Table 3

Sex-specific number of deaths, HR* and 95% CI for overall mortality by categories of lifetime alcohol use (g/day), by smoking status (never and current smokers), and type of alcoholic beverage

Never smokersCurrent smokersNever smokersCurrent smokers
DeathsHR†(95% CI)DeathsHR†(95% CI)DeathsHR†(95% CI)DeathsHR†(95% CI)
Never10091.34(1.24 to 1.45)1541.72(1.32 to 2.23)Never841.50(1.19 to 1.21)582.09(1.26 to 3.47)
0.1–4.930461Ref10211.53(1.23 to 1.90)0.1–4.94571Ref3671.62(1.04 to 2.53)
5–14.915501.04(0.98 to 1.11)8741.51(1.21 to 1.88)5–14.95380.93(0.82 to 1.06)7991.45(0.93 to 2.25)
15–29.93971.04(0.94 to 1.16)4351.74(1.38 to 2.19)15–29.93691.00(0.87 to 1.16)9271.66(1.06 to 2.58)
≥ 30821.29(1.03 to 1.61)1402.08(1.59 to 2.73)30–59.92541.22(1.23 to 1.43)8571.83(1.17 to 2.84)
pWald§<0.001<0.0010.150≥ 601071.56(1.25 to 1.95)5902.43(1.55 to 3.80)
 Wine useBeer use pdifference** Wine useBeer use pdifference**
DeathsHR¶(95% CI)DeathsHR¶(95% CI)DeathsHR¶(95% CI)DeathsHR¶(95% CI)
Never21561.15(1.09 to 1.22)50411.06(1.02 to 1.12)Never10641.21(1.12 to 1.30)9751.07(0.98 to 1.16)
3–9.928130.96(0.92 to 1.01)7871.15(1.07 to 1.24)3–9.921390.92(0.87 to 0.97)24861.04(0.98 to 1.10)
10–19.910571.00(0.93 to 1.07)1471.50(1.27 to 1.77)10–19.910400.96(0.89 to 1.03)12481.12(1.04 to 1.20)
≥ 203541.14(1.02 to 1.27)371.47(1.06 to 2.04)20–39.98141.03(0.95 to 1.13)8771.41(1.30 to 1.54)
pWald§<0.001<0.001<0.001≥406411.22(1.10 to 1.35)4191.86(1.66 to 2.09)
  • *Models were stratified by centre. Systematic adjustment was undertaken for age at recruitment, BMI and height, former drinking, time since alcohol quitting, smoking status, duration of smoking, age at start smoking, educational attainment and energy intake. In women adjustment was undertaken for menopausal status, ever use of replacement hormones and number of full-term pregnancies.

  • †Models included interaction terms between lifetime alcohol use and a smoking indicator (0=never smokers; 1=current smokers), keeping the reference category the group of moderate alcohol users (0.1–4.9 g/day) among never smokers, whereas former smokers and participants with unknown smoking status were excluded.

  • ‡Pheterogeneity: difference in HRs assessed comparing the log-likelihood of models with and without interaction terms between alcohol and smoking status to a four and five degrees of freedom (dof) χ2 distribution, in women and men, respectively.

  • §pWald: determined using a Wald test for contrasts according to a χ2 distribution with four and five degrees of freedom, in women and men, respectively.

  • ¶Models on wine and beer uses were mutually adjusted, and also included spirits/liquors use.

  • **pdifference expresses the difference of associations between wine and beer use, determined evaluating the significance of the parameter estimate γ2 in a model that included, other than the list of confounders, the terms γ1(X1+X2)/2+γ2(X1 − X2)/2, with X1=log(wine use+1) and X2=log(beer use+1).