Prevalence and Causes of Abnormal Liver Function in Patients With Coeliac Disease

Giovanni Casella; Elisabetta Antonelli; Camillo Di Bella; Vincenzo Villanacci; Lucia Fanini; Vittorio Baldini; Gabrio Bassotti


Liver International. 2013;33(7):1128-1131. 

In This Article


Liver abnormalities are a common extraintestinal manifestation in patients with classical CD and hypertransaminaseaemia could be the only sign of CD, in absence of gastrointestinal symptoms.[11] In the present series of 245 adult coeliac patients, the overall prevalence of increased AST or ALT levels at diagnosis was 17.5%. The frequency found in our study was relatively lower than those previously described (possibly because of differences in the populations under observation) and no correlation between serum transaminase levels and patient's age was observed.

The hepatic involvement in our CD patients was asymptomatic, as frequently described in these subjects [15], with the increase in liver enzymes being generally mild or moderate, even though in some patients it may be ten times higher than the upper normal limit.[4] The pathogenetic mechanisms underlying liver injury in CD patients are poorly understood; a widely accepted hypothesis is based on the increased intestinal permeability, which favours the absorption of toxins, antigens and inflammatory substances via the portal vein circulation.[16]

Before GFD was started we observed a mild (less than 3 times the upper limit of normal) increase in serum liver enzymes, with ALT more often abnormal than the AST. GFD is an effective medical therapy for hypertransaminaseaemia in patients with CD and this reversible gluten-induced liver damage has been described as "coeliac hepatitis".[17] Within a year of good adherence to the diet, serum transaminase normalized in 75–95% of CD patients, suggesting a relationship between gluten intake, intestinal damage and liver injury.[4] It remains to be established whether a long-term (several years) follow-up of these patients while on GFD would result in the complete normalization of the entire patients' group.

In our study, after 1 year of GFD we observed normal transaminase values in 88% of patients, which is in line with other literature data. In five patients, who had persistent hypertransaminaseaemia despite gluten exclusion, a different aetiology for liver injury was observed, namely PBC (2 patients), HCV infection (2 patients) and NASH (1 patient). On the contrary, in our series we did not identify patients with primary sclerosing cholangitis (PSC), which is reported to have a prevalence from 1.6 to 2.6% in CD patients[18,19] or patients affected by autoimmune hepatitis, reported to have a prevalence of 4.4% in CD subjects.[8]

Concerning the association between CD and PBC, this was first described by Logan et al.,[20] and a large study conducted on more than 8000 subjects subsequently confirmed the data revealing a prevalence of PBC 20-fold higher in CD patients than in the general population.[21] The frequency of diagnosed PBC in patients with CD is thought to be comprised between 0 and 3%. A recent epidemiological study, confirmed such evidence in a large cohort, estimating that CD patients have a four-fold risk of PBC vs the general population.[22] A shared susceptibility of biliary and small bowel epithelium to be attacked by autoimmune mechanisms could explain the association between CD and PBC.[23] Conversely, a clear association between CD and chronic hepatitis C is lacking and it is known that the prevalence of CD in HCV-related chronic hepatitis is not different from that of CD in the general population (1.2% vs 1%).[24] However, CD may present for the first time during treatment for hepatitis C, probably related to a type 1 helper T-cell immune response.[25,26] For this reason, it could be wise to rule out the presence of CD in a patient with HCV chronic hepatitis before starting the treatment with interferon.

Concerning the association between NASH and CD, a few studies describe various forms of fatty liver in patients with coeliac disease;[5,27–29] in all cases, patients had elevations in liver enzyme levels and evidence of hepatic steatosis on liver biopsy or ultrasound, with clinical and biochemical improvement after GFD. In our study the patient affected by NASH had persistent hypertransaminaseaemia even after GFD; however, the presence of fatty liver is a very common clinical finding in industrialized countries, affecting 10–24% of the general population[30] and some studies do not recognize a direct link between CD and NASH.[31] In addition, the mechanisms involved in fat deposition in the liver of CD patients are not defined; the deposition might be related to malabsorption secondary to advanced disease, leading to chronic deficiency of a lipotropic factor such as choline.[32] In this scenario, it is hard to draw definite conclusions from the available studies concerning the occurrence of NASH in coeliac patients. It seems that this association could be because of a coincidence rather than a true correlation or if an association exists, it is not strong. For this reason, the actual scientific evidence is not sufficient to warrant a serological screening for CD in NASH patients.

In conclusion, this study confirms that the presence of hypertransaminaseaemia at diagnosis of CD and the lack of normalization of liver enzymes after 12 months of GFD suggest the possibility of coexisting liver disease. In this contest, further evaluation is recommended, including serological testing and, eventually biopsy to exclude liver disease (even though the new non invasive methods seem able to reliably predict the presence of liver fibrosis.[33] Early recognition and treatment of CD in patients affected by liver disease are advisable to improve the liver function and prevent further complications.[22,23]