Chronic inflammation role in the obesity-diabetes association: a case-cohort study

Table 3 Hazards ratios for diabetes according to obesity levels, adjusting for insulin prior to inflammation markers

	Main exposure
	Overweight^a	Obesity^b	Waist circumference 4^thquartile^c
	HR (95%CI)	HR (95%CI)	HR (95%CI)
Model 1	2.07 (1.45–2.95)	6.43 (4.48–9.21)	8.30 (5.60–12.28)
Model 1’	1.27 (0.87–1.86	2.67 (1.73–4.13)	3.63 (2.28–5.78)
Model 2	1.23 (0.84–1.81)	2.51 (1.63–3.88)	3.43 (2.16–5.46)
Model 3	1.09 (0.74–1.60)	1.85 (1.20–2.86)	2.54 (1.58–4.08)
Model 4	1.29 (0.86–1.94)	2.48 (1.52–4.05)	4.10 (2.34–7.20)
Model 5	1.17 (0.78–1.76)	2.26 (1.37–3.73)	3.19 (1.58–6.45)

Model 1: Adjusted for age, study center, ethnicity, gender, and family history of diabetes.
Model 1’: Model 1 + ln-insulin.
Model 2: Model 1’ + adiponectin.
Model 3: Model 2 + inflammation score, C3; oxidized LDL, ICAM-1.
Model 4: Model 3 + leptin (gender-specific quartiles).
Model 5: Model 4 + hypertension, non-esterified fatty acids, ln-triglycerides, ln-triglycerides², HDL, BMI, BMI², waist-to-hip ratio, (those not already in the model).
^a Hazards ratio of overweight (25 kg/m² ≤ BMI < 30 kg/m²) versus BMI < 25 kg/m².
^b Hazards ratio of obese (BMI ≥ 30 kg/m²) versus BMI < 25 kg/m².
^c 4th vs. 1st quartile. In men: waist circumference >102 cm vs. <91 cm; in women: waist circumference >101 cm vs. <84 cm.

ISSN: 1758-5996