Table 3.

Hazard ratios (HRs) for incident metabolic syndrome according to the fatty liver status at baseline and at follow-up

	HR [95 % CI]a
	No fatty liver—no fatty liver (reference group) (n = 1344)	Fatty liver—no fatty liver(resolved fatty liver) (n = 112)	No fatty liver—fatty liver (incident fatty liver) (n = 251)	Fatty liver—fatty liver (persisting fatty liver) (n = 382)
Model 1	1	3.05 [1.47–6.36]	5.34 [3.29–8.65]	7.34 [4.77–11.28]
Model 2	1	3.10 [1.48–6.47]	5.31 [3.27–8.63]	7.42 [4.80–11.47]
Model 3	1	1.29 [0.60–2.80]	3.31 [1.99–5.51]	2.60 [1.61–4.20]

Model 1: adjusted for age and sex; model 2: adjusted for model 1 plus alcohol, smoking status, and exercise; model 3: adjusted for model 2 plus glucose, waist circumference, systolic blood pressure, triglyceride, and HDL-C

^aLogistic regression was used to determine hazard ratio (HR) for developing metabolic syndrome at follow-up