Determinants of multimorbidity in older adults in Iran: a cross-sectional study using latent class analysis on the Bushehr Elderly Health (BEH) program

Table 2 Comparison of latent class analysis models with different numbers of latent classes based on model selection statistics

Number of latent class	Number of parameters estimated	G²	df	AIC	BIC	X²	Entropy	Maximum log-likelihood
1	9	953	502	22,904	22,956	1584	4.71	-11443.10
2	19	691	492	22,661	22,771	857	4.66	-11311.95
3	29	510	482	22,501	22,669	668	4.63	-11221.81
4	39	457	472	22,468	22,694	575	4.61	-11195.19
5	49	425	462	22,456	22,740	528	4.61	-11179.19
6	59	389	452	22,440	22,782	507	4.60	-11161.22
7	69	366	442	22,437	22,837	491	4.59	-11149.83

Note LCA: latent class analysis; AIC: Akaike information criterion; BIC: Bayesian information criterion. df: degree of freedom; G²: the likelihood-ratio statistic

ISSN: 1471-2318