Dietary Ketosis Aids Memory in Mild Cognitive Impairment

[Tincup]

http://psychcentral.com/news/2014/12/08 ... 78345.html

A new study tested a ketogenic diet in people with mild cognitive impairment, sometimes a precursor to dementia. For six weeks, half of the 23 participants ate a ketogenic diet, with five to 10 percent of calories coming from carbohydrates, while the others ate a high-carbohydrate diet, 50 percent of calories from carbohydrates.

Those on the ketogenic diet showed significant improvements in verbal memory compared to the other group. The higher their ketone levels, measured in urine, the better their verbal memory.

[Gilgamesh]

George, thanks. Good news for our non-ε4 carrier friends and loved ones, but probably not relevant to us, alas: Krikorian didn't stratify by APOE status, and we know ketosis functions very diff'ly depending on the presence of an ε4 allele.

GB

[Julie G]

G, excellent point. I assume you contacted Krikorian and verified that? If not, I'd be happy to.

[Julie G]

Hey G, any word from Krikorian? I'm still intrigued with this one..

[Gilgamesh]

Circ- I accidentally deleted your post! Sorry! Serious senior moment. Well, no, I'm just rushed. I should get back to woodshed mode. I will after I report on ketosis and (if I hear back in time) betaine for people with bad BHMT SNPs.

Here's my post, responding to your question:

Hey circ, actually, ketosis may be less important for us -- or rather, less likely to have benefits.

See the first study here:

http://www.ncbi.nlm.nih.gov/pubmed/?term=ketosis+apoe

But, as Julie (and I think others) have pointed out, it might be that ketosis could help if other aspects of diet and lifestyle are good. Might, or might not. Just a theory, so far. I'm corresponding with a couple researchers and hope to have more solid info soon.

GB
ApoE 3/4, COM

[RichardS]

George, thanks. Good news for our non-ε4 carrier friends and loved ones, but probably not relevant to us, alas: Krikorian didn't stratify by APOE status, and we know ketosis functions very diff'ly depending on the presence of an ε4 allele.

GB

GB,

We know that a proprietary synthetic ketone intervention for mild to moderate AD did not seem to help E4's if we are to believe the Accera trial from Henderson. However, recall that the Kirkorian study was a dietary intervention for MCI - just low carb without any artificial means of adding ketones. I think the takeaway here is that if you are hyperinsulinemic (even if not diabetic), a 6-week low carb diet can be helpful for memory. The effect size is quite respectable. With only 12 in the low carb group, a breakdown by E4 status would probably not give us much direction statistically. Maybe I'm missing some important studies, but I don't think we know enough about diet-induced ketosis with E4's to say we won't benefit.

On a separate issue, I noticed from the Kirkorian paper that after only 6-weeks of low carbing, they lost an average of 4kg and supposedly were down to 1036 kcal/day which is in serious calorie deficit territory if we are to believe the diet diaries. Not a long-term sustainable intake. The researchers seem to think the positive effects were more about improving insulin than any ketone involvement. I would add calorie restriction to that. We know some research supports the use of intranasal insulin to improve cognitive impairment.

Richard

[Gilgamesh]

Hey Richard (and by the way, welcome!),

I'm definitely not ready to reject ketosis -- via low carb or MCTs or any means -- for ε4s. But after spending the last two years (my God, it's really been that long since I got my 23andMe results...) surfing PubMed, seeing so many tantalizing clues with very little outcome-based results backing them up, I tend much more towards skepticism than I used to.

GB

[Julie G]

I think the takeaway here is that if you are hyperinsulinemic (even if not diabetic), a 6-week low carb diet can be helpful for memory.

Amen, Brother Richard. There's more evidence supporting this- not new.

I'm definitely not ready to reject ketosis -- via low carb or MCTs or any means -- for ε4s. But after spending the last two years (my God, it's really been that long since I got my 23andMe results...) surfing PubMed, seeing so many tantalizing clues with very little outcome-based results backing them up, I tend much more towards skepticism than I used to.

I've often wondered if you might be applying different standards for outcome based results. For instance, have I missed the Cochrane Report proving that CR reverses Alzheimer's in E4 carriers?

I'm eager to hear Dr. Krikorian's response. I've been following up on him and finding some really interesting information- reminiscent of Dr. Gundry's approach: http://diabeticmediterraneandiet.com/ta ... krikorian/

[Stavia]

I'm definitely not ready to reject ketosis -- via low carb or MCTs or any means -- for ε4s. But after spending the last two years (my God, it's really been that long since I got my 23andMe results...) surfing PubMed, seeing so many tantalizing clues with very little outcome-based results backing them up, I tend much more towards skepticism than I used to.

I've often wondered if you might be applying different standards for outcome based results. For instance, have I missed the Cochrane Report proving that CR reverses Alzheimer's in E4 carriers?

I can chip in here, G does CR for many reasons and has been doing it for years, pre 23andme. The evidency for longevity in non human studies is compelling, and tantelising in shorter human studies. is I think its good for us to look beyond defining ourselves as apoe4 and not keeping a wide view. I think about cancer and cvd risk and diabetes as well, for instance. Not looking at anyone in particular, just thinking through the conclusion.

[RichardS]

Regarding the skepticism issue, I would like to point out that any dietary 6-week intervention with people who have cognitive impairment that shows a positive effect is impressive to me.

GB,

I know we E4's really want to find outcome based studies that break down the subject groups by apoe status. When a study finds benefit to E3's but not E4 carriers, we get disappointed. However, most of these studies do not try to balance the number of E4+ vs E4- carriers but rather just take them as they come in through the regular screening process. You sound like someone who probably gets this, but I want to point out for the less statistically minded that this has the effect of skewing the samples sizes where we typically see 20-25% E4+. This makes for an uphill battle in achieving statistical significance because for any given effect size, a larger sample will get you closer to statistical significance. A lot of the data I see reported on E4's suffers from inadequate sample sizes, meaning we can't typically say with confidence that the intervention works for E4- but not for E4+. A few studies balance E4+ and E4- but from my reading that is rare. I think we need to temper our skepticism on the supposed failures of studies to show benefit for E4's, not that we should ignore the data, just that looking at the effect sizes in the raw data can be instructive.