Relationship between BMI and risk of impaired glucose tolerance and impaired fasting glucose in Chinese adults: a prospective study

Table 2 Relationship between BMI and the risk of IGT and IFG

Outcome	BMI (kg/m²)				Pearson²	P
	18.5–23.9 (n = 2,856)	< 18.5	24.0–27.9	≥ 28.0
	18.5–23.9 (n = 2,856)	(n = 232)	(n = 1,770)	(n = 720)
IGT
Number of events	227	24	203	98
Follow-up year	17,633.2	1,413.1	10,709.3	4,297.5
Incidence rate (no./1,000 person-years)	12.9	17.0	19.0	22.8
Odds ratio (95% CI)
Model 1 [AIC = 3,409.76]	1.00 (Ref)	1.34 (0.86, 2.11)	1.55 (1.27, 1.90)	2.12 (1.64, 2.74)	24.93	0.096
Model 2 [AIC = 3,389.97]	1.00 (Ref)	1.33 (0.85, 2.09)	1.51 (1.23, 1.85)	2.12 (1.64, 2.75)	314.55	0.196
Model 3 [AIC = 3,378.10]	1.00 (Ref)	1.38 (0.88, 2.17)	1.35 (1.08, 1.70)	1.77 (1.27, 2.47)	1,116.15	0.618
IFG
Number of events	105	4	90	58
Follow-up year	17,633.2	1,413.1	10,709.3	4,297.5
Incidence rate (no./1,000 person-years)	6.0	2.8	8.4	13.5
Odds ratio (95% CI)
Model 1^* [AIC = 1,961.46]	1.00 (Ref)	0.48 (0.18, 1.33)	1.49 (1.11, 1.99)	2.69 (1.92, 3.77)	20.06	0.271
Model 2^* [AIC = 1,763.82]	1.00 (Ref)	0.58 (0.21, 1.61)	1.57 (1.15, 2.14)	3.00 (2.11, 4.26)	212.43	0.167
Model 3^* [AIC = 1,759.32]	1.00 (Ref)	0.61 (0.22, 1.69)	1.33 (0.94, 1.88)	2.22 (1.39, 3.54)	559.29	0.441

Note. Model 1 adjusts age and gender. Model 2 adjusts urban and rural areas, physical activity, and education on the basis of Model 1. Model 3 adjusts hypertension, abdominal obesity, and dyslipidemia on the basis of Model 2. Model 1^* adjusts age and gender. Model 2^* adjusts urban and rural areas, physical activity, and family history of diabetes mellitus on the basis of Model 1. Model 3^* adjusts hypertension, abdominal obesity, and dyslipidemia on the basis of Model 2^*
BMI body mass index, IGT impaired glucose tolerance, IFG impaired fasting glucose

ISSN: 1471-2458