Table 2.

Risk estimates and 95% CIs for the joint effects and interaction indexes between HBV and smoking

HBV−/Tob−HBV−/Tob+HBV+/Tob−*HBV+/Tob+Interaction index
AdditiveMultiplicative
All studies (n = 9)
No. of cases272960419168033313309
Random effect1.001.87 (1.30-2.69)15.8 (9.69-25.7)21.6 (15.2-30.5)1.44 (1.00-2.06)0.87 (0.58-1.29)
Adjusted random effects*1.001.59 (0.94-2.70)18.27 (14.5-23.0)21.7 (11.8-40.0)1.53 (1.34-1.75)0.77 (0.36-1.67)
Pheterogeneity0.0010.0110.0500.049<0.001
Egger's test for publication bias0.6090.105

NOTE: One cohort study in Asia reported 0 cases in the HBV+/Tob− category. Multiplicative interaction was not able to be calculated for this study.

Abbreviations: HBV−/Tob−, reference category, non-HBV infection and nonsmoker; HBV−/Tob+, non-HBV infection and smoker; HBV+/Tob−, HBV infection and nonsmoker; HBV+/Tob+, HBV infection and smoker.

  • *Three publications provided adjusted estimates (Wang et al., 2003, Jee et al., 2004, and Hassan et al., 2008). The estimate was based on male population. All three publications adjusted for age. Other adjustments included race (Hassan et al., 2008), education, residence, anti-HCV, (Wang et al., 2003 and Hassan et al., 2008), marital status (Hassan et al., 2008), alcohol, diabetes (Jee et al., 2004 and Hassan et al., 2008), family history of HCC (Wang et al., 2003), family history of cancer (Hassan et al., 2008), and liver function (Wang et al., 2003).