Table 2.

Association of baseline leukocyte mitochondrial DNA copy number with high-sensitivity C-reactive protein trajectory^a

Model		Model 1			Model 2
Variable	Effectsb	95% CI	P	Effectsb	95% CI	P
Intercept	0.01	-0.02 to 0.04	.688	-2.44	-2.88 to -2.00	< .001
Mitochondrial DNA copy number	-0.06	-0.09 to -0.03	< .001	-0.05	-0.08 to -0.02	.002
Age				0.01	0.01 to 0.01	< .001
Sex
Women					(Reference)
Men				-0.25	-.033 to -0.18	< .001
Smoking
Never					(Reference)
Former smoker				0.17	0.07 to 0.27	< .001
Current smoker				0.28	0.17 to 0.38	< .001
BMI				0.06	0.05 to 0.07	< .001
Serum triglycerides				0.001	0.001 to 0.002	< .001
Serum high-density lipoprotein cholesterol				-0.004	-0.007 to -0.002	< .001
Serum low-density lipoprotein cholesterol				0.004	0.003 to 0.005	< .001
Hypertension				0.02	-0.04 to 0.09	.74
Chronic kidney disease				0.09	-0.001 to 0.181	.052
Time	0.08	0.07 to 0.09	< .001	0.03	0.01 to 0.05	< .001
Mitochondrial DNA copy number x time	-0.0001	-0.0091 to 0.0090	.990	Not included
Sex x time
Women					(Reference)
Men				0.03	0.01 to 0.05	.001
Education x time
Illiteracy				0.08	0.05 to 0.12	< .001
Elementary school				0.04	0.02 to 0.06	< .001
Junior high school				0.04	0.01 to 0.07	.011
≥ High school					(Reference)
Chronic kidney disease x time				0.05	0.02 to 0.08	.001

^aThe results of the longitudinal part of the joint analysis, in which the association between the baseline leukocyte mitochondrial DNA copy number and the serum high-sensitivity C-reactive protein levels (both of which were natural log-transformed and standardized) was modeled using linear mixed regression with adjustment for other model variables

^bChange in high-sensitivity C-reactive protein levels (in 1 standard deviation)