Can avoiding certain foods reduce your dementia risk? One nutritional psychiatrist seems to think so, but the evidence is much messier than it looks at first glance.
Are you afraid of developing dementia in your golden years? If so, a few simple changes to your diet may be in order, according to Dr. Uma Naidoo, a nutritional psychiatrist at Harvard University and the author of “This Is Your Brain on Food." In a piece for CNBC over the weekend, Naidoo highlighted five troublesome foods: added sugars, fried foods, high-glycemic-load carbohydrates, alcohol and nitrates in cured meats. There is some evidence, she argued, that limiting consumption of these items may “fight inflammation and promote brain health, sharp thinking and good decision-making.”
There's no denying the connection between diet and brain health. It only makes sense given how important nutritious food is to our overall wellbeing. But here we're interested in a more specific question: are certain foods harmful enough to set us up for dementia? As this CNBC article unintentionally illustrated, the answer is still “we don't know.” Anyone who says that eating fewer chicken wings (or some other food) will cut your dementia risk is overstepping the evidence.
A plausible argument?
The primary problem is that Naidoo recommended limiting intake of foods and drinks we already know are harmful in excess, giving the argument a certain plausibility, without providing evidence that doing so will significantly benefit our brains. So while correct in a narrow, technical sense, it's unhelpful as medical advice. Imagine a patient diagnosed with depression, a potential risk factor for dementia and a complicated disease in its own right, getting such input from a psychiatrist.
“Doctor, it seems like my life has no purpose; my hobbies bring me no joy and I feel isolated from the world around me. What do I do?”
“Well, how much white bread do you eat in a week, Mr. Smith? Gut inflammation fueled by high-glycemic-load carbohydrates may contribute to your crippling disinterest in life, according to a new study in the Journal of Silly Epidemiology.”
That's a bit snarky on my part, but it's justified based on the CNBC article itself. “Existing studies point to the idea that we may be able to reduce the possibility of dementia [my emphasis] by avoiding foods that can compromise our gut bacteria and weaken our memory and focus,” Naidoo wrote. Any statement that needs to be so carefully qualified isn't based on solid data. Indeed, the review she linked to concluded that
"oral administration of drugs or therapeutic agents possessing immunoregulatory activities is simple, easy, and a hopeful way of regulating inflammatory diseases, though the detailed mechanisms of how intestinal immunity affects systemic immunity remain to be clarified."
The review looked mostly at animal studies, by the way, and it nowhere described a link between diet and dementia, depression or any other related disorder, except in the footnotes. These are the makings of an interesting hypothesis, not the kind of strong evidence that we should present to the public. The same sorts of limitations affect most the research in the CNBC story. For example,
"In 2018, in the British Medical Journal, [researchers] reported that people who had abstained from alcohol completely or who consumed more than 14 drinks per week had a higher risk of dementia compared to those who drank alcohol in moderation."
The researchers relied on the participants to self-report alcohol consumption five times over the course of roughly two decades. This sort of data is notoriously unreliable, because people struggle to remember how much they drink, or they just lie about it. As my colleague Dr. Chuck Dinerstein wrote in 2019, “... when this data forms the primary data of analysis, it weakens if not eliminates the certainty of the findings.” Moreover, out of more than 9,000 participants in the BMJ study, only 397 developed dementia, meaning the vast majority of people, regardless of how much alcohol they consumed, did not.
These associations are all over the place, in other words; you can pick the one that suits you best. For instance, low-carb diet advocates argue that high-fat, high-protein foods like bacon, salami and sausage (all on Naidoo's list) actually help reduce the risk of cognitive impairment later in life. Vegans make the opposite case, attributing the same benefit to avoiding animal products. Champions of the Mediterranean Diet fall somewhere between those two extremes, positing that fruits, vegetables, olive oil, legumes, whole grains and fish carry all the brain-boosting benefits.
My point isn't to deny that diet can impact dementia risk. It probably does. But science communicators and academics need to clearly outline the limitations of the existing evidence and refrain from proffering advice that goes beyond the data. Naidoo briefly mentioned the weaknesses in the research up front but failed to explain their significance to her readers. She's not alone; scientists and reporters, usually in a rush to attract attention, regularly oversimplify the evidence for popular audiences. This is troublesome no matter the topic, but it's especially unhelpful when the issue is something as life changing as dementia.