Additional analysis among subjects with tea consumption assessed at baseline and reassessed 2 to 3 years later (n = 63,737; 91.4%) found that those who reported consistently to be a regular tea drinker had the most pronounced reduction in CRC risk (multivariate RR, 0.43; 95% CI, 0.24-0.77). The RRs were 0.91 (95% CI, 0.53-1.56) and 0.81 (95% CI, 0.50-1.30), respectively, for women who quitted tea drinking or started tea drinking after the baseline recruitment (data not shown in tables).