Table 4.
Association between tomato intake (grams per day) and hepatocellular carcinoma by HBsAg serology, Singapore Chinese Health Study, 1993–2015
| Cases | Controls | OR (95% CI) | |
|---|---|---|---|
| Tomato Intake Overalla | |||
| 1st quartile (< 1.95) | 63 | 104 | 1.00a |
| 2nd quartile (1.95–4.81) | 40 | 108 | 0.51 (0.28–0.93) |
| 3rd quartile (4.81–8.83) | 57 | 130 | 0.58 (0.33–1.03) |
| 4th quartile (≥ 8.83) | 37 | 123 | 0.52 (0.29–0.94) |
| Ptrend | 0.0409 | ||
| Tomato Intake Stratified by HBV Status | |||
| HBsAg (−) | |||
| 1st quartile (< 1.95) | 41 | 100 | 1.00b |
| 2nd quartile (1.95–4.81) | 26 | 101 | 0.56 (0.30–1.02) |
| 3rd quartile (4.81–8.83) | 33 | 127 | 0.54 (0.30–0.96) |
| 4th quartile (≥ 8.83) | 27 | 118 | 0.51 (0.28–0.94) |
| Ptrend | 0.0300 | ||
| HBsAg (+) | |||
| 1st quartile (< 1.95) | 22 | 4 | 1.00b |
| 2nd quartile (1.95–4.81) | 14 | 7 | 0.21 (0.04–1.21) |
| 3rd quartile (4.81–8.83) | 24 | 3 | 0.90 (0.14–5.79) |
| 4th quartile (≥ 8.83) | 10 | 5 | 0.29 (0.05–1.64) |
| Ptrend | 0.3590 | ||
| Pinteraction, HBsAg and tomato intake | 0.75 |
a
Conditional logistic regression that retained case-control pairs where controls were matched to cases on age, gender, date of interview, dialect group, and date of sample collection. Model was adjusted for BMI, education, smoking status, alcohol intake, diabetes status, physical activity, total energy intake, and HBsAg serology.
b
Unconditional logistic regression adjusted for age, gender, dialect, education, BMI, smoking status, alcohol intake, year of enrollment, diabetes status, physical activity, and total energy intake.