Table 3.

Pooled multivariatea RRs (95% CI) for consumption of coffee, tea, and SSBs overall and by histologic subtype and risk factors for pancreatic cancer

Coffeeb, c (increment 237 g/d)Teab, c (increment 237 g/d)SSBb, d (increment 175 g/d)
No. of casesRR (95% CI)PhetePIntfNo. of casesRR (95% CI)PhetePIntfNo. of casesRR (95% CI)PhetePIntf
Total pancreatic cancer1,5951.01 (0.97–1.04)0.051,5951.00 (0.96–1.05)0.512,0471.06 (1.02–1.12)0.70
 Females7581.04 (0.97–1.11)0.010.167580.98 (0.91–1.05)0.200.391,0901.03 (0.93–1.13)0.760.38
 Males8370.98 (0.95–1.01)0.908371.02 (0.96–1.09)0.849571.08 (1.02–1.14)0.41
Adenocarcinomag1,1491.01 (0.98–1.05)0.121,2401.02 (0.96–1.08)0.201,5561.08 (1.03–1.14)0.84
Nondiabeticsh1,2440.98 (0.95–1.02)0.231,2441.00 (0.95–1.06)0.371,6891.07 (1.02–1.13)0.64
Smoking statusi
 Never5251.04 (0.95–1.15)<0.010.415250.96 (0.88–1.03)0.510.167241.08 (0.98–1.19)0.340.73
 Former4420.94 (0.89–1.01)0.194421.02 (0.95–1.10)0.556671.10 (1.01–1.19)0.88
 Current5911.00 (0.96–1.04)0.625911.09 (1.02–1.16)0.716331.06 (0.98–1.14)0.94
Alcohol consumption, g/day
 04051.06 (0.99–1.13)0.030.084051.05 (0.97–1.13)0.430.466261.14 (1.05–1.23)0.400.23
 >0–<158111.00 (0.96–1.04)0.308110.98 (0.92–1.04)0.489671.06 (0.99–1.13)0.90
 ≥153710.96 (0.89–1.04)0.113711.09 (1.01–1.18)0.874521.03 (0.93–1.15)0.94
Physical activityj
 Low6070.98 (0.94–1.02)0.390.686071.00 (1.00–1.00)0.630.668431.06 (0.99–1.14)0.790.98
 Medium4351.00 (0.94–1.06)0.264351.00 (0.99–1.00)0.095921.13 (1.04–1.22)0.44
 High3030.99 (0.94–1.04)0.593031.00 (1.00–1.01)0.544761.09 (1.00–1.19)0.90
BMI, kg/m2
 <256711.03 (0.98–1.07)0.190.056711.03 (0.96–1.10)0.210.368721.12 (1.03–1.22)0.260.08
 ≥258920.98 (0.95–1.01)0.438920.98 (0.93–1.04)0.781,1191.03 (0.97–1.10)0.97
Age at diagnosis, y
 <698171.01 (0.98–1.04)0.550.308141.01 (0.95–1.06)0.800.731,0081.04 (0.96–1.12)0.370.28
 ≥697780.99 (0.94–1.05)0.097781.02 (0.96–1.08)0.731,0391.10 (1.04–1.17)0.68
Follow-upk
 ≤54870.99 (0.95–1.03)0.710.745021.00 (0.93–1.08)0.310.536631.05 (0.98–1.13)0.700.56
 >51,0931.02 (0.97–1.08)0.011,0931.03 (0.97–1.09)0.311,3691.08 (1.02–1.15)0.96
  • aMultivariate RRs were adjusted for smoking status (never smokers; past smokers, pack-years <15 years; past smokers, pack-years ≥15 years; current smokers, pack-years <40 years, current smokers, pack-years ≥40 years), alcohol intake (0, 0.1–14.9, 15–29.9, ≥30 g/d), history of diabetes (no, yes), BMI (continuously), and energy intake (continuously); age in years and year of questionnaire return were included as stratification variable. In the smoking stratified analyses, past and current smoking analyses included pack-years (<15, ≥15 years for past smokers; <40, ≥40 years for current smokers) in the model; age in years and year of questionnaire return were included as stratification variables. For the other models, the stratification variable was excluded as a covariate.

  • bFor coffee and tea, 8 oz. weighs approximately 237g; for sugar-sweetened carbonated soft drinks, 12 oz. weighs approximately 355g.

  • cThe BCDDP Follow-up Cohort, the CPS II Nutrition Cohort, and CTS were not included in these analyses because they did not measure consumption of this particular beverage or the particular beverage was not assessed as a separate item.

  • dThe NYSC (males and females) were not included in these analyses because the beverage was not assessed as a separate item.

  • eP value, test for between-studies heterogeneity.

  • fP value, test for interaction.

  • gThe HPFS was not included in the analysis on pancreatic adenocarcinoma as they did not have the histology data available.

  • hBCDDP Follow-up Cohort, CNBSS, and NYSC were excluded from this analysis because they did not measure diabetes status at baseline.

  • iATBC Cancer Prevention Study was excluded from the never and past smoking analyses because this study only included current smokers. Because of the small case numbers (n < 10), The NLCS male cohort was excluded from never smoking analysis, NYSC female cohort was excluded from past smoking analysis, and MCCS (males and females) were excluded from current smoking analysis.

  • jCNBSS was excluded from the physical activity analysis.

  • kThe MCCS (males and females) were excluded from the follow-up less than 5 years analysis due to small case numbers (<10).