TABLE 2.
Association between dairy consumption and esophagus cancer risk in the PLCO trial.
| Variables (servings/day) |
Cases (n) | Person- years |
Incidence rate/10,000 person-years | Cox proportional hazards regression (HR, 95% CI) |
||
| Unadjusted model | Adjusted model 1a | Adjusted model 2b | ||||
| Total dairy products | ||||||
| Q1(≤0.6) | 39 | 226237.11 | 1.72 | Reference group | Reference group | Reference group |
| Q2(0.61–1.06) | 37 | 223812.13 | 1.65 | 0.96 (0.61–1.50), p = 0.852 | 0.89 (0.57–1.40), p = 0.608 | 0.96 (0.61–1.53), p = 0.866 |
| Q3(1.07–1.78) | 40 | 227364.74 | 1.76 | 1.02 (0.66–1.59), p = 0.929 | 0.90 (0.58–1.40), p = 0.628 | 0.92 (0.57–1.47), p = 0.716 |
| Q4(≥ 1.79) | 38 | 223488.36 | 1.70 | 0.99 (0.63–1.54), p = 0.952 | 0.81 (0.52–1.27), p = 0.355 | 0.83 (0.50–1.38), p = 0.477 |
| p for trend = 0.976 | p for trend = 0.387 | p for trend = 0.465 | ||||
| Milk | ||||||
| Q1(≤ 0.3) | 38 | 228373.45 | 1.66 | Reference group | Reference group | Reference group |
| Q2(0.31–0.66) | 41 | 223526.50 | 1.83 | 1.10 (0.71–1.71), p = 0.672 | 0.95 (0.61–1.48), p = 0.830 | 0.96 (0.61–1.51), p = 0.858 |
| Q3(0.67–1.3) | 37 | 224619.60 | 1.65 | 0.99 (0.63-1.55), p = 0.960 | 0.79 (0.50–1.25), p = 0.314 | 0.84 (0.53–1.34), p = 0.461 |
| Q4(≥ 1.31) | 38 | 224382.79 | 1.69 | 1.02 (0.65–1.59), p = 0943 | 0.77 (0.49–1.21), p = 0.258 | 0.76 (0.47–1.24), p = 0.278 |
| p for trend = 0.935 | p for trend = 0.184 | p for trend = 0.227 | ||||
| Yogurt# | ||||||
| Q1(0) | 98 | 401141.62 | 2.44 | Reference group | Reference group | Reference group |
| Q2(0.01–0.02) | 24 | 194938.22 | 1.23 | 0.50 (0.32–0.79), p = 0.003 | 0.68 (0.44–1.07), p = 0.098 | 0.68 (0.43–1.08), p = 0.103 |
| Q3(0.03–0.06) | 19 | 141768.02 | 1.34 | 0.55 (0.34–0.9), p = 0.016 | 0.86 (0.52–1.41), p = 0.536 | 0.94 (0.57–1.55), p = 0.808 |
| Q4(≥ 0.07) | 13 | 163054.48 | 0.80 | 0.33 (0.18–0.58), p<0.001 | 0.57 (0.32–1.03), p = 0.062 | 0.62 (0.35–1.13), p = 0.118 |
| p for trend <0.001 | p for trend = 0.051 | p for trend = 0.126 | ||||
| Cheese | ||||||
| Q1(≤0.1) | 46 | 239510.73 | 1.92 | Reference group | Reference group | Reference group |
| Q2(0.11–0.2) | 35 | 219272.34 | 1.60 | 0.83 (0.54–1.29), p = 0.411 | 0.83 (0.53–1.29), p = 0.408 | 0.80 (0.50–1.26), p = 0.329 |
| Q3(0.21–0.37) | 35 | 218078.51 | 1.60 | 0.84 (0.54–1.30), p = 0.426 | 0.81 (0.52–1.26), p = 0.354 | 0.76 (0.48–1.22), p = 0.260 |
| Q4(≥ 0.38) | 38 | 224040.76 | 1.70 | 0.88 (0.58–1.36), p = 0.576 | 0.79 (0.51–1.23), p = 0.299 | 0.75 (0.46–1.22), p = 0.246 |
| p for trend = 0.578 | p for trend = 0.302 | p for trend = 0.241 | ||||
PLCO, prostate, lung, colorectal, and ovarian cancer; HR, hazard ratio; 95% CI, 95% confidence interval; Q, quartile; BMI, body mass index.
aStratified by sex (male, female) due to proportional hazard assumption violation and adjusted for age (years).
bStratified by sex (male, female) due to proportional hazard assumption violation and adjusted for age (years), alcohol drinking status (never, current, former), smoking status (never, current, former), occupation (not working, working, retired, others), body mass index (<25 kg/m2, ≥25kg/m2), total calorie intake (kcal/day).
#The first category was non-consumers, with the remaining consumers classified as tertiles of distribution.