When then-NIH researcher Lon White, M.D., launched the Honolulu-Asian Aging Study (HAAS) in 1991 by cognitively testing 3,734 subjects who had submitted diet information in midlife as part of the Honolulu Heart Study (HHS), he did not know what he would find.

The discovery that those who ate tofu more than twice a week throughout midlife were more likely to have cognitive impairment and brain atrophy in old age surprised everyone.

"We went through the data to see what we could find in terms of patterns for risk factors, and then I stumbled upon this association," he told Psychiatric News.

The association to which he referred is that higher tofu consumption emerged as a statistically strong and independent predictor of both poorer cognitive function and smaller brain size.

White and his team measured cognitive functioning with the Cognitive Abilities Screening Instrument (CASI) and measured brain atrophy with neuroimaging and autopsy.

How is he so sure that the higher rates of cognitive impairment and brain atrophy couldn’t be linked to something else? White reports making a strong effort to identify all of the factors that "traveled with" higher tofu consumption, and testing each of these confounding factors for their correlation with cognitive function or brain size.

White, who initially thought that he might find a link with coffee or alcohol intake, said, "This finding essentially fell on me from out of the blue. Even though I used stratification and multivariate statistical analysis in an effort to attribute the effect to other factors, the tofu finding would not go away."

He added, "I and everyone else would have been ecstatic to have found an association in the other direction."

White also has plans to publish a paper with another shocking finding from this study—that rates of Alzheimer’s disease in those who consumed more tofu in midlife are 2.5 times higher.

The HAAS used diet information gathered decades before as part of the Honolulu Heart Study. In the latter study, the Japanese Americans submitted information about their diets in 1965 and again in 1971. Twenty years later, the HAAS researchers completed the cognitive testing and brain measurements in the same men, who were now in late life. According to White, those who reported eating the most tofu in midlife had lower CASI scores. "The results were equivalent to what they would be if [the men] were five years older," said White.

To measure brain atrophy, 574 men underwent MRI brain scans, and 290 of the deceased men were autopsied. White found brain atrophy in the men who had reported eating more tofu during midlife. While atrophy normally occurs with aging, White noted that the atrophy was exaggerated in those who ate the most tofu.

White pointed out that not all of the subjects with brain atrophy had poorer cognitive function. "What is clear is that the relationship of midlife tofu intake with atrophy did not depend on the appearance of cognitive impairment," he stated.

Researchers have known for years about the effects of phytoestrogens, which are isoflavone molecules produced in soy plants as they grow. As a self-preservation tactic, the phytoestrogens mimic estrogen, which may interfere with a predator’s estrogen-related physiology. In other words, plants that produce phytoestrogens may reduce the number of predators through what we might call birth control.

According to White, there is mounting empirical evidence that under certain circumstances, these dietary phytoestrogens may block estrogen receptors, thus inhibiting estrogen, which facilitates the repair of neural synapses and connections. Animal studies have shown, for example, that dietary intake of soy isoflavones, which mimic mammalian hormones, cause significant alterations in brain metabolism.

However, White, who worked with the National Institute on Aging when he began the study, told Psychiatric News, "The central focus of our study is not on dietary factors, but on trying to understand the determinants of late-life neurodegenerative problems and of the neural mechanisms that modulate the manifestations of those changes. We began the study hoping to find new risk factors, and this was certainly the most unexpected."

The president of the Hawaii Psychiatric Medical Association, Iqbal Ahmed, M.D., is familiar with White’s study and is intrigued by the findings. "This is a strong study, with thousands of subjects from a homogeneous population and solid measurement techniques. The findings are quite clear, but the question is, What do they mean?"

Ahmed, who specializes in geriatric psychiatry and is vice chair of the department of psychiatry at the University of Hawaii, continued by saying that epidemiological studies can point only to associations, not causality.

Predictably, the tofu connection has caused general alarm and concern throughout the health-conscious masses, although the media outside of Hawaii have shown little reaction. In fact, observed White, "Within days of my paper being published, several major city newspapers on the mainland—the Washington Post, the New York Times—had very similar full-page color spreads about wonderful recipes for soy foods and lots of information on how healthy they were."

However, White questions the much-touted benefits of soy—which has been thought to lower cholesterol and prevent breast cancer and heart disease. "The evidence supporting the benefits of soy are weak at best. Such evidence would never be enough to allow a prescription drug to be approved."

Ahmed recalls quite a different reaction from the media in Hawaii. "There was a great deal of publicity here in Hawaii, and tofu sales were said to drop by as much as 30 percent after White’s findings were released."

When White and his colleagues were interviewed, however, they did not advise people to change their diets, and they pointed to the need for further research.

Ahmed agreed wholeheartedly. "If other researchers make similar discoveries in another study, it would suggest there is definitely something behind these findings."