Table 2

Meta-regression models explaining study heterogeneity on the association between income inequality and poor SRH

Variables (number of studies)	Bivariate				Multivariate
	Bivariate				Model 1			Model 2			Model 3
	β	SE	p	τ²^*	β	SE	p	β	SE	p	β	SE	p
Intercept	–	–	–		−0.014	0.014	0.33	0.0131	0.005	0.006	0.023	0.010	0.02
Average population (1000)
Small: 8–820 (5)	Referent				Referent			Referent			Referent
Large: 2612–5000 (6)	0.005	0.009	0.57		−0.001	0.010	0.90	−0.0019	0.005	0.73	−0.001	0.010	0.91
Country: 23 700–30 900 (3)	0.088	0.021	<0.001	8.0⁻⁵	0.081	0.021	<0.001	0.0884	0.020	<0.001	0.078	0.022	<0.001
Continuous (per 500 000 population) (all)	0.002	0.0004	<0.001	8.6⁻⁵
Median GINI
≥0.30 (referent: <0.3) (12)	0.038	0.015	0.01	7.6⁻⁵	0.034	0.016	0.03
≥Median (=0.37) (referent: <0.37) (7)	−0.008	0.006	0.17	2.0⁻⁵
Time lag between income inequality and SRH
No (11)	Referent							Referent
Yes (3)	0.019	0.004	<0.001	0.0				0.0209	0.005	<0.001
Adjusting for area income
No (9)	Referent										Referent
Yes (5)	−0.021	0.009	0.02	8.4⁻⁵							−0.016	0.010	0.12
τ²^*	–	–	–	–	8.6⁻⁵			0.0			9.0⁻⁵

Intercepts in these meta-regression models represent the degree of the associations between income inequality and poor SRH when explanatory variables are in a referent category or zero (for continuous explanatory variables).
↵* τ² represents the magnitude of residual heterogeneity.
SRH, self-rated health.