Jump to comment:

• Chronic idiopathic neuropathy and coeliac disease
  Marios Hadjivassiliou
  Published on: 29 September 2005

• Published on: 29 September 2005

  Chronic idiopathic neuropathy and coeliac disease

  • Marios Hadjivassiliou, Consultant Neurologist
  • Other Contributors:
    • Richard A. Grunewlad

  Dear Editor,

  We are concerned that the recent article by Rosenberg et al.[1] contains numerous inaccuracies and methodological errors rendering their conclusion that “it is unlikely that chronic peripheral neuropathy… is associated with coeliac disease” unsound.

  The essence of their own contribution to the field is the screening of 16 patients with idiopathic neuropathy one of whom had positive antiglia...

  Show More

  Dear Editor,

  We are concerned that the recent article by Rosenberg et al.[1] contains numerous inaccuracies and methodological errors rendering their conclusion that “it is unlikely that chronic peripheral neuropathy… is associated with coeliac disease” unsound.

  The essence of their own contribution to the field is the screening of 16 patients with idiopathic neuropathy one of whom had positive antigliadin antibodies but no evidence of coeliac disease (CD) on duodenal biopsy. In the abstract of their article they state: “In 425 of the 478 patients, a cause other than coeliac disease was established. In the patients with no determined cause for neuropathy, one had abnormally increase IgA antigliadin antibodies but duodenal biopsy was normal.” The abstract is misleading, giving the impression that they have screened 53 (478 minus 425) patients with idiopathic neuropathy when in reality they have screened only 16 patients (as indicated in the text). This issue is important as the small number renders the study underpowered, making any sound conclusion on the prevalence of CD in patients with idiopathic neuropathy impossible.

  The issue of small numbers screened may well be what prompted the authors to perform a systematic review of the literature. We have concerns about a number of issues relating to this review:

  1. Study selection for systematic review

  For this systematic review they considered studies to be eligible if they included original data and excluded publications with “results presented in an editorial without details”. They included 10 studies that apparently fulfilled these criteria, claiming that the study design was described in them all. The authors should therefore explain the inclusion of an editorial (2) and a book chapter (3), both by the same authors. The study design and methodology are stated in neither of these publications and the data presented are apparently anecdotal. Furthermore both of these publications refer to the same study population, something that the authors acknowledge, yet still proceed to include this population twice, rendering their statistical calculation invalid.

  2. Assessment for the presence of neuropathy

  In 3 out of 4 of the studies that contribute to Group B (patients with CD investigated for polyneuropathy) the extent of neurological and/or neurophysiological investigation for a neuropathy of the whole cohort of patients with CD is not mentioned. Two of these studies[2,3] should not have been included for the reasons highlighted in point 1 above. Only in one of these studies[4] were systematic neurological and neurophysiological assessments performed in all of their patients with CD, giving the true prevalence of polyneuropathy in patients with CD of 23%. This is in marked contrast to the figure of 0.9% derived from wrongly combining all 4 studies.

  3. Exclusion of patients with CD

  Throughout their systematic review Rosenberg et al exclude all patients with CD who also have an additional autoimmune disease that may be aetiologically linked to a neuropathy. It is well known that patients with CD have a higher prevalence of other autoimmune diseases. This however does not mean that the aetiology of the neuropathy is more likely to be due to the additional autoimmune disease anymore than it is due to CD. Furthermore they also exclude all patients where the diagnosis of CD preceded the diagnosis of neuropathy. Just because CD preceded the neuropathy does not mean that CD should not be considered aetiologically linked to the neuropathy. By excluding such patients they greatly underestimate the prevalence of CD in patients with polyneuropathy.

  4. Inaccurate representation of data

  One of the studies[5] featured in Group A (patients in whom polyneuropathy preceded duodenal biopsy) has as its subject the neurological manifestations in a cohort of 144 patients with CD. Rosenberg et al do not include these data in their review but use the anecdotal report in the discussion section of 3 out of 204 patients with idiopathic neuropathy who were found to have CD. This is despite a clear statement by the authors of this paper that “it is probable that these calculations underestimate the frequency of CD because screening for CD in this group was not systematic”. Further problems arise with the interpretation of the third study[6] where Rosenberg et al make the assumption that the 400 patients screened all had idiopathic neuropathy.

  Yours sincerely,

  Marios Hadjivassiliou
  Richard Grünewald

  References

  1. Rosenberg NR, Vermeulen M. Should coeliac disease be considered in the work up of patients with chronic peripheral neuropathy? J Neurol Neurosurg Psychiatry 2005;76:1415-1419.

  2. Pengiran Tangah DS, Wills AJ, Holmes GK. Neurological complications of coeliac disease. Postgrad Med J 2002;78:893-8.

  3. Holmes GK. Neurological and psychiatric complications in coeliac disease. In:Epilepsy and other neurological disorders in coeliac disease. London:John Libbey 1997:251-64.

  4. Luostarinen L, Hinamen SL, Luostarinen M, Collin P, Pirttila T. Neuromuscular and sensory disturbances in patients with well treated coeliac disease. J Neurol Neurosurg Psychiatry 2003;74:490-494.

  5. Luostarinen L, Pirttila T, Collin P. Coeliac disease presenting with neurological disorders. Eur Neurol 1999;42:132-135.

  6. Chin RL, Sander HW, Brannagan TH, Green PHR, Hays AP, Alaedini A, Latov N. Celiac neuropathy. Neurology 2003.

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top