Hacker News Reader

Comment by paulpauper | Hacker News Reader

Consuming a drink with erythritol — an artificial sweetener used to add bulk to stevia and monk fruit and to sweeten low-carb keto products — more than doubled the risk of blood clotting in 10 healthy people, according to a new pilot study.

10 people. not exactly convincing. It reminds me that earlier study about a purported cure for colon cancer which had 8 participants, which no one talks about anymore.

View on HN

You're ignoring most of the data and mischaracterizing what you did mention.

Is this more convincing?

> That study also analyzed the blood of more than 4,000 people in the United States and Europe and found that those with the highest levels of erythritol were twice as likely to experience a heart attack or stroke.

This is not more convincing because the human body produces erythritol endogenously so high erythritol blood levels may be another effect of the health problems that cause heart attacks and strokes, rather than the cause. To prove causation they would have to track each subjects consumption of erythritol rather than their blood levels.

The referenced study's methods did NOT account for the fact that overweight people are more likely to be consuming erythritol and/or on a keto diet. Highly likely to be confounding.

How does one measure "double the risk"? Does that mean if one person in a control group happened to have a stroke in that time period, and now it's two people, that counts as double? Out of 4000? Surely that can't be right.

n=1 can be important if the effect size is good. What makes this n=10 (per group) is interesting and not insignificant is that there is 100% response for blood platelets "enhanced stimulus-dependent release of the platelet dense granule marker serotonin (P<0.0001 for TRAP6 [thrombin activator peptide 6] and P=0.004 for ADP) and the platelet α-granule marker CXCL4 (C-X-C motif ligand-4; P<0.0001 for TRAP6 and P=0.06 for ADP)"

There are other studies n = 1,157, n = 2,149, n = 833 and article links to some of them.

See my other comment to see better discussion and links to studies.

That person could have other unknown problems though... I find it hard to believe that 1 person can ever be enough to quantify anything.

If you give one person rare substance X and that person dies 40 minutes later, it's enough. P < 0.000001 or at least, because people dying in 40 min time period for unrelated reason is really rare.

This study was an experiment (P<0.0001) that directly observed blood coagulation markers. Drink -> coagulation.

It's still an improvement on top of studies that just used serum levels (your body can generate some sugar alcohols, so may not be as closely related to consumption as one would think).

Reply on news.ycombinator.com