Table 3.

Subgroup Analysis for the Associations Between Choline Metabolite Score and Risk of CVD

Characteristics	HR (95% CI) Per SD Increment	P Value for Interaction
Sex
Men (n=452)	2.31 (1.13–4.71)	0.83
Women (n=528)	2.23 (1.03–4.83)
Age, y
≤65 (n=338)	2.67 (1.06–6.69)	0.49
>65 (n=598)	2.42 (1.27–4.64)
Obesity, kg/m2
≤30 (n=535)	2.65 (1.31–5.35)	0.79
>30 (n=445)	2.04 (0.92–4.52)
Smoking status
Current/former smoking (n=402)	2.45 (1.19–5.06)	0.63
Ever smoking (n=578)	2.40 (1.09–5.28)
Family history of CHD
Yes (n=237)	2.19 (1.22–3.95)	0.15
No (n=743)	3.78 (0.84–5.42)
Baseline type 2 diabetes mellitus
Yes (n=486)	2.22 (1.00–4.92)	0.87
No (n=494)	2.37 (1.20–4.69)
Baseline hypertension
Yes (n=817)	1.94 (1.08–3.46)	0.36
No (n=163)	3.12 (0.75–5.65)
Baseline dyslipidemia
Yes (n=692)	2.55 (1.31–4.94)	0.73
No (n=288)	2.03 (0.86–4.78)

Multivariate adjusted HRs (95% CIs) of incident CVD for 1‐SD increment in the baseline choline metabolite score. Data were adjusted for age, sex, body mass index, physical activity (metabolic equivalent task units in min/d), family history of premature heart disease, smoking, hypertension, dyslipidemia, and diabetes mellitus and stratified by intervention group, except for the stratification variable in each model. P values for interaction were derived from Cox models adjusted as above, including an interaction term between the stratification variable and the choline metabolite score. CHD indicates coronary heart disease; CI, confidence interval; CVD, cardiovascular disease; and HR, hazard ratio.