Table 2.

Pooled Estimates of the Associations Between CMV Seropositivity or CMV IgG Antibody Quartiles with All-Cause and Cardiovascular Mortality, Compared to CMV-Seronegative Status

Cytomegalovirus	All-Cause Mortality, HR (95% CI)			Cardiovascular Mortality, HR (95% CI)
	Model 1b	Model 2c	Model 3c	Model 1b	Model 2c	Model 3c
Seronegativitya	1	1	1	1	1	1
Seropositivity all	1.05 (.97–1.14)	1.07 (.99–1.16)	1.07 (.99–1.16)	0.97 (.83–1.13)	1.00 (.82–1.21)	1.00 (.82–1.2)
IgG antibody quartile 1	0.98 (.84–1.15)	0.97 (.80–1.18)	0.99 (.82–1.20)	0.96 (.66–1.38)	0.94 (.56–1.59)	0.93 (.55–1.58)
IgG antibody quartile 2	1.07 (.96–1.19)	1.01 (.88–1.15)	1.02 (.90–1.16)	0.94 (.76–1.16)	0.85 (.65–1.11)	0.83 (.61–1.12)
IgG antibody quartile 3	1.08 (.92–1.27)	1.13 (.99–1.29)	1.13 (.99–1.29)	1.03 (.76–1.40)	1.17 (.90–1.51)	1.15 (.89–1.49)
IgG antibody quartile 4	1.16 (1.04–1.29)	1.13 (.99–1.29)	1.13 (.99–1.29)	0.97 (.75–1.25)	0.96 (.72–1.26)	0.96 (.72–1.26)

Cox regression analyses within individual cohorts (Supplementary Table 1) were performed in 3 models:

Model 1: adjustment for age and sex (for PROSPER, also country and statin use);

Model 2: adjustment for model 1 plus body mass index, education, smoking status, number of comorbidities, and medications;

Model 3: adjustment for model 2 plus log transformed C-reactive protein.

Abbreviations: CMV, cytomegalovirus; CI, confidence interval; HR, hazard ratio; IgG, immunoglobulin G; PROSPER, Prospective Study of Pravastatin in the Elderly at Risk.

^aCMV-seronegative individuals were the reference group.

^bAll cohorts included.

^cAll cohorts except for Leiden Longevity Study (LLS) F1 and F2.