Is Soy Good or Bad for the Brain?

The association between legume consumption and improved cognitive performance has been used to try to explain why dementia prevalence is lower in East Asia, where they eat 10 to 40 times more soybean products, compared to the West. But what about headlines like “Tofu may raise risk of dementia” in response to a study out of Indonesia that found that high tofu intake was associated with worse memory in the elderly? In the same analyses, however, they found tempeh—another soy food—was associated with a better memory, so it couldn’t have been the soy itself. It turns out that it was widespread practice to add formaldehyde to tofu in the country (but not tempeh) to preserve its freshness. Indeed, years later after the formaldehyde practice appears to have been stopped the negative association with tofu in the same population seems to have disappeared. Mystery solved, right? Well, the epidemiological evidence connecting soy and cognition is a mess of contradictions. 
 
Overall, three studies link some form of soy to worse cognitive outcomes, three to better outcomes, and four found mixed results or no significant association either way.  
 
In addition to the original Indonesian study, a study of Japanese-American men living in Hawaii found that mid-life tofu consumption was associated with four years of accelerated brain aging later in life, and a cross-section study of hundreds in China found higher odds of cognitive impairment among tofu consumers. But a much larger study of thousands of older people in China found the opposite—a significantly lower risk of cognitive impairment among those who ate tofu on a regular basis. In Taiwan, daily tofu consumption was associated with 55% lower odds of cognitive impairment, similar to what was found in Japanese women (though no significant effect was found in Japanese men). Importantly, the protective association with soy foods on cognition controlled for dairy intake, so the lower risk of cognitive impairment wasn’t just because soy milk drinkers were avoiding milk, as 3 out of 4 large prospective studies associate milk and dairy consumption with cognitive decline. 
 
OK, so the observational data is mixed. What happens when soybeans, soy foods, and soy compounds are put to the test in interventional trials? Some are more persuasive than others. Men fed five cups of soybeans over a week experienced an improvement in spatial cognitive abilities, but had no control group, so no conclusions can be made on what role the beans played, if any. Other studies have been much more rigorous, disguising soybeans in chili for example, to randomize people to higher or lower soy diets, and can demonstrate significant improvements in short- and long-term memory within 10 weeks, suggesting significant cognitive improvements can arise from a relatively brief dietary intervention. There have been sixteen such randomized controlled trials, involving more than a thousand participants, and overall, soy or soy compound interventions improve overall cognitive function and memory.