Table 5.

Proportions of latent subgroups by diagnosis

Class Name, N (%)	AN-R‡ N (%)	AN-BP N (%)	BN N (%)	ARFID N (%)	BED N (%)	OSFED N (%)
“Low ACEs” † 503 (47%)	145† (52%)	147† (49%)	62† (46%)	68† (46%)	29† (37%)	52† (41%)
“Household ACEs” 347 (33%)	81‡ (29%)	99 (33%)	47 (35%)	48 (33%)	24 (31%)	48 (38%)
“Abuse ACEs” 99 (9%)	25‡ (9%)	20 (7%)	10 (7%)	17 (12%)	14 (18%)	13 (10%)
“All ACEs” 112 (10%)	27‡ (10%)	32 (11%)	15 (11%)	13 (9%)	11 (14%)	14 (11%)

^†Reference group of “Low ACEs” latent subgroup in categorical outcome of the multinomial logistic regression. ^‡Reference group of patients with AN-R in the categorical predictor of the multinomial logistic regression. Bolded numbers represent a significant increase in the risk ratio of being in the identified latent subgroup row (vs. the “Low ACEs” subgroup) for patients with the identified diagnosis column (vs. patients with AN-R)—a comparison across four total prevalence rates. No inferential comparisons have been made between any two prevalence rates. Latent subgroup row percentages add up to 100% down each diagnosis column. Analysis was multinomial logistic regression