From Nature Clinical Practice Gastroenterology & Hepatology
Is Coffee or Tea Good for Your Liver? Posted 10/02/2006
W Ray Kim
Synopsis Background It has been postulated that the consumption of coffee and tea is associated with reduced liver disease; however, this relationship has not been assessed in the US on a population-wide basis.
Objectives To investigate the association between chronic liver disease (CLD) and consumption of coffee and tea.
Design and Intervention This study employed data from the US National Health and Nutrition Examination Survey (NHANES) 1971-1975 on coffee, tea, and alcohol consumption, and also that from the NHANES Epidemiologic Follow-Up Study. Data from a separate analysis (1982-1984) assessing more detailed drinking habits of some participants from the NHANES were also included. In both analyses, daily consumption was categorized as <1 cup, 1-2 cups, or >2 cups. Exclusion criteria included lack of data on coffee and tea consumption, jaundice, hepatomegaly, splenomegaly, and serum albumin >30 g/l. BMI was calculated for all participants and laboratory tests, including transferrin saturation, serum iron, and iron-binding capacity, were performed. Participants were followed up until 1992 or 1993; follow-up included interview and examination of death certificate or hospital records. CLD was determined by death certificate or hospital discharge record. A number of statistical tests were performed on the available data.
Outcome Measures The main outcome measures were coffee and tea consumption and the incidence of CLD.
Results In total, data from 9,849 NHANES participants (mean follow-up 19 years [range 0.2-22.1 years]) and 9,650 participants from the separate analysis were included. Mean daily intake of coffee and tea in the NHANES group was 2 cups (range 0-16 cups), and the mean daily consumption (in cups) for each category was 0.2 for <1 cup, 1.5 for 1-2 cups, and 4.0 for >2 cups. The risk of death or hospitalization from CLD was 1.4% at 20-year follow-up. The unadjusted risk of CLD at 20 years was 1.8% (hazard ratio [HR] 1.0) for those who drank <1 cup per day, 1.6% (HR 0.84, 95% CI 0.52-1.40) for those who drank 1-2 cups per day, and 1.1% (HR 0.46, 95% CI 0.26-0.80) for those who drank >2 cups per day (P = 0.002). The incidence of CLD was higher in participants with a BMI ¡Ý30 (1.8%; HR 1.5, 95% CI 0.87-2.50) than those with a BMI of 25 to <30 (1.6%; HR 1.5, 95% CI 0.99-2.30) or <25 (1.2%; HR 1.0) Multivariate analysis revealed that drinking >2 cups of coffee per day reduced the risk of CLD by more than 50% compared to those who drank <1 cup per day (HR 0.43, 95% CI 0.24-0.75 vs HR 1.0; P = 0.003). The incidence of CLD in high-risk individuals (such as those with diabetes or a high alcohol consumption) drinking >2 cups per day was less than in those drinking ¡Ü2 cups per day (1.1%; HR 0.40, 95% CI 0.23-0.68 vs 2.3%; HR 1.0; NS).
Conclusion The consumption of coffee and tea is associated with a reduced risk of CLD.
Commentary Is coffee or tea good for your liver? The study by Ruhl and Everhart seems to indicate that drinking coffee or tea does decrease the risk of clinically significant CLD. This work extends their previous observation that coffee and caffeine consumption is associated with lower serum alanine aminotransferase (ALT) levels.[1] In this earlier study, which included subjects who participated in the third NHANES, and who were deemed at high risk of liver disease, coffee and caffeine intake was inversely correlated with serum ALT levels.
The current study cleverly uses longitudinal data made available through the NHANES Epidemiologic Follow-up Study. The main strength of this analysis is that the study was based on prospective collection of exposure history (coffee or tea consumption) and clinical outcome (hospitalization or death due to CLD) in a large population-based sample (NHANES-I). Methodologically, however, this study was limited for a number of reasons. Firstly, a relatively large proportion (25%) of subjects were excluded from the analysis because of a lack of data on coffee or tea consumption; secondly, ascertainment of liver disease was based on hospital discharge records and death certificates, as opposed to verifiable clinical records; and thirdly, there was a lack of detail regarding the amount and type of beverage consumed. Weighing up these strengths and limitations, however, the reader becomes intrigued (if not reasonably convinced) that coffee or tea consumption might, indeed, be good for the liver.
Caution must be exercised, however, before physicians begin to advise patients with liver disease to consume more tea or coffee. Although these observational data show a consistent association between coffee or tea consumption and CLD, it is premature to conclude a causal relationship between the two (i.e that these beverages reduce the risk of liver disease).
Firstly, no known ingredients of coffee or tea have been linked with a protective effect in the pathogenesis of CLD. Caffeine might not be responsible, as caffeine-containing beverages other than coffee did not show any benefit in a study by Corrao et al.[2] Despite the recent interest in the antioxidant and other potentially beneficial properties of catechins in tea, a protective effect against CLD remains to be determined.[3]
Secondly, it is possible that the association revealed in this study might have been confounded with other dietary or behavioral factors that are indeed responsible for the reduced risk of CLD. For example, coffee or tea consumption was inversely associated with BMI. It could be that the consumption of these beverages is associated with healthier dietary practices, which might reduce the risk of the metabolic syndrome¡ªitself associated with nonalcoholic fatty liver disease, the most common cause of CLD in the US. Similarly, consumption of coffee or tea might be inversely correlated with heavy alcohol consumption. The lack of a consistent pattern between coffee or tea consumption and alcohol intake in the Ruhl and Everhart study might be attributable to the fact that all levels of alcohol consumption >2 drinks per day were lumped together; patients with alcoholic liver disease who consumed >2 alcoholic drinks per day could not be separately analyzed. Although some of these questions were addressed in European studies, differences in culture and liver disease epidemiology make it difficult to directly extrapolate those results.[2,4]
In summary, Ruhl and Everhart provide as strong evidence as observational data possibly can for an inverse relationship between coffee or tea intake and CLD. Before coffee or tea can be 'prescribed' in patients with liver disease, however, interventional data are strongly warranted.
Practice Point There might be an inverse relationship between tea, coffee, or caffeine intake and liver disease; however, further data are needed before tea or coffee can be adopted as a treatment in patients with chronic liver disease.
Acknowledgements
The synopsis was written by Rachel Jones, Associate Editor, Nature Clinical Practice.
Reprint Address
Gastroenterology and Hepatology, Plummer 6 Mayo Clinic College of Medicine 200 First Street, SW Rochester, MN 55905 USA. Email: [email protected]
|