Skip to main content
. 2018 Dec 3;173(2):153–159. doi: 10.1001/jamapediatrics.2018.4208

Table 4. Weighted Binary Logistic Regression Models for Dichotomized Treatment Types in US Children With Current Autism Spectrum Disorder (ASD).

Characteristic Behavioral Treatmenta Medication Treatmenta Neither Treatmenta
Odds Ratio (95% CI)b P Value Odds Ratio (95% CI)b P Value Odds Ratio (95% CI)b P Value
Age, y
3-5 1 [Reference] NA 1 [Reference] NA 1 [Reference] NA
6-11 0.63 (0.28-1.41) .26 3.73 (1.11-12.49) .03 1.22 (0.55-2.72) .62
12-17 0.33 (0.15-0.71) .005 4.02 (1.20-13.41) .02 1.81 (0.87-3.78) .11
Sex
Male 1 [Reference] NA 1 [Reference] NA 1 [Reference] NA
Female 1.36 (0.77-2.42) .29 0.56 (0.30-1.03) .06 0.90 (0.51-1.59) .72
Race/ethnicity
Non-Hispanic white 1 [Reference] NA 1 [Reference] NA 1 [Reference] NA
Hispanic 0.97 (0.47-2.01) .93 0.47 (0.20-1.10) .08 0.86 (0.40-1.85) .71
Non-Hispanic black 1.17 (0.55-2.50) .69 1.44 (0.71-2.90) .31 0.68 (0.35-1.33) .26
Non-Hispanic Asian 1.73 (0.73-4.11) .21 0.49 (0.21-1.11) .09 0.63 (0.26-1.54) .31
Other 0.54 (0.27-1.08) .08 0.90 (0.43-1.86) .77 1.88 (0.90-3.92) .09
Family income to poverty ratio
<1.0 1 [Reference] NA 1 [Reference] NA 1 [Reference] NA
1.0-1.9 0.74 (0.35-1.56) .42 1.17 (0.55-2.50) .69 1.38 (0.66-2.88) .40
2.0-3.9 1.00 (0.48-2.09) .99 0.90 (0.45-1.83) .78 1.20 (0.57-2.56) .63
≥4.0 0.83 (0.37-1.84) .65 1.20 (0.54-2.69) .66 1.37 (0.59-3.17) .46
Insurance type
Private only 1 [Reference] NA 1 [Reference] NA 1 [Reference] NA
Public only 1.12 (0.60-2.10) .71 1.87 (0.98-3.55) .06 0.97 (0.50-1.88) .93
Private and public 0.77 (0.40-1.50) .45 1.24 (0.53-2.93) .62 1.00 (0.52-1.94) .99
Unspecified 0.04 (0-0.36) .004 2.15 (0.46-9.92) .33 12.31 (1.45-104.45) .02
Not insured 0.12 (0.04-0.42) <.001 0.54 (0.09-3.11) .49 5.92 (1.49-23.58) .01

Abbreviation: NA, not applicable.

a

Based on nonmutually exclusive categories. For example, in the analysis on risk factors for behavioral treatment, the outcome variable was dichotomized as receiving behavioral treatment vs not receiving behavioral treatment.

b

Adjusted simultaneously for all other characteristic variables in the table as well as the state.