Type 2 diabetes and gender differences in liver cancer by considering different confounding factors: a meta-analysis of cohort studies.

PURPOSE: Questions remain uncertainty regarding the gender differences in the relationship between type 2 diabetes (T2DM) and liver cancer risk. By considering several confounding factors, we aimed to identify this issue according to a meta-analysis of cohort studies.

METHODS: We searched EMBASE and MEDLINE for studies on the association between T2DM and risk of liver cancer up to November 30, 2014. A random-effects model was performed to calculate summary relative risks (SRRs) with corresponding 95% confidence intervals (CIs).

RESULTS: A total of 24 cohort studies (including more than 20,000 liver cancer cases) were recruited. T2DM was associated with an elevated liver cancer incidence in both men (SRR = 2.16; 95% CI, 1.74-2.69) and women (SRR = 1.85; 95% CI, 1.40-2.44). Stratified analyses showed that the risk associations were significantly stronger in non-Asian than those in Asian for both men and women. Both tobacco smoking and body mass index were significant confounding factors for the T2DM-liver cancer association in men, whereas alcohol use was not the case. The SRR estimates of liver cancer mortality with T2DM were statistically significant in both men and women (men: SRR = 2.26; 95% CI, 1.60-3.19 and women: SRR = 2.01; 95% CI, 1.45-2.74).

CONCLUSIONS: Results of this meta-analysis indicate that the T2DM-liver cancer correlation is confounded by smoking and body mass index in both men and women. Results also suggest a significantly stronger T2DM-liver cancer correlation in non-Asian than that in Asian for both men and women.