Gut 60:1602-1603 doi:10.1136/gut.2010.230656
  • PostScript
  • Letter

Risk of all-cause and cardiovascular mortality in patients with chronic liver disease

  1. Christopher P Day2
  1. 1Section of Endocrinology and Metabolism, Department of Medicine, University of Verona, Verona, Italy
  2. 2Institute of Cellular Medicine, Newcastle University, Newcastle upon Tyne, UK
  1. Correspondence to Dr Giovanni Targher, Section of Endocrinology and Metabolism, Department of Medicine, University of Verona, Ospedale Civile Maggiore, Piazzale Stefani 1, 37126 Verona, Italy; giovanni.targher{at}

We read with interest the article by Stepanova et al in a recent issue of Gut. These authors used death certificate data from the third National Health and Nutrition Examination survey (NHANES-III 1988–1994) in order to assess the impact of metabolic syndrome components on all-cause and cause-specific mortality in patients with chronic liver disease (CLD).1 They concluded that components of the metabolic syndrome are independently associated with overall and liver-related mortality in patients with CLD (principally among those with viral hepatitis C, alcohol-related liver disease (ALD) or non-alcoholic fatty liver disease (NAFLD)).1

Overall, the article is well written and the authors have discussed their results cautiously and in a balanced way, given the well-recognised limitations of using the NHANES-III dataset to study the prevalence and the possible aetiologies of CLD. Important limitations that have not been sufficiently emphasised by the authors in this article are the relatively low number of mortality outcomes (especially for liver disease and diabetes) and the lack of validation of cause of death. Although ascertainment of vital status using the National Death Index is very high, assigning cause of death based on death certificate diagnoses may lead to misclassification.2 Such inaccuracies might have resulted in misclassification of some participants.

In this analysis, ALD was presumed in subjects who reported excessive alcohol consumption (ie, >20 g/day for men and >10 g/day for women) and had elevated serum aminotransferases (ie, ALT>40 U/l or AST>37 U/l in men and ALT or AST>31 U/l in women). Based on this definition, the prevalence of ALD in the NHANES-III cohort appears to be very low (ie, ∼1.4%). If correct, it would be interesting to know how the authors have ‘handled’ in their analyses the subjects who reported an excessive alcohol consumption but had normal serum aminotransferases as well as those who had a combination of potential causes of CLD (including alcohol abuse, viral hepatitis or iron overload).

In this analysis, the authors also reported that none of the CLD (including NAFLD, which had a prevalence of 5.5% in the NHANES-III cohort and was suspected in patients who had elevated aminotransferases in the absence of any other evidence of CLD) was associated with cardiovascular (CVD) mortality, and that none of the components of metabolic syndrome (except for diabetes) was associated with CVD mortality in patients with NAFLD. At first glance, these findings might appear somewhat unexpected. There is a mounting body of evidence suggesting that NAFLD patients have multiple risk factors for CVD; CVD is much more common than liver disease as a cause of death in NAFLD patients, especially in those with more advanced stages of disease; and NAFLD is linked to an increased risk of incident CVD events.3

To gain a better understanding of these unexpected findings, we believe that the authors should provide the number of deaths and the age-, and sex-standardised rates for all-cause and CVD mortality for the suspected NAFLD group and the control group. Moreover, based on the results published by Dunn et al, who demonstrated a significant age interaction for both all-cause mortality and CVD mortality using the NHANES-III database (ie, suspected NAFLD, especially in the 45–54-year-old age group, was a strong and independent risk factor for CVD death),4 it would be interesting to know if all-cause and CVD survival curves for subjects with and without suspected NAFLD were different after stratification for age decades. Finally, it would be also interesting to know if the results observed by the authors were different if they used more stringent criteria to diagnose NAFLD (ie, ALT >30 U/l for men and ALT >19 U/l for women).


  • Competing interests None.

  • Patient consent Obtained.

  • Provenance and peer review Not commissioned; not externally peer reviewed.


Relevant Article