TABLE 4.
RRs, 95% CIs, and regression coefficients for risk of hepatocellular carcinoma associated with coffee intake (>3 compared with ≤3 cups/d) and percentage change in regression coefficients with adjustment for individual biomarkers: EPIC1
| Adjustment for biomarkers | RR (95% CI) | β Coefficient2 | Effect change3 (95% CI),4 % |
| Multivariable model | 0.35 (0.17, 0.76) | −1.03 | |
| Multivariable model with additional adjustment for individual biomarkers | |||
| Immune and inflammatory reaction | |||
| CRP, mg/L | 0.37 (0.16, 0.83) | −0.99 | 4 (−15, 42) |
| IL-6, pg/mL | 0.53 (0.23, 1.23) | −0.62 | 40 (7, 125) |
| Metabolic dysfunction | |||
| C-peptide, ng/mL | 0.37 (0.16, 0.84) | −0.99 | 4 (−27, 45) |
| Fetuin A, μg/mL | 0.36 (0.17, 0.79) | −1.01 | 2 (−14, 17) |
| Adiponectin, μg/mL | 0.37 (0.16, 0.80) | −0.99 | 3 (−13, 25) |
| HMW adiponectin, μg/mL | 0.39 (0.18, 0.86) | −0.93 | 3 (−12, 20) |
| Non-HMW adiponectin, μg/mL | 0.42 (0.19, 0.92) | −0.87 | 8 (−11, 41) |
| Leptin, ng/mL | 0.37 (0.17, 0.82) | −0.98 | 4 (−16, 34) |
| Hepatocellular/necroinflammatory/injury | |||
| GLDH, μmol · s−1 · L−1 | 0.45 (0.19, 1.08) | −0.77 | 24 (0, 102) |
| ALT, U/L | 0.46 (0.19, 1.05) | −0.78 | 24 (0, 84) |
| AST, U/L | 0.63 (0.25, 1.61) | −0.45 | 56 (9, 182) |
| LDH, U/L | 0.42 (0.19, 0.93) | −0.86 | 17 (−5, 65) |
| Cholestatic injury | |||
| GGT, U/L | 0.66 (0.25, 1.78) | −0.40 | 60 (7, 190) |
| ALP, U/L | 0.37 (0.14, 0.94) | −1.00 | 3 (−57, 71) |
| Global decrease in liver synthesizing capacity | |||
| Albumin, g/L | 0.39 (0.16, 0.94) | −0.92 | 10 (−16, 61) |
| Total bilirubin, μmol/L | 0.45 (0.19, 1.04) | −0.79 | 23 (2, 72) |
| Total protein, g/L | 0.40 (0.18, 0.88) | −0.90 | 12 (−3, 47) |
| Hepatocarcinogenesis | |||
| AFP, kUI/L | 0.45 (0.20, 1.05) | −0.79 | 23 (−2, 81) |
| Iron metabolism5 | |||
| Iron, μmol/L | 0.30 (0.13, 0.67) | −1.19 | −15 (−64, 7) |
| Ferritin, μmol/L | 0.33 (0.15, 0.75) | −1.08 | −4 (−33, 18) |
The multivariable model accounted for matching factors: age, sex, study center, follow-up time since blood collection, time of day of blood collection, and fasting status plus adjustment for education (no school degree or primary school, technical or professional school, secondary school, university degree, or unknown), smoking status (never, past, current, or unknown), alcohol intake (mL/d; continuous), nondrinking (categorical), hepatitis B surface antigen/antibodies to hepatitis C virus infection (positive, negative, or unknown), fruit and vegetable intake (g/d; continuous), physical activity (inactive, moderately inactive, moderately active, active, or missing), diabetes (yes, no, or missing), tea intake (mL/d), BMI (in kg/m2; continuous), and waist circumference adjusted for BMI by using the residual method (cm; continuous). Women were further matched by menopausal status and phase of menstrual cycle at blood collection; postmenopausal women were matched on use of hormone replacement therapy. 1 cup = 250 mL. AFP, α-fetoprotein; ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; CRP, C-reactive protein; EPIC, European Prospective Investigation into Cancer and Nutrition; GGT, γ-glutamyltransferase; GLDH, glutamate dehydrogenase; HMW, high molecular weight; LDH, lactatate dehydrogenase.
The β coefficient (regression coefficient) is the natural log of the RR estimate.
The percentage change in the regression coefficient with adjustment for each additional biomarker compared with the multivariable model.
The 95% CI was calculated based on Fieller’s theorem (28).
The biomarker of iron metabolism, transferrin, was excluded from the analysis because it was not associated with coffee intake or risk of hepatocellular carcinoma and therefore did not meet the criteria for being a potential mediator.