Income inequality and cardiovascular disease risk factors in a highly unequal country: a fixed-effects analysis from South Africa

Table 3 Risk ratios and 95% confidence intervals for the association between income inequality and behavioral CVD risk factors

	Longitudinal Fixed-Effects Models
	Model 1^a	Model 2^b	Model 3^c
Smoking (risk ratio)	0.99 (0.84, 1.16)	0.97 (0.83, 1.15)	0.94 (0.74, 1.19)
High alcohol (risk ratio)	1.36 (0.93, 1.97)	1.35 (0.92, 1.97)	1.22 (0.54, 2.80)
Physical inactivity (risk ratio)	0.91 (0.82, 1.02)	0.92 (0.82, 1.01)	0.93 (0.77, 1.13)

Estimates correspond to a change of 0.10 in the Gini coefficient from conditional (fixed-effects) Poisson regression models. Standard errors are clustered by district
^aControls for survey wave
^bControls for: wave; marital status, employment status; and household log household income, size, and receipt of government grants
^cAdds district-level variables to Model 2: mean age; log mean monthly equivalized household income; and percents female, African, unemployed, with no education, with tertiary education, and rural

ISSN: 1475-9276