Book review: Nutrient Power: Heal Your Biochemistry and Heal Your Brain, by William J. Walsh.
Nutrient Power is an eccentric book about nutritional problems and their effects on the brain. It’s full of information that’s somewhat at odds with conventional wisdom.
It’s a short book, and I wasn’t tempted to read all of it. I usually don’t review books unless I’m willing to read the whole thing, yet this time I can’t resist the temptation.
I expect that it’s important reading if you’re building your own model of how nutrition affects cognition, you’re frustrated about how little you’ve found in peer-reviewed publications, and you’re interested enough to treat this as something closer to a career than a hobby. If, like me, you’re less ambitious than that, you should expect to find at least parts of the book frustrating. And if you just want easy-to-follow or rigorously proven advice, this is definitely not the book you want.
I’ll guess that a bit more than half of the unusual ideas are correct and valuable, and that less than 10% of the others are harmful. Don’t expect it to be easy to distinguish the good ideas from the bad.
Walsh’s approach to science writing
Think of the book as more of a teaser than as a rigorous description of techniques. Walsh clearly did not devote much effort into overcoming readers’ doubts. His message is somewhere in between “trust me” and “you’ll want to investigate this further”.
Walsh sounds like what a fair number of scientists sounded like a century ago. He expects his reputation as a scientist to be established by something other than publications. I’m not so clear on what that something is. Maybe his reputation among the people he works with? Or maybe among his patients?
Whatever he’s doing, it’s something very different from either the kind of writing that would get respect from the modern scientific establishment, or attract many consumers.
Is Walsh really one of those ideal scientists who we can trust to pursue the truth without biases? Maybe. I’m pretty sure he has avoided the standard medical biases, but I still suspect he’s subtly trying to impress some non-obvious audience.
He’s not making a serious enough attempt at earning enough money from book sales to justify the effort of writing the book. He doesn’t mention in the book that his institute is seeking charitable donations, so it’s hard to claim that’s his main motive. He seems somewhat motivated by a desire to market his training programs, but that’s awfully subdued marketing – the only reference I see to those programs is in the “about the author” paragraph.
A number of Walsh’s ideas involve the poorly explained terms undermethylation and overmethylation. These are influenced by nutrients such as folate and SAMe.
Walsh’s description almost leads me to believe that he’s talking about some general shortage or surplus of methyl groups. At other times, it sounds like he’s talking more specifically about DNA methylation, which has important epigenetic effects.
Walsh divides depression into several subtypes, and gives long lists of symptoms for two most common subtypes: undermethylated depression, and low folate depression.
I’ve had mild depression at times, so I tried guessing which subtype described me. I seem to have slightly less than half the symptoms of both undermethylated depression, and low folate depression. In both cases, a few of the symptoms seem to be pretty clearly the opposite of what I have. And Walsh’s other depression subtypes seem to describe me less well. So I’m confused. Are they bogus categories? I see a bunch of weak hints that Walsh would diagnose me as undermethylated. But I guess he’d want to discourage me from jumping to conclusions before I measure my whole-blood histamine and SAMe/SAH ratio.
My initial reaction was that under- and overmethylation were likely to be confused concepts. What research has been published on these ideas? His main citations for this topic are an unpublished(?) reference (Walsh WJ, Rehman F. (1997). Methylation syndromes in mental illness. Abstracts: Society for Neuroscience 27th Annual Meeting (pt 2)), and a book by an author whose credentials seem less clear than Walsh’s hard-to-evaluate credentials.
For some strange reason, I took Walsh seriously enough to check for other publications about methylation.
It turns out there are a bunch of peer-reviewed papers (e.g. DNA Methylation and Psychiatric Disorders) about DNA methylation, mostly published more recently than the book, which describe correlations between various disorders and global hypo- or hypermethylation of DNA. Those tend to confirm some of Walsh’s claims about how methylation levels influence specific disorders.
They even sort of confirm something resembling the hypothesis that a general shortage of methyl groups causes problems (from here):
This study suggests that maternal dietary and supplemental intake of methyl-group donors, especially in the periconception period, can influence infant’s buccal DNA methylation in genes related to metabolism, growth, appetite regulation, and maintenance of DNA methylation reactions.
Or, for a different, and more practical, take on methylation, see Chris Masterjohn (mainly in the cheat sheet that he sells there, and Why I Don’t Believe in “Undermethylators” and “Overmethylators”) (Masterjohn seems generally competent about nutrition, but I haven’t put much effort into deciding whether he’s right about methylation).
I changed my mind more than once while writing and researching this review, and I now believe that Walsh’s methylation ideas are about half right, although I’m still fairly confused about the details.
Walsh includes a one-page explanation of how his patented autism treatment also improves memory in 70% of Alzheimer’s patients, and that the improvement is maintained for several years. It seems kind of weird to treat almost as an afterthought something that would, if true, likely be one of the top 5 medical advances of the decade.
Instead of citing a peer reviewed study, Walsh cites his patent that covers the treatment.
Walsh’s Alzheimer’s treatment sounds somewhat like the approach that Bredesen later developed. Did Bredesen get some ideas from Walsh? Maybe. But Walsh shows few signs of having studied Alzheimer’s, and I expect that his understanding of the causes has more gaps than does Bredesen’s. I’m guessing that Bredesen’s approach is better than Walsh’s, in ways that render Walsh’s approach mostly obsolete.
Walsh’s approach also overlaps a bit with ideas about Alzheimer’s that Steve Fowkes developed around the same time (Fowkes is eccentric in ways that are similar to Walsh; I read Nutrient Power due to Fowkes’ recommendation).
Walsh says elsewhere that he’s cautious in order to minimize FDA objections:
Well, we have learned a way to get around it. What we do is be very careful to say, “We don’t treat depression. We don’t treat anxiety. We don’t treat violent behavior. We treat chemical imbalances.”
Yet he comes close enough to claiming to cure diseases that I expect he’d still be at risk for book burning or something if the book contained much that resembled hype. But I expect few consumers will react to the book by doing anything more than asking more questions of doctors, which likely limits the extent to which regulators will care.
Walsh suggests that some of the more famous school shootings were caused by people reacting badly to SSRIs. Walsh believes he can identify, via a simple blood test, a specific type of person who will react nastily to SSRIs. He’s a little vague about the specifics, but it includes a histamine test. Probably some experts can find enough evidence to evaluate such a claim, but it looks too hard for me to analyze.
Walsh has some complaints about mainstream medicine that sound like arguments against straw men. Here’s the most conspicuous example:
Despite brilliant advances in brain science, mainstream medicine continues to regard clinical depression as a single entity rather than a collection of different disorders. Dysthymia, postpartum depression, seasonal affective disorder, and other types of depression are believed to be variations within a central theme: low serotonin neurotransmission kinetics.
See SlateStarCodex for arguments that Walsh is perpetuating a myth (“Further, anyone who said that depression was caused solely by serotonin deficiency wouldn’t just be failing as a scientist, but also failing as a drug company shill.”)
How much risk of harm do Walsh’s ideas pose? Some of his suggestions look pretty low risk. For others, it’s somewhat hard to tell because he doesn’t indicate what doses he uses. I got some dosage info from his patent, and I’m pretty concerned about using “from about 150 to about 750 mg of vitamin B6” unless he’s doing frequent B6 testing, but the rest of his protocol looks fairly cautious.
The other place where I noticed advice that worries me is that he approves of treating autism with ABA. I checked Wikipedia, and saw that it portrayed ABA more favorably than I expected, albeit with decent references to explanations of why it’s controversial. My impression is that ABA is designed to help parents of autistic kids, without adequate regard for whether the kids benefit. But Walsh doesn’t claim much expertise in this kind of treatment, so I can’t fault him much here.
In sum, the book is full of possibly valuable ideas, which somebody ought to subject to rigorous research, but which most people should feel confused about.