Table 4. Relationship between volume of LTPA and incident T2DM in residents with IFG in Taiwan.

	Volume of LTPA
	Inactive	Low	Moderate	High
Cox proportional hazards method was used with inactive LTPA as the reference category and adjustment for the confounders at baseline, including age, sex, marital status, education, physical labour at work, smoking, alcohol drinking, sleep duration, vegetable intake, systolic blood pressure, heart rate and total cholesterol. Volume of LTPA was classified as inactive (< 3.75 MET·h/week, low (3.75–7.5 MET·h/ week), moderate (7.5–15.0 MET·h/week) and high (≥ 15.0 MET ·h/week). IFG: impaired fasting glucose; LTPA: leisure-time physical activity.
Numbers of incident cases	2,535	731	542	612
Annual incidence/%	3.12	2.96	2.54	2.63
HR (95% CI )	1.00	0.88 (0.80–0.98)	0.80 (0.71–0.90)	0.75 (0.67–0.83)