Longitudinal association of C-reactive protein and Haemoglobin A1c over 13 years: the European Prospective Investigation into Cancer - Norfolk study

Table 2 Cross-sectional and longitudinal association of Haemoglobin A_1c with baseline levels of C-reactive protein in the EPIC-Norfolk cohort study (1993–2011)

	Cross-sectional association^a		Longitudinal association^b
	n = 7,485		n = 4,595
CRP levels (mg/L)	Age-sex adjusted	Multivariable adjusted	Age-sex adjusted	Multivariable adjusted
	Mean difference (95 % confidence interval)	Mean difference (95 % confidence interval)	Annual change (95 % confidence interval)	Annual change (95 % confidence interval)
≤1	Reference	Reference	0.043 (0.041, 0.045)	0.039 (0.031, 0.046)
1.1 – 3	0.065 (0.032, 0.098)	0.040 (0.007, 0.074)	0.045 (0.043, 0.047)	0.040 (0.032, 0.047)
3.1 – 10	0.198 (0.159, 0.238)	0.143 (0.101, 0.185)	0.048 (0.044, 0.051)	0.040 (0.031, 0.047)
>10	0.308 (0.236, 0.379)	0.256 (0.183, 0.328)	0.052 (0.045, 0.059)	0.041 (0.028, 0.054)
1-SD increase in CRP	0.086 (0.071, 0.10)	0.059 (0.046, 0.072)	0.003 (0.0002, 0.005)^c	0.002 (−0.0007, 0.004)^c
P-value	<0.001	<0.001	0.02	0.15

CRP C-reactive protein, EPIC-Norfolk European Prospective Investigation into Cancer in Norfolk, SD standard deviation
^aMean difference in HbA_1c (%) in each category of CRP and the reference category (CRP ≤ 1 mg/l). Multivariable cross-sectional analyses are adjusted for baseline covariates: age, sex, body mass index, waist circumference, smoking, physical activity, alcohol intake, medical history of cancer, myocardial infarction, and stroke, corticosteroid medication, and in women only menopausal status and hormone replacement therapy
^bAnnual change calculated as an interaction of baseline CRP categories with time in participants who had at least one follow-up health examination. Adjusted for baseline age and sex. Multivariable analyses are additionally adjusted for body mass index, waist circumference, smoking, physical activity, alcohol intake, medical history of cancer, myocardial infarction, and stroke, corticosteroid medication, and in women only menopausal status and hormone replacement therapy, all treated as time varying variables. Analysis is restricted to individuals who had at least two measurements of HbA_1c
^cValues are change in HbA_1c over 13 years

ISSN: 1475-2840