TABLE 3.

Association between vitamin C and gout according to subgroups¹

	n Placebo/n active	HR (95% CI)	P-interaction2
Age, y
<60	2938/2953	0.80 (0.63, 1.01)	0.66
60–69	2348/2348	0.90 (0.73, 1.10)
≥70	2026/2028	0.92 (0.75, 1.14)
Black
No	7239/7249	0.87 (0.77, 0.99)	0.54
Yes	73/80	1.09 (0.44, 2.70)
Prevalent gout
No	6823/6865	0.90 (0.76, 1.07)	0.58
Yes	489/464	0.87 (0.71, 1.05)
BMI, kg/m2
<25	3080/3060	0.74 (0.59, 0.92)	0.01
25 to <30	3487/3505	0.85 (0.71, 1.01)
≥30	743/763	1.29 (0.95, 1.74)
Rare/no alcohol use
No	5896/5920	0.90 (0.79, 1.04)	0.25
Yes	1366/1364	0.73 (0.52, 1.03)
Aspirin use in the past 12 mo
No	1623/1638	0.73 (0.54, 0.97)	0.18
Yes	5562/5605	0.91 (0.79, 1.05)
History of hypertension
No	4154/4252	0.91 (0.74, 1.13)	0.76
Yes	3117/3039	0.88 (0.75, 1.03)
History of high cholesterol
No	4472/4494	0.83 (0.70, 0.99)	0.41
Yes	2566/2624	0.93 (0.77, 1.12)
History of diabetes
No	6841/6880	0.86 (0.76, 0.99)	0.47
Yes	463/442	1.04 (0.68, 1.57)
History of myocardial infarction or stroke
No	6942/6945	0.89 (0.78, 1.02)	0.32
Yes	370/384	0.71 (0.46, 1.09)
eGFR <60 mL/min/1.73 m2
No	2258/2215	1.04 (0.82, 1.32)	0.24
Yes	396/418	0.79 (0.52, 1.18)

¹HRs were generated using Cox proportional hazards models that included age at randomization, recruitment source (prior PHS I or newly recruited PHS II participant), β-carotene assignment, multivitamin assignment, and vitamin C or E assignment. Models included baseline gout status as a stratification term. Models were repeated with multiplicative interaction terms to enable comparisons across strata. eGFR, estimated glomerular filtration rate; PHS, Physicians’ Health Study.

²Based on a joint test for an effect using Wald statistics that test whether the parameters associated with that effect are zero.