Table 2.

Association of coffee consumption and telomere length in multivariable linear regression.

Exposure	Model 1 1			Model 2 2			Model 3 3
	n	Effect in Years (95% CI) 4	p	n	Effect in Years (95% CI) 4	p	n	Effect in Years (95% CI) 4	p
Coffee intake	467,329	−0.22 (−0.28, −0.16)	<0.001	347,490	−0.13 (−0.20, −0.06)	<0.001	340,334	−0.12 (−0.19, −0.05)	<0.001
Instant coffee intake	110,610	−0.58 (−0.78, −0.38)	<0.001	85,829	−0.39 (−0.62, −0.16)	<0.001	84,842	−0.38 (−0.61, −0.15)	0.001
Filtered coffee intake	58,434	0.01 (−0.37, 0.40)	0.952	46,843	0.01 (−0.42, 0.44)	0.997	46,504	−0.04 (−0.47, 0.39)	0.862

¹ Covariables in model 1: age, sex, ethnicity, BMI. ² Covariables in model 2: Model 1 + Townsend deprivation index, total MET, WBC, CRP, qualification, smoking status, alcohol intake frequency, vascular/heart problems diagnosed by doctor, cancer diagnosed by doctor, diabetes diagnosed by doctor. ³ Covariables in model 3: Model 2 + oily fish intake, processed meat intake, beef intake, mutton/lamb intake, pork intake, cooked vegetable intake, salad/raw vegetable intake, fresh fruit intake, dried fruit intake. ⁴ Effect in years represents the year of age-related change in telomere length for each additional cup of coffee consumption. It was calculated by dividing β coefficient of multiple linear regression by the β coefficient of age-related telomere length decrease (0.023 per year).