NEW YORK - Adding soy supplements to the diet may not improve blood sugar control in older women who are at high risk of or in the early stages of type 2 diabetes, a new study suggests.

The findings, reported in the American Journal of Clinical Nutrition, add to the conflicting body of research on soy and diabetes risk.

Lab research has suggested that soy proteins and soy isoflavones -- "phytoestrogen" compounds that are structurally similar to human estrogen -- may help control blood sugar levels. But so far, the few small clinical trials that have been done have reached different conclusions as to whether soy foods or soy-protein supplements are beneficial to people with diabetes.

For this latest study, researchers in Hong Kong recruited 180 postmenopausal women who either had "pre-diabetes" -- blood sugar levels that are elevated but not high enough for a type 2 diabetes diagnosis -- or were in the early stages of diabetes and had not yet started any treatment.

Zhao-min Liu and colleagues at the Chinese University of Hong Kong randomly assigned the women to one of three groups: one that took a powdered supplement containing milk protein; one that took a supplement of milk protein plus soy isoflavones; and a third that took a soy-protein supplement with additional isoflavones.

All three groups were told to mix the supplement with a drink each morning but otherwise maintain their normal diets.

After six months, Liu's team found no clear benefit of either the soy or the milk-plus-isoflavone supplement in the women's blood sugar control or levels of the blood-sugar-regulating hormone insulin.

One exception, the researchers write, was that there was a "marginally favorable effect" of the soy supplement on women's blood sugar levels two hours after eating.

Still, they conclude, the findings do not support the theory that soy or isoflavone supplements benefit overall blood sugar control in women like those in the study.

However, the findings should not be seen as the final word on soy and diabetes, according to Liu's team. They note that this appears to be the first controlled clinical trial designed to specifically look at the effects of soy supplements in people with pre-diabetes or early, untreated type 2 diabetes.

"Additional studies that are based on longer trials and using different soy and phytoestrogen regimens are warranted," the researchers conclude.