Tofu Study Update

Ian Goddard (
Thu, 09 Dec 1999 02:33:39 -0500

The following contains much information about the "tofu study" reported:

I contacted Lon White, lead researcher in the tofu study to ask some questions. I asked him about the statistical significance of the findings in the study. He informed me that there were four major endpoints in the study:

  1. cognitive impairment of men (number of men=3734)
  2. cognitive impairment of sample of wives (n=502)
  3. brain atrophy by weight of men who died (n=290)
  4. brain atrophy by volume measured via MRI (n=574)

The probability (p) value was less than .05 (significant) for a correlation between each endpoint and tofu intake (there was no significant correlation for other foods). In some cases the p value is less than .001 (if the p value is more than .05, it's not significant, if it's .05 or less, it's significant). Here's what this means:

 IF  p = .05  there's 95% chance the correlation is true
 IF  p = .01  there's 99% chance the correlation is true
 IF  p = .001 there's 99.9% chance the correlation is true

So the correlation between the four endpoints listed above and tofu consumption had an over 95% probably of being true, with a probability of truth as high as 99.9%. Of course this cannot be read as proof that tofu causes those problems, but at the least it signals the need for further investigation. What is perhaps most significant is that Dr White informed me that there is a dose-dependent relationship between tofu consumption and the endpoints, ie, the more tofu was eaten, the more impairment and/or atrophy is found. The odds that such findings would be chance are extremely low!

Here are two papers Dr. White, et al., have had published:

White, L., Petrovitch, H., Ross, G.W., & Masaki K.H. (1996) Association of mid-life consumption of tofu with late life cognitive impairment and dementia: The Honolulu-Asia Aging Study. The Neurobiology of Aging, 17 (suppl 4), S121.

White, L., Petrovich, H., Ross, G. W., Masaki, K. H., Abbot RD, et al. (1996) Prevalence of dementia in older Japanese-American men in Hawaii. JAMA, 276, 955-960.


Here's a brief review of those two papers found in a letter to the FDA from Daniel Sheehan, PhD, Director of the Estrogen Base Program Division of Genetic and Reproductive Toxicology, and Daniel Doerge, PhD, Division of Biochemical Toxicology:

On 02/18/99, Dr. Sheehan & Dr. Doerge comment on White's work

Finally, initial data from a robust (7,000 men) long-term (30+ years) prospective epidemiological study in Hawaii showed that Alzheimer's disease prevalence in Hawaiian men was similar to European-ancestry Americans and to Japanese (White, et al, 1996a). In contrast, vascular dementia prevalence is similar in Hawaii and Japan and both are higher than in Europeanancestry Americans. This suggests that common ancestry or environmental factors in Japan and Hawaii are responsible for the higher prevalence of vascular dementia in these locations. Subsequently, this same group showed a significant dose-dependent risk (up to 2.4 fold) for development of vascular dementia and brain atrophy from consumption of tofu, a soy product rich in isoflavones (White, et al, 1996b). This finding is consistent with the environmental causation suggested from the earlier analysis, and provides evidence that soy (tofu) phytoestrogens causes vascular dementia. Given that estrogens are important for maintenance of brain function in women; that the male brain contains aromatase, the enzyme that converts testosterone to estradiol; and that isoflavones inhibit this enzymatic activity (Irvine, 1998), there is a mechanistic basis for the human findings.

Full Letter:

The following is the abstract to the recent study, which should be published by April. Below that is a memorandum sent out by Doctor White after the newspaper publications:


Association of high midlife tofu consumption with accelerated brain aging . Lon White, MD, MPH (From the Pacific Health Research Institute, Honolulu, HI.)

This investigation utilized the resources of the Honolulu Heart Program, a longitudinal study of Japanese-American men established in 1965 for research on heart disease and stroke. Questions regarding frequency of consumption of tofu and 26 other foods were asked at interviews in 1965-67 and again in 1971-74. Cognitive testing was done (n=3734) and cases of dementia identified (n=225) at the 1991-93 examination, when participants were aged 71-93 years. Atrophy was assessed by neuroimaging (n=574) or autopsy (n=290). Cognitive test data were also analyzed for wives of a sample of study participants (n=502) who had been living with the participants when their dietary interviews were done. Poor cognitive test performance in late life was associated with higher midlife tofu consumption. An independent association of similar magnitude and direction was apparent among wives of cohort members, using the husband's answers as proxy for the wife's consumption. Midlife tofu consumption was independently associated with low brain weight and with ventricular enlargement. Independent associations of more frequent midlife tofu consumption with clinically diagnosed Alzheimer's disease and with poor cognitive functioning among non-demented subjects were demonstrated. Associations generally followed a dose-response pattern, were statistically significant after controlling for all relevant and potentially confounding factors, and remained apparent after stratifying for age or obesity. These data suggest that regular consumption of tofu over many years in middle life may have an adverse influence on brain aging manifest as accelerated atrophy, cognitive decline, and a lowering of the threshold for the clinical manifestations of Alzheimer's disease. We speculate that these may reflect chronic sub-optimal neuronal plasticity caused by isoflavone inhibition of tyrosine kinase activity and/or by interference with estrogen-related mechanisms.


FROM: Lon White, M.D., M.P.H., Senior Neuroepidemiologist, HAAS and PHRI DATE: November 30, 1999
RE: Responding to questions about tofu

The articles that recently appeared in the Star Bulletin and the Advertiser have caused a great deal of distress locally, and have elicited a number or requests for information. While we cannot give advice and no definitive statements are possible at this time, the following can be said:

  1. We observed a consistent pattern of association between answers given by men interviewed in 1965 and in 1971 regarding the number of servings of tofu consumed per week - and their scores on a test of cognitive function when examined 1991-93. Higher frequencies of tofu consumption were associated with poorer test scores. This observation remained strong after controlling for all relevant factors (such as age, education, occupation, etc.) and no other foods or drink showed such an association.
  2. Separately, we observed an association of high midlife tofu intake with low brain weight determined at autopsy, again after controlling for all other relevant factors.
  3. Separately, we observed an association of high midlife tofu intake with enlarged ventricles identified by MRI.
  4. Although we did not find an association of high midlife tofu intake with the brain lesions that characterize Alzheimer's disease in our autopsy study, we did find high midlife tofu intake to be a risk factor for the clinical diagnosis of Alzheimer's disease.

Taken together, we interpret these as suggesting that consumption of tofu over many years during middle life is linked to a mild to moderate acceleration in brain aging, perhaps similar to that which might occur in a woman who did not receive hormone replacement therapy after menopause. We cannot be sure that the causal factor was really the tofu, but no other alternative is apparent at this time.

We believe that the most reasonable mechanism would be an interference with the normal mechanisms in the brain that maintain the connections between neurons during aging (i.e., brain plasticity mechanisms), caused by the isoflavone phytoestrogens that are known to be present in pharmacologically significant concentrations in most soy foods.

Although the findings are very consistent, it is never proper to draw definitive conclusions from a single study. It would be premature to advise anyone that they should change their diets based on a single research study. In addition, there is evidence that consumption of soy foods may have beneficial effects related to improving blood lipid levels, and reducing risks for breast cancer.

Unfortunately it is likely that independent confirmation or refutation of our findings will require 1-3 years. In the meantime no definitive statements can be made. Individuals will have to weigh the evidence and make their own decisions concerning their consumption of soy foods and their isoflavone derivatives.

LR White