Table 2.

Statistical methodological approaches for the determination of associative multimorbidity.

	Author, year, country	Statistical analysis	Patterns of multimorbidity	Stratification variables	Proximity measures	Type of clustering algorithm	Pattern number determination criteria (statistical)
1	Garin et al., 2015, China and India [17]	Exploratory factor analysis	Disease clusters	No	Tetrachoric correlation matrix	NS	Scree plot test and parallel analysis
		Exploratory factor analysis	Disease clusters	No	Tetrachoric correlation matrix	NS	Scree plot test and parallel analysis
2	Hussain et al., 2015, Indonesia [18]	Observed/expected ratio	Dyads and triads	Yes (by age and sex)	NA	Conditional probability	Chi-square test
3	Wang et al., 2015, China [19]	Observed/expected ratio	Dyads	No	NA	NS	Logistic regression
		Exploratory factor analysis	Disease clusters	No	Tetrachoric correlation matrix	NS	Eigenvalues ≥ 1.0
4	Wang et al., 2015, China [20]	Logistic regression	Dyads	Yes (by age and sex)	NA	Conditional probability	Adjusted odds ratio > 3.0
5	Gu et al., 2017, China [21]	Observed/expected ratio	Dyads	Yes (by age and sex)	NA	Conditional probability	Logistic regression
		Exploratory factor analysis	Disease clusters	Yes (by age and sex)	Correlation matrix	Principal factor method	Eigenvalues ≥ 1.0
6	Gu et al., 2018, China [22]	Exploratory factor analysis	Disease clusters	No	Tetrachoric correlation matrix	Principal factor method	Eigenvalues ≥ 1.0
7	Aoki et al., 2018, Japan [23]	Exploratory factor analysis	Disease clusters	No	Tetrachoric/polychoric correlation matrix	NS	Scree plot test and parallel analysis
8	Yao et al., 2019, China [24]	Hierarchical cluster analysis	Disease clusters	Yes (by sex and residential regions)	Yule's Q distance	Average linkage	NS

NS: not stated.