Perspectives in carbohydrate toxicology with special reference to carcinogenicity.

All chemicals, even water and salt, can cause toxic effects if they are given to humans or laboratory animals in high enough doses. Similarly, the incidences of various kinds of neoplasm may be increased non-specifically in animals by the administration of innocent chemicals by an inappropriate route or in doses that are excessive enough to disturb normal nutritional, or hormonal status or interfere with mineral balance. High dietary concentrations of sorbitol or xylitol, if fed to laboratory rats cause enlargement of the caecum, increased absorption of calcium from the gut, increased urinary excretion of calcium, pelvic and corticomedullary nephrocalcinosis, acute tubular nephropathy, urinary calculus formation and both hyperplasia and neoplasia of the adrenal medulla. High dietary concentrations of lactose give rise to a similar spectrum of effects when given in excessive dosage to laboratory rats. Recent evidence suggesting that in the rat, but not in the mouse or in man, excessive calcium absorption stimulates the adrenal medulla is reviewed. In the mouse, but not in the rat or in man, a biologically significant amount of glycolic acid, which is a minor metabolite of xylitol in all three species, is converted to oxalate which then appears in the urine. Although the increase in urinary oxalate in the mouse is only about 20% of normal, this is enough in animals fed on diets containing 10% or 20% xylitol to predispose to bladder stone formation, and the prolonged presence of stones in the bladders, particularly of mice, in turn, predisposes to bladder tumour development. Neither bladder stones nor bladder tumours are seen in rats because biologically significant conversion of glycolate to oxalate does not occur. Studies in humans exposed up to 1 g/xylitol/kg body weight/day have revealed no evidence of increased urinary oxalate excretion. It is concluded that both the bladder tumours seen in mice, in response to 10% or 20% xylitol in the diet, and the adrenal tumours seen in rats, in response to 20% sorbitol or 20% xylitol in the diet, are laboratory artefacts. In other words, humans exposed to "normal" levels of these agents would be at no risk of developing either of these kinds of neoplasm.