Effects of soy or rye supplementation of high-fat diets on colon tumour development in azoxymethane-treated rats.

Abstract

Evidence is accumulating that a diet high in plant-derived foods may be protective against cancer. One class of plant component under increasing investigation is the phytoestrogens of which there are two main groups: the isoflavones, found mainly in soy products, and the lignans, which are more ubiquitous and are found in fruit, vegetables and cereals with high levels being found in flaxseed. In this study, we have used carefully balanced high-fat (40% energy) diets: a control diet (containing low isoflavone soy protein as the sole protein source), a rye diet (the control diet supplemented with rye bran) and a soy diet (containing as protein source a high isoflavone soy protein). The effect of these diets on the development of colonic cancer was studied in F-344 rats treated with the carcinogen, azoxymethane (two doses of 15 mg/kg given 1 week apart). Colons from treated animals were examined for aberrant crypt foci (ACF) and tumours after 12 and 31 weeks. Results after 12 weeks showed no differences in the total number of ACF in the control, soy or rye bran groups. However, the soy group had increased numbers of small ACF (less than four crypts/focus) while the rye group had decreased numbers of large ACF (greater than six crypts/focus). Examination of colons after 31 weeks gave similar low numbers of ACF in each group with no differences in multiplicity. There were no differences in the number of tumours between the control (1.36 tumours/rat) and soy (1.38 tumours/rat) groups. However, there was a significant decrease in the number of tumours in the rye group (0.17 tumours/rat). These results suggest that soy isoflavones have no effect on the frequency of colonic tumours in this model while rye bran supplementation decreases the frequency of colon cancer. This effect is due not to a decrease in early lesions but in their progression to larger multi-crypt ACF. The study also supports the hypothesis that larger ACF are more predictive of subsequent tumorigenicity.