Reply

Authors

  • Nazim A. Ghouri M.D.,

    1. Institute of Cardiovascular & Medical Sciences, British Heart Foundation Glasgow Cardiovascular Research Centre University of Glasgow Glasgow, UK
    Search for more papers by this author
  • David Preiss M.D.,

    1. Institute of Cardiovascular & Medical Sciences, British Heart Foundation Glasgow Cardiovascular Research Centre University of Glasgow Glasgow, UK
    Search for more papers by this author
  • Naveed Sattar M.D.

    Corresponding author
    1. Institute of Cardiovascular & Medical Sciences, British Heart Foundation Glasgow Cardiovascular Research Centre University of Glasgow Glasgow, UK
    • Institute of Cardiovascular & Medical Sciences, British Heart Foundation Glasgow Cardiovascular Research Centre University of Glasgow Glasgow, UK
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

Reply:

We thank Musso and colleagues for their useful comments on our article.We agree that a comprehensive approach to patients with nonalcoholic fatty liver disease, namely one which considers their diabetes and cardiovascular risk, as well as their long-term liver-related risk, is needed. We wish only to make some points of clarification which will help hepatologists to efficiently assess cardiovascular and diabetes risk in their patients. First, there have been recent calls, recently verified by the World Health Organization, to diagnose diabetes by measurement of hemoglobin A1c (HbA1c), and this measure has the advantage of being measurable on nonfasting samples.1 Of interest, HbA1c is at least as good as fasting glucose at assessing microvascular risk and appears to be the best glycemic predictor of macrovascular risk, performing even better than the 2-hour glucose tolerance test.2 We would therefore argue against routine oral glucose tolerance testing in patients with nonalcoholic fatty liver disease, because of both clinical and economical concerns. Second, the benefit of testing for insulin as a prognostic factor, we believe, remains to be proven. At present, we would advise against measuring insulin routinely, at least until this approach has been shown to be cost-effective. Third, lipid tests to enhance prediction of cardiovascular disease (as part of national cardiovascular risk guidelines or algorithms) also do not require fasting, as recently shown by the Emerging Risk Factor Collaboration group.3 Rather, nonfasting total cholesterol (or low-density lipoprotein cholesterol or non–high-density lipoprotein cholesterol) and high-density lipoprotein cholesterol adequately reflect lipid-associated vascular risk and perform as well as fasting lipids3 for risk prediction. Fasting is therefore not required to test for diabetes, or indeed cardiovascular risk, in the majority of patients. Finally, we have recently proposed4 how combined vascular and diabetes screening may be conducted by simply adding a nonfasting HbA1c test to nonfasting lipids as part of a cardiovascular risk screen, a simple process which may be easily added to other routine tests conducted in most hepatology and hepatology-related clinics.

Nazim A. Ghouri M.D.*, David Preiss M.D.*, Naveed Sattar M.D.*, * British Heart Foundation Glasgow Cardiovascular Research Centre, University of Glasgow, Glasgow, UK.

Ancillary