Comment on: Harte et al. High fat intake leads to acute postprandial exposure to circulating endotoxin in type 2 diabetic subjects. Diabetes Care 2012;35:375-382.

We read with great interest the article by Harte et al. (1), which reports that high intake of dietary fat results in increased levels of endotoxin in the blood of both type 2 diabetic patients and control subjects, thereby supporting an earlier article by Erridge et al. (2) who studied healthy subjects.

Low-grade inflammation is associated with obesity, and much evidence suggests that such chronic tissue inflammation is a key factor contributing to development of type 2 diabetes. Some of this inflammation may be caused by endotoxin from the gut microbiota, from where endotoxin may be absorbed together with dietary fat. In such a case, endotoxin should increase in the plasma after a fat-rich meal. Endotoxin in the circulation associates with lipopolysaccharide-binding protein, and this complex is responsible for activating Toll-like receptors on leukocytes.

Endotoxin (also called lipopolysaccharide) is a major constituent of the outer cell wall of Gram-negative bacteria, which can also be found in the gut microbiota. Endotoxin levels are widely measured by the use of the commercially available limulus ameboecyte lysate assay that was also used by the authors (1). As endotoxin does not have a well-defined molecular weight, its activity is usually reported as endotoxin units (EUs), as is also done by the authors. For the World Health Organization international standard of endotoxin, 1 EU corresponds to 100 pg of E.coli lipopolysaccharide (NIBSC code: 94/580). Lipopolysaccharide is a potent inflammatory compound that activates whole-blood leukocytes in the pg/mL range (3). Harte et al. have reported levels of endotoxin to be between 3.3 and 14.2 EU/mL, but these concentrations are known to increase levels of tumor necrosis factor α (TNFα) in plasma and induce a massive inflammatory response in humans. Low-grade inflammation has been induced in human volunteers by intravenous injection or infusion of lipopolysaccharide (4,5), and as little as 60 pg/kg body weight given as a bolus resulted in a significantly increased plasma level of TNFα (4). This dose would be expected to result in a peak concentration of endotoxin in a 70 kg human of 1 pg/mL being equivalent to around 0.003 EU/mL. Furthermore, a concentration of >0.25 EU/mL is indicative of endotoxemia in humans, and endotoxin levels of around 450 pg/mL (= 4.5 EU/mL) are found in patients with Gram-negative infection (6).

In the study by Harte et al. (1), endotoxin was reported to increase from 3.3 to 6.3 EU/mL in normal subjects and from 5.3 to 14.2 EU/mL in the type 2 diabetic group. At the same time, there was no increase in levels of the inflammatory marker TNFα in plasma in both groups. So whatever the authors are measuring with the limulus amebocyte lysate assay, it does not represent endotoxin that is bioactive in humans. Furthermore, the lack of such bioactive endotoxin questions the suggestion that gut-derived endotoxin is a contributing factor for development of low-grade inflammation in adipose tissue or in other tissues that seems to be a hallmark for development of type 2 diabetes.