CHICAGO, July 29 -- Diabetics who took both insulin and other medications for the disease had fewer plaques associated with Alzheimer's disease than other patients, researchers said here.

In a postmortem study, patients on combination treatment had significantly fewer beta-amyloid plaques (P=0.014) than diabetics who were on one treatment alone or patients who did not have diabetes, Michal Schnaider Beeri, Ph.D., of Mount Sinai School of Medicine in New York, reported at the International Conference on Alzheimer's Disease.

Overall, the combination group had about 80% fewer plaques compared with the other groups combined.

However, there were no significant between-group differences in the occurrence of neurofibrillary tangles.

"These results suggest that the combination [of insulin and other diabetes medications] … may beneficially influence Alzheimer's-related brain changes," Dr. Beeri said. "This also points to biological pathways in the brain, such as insulin signaling, that might be a focus for developing new treatment strategies."

Because diabetes has been associated with a greater risk of mild cognitive impairment and Alzheimer's disease in epidemiological studies, the researchers expected to see more plaques when they examined the brains of diabetics.

In a previous study by Dr. Beeri's group, however, that wasn't the case. In fact, they found fewer plaques. Other neuropathological studies have failed to find any association between diabetes and the number of plaques or neurofibrillary tangles.

Dr. Beeri and her colleagues hypothesized that the treatments for diabetes may influence the neuropathology of Alzheimer's.

To test the hypothesis, the researchers studied 248 brains -- 124 from diabetics and 124 from non-diabetics -- from the Mount Sinai School of Medicine Brain Bank. Most of the specimens came from the Jewish Home and Hospital in Bronx, N.Y., and the two groups were matched by age, sex, and severity of dementia.

The mean age of the patients at death was 81.2 and 57.3% were female. The mean clinical dementia rating was 2.4, indicating moderate to severe dementia.

Non-diabetics had slightly lower body mass indices but the difference was not statistically significant.

Of the diabetics, 29 were not on any medication, 49 were taking insulin only, 28 were taking medications other than insulin -- like glyburide or metformin -- and 18 were taking both insulin and another diabetes medication.

The researchers examined the extent of plaques and neurofibrillary tangles in several neocortical regions and in the hippocampus, entorhinal cortex, and amygdala using the CERAD (Consortium to Establish A Registry for Alzheimer's Disease) neuropathological battery.

The group that was on combination treatment for diabetes had significantly fewer beta-amyloid plaques in the entorhinal cortex (P=0.003), amygdala (P=0.009), and overall (P=0.014) compared with the other groups, which did not differ significantly from each other.

The differences approached statistical significance in the hippocampus (P=0.057) and the combined neocortical measure (P=0.052).

When the researchers additionally controlled for APOE e4 genotype, BMI, and fasting glucose at the time of nursing home admission, there was no change in the findings.

Dr. Beeri acknowledged some limitations of the study, including the inability to determine causation and potential confounding. Nor could the survivor effect be ruled out, she said.

Nevertheless, Dr. Beeri said, the results suggest that diabetes medications may influence biological pathways involved in beta-amyloid processing.