The association between leukocyte telomere length and chronic obstructive pulmonary disease is partially mediated by inflammation: a meta-analysis and population-based mediation study

Table 4 Mediation of 5 inflammatory factors for the associations between telomere length and COPD

	M:CRP	P value	M: Fibrinogen	P value	M:WBC	P value	M:B-Neu	P value	M:B-Eos	P value
	β(95%CI)	P value	β(95%CI)	P value	β(95%CI)	P value	β(95%CI)	P value	β(95%CI)	P value
X → M	−0.337 (−0.490, −0.183)	< 0.001	0.006 (−0.022, 0.033)	0.696	−0.069 (−0.106, −0.032)	< 0.001	−0.060 (−0.109, −0.011)	0.016	−0.049 (−0.128, 0.031)	0.232
M\|X → Y	0.193 (0.090, 0.296)	< 0.001	0.941 (0.367, 1.516)	0.001	0.670 (0.242, 1.099)	0.002	0.532 (0.200, 0.864)	0.002	0.152 (−0.054, 0.358)	0.148
X\|M → Y	−0.692 (−1.251, −0.133)	0.015	−0.767 (−1.326, −0.209)	0.007	−0.700 (−1.260, −0.141)	0.014	−0.723 (−1.283, −0.163)	0.011	−0.763 (−1.335, −0.191)	0.009
X → M → Y	−0.065 (−0.124, −0.026)	–	0.005 (−0.021, 0.036)	–	−0.046 (−0.094, −0.017)	–	−0.032 (−0.077, −0.007)	–	−0.007 (−0.038, 0.003)	–

X = telomere length; M = mediator (log-transformed CRP, fibrinogen, WBC, B-Neu or B-Eos); Y = COPD
Analytical process includes: 1) an effect of X on M (X → M); 2) an effect of M on Y controlling for X (M|X → Y); 3) a direct effect of X on Y, i.e. an effect of X on Y controlling for M (X|M → Y); and 4) an indirect effect of X on Y, i.e. an effect of X on Y mediated by M (X → M → Y). Each process was adjusted for age, sex, education, race, poverty income ratio, body mass index, smoking status, alcohol-use status

ISSN: 1471-2466