Fruit and vegetable intake and liver cancer risk: a meta-analysis of prospective cohort studies

The associations of vegetable and fruit intake with liver cancer risk have been inconsistent based on epidemiological studies. The present study aimed to quantitatively evaluate these associations with prospective cohort studies. A systematic literature search was performed with PubMed and Scopus databases up to June 2019. Multivariate-adjusted relative risks (RRs) with a corresponding 95% confidence interval (CI) for the highest versus lowest category were pooled by using a random-effects model. Pre-specified subgroup and univariate meta-regression analyses were performed to identify the sources of heterogeneity. Dose-response analysis was conducted by using the variance weighted least squares regression model. Nine independent prospective cohort studies with 1703 liver cancer events and 1 326 176 participants were included for data synthesis. The summary estimates showed that higher vegetable intake was associated with a 39% (95%CI: 0.50, 0.75) reduction in liver cancer risk, with no significant between-study heterogeneity (P = 0.057). Dose-response analysis indicated that the risk of liver cancer was reduced by 4% (95%CI: 0.97, 0.95; P for trend <0.001) with a 100 gram per day increment of vegetable intake. Subgroup analysis showed that higher intakes of vegetables were associated with a 50% (95%CI: 0.35, 0.72) reduction of liver cancer risk in males, but not in females. However, a non-significant association was found between fruit intake and liver cancer risk. The present study provides strong evidence that higher intakes of vegetables would have beneficial effects on the prevention of liver cancer, especially for males.