GlcN is a widely used dietary supplement that is described as efficacious and safe for many individuals with osteoarthritis, especially of the knees 66
. Concerns that GlcN consumption may worsen glucose tolerance and induce insulin resistance were not based on clinical observations, but on in vitro
studies by Marshall et al.
showing that exogenous GlcN could increase the activity of the hexosamine biosynthesis pathway, a metabolic process that is believed to function as a nutrient sensor modulating insulin sensitivity and glucose uptake in peripheral tissues 4
(). The end product of this pathway is UDP-N
-acetylGlcN, a substrate for O
-GlcNAc transferase, which mediates the addition of β-N
-acetylGlcN to the hydroxyl groups of serine and/or threonine residues on a wide variety of proteins. This post-translational modification regulates a wide range of biological processes, including signal transduction/metabolic proteins that modulate glucose metabolism and insulin sensitivity 5
. As reviewed by Copeland et al. 5
, there are strong associations between elevated GlcN acylation of proteins with glucose toxicity and impaired insulin signalling; excessive flux of sugars through the hexosamine signalling pathway has therefore been implicated as a causative factor in the development of type 2 diabetes.
Figure 1 Glucosamine and the hexosamine biosynthesis pathway. Glucose transporters are indicated by arrow and major enzymes are included in ellipses. GLUT1, GLUT2, GLUT4, glucose transporters; GlcN-6-P, glucosamine-6-phosphate; Glc-NAc-6-P, N-acetyl-glucosamine-6-phosphate; (more ...)
There is a large body of evidence from in vitro
studies using human- and rodent-derived cells that high concentrations of GlcN in the incubation media impairs insulin-mediated glucose uptake. This has led mistakenly to GlcN being described as a diabetogenic agent. Interference with glucose metabolism occurs only at concentrations comparable with those of glucose, i.e. within the 2–50 mmol/L range, concentrations that are several hundred- to a thousand-fold greater than plasma concentrations that occur during oral supplement use. At these concentrations (<10 µmol/L), GlcN neither augments the hexosamine biosynthesis pathway nor reduces insulin-mediated glucose uptake 67,68
. Additionally, in vitro
observations in human adipocytes have shown that in the presence of normal glucose concentrations, the uptake of GlcN is almost completely inhibited 67
Animals-administered high intravenous concentrations of GlcN have shown that rodents are especially sensitive to its diabetogenic effects. During euglycaemic–hyperinsulinaemic clamping, the infusion of GlcN in rodents producing plasma GlcN concentrations of between 800 and 1200 µmol/L results in glucose intolerance and insulin insensitivity 32,37
. In contrast, the consumption of GlcN at doses many times greater than are used clinically was not associated with any adverse effects on glucose metabolism. For example, Echard et al. 69
reported that in strains of rodents that are highly sensitive to sugar-induced insulin resistance, the consumption of GlcN in large amounts (9% in the diet; ~4.5 g/kg body weight) for 9 months had no effect on fasting glucose concentrations or glucose tolerance. Similar observations have been made in other species (dog and rabbit) in which GlcN was administered orally, at doses greatly exceeding those used clinically 70,71
. In none of these studies were sensitive insulin response monitoring methods employed or plasma GlcN levels measured; nevertheless, the absence of a diabetogenic effect in animal feeding studies is consistent with its low bioavailability and its lack of biological effect on glucose metabolism.
In clinical trials investigating the effect of GlcN on osteoarthritis, GlcN has been administered to many patients with type 2 diabetes or impaired glucose tolerance, without specifying the numbers. In long-term randomized placebo-controlled studies, daily GlcN use was not associated with increased incidences of type 2 diabetes, and a non-significant fall in fasting blood glucose values over 3-year periods of daily GlcN use was reported in one study, an effect that could be explicable by the concept of regression to the mean as it was observed only in those with the highest blood glucose values at the start 51,52
. Overall, the data from randomized placebo-controlled osteoarthritis trials have not shown any adverse effects on fasting blood glucose levels, glucose metabolism, or insulin sensitivity from oral GlcN, at any dose level.
Six studies that investigated diabetes-related outcomes as primary endpoints were identified, and included participants with type 2 diabetes or obese subjects with apparent insulin resistance 7,15,44–46
. In most short-term studies, there were no significant changes in fasting plasma glucose or insulin concentrations or in HbA1c
after consumption of GlcN in usual doses. Nor was there any change in insulin sensitivity determined by the use of a hyperglycaemic clamp. One short-term clinical study 7
, which concluded that oral GlcN use in ‘pre-diabetic’ subjects may adversely affect insulin sensitivity and glucose tolerance, is flawed by its small sample sizes (n
= 3), the use of post hoc
analysis, failure to allocate subjects to treatment groups in a randomized fashion, and the inherent limitations of uncontrolled intervention studies.
The only diabetogenic effect attributed to GlcN in humans was in response to high-dose GlcN infusion, which resulted in a slight elevation in plasma glucose concentrations following an IVGTT 40
. However, this observation was not due to the development of insulin resistance, and was consistent with GlcN's well-established inhibitory activity towards glucokinase, an effect that in humans would only occur when plasma levels of GlcN approach normal glucose concentrations. This concentration (5 mM) is roughly 500- to 1000-fold above plasma levels that are reasonably expected following supplemental GlcN use.
In conclusion, the available evidence implicating GlcN as a diabetogenic agent are limited to rodent infusion studies and in vitro observations, experimental models that were determined not to be relevant to humans. A comprehensive and critical review of the clinical literature indicated that the consumption of GlcN at usual doses was well tolerated by normal, diabetic, and ‘pre-diabetic’ subjects. Thus, based on the overall weight of scientific evidence, there currently appears to be no reason to restrict the use of oral GlcN for individuals at risk for diabetes, or those with type 1 or 2 diabetes, or normoglycaemics with respect to any adverse effects on sugar metabolism.