Impact of Coffee on Liver Diseases

A Systematic Review

Sammy Saab; Divya Mallam; Gerald A. Cox II; Myron J. Tong


Liver International. 2014;34(4):495-504. 

In This Article


We searched MEDLINE and PubMed for all studies published on coffee and liver diseases from 1986 to September 2012. We used a combination of the keywords 'coffee', 'caffeine', 'liver disease', 'cirrhosis', 'fibrosis', 'hepatitis B', 'hepatitis C', 'non-alcoholic fatty liver disease', 'fatty liver', 'hepatocellular carcinoma', 'liver cancer', 'alcoholic liver disease', and 'alcoholic hepatitis'. We analyzed all studies published in scientific journals including observational studies and case-controlled studies. Bibliographies of identified studies were searched for relevant articles.

Liver-associated Laboratory Tests

A number of studies noted a beneficial effect of increased coffee consumption on liver-associated laboratory tests. This benefit was reported in a variety of populations at risk for liver disease, including those with excessive alcohol intake, obesity, smokers, and those with chronic viral hepatitis. These studies demonstrated that increased coffee consumption was associated with aspartate-aminotransferase (AST), alanine-aminotransferase (ALT), and Gamma-glutamyltransferase (GGT) levels in a dose-dependent manner.

The first studies to demonstrate a relationship between coffee intake and liver-associated test values were two Norwegian reports which revealed an inverse association between coffee consumption and serum GGT values.[16,17] The results were further illustrated in a 7 year longitudinal study of a subset of the Tromsø Study by Nilssen et al..[18] Their findings have since been confirmed in multiple population studies in Japan,[19–22] Italy,[24,25] and Finland[26] ( Table 1 ). Another study from Norway also found coffee consumption inversely related to serum GGT.[26] The results of a randomized study, however, showed that cafetière increased liver enzymes.[28]

In one of the largest studies including over 12 000 health examinees in Japan, Tanaka et al. described that increased coffee consumption had a strong and independent association with decreased GGT activity in male alcohol drinkers (P < 0.0001).[19] However, consumption of coffee was only weakly associated with lowered GGT levels among women. Using the data from the Self-Defense Forces Fukuo Hospital, Honjo et al. further confirmed earlier observations which demonstrated that coffee consumption was associated with lowered serum GGT levels.[20–22]

Population studies in Italy, Japan, and the United States also have reported an inverse relation of coffee consumption with serum aminotransferase levels. In a cohort study of over 2000 Italian patients aged 65 or older, Casiglia et al. observed that ALT values were 10.3% lower in those who drank three or more cups of coffee daily (P = 0.0160).[24] In a Japanese study of 12,020 middle-aged and elderly male participants, a strong inverse association between coffee consumption and elevated ALT values also were noted.[29]

The results of the third National Health and Nutrition Examination Survey (NHANES) found that coffee consumption and caffeine were associated with a decreased risk of elevated ALT levels amongst persons at high risk for liver injury in the United States, such as persons who were overweight, had viral hepatitis, impaired glucose metabolism, iron overload or excessive alcohol intake.[30] In an unadjusted analysis, a lowered ALT activity was associated with an increasing consumption of coffee (P = 0.001) and caffeine (P = 0.001).

Chronic Liver Disease and Cirrhosis

Chronic Liver Disease. Coffee intake has also been associated with a decreased incidence of chronic liver disease ( Table 2 ). The histologic benefit of coffee was first demonstrated by Modi et al. who examined the association between daily caffeine intake and severity of hepatic fibrosis amongst persons with chronic liver diseases.[31] The authors described that daily caffeine intake above two cups of coffee was associated with lower rates of hepatic fibrosis (OR 0.33, 95% CI .14–0.8, P = 0.015). The observed protective relationship persisted after controlling for age, sex, race, liver disease, BMI, alcohol intake, and hepatitis C viral infection (HCV) infection. Of note was the fact that caffeine intake from non-coffee sources was not associated with decreased hepatic fibrosis. In a large cross sectional US study, intake of regular ground coffee and caffeine intake was associated with a decreased risk of chronic liver disease.[32] Persons included in the report were those with hepatitis B, hepatitis C, iron overload, impaired glucose metabolism, and excessive alcohol intake (>2 alcoholic beverages per day). In the study, participants were followed for a median of 19 years, and those who drank greater than two cups of coffee daily had less than half the rate of chronic liver disease than those who drank less than one cup daily.

Cirrhosis and Mortality. In addition to the decreased risk of chronic liver disease, coffee consumption has been associated with a decreased risk of cirrhosis ( Table 2 ). Two Italian hospital-based case-control studies found that coffee may delay the development of cirrhosis. The first study by Corrao et al. found a dose-dependent inverse relationship between caffeine intake and risk of cirrhosis.[33] The odds ratios of cirrhosis development decreased from 1.0 (lifetime non-coffee drinkers) to 0.47 (95% CI 0.2–1.10), 0.23 (0.10–0.53), 0.21 (0.06–0.74), and 0.16 (0.05–0.50) in those ingesting 1, 2, 3, and 4 or more cups of coffee daily respectively. A longer term follow up in the same cohort confirmed earlier observations.[35] The second study by Gallus et al. found an inverse relationship between coffee consumption and cirrhosis [odds ratio (OR) 0.54 for coffee drinkers compared to non-coffee drinkers; OR 0.29 for 3 or more cups daily]. An inverse relationship with duration of coffee consumption and cirrhosis also was observed (OR 0.45 for 40 years or more of coffee consumption).[34]

The results of a large 10-year cohort follow-up study in the US revealed that coffee may be protective against both hospitalization and death from alcoholic cirrhosis.[36] Subsequently, Klatsky et al. reported that coffee consumption may lead to a decreased mortality risk from nonalcoholic and alcoholic cirrhosis; the relative risk per cup of coffee daily was 0.77 (95% CI 0.67, 0.89).[37] Further expanding upon their original cohort study from 1992, Klatsky et al. presented a 22-year follow-up report in which they provided further evidence that increased coffee intake led to a decreased risk of alcoholic cirrhosis; with greater coffee intake associated with a lower relative risk of cirrhosis [1–3 cups, 0.6 (95% CI, 0.4–0.8; P < 0.001; 4 or more cups, 0.2 (95% CI, 0.1–0.4, P < 0.001)].[38] There was no relationship between tea intake and the development of alcohol-related cirrhosis.

The inverse association between coffee intake and mortality from cirrhosis also was supported by a Norwegian cohort study by Tverdal et al.[39] In the study, mortality rates were lower for persons drinking 3 or more cups of coffee daily compared to those drinking 2 or less cups. After adjusting for age, sex, alcohol use and other cardiovascular risk factors, similar benefits of coffee were also observed among patients with alcoholic cirrhosis. The relative risk of cirrhosis associated with an increase of two cups of coffee was 0.6 (95% CI, 0.5–0.8).[39]

Chronic Hepatitis B and Hepatitis C

There are limited published data on the association between caffeine consumption and chronic viral hepatitis ( Table 3 ). Only one report assessed the association in chronic hepatitis B, and found no benefit from coffee consumption on severity of hepatitis B as measured by transient elastography.[40] Since most participants who consumed alcohol were also mostly coffee drinkers, the benefit of caffeine intake may have been confounded by the deleterious effects of alcohol consumption. Furthermore, the use of transient elastography, while proven to be accurate in diagnosing histological advanced fibrosis in chronic Hepatitis B viral infection (HBV)-infected patients, may be of limited utility in patients with elevated serum transaminase levels.[41]

Several studies have examined the impact of coffee consumption on fibrosis severity in patients with chronic hepatitis C ( Table 3 ). Modi et al. demonstrated the beneficial impact of coffee consumption on fibrosis severity in patients with hepatitis C (OR 0.19, 95% CI 0.05–0.66, P = 0.009).[31] The patients in this study completed detailed caffeine questionnaires on three occasions over a six-month period of time. Freedman et al. also found that regular coffee intake was associated with decreased rates of liver disease progression amongst chronic hepatitis C patients enrolled in the HALT-C Trial.[42] The rates of liver disease progression declined with increasing coffee intake (P = 0.0011). Compared to non-coffee drinkers, the relative risks for reaching pre-defined endpoints indicating disease progression were 1.11 for less than 1 cup daily (CI 0.76–1.61), 0.7 for 1 to <3 cups daily (CI 0.48–1.02), and 0.47 for 3 or more cups daily (CI 0.27–0.85, P = 0.0003). Utilizing data from the HALT-C trial, Freedman et al. also explored the relationship between coffee consumption and response to antiviral therapy.[43] The authors demonstrated that greater than 3 cups of coffee daily was an independent predictor of improved virologic response to retreatment with peg-interferon plus ribavirin in patients with hepatitis C who failed initial treatment. Consumption of 3 or more cups daily was associated with a higher tolerance for the full dose of peg-IFN (60.6% compared to 50.4% of noncoffee drinkers).

A prospective cohort study done in France by Costentin et al. evaluated the effect of coffee consumption on treatment-naïve chronic hepatitis C patients.[44] In the report, multivariate analysis showed that daily caffeine consumption equivalent to 3 cups of coffee was associated with a decreased necroinflammatory activity [OR 0.32, Confidence Interval (CI) 0.12–0.85]. Of note, caffeine intake in this study was not limited to coffee alone, but included caffeinated coffee, tea, and caffeine-containing sodas. Carrieri et al. evaluated whether coffee intake improved the tolerability of peginterferon alfa-2a and ribavirin in HIV-HCV co-infected patients.[45] The authors found that those patients drinking 3 or more cups of coffee daily were less likely to report adverse effects compared to coffee nondrinkers (OR 0.19, CI .05–0.78, P = 0.02). These findings remained significant after adjustments for gender, age, cirrhosis, and history of opioid use.

Nonalcoholic Fatty Liver Disease

Several studies have assessed the association between coffee consumption and NAFLD ( Table 4 ). In an analysis of four continuous cycles (2001–2008) of the Nat-ional Health and Nutrition Examination Survey (NHANES), a dietary intake questionnaire collected by the National Center for Health Statistics of the Centers for Disease Control and Prevention revealed that caffeine intake was independently associated with a decreased risk of development of NAFLD (OR 0.931, CI 0.900–0.964).[46]

Two case–control studies have suggested the beneficial effect of coffee on the risk of NAFLD as defined by abdominal imaging.[47,48] Catalano et al. found that a decrease in fatty liver severity in coffee drinkers as compared to non-coffee drinkers (β = −2.585, P = 0.011, CI −0.133 to 0.018). Also, coffee drinking was inversely associated with obesity and insulin resistance.[47] A case control study performed in Mexico showed similar results with a dose-dependent reduction in coffee intake with increasing severity of hepatic steatosis.[48] In addition, a cross-sectional study demonstrated that coffee consumption was associated with a significant decrease in the risk of hepatic fibrosis among patients with nonalcoholic steatohepatitis (NASH).[49] In the study, an inverse relationship was found between coffee consumption and hepatic fibrosis (r = −9.215, P = 0.035). There was a significant difference between coffee consumption in patients with bland steatosis/not-NASH (P = 0.005), NASH Stage 0–1, and between NASH stage 0–1 and NASH stage 2–4 (P = 0.005).

The results of a recent European study provided further evidence regarding the protective effects of coffee in morbidly obese persons with NAFLD.[50] Regular coffee consumption was associated with decreased liver fibrosis in morbidly obese women with NAFLD undergoing bariatric surgery (OR 0.752, CI 0.578–0.98, P = 0.035).[50] However, regular filtered coffee, but not espresso coffee, was found to be associated with the decreased likelihood of fibrosis.

Hepatocellular Carcinoma

Decreased Risk of Hepatocellular Carcinoma. The relationship between coffee and the development of HCC was initially unsettled ( Table 5 ). Two early case-control studies found no association between coffee consumption and the risk of HCC.[51,52] However, the results of a follow up study combining both data sets indicated that coffee was indeed protective against HCC.[50,52,53] Additional more recent reports indicate a potential benefit of coffee on the incidence of hepatocellular carcinoma. Several case-control studies also demonstrated that coffee drinking was associated with decreasing HCC risk with a dose-effect relationship. Gelatti et al. found that compared to non-coffee drinkers, the odds ratio for HCC development was 0.4 (95% CI 0.2–0.8) for those consuming 3–4 cups daily regardless of the underlying liver disease cause.[54] Tanaka et al. reported on coffee use either during the last 1–2 years or 10 years using three different control groups (hospital, community, and patients with chronic liver disease).[55] Coffee use during the last 1–2 years was associated with decreased risk of HCC against all three control groups. Coffee use over the last 10 years was associated with decreased risk of HCC in reference to community controls or patients with chronic liver disease. The results of another case-control study by Montella et al. also found a dose-effect relationship between coffee intake and risk for HCC for persons who consumed 4 or more cups daily (OR = 0.4, 95% CI 0.2–1.1); the inverse relationship was maintained amongst both HBV and HCV-infected individuals.[56]

Several studies support the hypothesis that coffee consumption leads to decreased risk of liver cancer.[57] In a Japanese pooled analysis of two prospective cohort studies, a significant inverse relationship was found between coffee consumption and risk of liver cancer in patients with liver disease.[58] Compared to nondrinkers, those with coffee intake of 1 or more cups daily had a relative risk 0.58 (CI 0.36–0.96).[58] Another Japanese prospective cohort of 18,815 subjects with 110 incident cases of liver cancer found increased coffee consumption to be associated with reduced liver cancer risk (hazard ratio for <1, 1–2, 3 or more cups daily 0.67, 0.49, 0.54, p trend 0.025).[59,60] A similar risk tendency was observed in patients with HBV and/or HCV infections.

The results of a large population-based cohort study by Hu et al. performed in Finland supported the finding that coffee drinking led to a decreased liver cancer risk in a dose-response manner.[61] Multivariable-adjusted hazards ratios of liver cancer in participants who drank 0–1, 2–3, 4–5, 6–7, and 8 or more cups of coffee daily were 1.00, 0.66, 0.44, 0.38, and 0.32 (P = 0.003), respectively. The operational definition of 'liver cancer' in the study included a diagnosis of HCC, cholangiocarcinoma, adenocarcinoma, and primary liver cancer of unspecified etiology.

A Japanese cohort study also showed that those who drank coffee on a daily or almost daily basis had a lower HCC risk than those who almost never drank coffee with a dose-response effect [HR for 3–4 cups daily: 0.48 (95% CI 0.28–0.83)].[60] The risk of HCC in subjects who never or occasionally drank coffee was 547.2 cases per 100 000 over 10 years, but was 214.6 cases per 100 000 in those who drank coffee regularly. Findings from a Singapore prospective cohort study found that compared with coffee non-drinkers, persons who drank 3 or more cups of coffee daily had a 44% risk reduction of HCC (HR 0.56, 95% CI 0.31–1.00, P = 0.049), after adjustment for confounding variables and tea intake.[62] A Hong Kong case-control study found that moderate coffee intake led to a reduction in HCC risk by almost half in daily coffee drinkers with chronic HBV infection compared to non-drinkers (OR 0.54, 95% CI 0.3–0.97), with a dose-response effect (P = 0.02).[63]

The results of several studies analyzing data from the Japan Collaborative Cohort Study indicated the beneficial effects of coffee consumption on the incidence of HCC deaths.[64–66] All three studies found a statistically significant decrease in risk of death from HCC in persons consuming one or more cups of coffee daily. Kurozawa et al. found that in persons aged 60–79 with a history of liver disease, drinking one or more cups of coffee daily had a significant inverse relationship with mortality due to HCC (men: HR 0.44, CI 0.19–0.90, women: 0.30, 0.10–0.89). No significant relationship was found for men and women aged 40–59.[64] In another case-control study by Kurozawa et al., coffee was again found to be protective against HCC mortality.[65] Hazard ratio for one or more cups per day compared to non-coffee drinkers was 0.5 (95% CI 0.31–0.79). The HCC mortality Hazard Ratio was significantly reduced in men who drank one or more cups of coffee daily, but not in women. The authors did not control for HBV or HCV infection.

In a nested case-control study, the multivariate-adjusted OR(with 95% CI) for HCC mortality in daily coffee drinkers (1 or more cups daily) in contrast to nondrinkers was 0.49(0.25–0.96) in all participants, 0.31 (0.11–0.85) in HCV-positive patients, and 0.75 (0.29–1.92) in HCV-negative patients.[66] This study supports previous findings in case-control studies by Gelatti et al. and Ohfuji et al. that coffee has a protective effect amongst HCV-infected persons.[54,67] However, another report did not find a protective effect of coffee consumption among HCV-infected individuals.[56]