Advertisement
Mayo Clinic Proceedings Home

Regarding L-Carnitine and Cardiovascular Disease

      To the Editor:
      We read with interest the recent meta-analysis performed by DiNicolantonio et al
      • DiNicolantonio J.J.
      • Lavie C.J.
      • Fares H.
      • Menezes A.R.
      • O'Keefe J.H.
      L-Carnitine in the secondary prevention of cardiovascular disease: systematic review and meta-analysis.
      on L-carnitine and cardiovascular disease published in the June 2013 issue of Mayo Clinic Proceedings, and we have several concerns regarding the analysis and conclusions.
      Although the authors evaluated the quality of the included studies and reported their risk of bias, we believe that they underestimated this risk and that this body of evidence is, in fact, of very low quality. We also believe that the benefit reported is not believable, in part because it is greater than other well-studied interventions in cardiovascular disease.
      Regarding the risk of bias, several of the weaknesses were noted by the authors. However, other weaknesses were not identified. Of the 13 articles included in the 5 meta-analyses performed, only 4 were definitely blinded to the patient, health care practitioner, and outcome assessor (Supplemental Table 3 in the article), and only 6 were of good quality by the authors' own criteria (Jadad score >2). The authors indicate in their Supplemental Table 1 that the articles by De Pasquale et al
      • De Pasquale B.
      • Righetti G.
      • Menotti A.
      L-carnitine for the treatment of acute myocardial infarct.
      and Davini et al
      • Davini P.
      • Bigalli A.
      • Lamanna F.
      • Boem A.
      Controlled study on L-carnitine therapeutic efficacy in post-infarction.
      have “unclear bias risk.” Both of these articles found a much greater effect of L-carnitine on mortality than the other articles in the meta-analysis. The article by De Pasquale et al
      • De Pasquale B.
      • Righetti G.
      • Menotti A.
      L-carnitine for the treatment of acute myocardial infarct.
      was not a randomized trial, and the investigators gave L-carnitine when it was “available.” A study such as this is at high risk for bias and should not have been included in the meta-analysis. The study by Davini et al
      • Davini P.
      • Bigalli A.
      • Lamanna F.
      • Boem A.
      Controlled study on L-carnitine therapeutic efficacy in post-infarction.
      was not blinded. Results indicate that postinfarction angina developed in 77 of the 79 patients in the control group, a proportion far greater than the literature reports in this situation.
      This study also found that L-carnitine reduced heart rate, systolic blood pressure, and blood lipid levels—findings not reported elsewhere in the literature. Thus, this study is also at high risk for bias and should not have been included in the meta-analyses.
      The authors acknowledged that there have been problems noted with research conducted by Dr Singh, but they rate the article by Singh et al
      • Singh R.B.
      • Niaz M.A.
      • Agarwal P.
      • Beegum R.
      • Rastogi S.S.
      • Sachan D.S.
      A randomised, double-blind, placebo-controlled trial of L-carnitine in suspected acute myocardial infarction.
      as the second highest-quality article in their meta-analysis. They included the article because Singh et al have not retracted their article. Dr Singh has not retracted any of his many articles, but when asked to produce his data for the BMJ he replied that they were “eaten by termites.”
      • White C.
      Suspected research fraud: difficulties of getting at the truth.
      As a result, both the Lancet and the BMJ have retracted all the articles he has published in their journals. The Postgraduate Medical Journal (where the article on carnitine by Singh et al
      • Singh R.B.
      • Niaz M.A.
      • Agarwal P.
      • Beegum R.
      • Rastogi S.S.
      • Sachan D.S.
      A randomised, double-blind, placebo-controlled trial of L-carnitine in suspected acute myocardial infarction.
      cited by DiNicolantonio et al
      • DiNicolantonio J.J.
      • Lavie C.J.
      • Fares H.
      • Menezes A.R.
      • O'Keefe J.H.
      L-Carnitine in the secondary prevention of cardiovascular disease: systematic review and meta-analysis.
      was published) does not have the same resources as the Lancet to investigate fraud. It is clear from the article by White
      • White C.
      Suspected research fraud: difficulties of getting at the truth.
      in the BMJ (cited by the authors) that the investigators from the BMJ believe that Dr Singh's research is fraudulent. This article should not have been included in the meta-analyses.
      The authors note that the article by Iliceto et al
      • Iliceto S.
      • Scrutinio D.
      • Bruzzi P.
      • et al.
      Effects of L-carnitine administration on left ventricular remodeling after acute anterior myocardial infarction: the L-carnitine Ecocardiografia Digitalizzata Infarto Miocardico (CEDIM) Trial.
      had complete follow-up (Supplemental Table 3). This article studied the effect of L-carnitine on echocardiography after 1 year (negative study), not mortality. The authors of this study reported on mortality but made no effort to search death indexes. There were nearly as many dropouts as deaths, which suggests that, had death indices been searched, the results could have been considerably different.
      The authors noted that 3 of their meta-analyses gave significant results. The angina meta-analysis had only 2 studies, one by Davini et al
      • Davini P.
      • Bigalli A.
      • Lamanna F.
      • Boem A.
      Controlled study on L-carnitine therapeutic efficacy in post-infarction.
      and one by Singh et al.
      • Singh R.B.
      • Niaz M.A.
      • Agarwal P.
      • Beegum R.
      • Rastogi S.S.
      • Sachan D.S.
      A randomised, double-blind, placebo-controlled trial of L-carnitine in suspected acute myocardial infarction.
      This meta-analysis should not have been conducted because of the poor quality of these 2 articles. Furthermore, the number needed to treat for this meta-analysis is 3, which is difficult to believe (it far exceeds the benefits of β-blockers, angioplasty, and other currently used interventions). The analysis on clinically important ventricular arrhythmias examined 5 studies. The article by Singh et al
      • Singh R.B.
      • Niaz M.A.
      • Agarwal P.
      • Beegum R.
      • Rastogi S.S.
      • Sachan D.S.
      A randomised, double-blind, placebo-controlled trial of L-carnitine in suspected acute myocardial infarction.
      was the study with the highest weight. The 2 studies with significant results
      • Rizzon P.
      • Biasco G.
      • Di Biase M.
      • et al.
      High doses of L-carnitine in acute myocardial infarction: metabolic and antiarrhythmic effects.
      • Martina B.
      • Zuber M.
      • Weiss P.
      • Burkart F.
      • Ritz R.
      Anti-arrhythmia treatment using L-carnitine in acute myocardial infarct.
      examined the reduction in ventricular premature beats on day 2 after myocardial infarction, which is not considered a serious ventricular arrhythmia. The number needed to treat for this meta-analysis was 4, which is similarly extraordinarily low and difficult to believe.
      The authors did not perform a sensitivity analysis based on study quality, which would have been helpful. The mortality analysis used the I2 statistic and suggested low heterogeneity. This statistic is misleading with so many small studies. In fact, heterogeneity is likely in this analysis, considering that the χ2 statistic for heterogeneity was greater than its degrees of freedom, which makes combining these studies concerning.
      As outlined above, there are too many flaws with the articles included in this systematic review to conclude that L-carnitine is associated with a 27% reduction in mortality, a 65% reduction in ventricular arrhythmias, or a 40% reduction in angina. Thus, it is premature to conclude from this weak evidence that “considering its low cost and excellent safety profile, L-carnitine could be considered in selected patients.”

      References

        • DiNicolantonio J.J.
        • Lavie C.J.
        • Fares H.
        • Menezes A.R.
        • O'Keefe J.H.
        L-Carnitine in the secondary prevention of cardiovascular disease: systematic review and meta-analysis.
        Mayo Clin Proc. 2013; 88: 544-551
        • De Pasquale B.
        • Righetti G.
        • Menotti A.
        L-carnitine for the treatment of acute myocardial infarct.
        Cardiologia. 1990; 35 ([in Italian]): 591-596
        • Davini P.
        • Bigalli A.
        • Lamanna F.
        • Boem A.
        Controlled study on L-carnitine therapeutic efficacy in post-infarction.
        Drugs Exp Clin Res. 1992; 18: 355-365
        • Singh R.B.
        • Niaz M.A.
        • Agarwal P.
        • Beegum R.
        • Rastogi S.S.
        • Sachan D.S.
        A randomised, double-blind, placebo-controlled trial of L-carnitine in suspected acute myocardial infarction.
        Postgrad Med J. 1996; 72: 45-50
        • White C.
        Suspected research fraud: difficulties of getting at the truth.
        BMJ. 2005; 331: 281-288
        • Iliceto S.
        • Scrutinio D.
        • Bruzzi P.
        • et al.
        Effects of L-carnitine administration on left ventricular remodeling after acute anterior myocardial infarction: the L-carnitine Ecocardiografia Digitalizzata Infarto Miocardico (CEDIM) Trial.
        J Am Coll Cardiol. 1995; 26: 380-387
        • Rizzon P.
        • Biasco G.
        • Di Biase M.
        • et al.
        High doses of L-carnitine in acute myocardial infarction: metabolic and antiarrhythmic effects.
        Eur Heart J. 1989; 10: 502-508
        • Martina B.
        • Zuber M.
        • Weiss P.
        • Burkart F.
        • Ritz R.
        Anti-arrhythmia treatment using L-carnitine in acute myocardial infarct.
        Schweiz Med Wochenschr. 1992; 122 ([in German]): 1352-1355
      1. Cochrane Handbook for Systematic Reviews of Interventions. Wiley-Blackwell, West Sussex, England2008

      Linked Article