April 2000, Lon White and others reported a dose-dependent positive
correlation between tofu consumption and brain atrophy in a large
sample of men over several decades.1
While correlation does not prove causation, study size and duration
along with the robust dose-dependent relationship caused me, even
as a vegetarian, to avoid tofu and other soy products.
hypotheses should be tested against the availability of possible
causal mechanisms. In addition to possible causal mechanisms previously
cited by this author,2 recent findings
significantly increase the case for a causal mechanism of soy-induced
neurotrophic factor (BDNF) facilitates the survival and genesis
of brain cells.3-4 The neuroprotective
effects of caloric restriction are attributed in part to increased
BDNF.5 On the other hand, reduced BDNF
is known to cause brain-cell atrophy and is associated with Alzheimer's
disease.6-7 Now, a study in "Neuroscience
Letters" reports that soy significantly reduced BDNF in the
hippocampus and cerebral cortex of male rats.8
Since reduced BDNF can cause neural atrophy, these findings appear
to provide compelling evidence for a causal mechanism that might
explain the positive correlation between tofu (soy) consumption
and brain atrophy demonstrated by White et al.1
For Boys, Good For Girls?
soy appears to reduce BDNF in male rats, it has also been shown
to increase BDNF in female rats.9 In
fact, soy appears to affect neurological parameters in a sex-defined
fashion wherein females benefit and males suffer.10-13
There is little doubt among researchers that this is because soy
is high in phytoestrogens, which are plant-derived substances that
act like the female hormone estrogen.
that sex-defined difference fails to explain the findings regarding
the wives of male subjects in White et al., who reported: "A
similar association of midlife tofu intake with poor late life cognitive
test scores was also observed among wives of cohort members, using
the husband's answers to food frequency questions as proxy for the
wife's consumption."1 White et
al. proposed that long-term consumption of weaker soy estrogens
may displace the body's own stronger estrogen along with its benefits.
Against Soy-Dementia Hypothesis?
possible signal contrary to a soy-dementia link is the low prevalence
of dementia14 and high consumption of
soy in Okinawa, Japan.15 However, that
negative correlation, like any correlation, does not prove causation.
For example, perhaps soy does cause dementia but other factors in
Okinawa offset the effect.
White et al. explored correlations of a range of foods to neurological
parameters, whereas this Okinawa analysis is a sweeping generalization
of only tofu to all of Okinawa. In other words, it stands to reason
that the study by White et al. finding a positive tofu-dementia
correlation has the greater likelihood of providing the more accurate
picture. Nevertheless, in my view this Okinawa data warrants further
examination as a possible route to falsifying the soy-dementia hypotheses.
closing, the findings of soy-induced BDNF reduction in male rat
brain regions that are central to the onset of dementia, in addition
to previous findings,2 appear to provide
compelling evidence of a possible causal mechanism that might explain
the soy-dementia correlation reported by White et al.1
Obviously further research is necessary before a clear picture emerges
regarding the effects of long-term soy consumption on the brain.
But in the meantime, my inclination is to play it safe and avoid
et al.: "In this population, higher midlife tofu consumption
was independently associated with indicators of cognitive impairment
and brain atrophy in late life."
(scroll to): "Is There Reason to Believe Tofu May Cause Brain
(Sep 2002): "Neurotrophic factors have long been known
to promote neuronal survival and differentiation."
Neurochem (Sep 2002): "These findings suggest that
BDNF plays an important role in the regulation of the basal level
of neurogenesis in dentate gyrus of adult mice […]."
(Jun 2003): "Recent studies have shown that DR [dietary
restriction] stimulates the production of brain-derived neurotrophic
factor (BDNF) in brain cells, which may mediate neuroprotective
and neurogenic actions of DR."
(Jul 1997): "stress can decrease the expression of brain-derived
neurotrophic factor and lead to atrophy of these same populations
of stress-vulnerable hippocampal neurons."
Res Mol Brain Res
(Oct 3, 1997): "a reduction in BDNF mRNA expression has
been observed in human post-mortem Alzheimer's disease hippocampi.
[…] These results support and extend previous findings that
BDNF mRNA is reduced in the human Alzheimer's disease hippocampus
and temporal cortex, and suggest that loss of BDNF may contribute
to the progressive atrophy of neurons in Alzheimer's disease."
(Feb 27, 2003): "significant reductions were found in
brain-derived neurotrophic factor (BDNF) mRNA expression in the
CA3 and CA4 region of the hippocampus and in the cerebral cortex
in the [male] rats fed the diet containing phytoestrogens, compared
with those on the soya-free diet."
(Feb 1999): "soy phytoestrogens significantly increased
the mRNA levels of BDNF […in] female rats."
(Jan-Feb 2002): "when learning and memory parameters
were examined in a radial arm maze testing visual-spatial memory
(VSM), the diet treatments significantly changed the typical sexually
dimorphic pattern of VSM. Specifically, adult Phyto-rich fed females
outperformed Phyto-free fed females, while in males on the same
diets, the opposite pattern of maze performance was observed."
(2001 2(1):20): "Female rats receiving lifelong exposure
to a high-phytoestrogen containing diet (Phyto-600) acquired the
maze faster than females fed a phytoestrogen-free diet (Phyto-free);
in males the opposite diet effect was identified. […] These
findings suggest that dietary soy derived phytoestrogens can influence
learning and memory and alter the expression of proteins involved
in neural protection and inflammation in rats."
BMC Neurosis (2001 2(1):21): "When a diet change
was initiated in adulthood,
control phytoestrogen-rich fed females outperformed control females
switched to a phytoestrogen-free diet. Whereas, in control males
the opposite diet effect was identified."
Neurosci Lett (May 15, 2003): "This study is the
first to show that
lifelong consumption of dietary phytoestrogens alters the HPA
response in male rats."
Rates in Okinawa vs Japan & US.
Phytochemical Intake in Okinawa.
Some alternative views on soy: