Table 2. Associations Between Sugar-Sweetened and Artificially Sweetened Beverage Intake and Liver Cancer Risk in the Women’s Health Initiative, 1993-2020.
| Beverages | Consumption categoriesa | P value for trendb | ||
|---|---|---|---|---|
| Never to ≤3 servings/mo | 1-6 servings/wk | ≥1 serving/d | ||
| Sugar-sweetened beverages | ||||
| Cases of liver cancer, No. | 153 | 31 | 23 | |
| No. of participants | 77 173 | 14 921 | 6692 | |
| Unadjusted rate/100 000 person-years | 10.3 | 10.8 | 18.0 | .02 |
| Age and energy-adjusted HR (95% CI)c | 1 [Reference] | 1.12 (0.76-1.65) | 2.03 (1.30-3.18) | .003 |
| Multivariable-adjusted HR (95% CI)d | 1 [Reference] | 1.10 (0.74-1.64) | 1.85 (1.16-2.96) | .01 |
| Soft drinks | ||||
| Cases of liver cancer, No. | 169 | 22 | 16 | |
| Unadjusted rate/100 000 person-years | 10.7 | 9.7 | 16.9 | .18 |
| Age and energy-adjusted HR (95% CI)c | 1 [Reference] | 0.96 (0.61-1.50) | 1.86 (1.10-3.14) | .06 |
| Multivariable-adjusted HR (95% CI)d | 1 [Reference] | 0.94 (0.60-1.48) | 1.66 (0.97-2.85) | .14 |
| Fruit drinks | ||||
| Cases of liver cancer, No. | 189 | 12 | 6 | |
| Unadjusted rate/100 000 person-years | 10.5 | 15.0 | 19.3 | .06 |
| Age and energy-adjusted HR (95% CI)c | 1 [Reference] | 1.51 (0.84-2.72) | 1.93 (0.85-4.36) | .04 |
| Multivariable-adjusted HR (95% CI)d | 1 [Reference] | 1.38 (0.76-2.51) | 1.71 (0.75-3.92) | .11 |
| Artificially sweetened beveragese | ||||
| Cases of liver cancer, No. | 83 | 30 | 20 | |
| No. of participants | 41 332 | 14 949 | 8506 | |
| Unadjusted rate/100 000 person-years | 10.2 | 10.1 | 11.8 | .70 |
| Age and energy-adjusted HR (95% CI)c | 1 [Reference] | 1.06 (0.70-1.61) | 1.37 (0.83-2.24) | .29 |
| Multivariable-adjusted HR (95% CI)d | 1 [Reference] | 0.99 (0.65-1.51) | 1.17 (0.70-1.94) | .66 |
Abbreviation: HR, hazard ratio.
One serving defined as 12 fl oz or 355 mL.
The median intake of each category was modeled as a continuous variable to calculate the P value for trend.
Results were from the Cox proportional hazards model adjusting for age at entry (continuous) and total energy intake (quartile).
Results were from the Cox proportional hazards model additionally adjusting for race and ethnicity (Hispanic, non-Hispanic Black, non-Hispanic White, other), education (<high school, high school or some college, >college), smoking status (non-smoker, past smoker, current smoker), alcohol consumption (nondrinker, past drinker, <1 drink/month, <1 drink/week, 1-<7 drinks/week, ≥7 drinks/week), body mass index (<18.5, 18.5-<25, 25-<30, 30-<35, 35-<40, ≥40 [calculated as weight in kilograms divided by height in meters squared]), physical activity (quartile), nonsteroidal anti-inflammatory drugs use (yes, no), family history of cancer (yes, no), prior oral contraceptive use (yes, no), postmenopausal hormone therapy (yes, no), and self-reported diabetes (yes, no).
For artificially sweetened beverage analyses, we used the 3-year follow-up as the baseline in the observational study (64 787 participants).