Skip to main content
. 2022 Mar 3;47(6):1263–1270. doi: 10.1038/s41386-022-01295-4

Table 2.

Model-estimated Mean changes from baseline (95% Confidence Intervals) for study outcomes at 10 weeks.

Mirtazapine (N = 20)1 ES2 p Placebo (N = 10) ES2 p Mirtazapine-Placebo ES2 p
Pediatric Anxiety Rating Scale, 5-Item Total3 −4.9 (−7.3, −2.6) 1.76 <0.001 −3.2 (−6.5, 0.2) 1.13 0.07 −1.8 (−5.9, 2.4) 0.63 0.64
Screen for Childhood Anxiety Related Emotional Disorders −9.7 (−15.4, −4.0) 0.77   0.003 −5.1 (−12.6, 2.4) 0.41 0.17 −4.6 (−13.2, 4.1) 0.36 0.29
Child and Adolescent Symptom Inventory Anxiety Items −9.4 (−14.1, −4.8) 0.98 <0.001 −5.5 (−11.9, 0.8) 0.58 0.09 −3.9 (−11.6, 3.8) 0.41 0.31
Aberrant Behavior Checklist
I - Irritability −3.7 (−7.5, 0.1) 0.43   0.053 −3.3 (−8.5, 2.0) 0.38 0.21 −0.5 (−6.9, 6.0) 0.06 0.88
II - Lethargy/Social Withdrawal −4.9 (−7.6, −2.1) 0.59   0.001 −2.0 (−5.7, 1.7) 0.24 0.28 −2.9 (−7.2, 1.5) 0.35 0.19
III - Stereotypic Behavior −2.8 (−4.3, −1.2) 0.50   0.001 −1.9 (−4.0, 0.2) 0.34 0.08 −0.9 (−3.3, 1.5) 0.16 0.46
IV - Hyperactivity/Noncompliance −5.0 (−8.1, −1.9) 0.43   0.003 −1.0 (−5.4, 3.3) 0.09 0.62 −3.9 (−9.3, 1.4) 0.34 0.14
V - Inappropriate Speech −0.7 (−1.9, 0.5) 0.19   0.26 −0.5 (−2.1, 1.2) 0.13 0.56 −0.2 (−2.2, 1.8) 0.06 0.84
Children’s Sleep Habit Questionnaire −4.3 (−7.1, −1.4) 0.49   0.004 −3.9 (−7.5, −0.2) 0.44 0.04 −0.4 (−4.6, 3.8) 0.05 0.85
Developmental Disability-Child Global Assessment Scale 4.6 (1.9, 7.3) 0.76   0.002 1.8 (−1.9, 5.5) 0.30 0.32 2.8 (−1.7, 7.4) 0.46 0.21

1One participant assigned to mirtazapine who withdrew due to an adverse effect was missing week 6 and week 10 measurements.

2Effect sizes were calculated by dividing the absolute value of the model-estimated mean 10-week change from baseline (mirtazapine and placebo) or the model-estimated mean treatment difference at 10 weeks (mirtazapine-placebo) by the pooled standard deviation at baseline.

3The Pediatric Anxiety Rating Scale (PARS) 5-item total score was a co-primary outcome measure for the study. The corresponding confidence intervals and p-value have been adjusted to reflect a Bonferroni correction accounting for the study’s two co-primary outcomes, the PARS and the Clinical Global Impressions-Improvement subscale.