Hacker News new | past | comments | ask | show | jobs | submit login
Diabetes drug 'significantly reverses memory loss' in mice with Alzheimer's (sciencedirect.com)
101 points by dsego on Jan 1, 2018 | hide | past | favorite | 26 comments

This would be great news if we have not cured mice many times before of Alzheimer's. Unfortunatly all the drugs that have worked in mice failed when tested in humans.

It appears that the mouse model is a very poor model for the human form of the disease. I would go so far as to say it is worse than useless and actively leads us astray.

I wonder if we had similar success with curing Alzheimer (or cancer for that matter) in humans if experimenting on humans were possible.

Experimenting on humans is possible; that’s what clinical trials are for

But you can't try a couple of drugs and then dissect the humans to see how the plaque in their brain looks like. Genetically engineering humans to develop Alzheimer more quickly is also generally frowned upon.

The problem is the FDA won’t let you in practice test in humans until you have shown effectiveness in animal models.

Of course if the animal model is not similar to the human form then what you end up doing is testing a drug that worked for one disease on a totally different one. In someways it is like using a drug that treats high blood pressure and then testing to see if cures bladder cancer. You might get lucky, but it is a pretty longshot.


You are funny. Please show me the regs via which this is as possible as you suggest.

Do you mean mouse & rat?

> I would go so far as to say it is worse than useless and actively leads us astray.

A notable example is when a strain (or perhaps race is a better terminology) is used which is known to be very sensitive to cancer. Using that to prove something causes cancer is a slippery slope, especially when its extrapolated to other conclusions.

There is no animal good model for Alzheimers.

You raise a very good point which is almost all the animal models we have are terrible models for human disease that is supposedly being studied.

Cancer models of cancer are just as bad. Interestingly the only good model of human cancer are spontaneous cancers in human pet dogs [1]. Of course we hardly use this models despite most dog owners being willing to volunteer their pets for study.

1. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3130881/

I recall reading an article saying that doctors were nearly ready to begin calling Alzheimer's disease Type III diabetes. This result would lend credence to that notion.

I predict that in 5-10 years the evidence will be overwhelming that sugar/carbohydrates for some subset of the population are essentially poison. Cancer, heart disease, Alzheimer’s - I think we will start to see more and more mainstream connections of all three with carbohydrate intake.

Unless you're talking about an extremely small subset of the population, like 5% or less, I think that would be a pretty bad prediction. Whole fruits, vegetables, legumes, leafy greens are all recommended by virtually all doctors and official dietary guidelines. And each of those are all high in carbohydrates.

I'm so jaded by expert opinion these days. For decades doctors have been recommending a low-fat, high carb diet and saying things like saturated fats are bad for you. It's becoming very clear that there's no evidence this is considered healthy. If anything, the demonization of fats around the 80s coincides with increasing obesity rates in the US.

The American Heart Association literally has an infographic on their website that tells people to eat more polyunsaturated fats (i.e. vegetable oils) and completely avoid coconut oil (https://healthyforgood.heart.org/eat-smart/infographics/the-...). Anyone who has studied the science behind ketosis knows this is completely bonkers.

Leafy greens have a high percentage of their calories in carbohydrates, but they are not high in carbohydrates (they are recommended because they have a high ratio of valuable vitamins, etc., to calories, and recommended for weight loss because they have very low calorie density. They are typically recommended even on specifically low-carb diets.)

Also important to note the different between carbs and net carbs. Leafy greens are high in fiber, which is a carb, but you need to subtract that value from the total carbs per serving to get to a net carb count (i.e. total carbs your body can actually turn into energy)

I do not think that carbohydrates all together are the problem. Its just that our capabilities in understanding (poly-) saccharides are 20 to 30 years delayed compared to other important biomolecules like nucleobases or aminoacids while playing a very important role (like the polysaccharide modifications of proteins). The problem is not that we cannot detect whether there is any sugar molecule present, its just hard to say which one (of the thousands of possible combinations of e.g C6H12O6). We do not know how much is decoded in there.

Hopefully, but probably not. For some reason, the subset of the population that will have the negative effects, more often gets forgotten.

This sounds interesting. Do you remember why they were making this allusion to diabetes?

I read an article just a few days ago that discussed how sleep is important for cleaning out plaques from the brain like those associated with Alzheimer's. Unfortunately I flush my browser history frequently or I would post it.

I bought and read Matthew Walker's book Why We Sleep (which is referred to in PP's thirs reference) and would thoroughly recommend it to all humans. I now take some solace in my grumpiness when tired having a biological explanation.

stupid question, but how do they got the mice with alzheimer's?

You buy mice that are likely to develop it, for example at https://www.jax.org/mouse-search (I notice they offer a service to use CRISPR to design a mouse strain for you, too)

Not a stupid question. Mice (and all other animals besides maybe nonhuman primates) don't get Alzheimer's. There are mice bred to display symptoms of Alzheimer's. Because nobody knows the "root cause" it's a bit of a catch-22. The same holds true for many other neurodegenerative/psych disorders.

I don't know in detail, but suffice it to say that for many years we've had mouse strains with particular genetic alterations that appear to mimic Alzheimer's. Not my field, so google for more info.

Not a stupid question at all. The answer is the mice are genetically engineered to overexpress beta-amaloid (or sometimes Tau).

Alzheimer's is from the environment

There are also quite a few genes that are tied to increased odds of Alzheimer’s.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact