Insulin Assay Standardization: Leading to Measures of Insulin Sensitivity and Secretion for Practical Clinical Care

Response to Staten et al.

  1. Lutz Heinemann, PHD
  1. From the Profil Institut für Stoffwechselforschung Neuss, Germany; and the Profil Institute for Clinical Research, Chula Vista, California.
  1. Corresponding author: Lutz Heinemann, lutz.heinemann{at}

In their recent commentary, Staten et al. (1) describe their very valuable attempt to standardize insulin assays. Their aim is to allow quantitative comparison of the results obtained in one lab with those measured by a different assay in another lab. I'd like to highlight an issue when it comes to reporting the results. The insulin concentration must be stated in SI units nowadays, i.e., in pmol/l instead of μU/ml. The latter unit refers to the biologic action of this hormone (i.e., its blood glucose–lowering activity) and the former to the number of insulin molecules in a given volume. A fixed conversion factor is used to convert the microunits into picomoles: 1 μU/ml insulin = 6 nmol/l. This conversion factor is stated in the SI units table of this journal, available online in the instructions for authors, and also mentioned in the commentary of Staten et al.

One citation backs up this conversion factor. In his letter to the editor in 1993, Vølund (2) stated that an incorrect conversion factor (7.175) was used in the SI unit table published in the American Journal of Clinical Nutrition. That factor relied on the potency of the 4th International Standard of Insulin (24,000 units/g) from 1959. Taking the molecular weight of insulin (5,808 Da) into account, 1 unit = (1/24,000)/5,808 = 7.174 nmol or 1 μU/ml = 7.174 pmol/l. The author clarified that the new World Health Organization insulin standard from 1987 has a potency of 26,000 units/g, and by taking additional factors into account the correct conversion factor is 1 μU/ml = 6.00 pmol/l.

Recently, I learned, however, that many, if not all, SI unit tables give a different conversion factor of 6.945. See, for example: (JAMA!),∼rowlett/units/scales/clinical_data.html, or Manufacturers and analytical labs also appear to use this conversion factor, which is ∼15% higher (35)!

I was unable to find any explanation for this difference on the Internet. When I asked a number of colleagues, none were able to provide a good reason. In addition, I contacted one large insulin manufacturer to see if they were aware of the difference in conversion factor or knew the reason (perhaps a new insulin standard was established). However, they were also not aware of any such change and could provide no answers. They explained that, in the old days, the metabolic activity of insulin varied with batches and bioassays were run to determine the biological activity of each batch; however, there is no longer any need for that step. They suggested that the conversion factor of 6.945 might be a vestige from that time.

In view of the efforts to have more precise and reliable insulin measurements established, I believe we can't accept the 15% difference in measurement results that is introduced by a difference in calculation. This would not only hamper comparison of insulin data from different sources, but also make difficult the comparison of data from different publications. The question would always be: which conversion factor was used?


No potential conflicts of interest relevant to this article were reported.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. See for details.


| Table of Contents

Navigate This Article