More on Diet and Dementia

Thanks to Oregon, another diet and dementia study hit the press (or internet-prior-to) last week.  A modest number of seniors were tested for levels of all sorts of vitamins and fatty acids, given cognitive testing, and then some had MRIs as well.  The punch line is that those seniors with the highest levels of B vitamins (B1, B2, B6, B12, and folate) and vitamins C, D, and E had the best scores on cognitive function and some mighty fine lookin' brains.  Those with high omega3 marine fatty acids also did well.  The cohort with high trans fats did poorly.  Exactly what we might expect!  Imagine that.  What we eat might impact the brain after all.

The seniors tested were all white, educated Oregonians who joined an aging cohort study back in 1989.  293 healthy older folks agreed to be poked, prodded, questioned, imaged, and measured until their deaths.  At the time of this study, circa 2006, 104 subjects were left, with an average age of 82, 62% female.  The population was originally chosen to be healthy, and after 16 years those that remained were pretty healthy, with a relatively low amount of comorbid illness such as high blood pressure.  Only 10% were ApoE4 carriers. Only 7% had B12 deficiency (<200 pg/ml) and 25% were 25 (OH) vitamin D deficient (<20 ng/ml) - meaning that as a whole, the population was pretty well stocked up on vitamins.  Eating well is exactly what I would expect from a bunch of older, educated Oregonians.

The weaknesses of the study are the small number and the lack of ethnic or socioeconomic diversity and the observational nature.  The strength is that multiple nutrients were measured via blood tests rather than relying on the more classic (and classically unreliable) food frequency questionnaires (FFQs).  The multiple nutrients were then statistically arranged into certain nutrient patterns - on one hand, this makes it easier to apply the results to real life diets, as people eat food, not isolated nutrients.  On the other hand, this type of statistical manipulation can lend itself to data mining.

With imaging, various memory and cognitive tests, 30 plasma biomarkers and 8 biomarker patterns (for example, the B, C, D, E vitamins, marine omega3s, omega6 + retinol, lutein + HDL, saturated fat, trans fat (mostly linolelaidic acid (18:2omega6t)), carotenoids, etc.) there were quite a lot of associations to be had.  The most interesting are the previously mentioned generally better global findings in the BCDE group and the omega3 group, and the generally worse findings in the trans fat group.  In addition, folks in the lutein-HDL group had better memory, and those in the omega6 + retinol group had poorer memory and language scores.  In adjusting for comorbid factors, age, education, gender, APOE4 status, and other health factors didn't seem to impact these findings that much.  The only major link between health problems and nutrients was a link between hypertension and trans fats and low B vitamins, which makes physiologic sense (remember, lousy B vitamin status leads to high homocysteine, which is likely a cause of hypertension).

Interestingly, if the study subject had depression, the link between omega3 status and white matter density in the brain was lost.  In subjects without depression, there was a significant link between omega3 levels and a healthier white matter density compared to whole brain volume.

So when the researchers broke it all down, they found that age, gender, APOE4 status, education years, hypertension and depression accounted for 46% of the variation in cognitive scores, and the nutrient status was responsible for an additional 17%.  They found nutrient status to account for 37% of the total brain volume and 9% of the white matter density compared to whole brain volume variation.

What, then, would constitute the supposed diet of the healthiest brains?  Dark leafy green and cruciferous vegetables, fruits, and fish were on the good side, bakery foods such as cookies, pies, doughnuts, fried foods, margarine, red meat, and offal were on the "trans fat" side.  (Red meat and offal seemed to be added to this group due to the previous dietary pattern study done in Manhattan implicating offal consumption in worsening cognitive status.  Chris Masterjohn had something to say about that study here.  "When holism goes horribly wrong -- the perils of dietary pattern analysis" will give you a clue as to his thoughts!)  There are, indeed, trans fats in red meat and offal, but they tend to be CLA: the 18:2 isomers implicated in this newest study are the baked goods/margarine variety of trans fats.  As you know, red meat and offal are fantastic sources of B vitamins.

The mechanisms by which the specific nutrients may be helpful or harmful are ones we have discussed many times on this blog.  The B vitamins are necessary for proper functioning of the folate cycle, which has all sorts of downstream effects with relation to neurotransmitters, oxidative stress, inflammation, etc.  I've also discussed vitamin D a number of times, and while I haven't covered much on vitamin E or C, I suppose I should get to that.  Omega3s have also been reviewed at length.

Trans fats, on the other hand, may well replace omega3s in the cell membranes, and are associated with systemic inflammation, cardiovascular disease, and endothelial dysfunction.  All those processes could have an impact on memory, brain structure, and cognition.

To break it down, eat "real food," not cookies and doughnuts.  How hard is that?