Tofu Intake and Worse Memory in Elderly Indonesians: Was Professor Eef Hogervorst of UK’s Loughborough University in Scientific Pursuit of the Universal Truth?

Universal truth is about a description of reality that applies to all things of the same kind, or in terms of humans, to all similarly situated individuals, wherever and whenever they are. In the pursuit of scientific truth, scientists set up hypotheses based on prior facts and preconceived assumptions, and then perform experiments to gather factual evidence or data to support or reject the hypotheses. In order to get clear, meaningful interpretation of the data without unnecessary complication or ambiguity, they will control the distracting variables to focus on the variables of interest, and they will ensure that they are really measuring the things they set out to measure, and not something else, to establish validity. Then they often repeat the experiments to test the reliability of the results, and they will also repeat the experiments with different groups of test objects, if they are biologically variable, in different places and at different times, to establish universality of the results. Scientific discussion is based on facts, and certainly not on afterthought rationalization of overlooked assumptions. It is indeed a disciplined journey on the way to universal truth. 

 

If we want to test whether tofu intake hurts memory in the elderly, we would first ensure that we have a normal, generic tofu, without extraneous complicating toxins like aflatoxin from molds or deliberately-added formaldehyde, monitor tofu intake by the elderly, with a large and representative sample population in a country, and then measure individual memory by an objective, valid, and reproducible test. Once the tests are repeated in similarly situated individuals across different countries, and we find that invariably tofu intake hurts memory in the elderly, then we may be on our way to some universal truth. 

 

Unfortunately, when I was reading the abstract of the recent paper on "High Tofu Intake Is Associated with Worse Memory in Elderly Indonesian Men and Women," authored by E. Hogervorst, T. Sadjimim, A. Yesufu, P. Kreager, T.B. Rahardjo, that will appear in 2008's first issue of the Dementia and Geriatric Cognitive Disorder Journal, pages 50-57, I did not feel the researchers were disciplined in the scientific pursuit of universal truth.

 

First, the title of the research paper, "High Tofu Intake Is Associated with Worse Memory in Elderly Indonesian Men and Women," is, to say the least, scientifically confusing if not misleading, and not valid by its implication. Tofu is a universal generic term understood, popular, and relevant in many countries, that is why the research has captured so much attention worldwide, whereas the researchers were actually referring only to the specific tofu consumed in Indonesia and that, because of formaldehyde, a potent carcinogen and protein cross-linking agent, which could have been widely used in preserving tofu (and making tofu firmer) in that country, could be very different from the generic, formaldehyde-free tofu consumed elsewhere. When you try to establish a generic link between tofu (or tempe, or any other soy food) and memory, that tofu must not be a very different kind of tofu with extraneous toxin added, which is not normally desired or consumed by choice, but must be a normal, generic tofu without an undue complicating factor like formaldehyde. To regain validity, either the experiment is re-designed to factor out formaldehyde (like conducting the research in another country without the formaldehyde issue), or the title can be reworded to read, "High Indonesian Tofu Intake Is Associated with Worse Memory in Elderly Indonesian Men and Women," or "High Tofu Intake Is Associated with Worse Memory in Elderly Indonesian Men and Women, with a Caveat on Formaldehyde." The replacement of a specific tofu by generic tofu in the title was inexcusable concept swapping. Choosing the invalid title indicated either the authors had no idea that they had an invalidating complicating factor called formaldehyde on their hands until after the research had been done, or they decided to go with sensational journalism, which no serious, truth-seeking, research scientist should ever get involved and discredit themselves, their publishing journal, and their institutions. 

 

Secondly, even the abstract alone is confusing. Because the full paper is not yet available until later, we do not have the benefit of the full text but the abstract. English is probably the most clear and precise language for use in science or law, and yet the abstract was confusing enough to prompt science writers at the very respectable Nutraingredients.com (http://www.nutraingredients.com/news/ng.asp?n=86342&c=gPTKXx15OjUj%2BpBkIE9URw%3D%3D) to quote either precisely aligned with, or exactly contrary to, what the researchers meant, by saying, "Interestingly, increased consumption of fruit was also positively linked to dementia," when the researchers had written that "Fruit consumption also had an independent positive association." In scientific language dealing with statistical interpretation, "positive association," to a fellow scientist, would mean a positive statistical relationship, with the variables moving in the same direction, and not an association with positive memory. So what did the researchers really mean? If they meant to say that fruits hurt memory in the elderly, then they had better explain why, because nobody would expect fruits, in generic terms, to hurt memory. And that was what made the science writers at Nutraingredients.com deem the finding "interesting." But if they meant to say that fruits help memory, it would not be sensationally newsworthy.

 

Thirdly, the researchers were treating tempe and tofu as if they were good guys and villains miles apart, when in fact the two foods share attributes that are more similar than different. For readers not familiar with the products, tofu comes from soaking, cooking, and grinding up whole soybeans to make soymilk and then coagulating the high-protein cheese-like curd, whereas tempe is a mold-fermented mass of cooked, whole soybeans. Tofu, because of the relatively high moisture content, is much more easily spoiled microbiologically than tempe in Indonesia's hot climate; that is why it was reported that formaldehyde is widely used as a tofu preservative in Indonesia. (Here in North America, tofu is generally pasteurized and kept fresh under refrigeration). When the researchers talked about the good attributes of tempe like folate, which they said possibly exerted a positive effect on memory, firm tofu is also a good source of folate, at about 80% the amount found in tempe (http://www.tempeh.info/health/tempeh-nutritional-facts.php, and http://www.womenfitness.net/tofu_health.htm).And while the researchers were speculating that soy phytoestrogens might be causing memory loss in the elderly, they failed to mention that phytoestrogens were in fact more biologically available in tempe because the fermentation changed the phytoestrogens to the more readily absorbed aglycone form[1] (phytoestrogen without the carbohydrate moiety). And in terms of the content of the key soy phytoestrogen, genistein, the content in tofu is about 75% that in tempe (http://www.soyfoods.com/nutrition/isoflavoneconcentration.html). While the authors were speculating that phytoestrogens might cause memory loss because they could promote oxidative stress through generating cell-damaging free radicals, in fact phytoestrogens were known to act as scavengers of free radicals[2], and to offer protection against hydrogen peroxide (highly potent oxidizing agent) greater than that offered by the known antioxidant vitamins ascorbic acid and α-tocopherol[3]. No wonder the Loughborough University press release reported, "Intriguingly, the researchers also found that consuming tempe, a fermented soy product made from whole soy bean, is associated with better memory." (http://www.lboro.ac.uk/service/publicity/news-releases/2008/104_tofu.html). Instead of emphasizing the disabling complication raised by formaldehyde, Prof. Eef Hogervorst, the lead researcher from Loughborough University, chose to defuse the issue by saying, "Tempe consumption very likely offsets tofu's negative associations with memory," because of tempe's higher folate content through fermentation. According to this rationale, the high folate benefits of tempe would offset the high phytoestrogen "harms" of tempe. Similarly, the moderately high folate benefits of tofu should also offset the moderately high phytoestrogen "harms" of tofu. Prof. Eef Hogervorst's rationale is therefore "intriguing". Prof. Eef Hogervorst then went on to cover all possibilities by saying, "It may be that the interaction between high levels of both folate and phytoestrogens protects against cognitive impairment." We could only hope that she had done her homework before making self-contradictory rationalizations. In addition, either because she was not aware of the formaldehyde issue when she was making the statement, or she just got into the habit of concept-swapping between generic and specific tofus, she mentioned in conclusion that "eating tofu in moderation posed no problem" (http://www.zeenews.com/articles.asp?aid=453374&sid=env&ssid=28). Knowing that formaldehyde is such a potent carcinogen, I am not so sure I could agree with her that eating the specific formaldehyde-preserved tofu in moderation posed no problem. And the authors even stayed away from mentioning the word formaldehyde in the all-important abstract that most people read, but wrote instead, "It is unclear whether these negative associations could be attributed to potential toxins or to its phytoestrogen levels." They even failed to be specific whether the potential toxins were inherent in or extraneous to the tofu. The language they used was so scientifically imprecise as to defy unambiguous interpretation. Apparently, the researchers were trying to salvage a poorly designed experiment, and to come up with every imaginable tactic to try to hide or deflect the toxic and disabling formaldehyde issue that, being an afterthought, reflected so badly on the actual counter-scientific behavioral conduct of the researchers themselves. They could not establish experimental validity on two supposedly generic products in a single-country setting, how could they go on to establish universal truth and regain our trust? 

 

Finally, the research findings were based on association and not cause-and-effect. Even if formaldehyde is incriminated, it still needs cause-and-effect, hypothesis-based, and valid and reliable experimental evidence. If I were Prof. Eef Hogervorst, as a responsible researcher, that is what I will do in order to dispel validity doubts. And if I cannot feed formaldehyde to humans, at least I can perform the experiment using an animal model to gather valid initial evidence. There are certainly valid, objective tests on the memory of lab mice, for example. The same applies to studies involving isoflavones and folate. 

 

If I were Prof. Eef Hogervorst, the lead author, I would retract the research paper from the journal and issue an official apology to all parties concerned, including Dementias and Geriatric Cognitive Disorders that published the research, Alzheimer's Research Trust that funded the research, and the public thus misinformed. And if I were the Indonesian government, I would be obliged to take immediate remedial steps to protect Indonesian citizens from the harm of extraneous toxic agents, first from formaldehyde in tofu and other foods, and then from toxins in all foods. That would be a logical first step towards establishing universal food safety in Indonesia. The next universal step, of course, is for all countries to step up concerted efforts toward truly protecting public food safety. 

---