Table 3.
Multivariable HRs and 95% CIs of total coffee, caffeinated coffee and decaffeinated coffee consumption in relation to endometrial cancer by body mass index status in the Women’s Health Initiative Observational Study (N = 45,696).
| Beverage consumption | 0 or <1 cup/day | 1 cup/day | ≥2 cups/day | p-val a (trend) b | P-int c |
|---|---|---|---|---|---|
| Total Coffee | 0.38 | ||||
| Normal-weight (BMI < 25 kg/m2) | |||||
| N = 20,039 | |||||
| Incident cases | 50 | 27 | 99 | ||
| Person-years | 44,601 | 22,450 | 84,986 | ||
| HR (95% CI) ‡ | 1.00 (ref) | 1.03 (0.64-1.64) | 1.02 (0.72-1.45) | 0.99 (0.90) | |
| Overweight (BMI 25 kg/m2-29.9 kg/m2) | |||||
| N = 15,159 | |||||
| Incident cases | 26 | 17 | 60 | ||
| Person-years | 30,127 | 15,619 | 67,741 | ||
| HR (95% CI) ‡ | 1.00 (ref) | 1.21 (0.66-2.32) | 1.01 (0.63-1.60) | 0.79 (0.93) | |
| Obese (BMI ≥ 30 kg/m2) | |||||
| N = 10,498 | |||||
| Incident cases | 50 | 27 | 71 | ||
| Person-years | 23,951 | 10,939 | 42,511 | ||
| HR (95% CI) ‡ | 1.00 (ref) | 1.20 (0.70-1.79) | 0.72 (0.50-1.04) | 0.08 (0.06) | |
| Caffeinated Coffee | 0.3 | ||||
| Normal-weight (BMI < 25 kg/m2) | |||||
| N = 20,039 | |||||
| Incident cases | 70 | 34 | 72 | ||
| Person-years | 66,860 | 26,212 | 58,965 | ||
| HR (95% CI) ‡ | 1.00 (ref) | 1.23 (0.81-1.87) | 1.17 (0.83-1.64) | 0.53 (0.37) | |
| Overweight (BMI 25 kg/m2-29.9 kg/m2) | |||||
| N = 15,159 | |||||
| Incident cases | 44 | 16 | 43 | ||
| Person-years | 46,985 | 18,717 | 47,785 | ||
| HR (95% CI) ‡ | 1.00 (ref) | 0.88 (0.50-1.58) | 0.99 (0.65-1.53) | 0.91 (0.99) | |
| Obese (BMI ≥ 30 kg/m2) | |||||
| N = 10,498 | |||||
| Incident cases | 71 | 31 | 46 | ||
| Person-years | 34,215 | 12,259 | 30,927 | ||
| HR (95% CI) ‡ | 1.00 (ref) | 1.16 (0.75-1.78) | 0.66 (0.45-0.97) | 0.03 (0.05) | |
| Decaffeinated Coffee | 0.51 | ||||
| Normal-weight (BMI < 25 kg/m2) | |||||
| N = 20,039 | |||||
| Incident cases | 129 | 25 | 22 | ||
| Person-years | 102,756 | 23,973 | 25,307 | ||
| HR (95% CI) ‡ | 1.00 (ref) | 0.77 (0.50-1.18) | 0.67 (0.43-1.06) | 0.15 (0.06) | |
| Overweight (BMI 25 kg/m2-29.9 kg/m2) | |||||
| N = 15,159 | |||||
| Incident cases | 68 | 17 | 18 | ||
| Person-years | 76,691 | 16,583 | 20,212 | ||
| HR (95% CI) ‡ | 1.00 (ref) | 1.10 (0.64-1.87) | 0.93 (0.55-1.58) | 0.90 (0.86) | |
| Obese (BMI ≥ 30 kg/m2) | |||||
| N = 10,498 | |||||
| Incident cases | 102 | 26 | 20 | ||
| Person-years | 55,198 | 10,840 | 11,362 | ||
| HR (95% CI) ‡ | 1.00 (ref) | 1.17 (0.76-1.82) | 0.80 (0.49-1.30) |
‡ Multivariable adjusted hazard ratios from Cox-proportional hazard regression models: adjusted for age (continuous), ethnicity (white, black and other), unopposed estrogen use (never, past, current), combined estrogen-progestin use (never, past, current), and smoking (never, past, current). When we examined caffeinated coffee or decaffeinated coffee in relation to endometrial cancer risk, each beverage was mutually adjusted for each other; a Represents p-value from likelihood ratio test for categorical coffee variables; b Represents p-value from linear trend test across coffee categories; c Represents p-value for interaction.