis of AS, higher SRS score (when available), presence of ID, lack of autistic regression, non-Hispanic ethnicity, initial evaluation by a psychiatrist or psychologist, and history of maternal psychiatric illness were all associated with increased risk of psychiatric comorbidity (chi-square P < .01).
Bivariate chi-square analysis of comorbidity by race was significant only for anxiety (P < .02). Hispanic children were less likely to have an anxiety or depression diagnosis, compared with their non-Hispanic peers (P < .01). Chi-square analysis of depression showed that children of women with college education but not graduate degrees were more likely than other children to have a depression diagnosis (13.0% versus 9.5% and 10.0%, P = .03). Children of college-educated mothers were less likely to have bipolar disorder diagnosis (P = .011).
As shown in , logistic regression of anxiety, depression, and bipolar disorder cooccurrence were high among those with at least one comorbidity, adjusted for age, P < .001); the odds of having any anxiety disorder was two times higher for those with depression compared to those without depression, for example. AD/HD or ADD, however, was negatively correlated with anxiety (OR 0.1, 95% CI, 0.1, 0.2) and depression (OR 0.7, 95% CI, 0.5, 0.9).
Odds ratio (OR) of lifetime history of comorbidity among study participants with ASD, with at least one psychiatric comorbidity (n = 2133).
shows the results of five separate multivariate logistic regressions for any comorbidity and separately for anxiety, depression, bipolar disorder, and AD/HD or ADD, adjusting for maternal education, ethnicity, initial evaluator, location, and maternal psychiatric illness.
Multivariate logistic regression odds ratios (95% CI) of overall and individual parent-reported lifetime psychiatric comorbidities in ASD (n = 2219), by statistically significant factor (P < .001 unless otherwise noted).
Female gender was moderately protective for both bipolar disorder (OR 0.5, 95% CI, 0.3, 0.9) and for AD/HD or ADD (OR 0.8, 95% CI, 0.6, 1.0). PDD-NOS and AS were significantly correlated with increased odds of each comorbidity and for overall comorbidity. SRS score was positively correlated with increased odds of each and overall comorbidity (P < .001). For every 10-point increase in SRS score, for example, odds of any comorbidity was at least 20% higher (OR 1.2, 95% CI, 1.1, 1.3) than for those without comorbidity. Presence of ID was not associated with any comorbidity except as a negative correlate of depression (OR 0.6, 95% CI, 0.4, 0.9; P < .01). Autistic regression was associated with lower risk of any comorbidity (OR 0.8, 95% CI, 0.6, 1.0), any mood disorder (OR 0.7, 95% CI, 0.5, 1.0), and AD/HD or ADD (OR 0.7, 95% CI, 0.5, 0.8).