Table 2.
Logistic regression model investigating ADHD screening status as a predictor of self-harm, adjusted for age, biological sex, SES, symptoms of depression and conduct disorder, as well as familial history of self-harm and suicide attempts (N = 8839)
| Dependent variable: Self-harm | ADHD-SC+ (N = 2117) 1 OR (95%CI) |
ADHD-SC- (N = 6722) 1 OR (reference) |
|---|---|---|
| Model 1: Unadjusted | 3.02 (2.57–3.54) | 1.00 |
| Model 2: Demographics 2 | 2.75 (2.34–3.25) | 1.00 |
| Model 3: Demographics 2 + familial history of self-harm and suicide | 2.70 (2.29–3.19) | 1.00 |
| Model 4: Demographics 2 + YCD ≥ 1 | 2.40 (2.03–2.84) | 1.00 |
| Model 5: Demographics 2 + SMFQ ≥ 12 | 1.72 (1.44–2.06) | 1.00 |
| Model 6: Demographics 2 + SMFQ ≥ 12 + YCD ≥ 1 + familial history of self-harm and suicide | 1.58 (1.31–1.89) | 1.00 |
1 Includes only participants with no missing values on included variables. 2Age + biological sex + parent’s level of education. ADHD-SC+: Screen positives for ADHD, ADHD-SC-: Screen negatives for ADHD, OR: Odds ratio, CI: Confidence interval, SMFQ: Short Mood and Feelings Questionnaire, YCD: The Youth Conduct Disorder Scale