Table 3.

Model fitting from the three-class multi-level latent class model

Disorder	Class 1		Class 2		Class 3
	Prob	95% CI	Prob	95% CI	Prob	95% CI
GAD	0.12	0.05–0.25	0.56	0.42–0.69c	0.67	0.58–0.76a
SAD	0.11	0.05–0.20	0.26	0.17–0.37d	0.41	0.33–0.50a
PD	0.10	0.06–0.17	0.03	0.00–0.21	0.48	0.38–0.59af
Tics	0.13	0.07–0.23	0.41	0.30–0.52c	0.27	0.20–0.35b
MDD	0.27	0.20–0.36	0.24	0.15–0.35	0.68	0.56–0.77ae
Mania	0.03	0.01–0.09	0.00	0.00–1.00	0.19	0.13–0.27a
GrD	0.16	0.08–0.28	0.48	0.37–0.59c	0.59	0.50–0.68a
Som	0.06	0.02–0.13	0.16	0.09–0.26	0.53	0.43–0.63af
Average prevalence	0.34	0.27–0.52	0.32	0.20–0.46	0.30	0.22–0.39
Intra-cluster correlation	0.44 (0.30–0.59)

Prob, Probability; CI, confidence interval; GAD, generalized anxiety disorder; SAD, separation anxiety disorder; PD, panic disorder; MDD, major depressive disorder; GrD, grooming disorders (trichotillomania, pathological skin picking); Mania, bipolar disorder (mania); Som, somatic disorders (body dysmorphic disorder and hypochondriasis).

^a3 v. 1; p<0.001;

^b3 v. 1; p<0.05;

^c2 v. 1; p<0.001;

^d2 v. 1; p<0.05;

^e3 v. 2; p<0.001;

^f3 v. 2; p<0.05.

(1) The results in Table 3 are exponentiated coefficients. The interpretation for 0.517 in Table 1 model 1 is: Pr(C=2|female)/Pr(C=1|female)=0.52×Pr(C=2|male)/Pr(C=1|male).

(2) The model only includes 2 v. 1, 3 v. 1 parts. The 3 v. 2 results were calculated after fitting the models.

(3) Male is the reference group for gender.

(4) The number of subjects in the various models changes depending upon the subjects, with available clinical data for the variables entered into the model.