Scientists Pit Sourdough Against White Bread | Hacker News Reader

Scientists Pit Sourdough Against White Bread

theatlantic.com

25 pointsprostoalex12 comments

Sourdough is purported to be healthier, have a lower glycemic index, etc. but this effect was not observed. They speculate that it may differ from person to person based on microbiome.

If I'd known that was the result I wouldn't have clicked through. The title is fucking idiotic.

Sounds like the title served its purpose perfectly. Annoying and effective.

Plus, other researchers are skeptical of the results due to the low sample size (n = 20).

It exposes the high amount of variability in even simple nutritional experiments. At least it is honest. Unlike some research that suggest x is good by testing on similarly small samples.

The journalists do include two very frank quotations from peers criticizing the experiment and what conclusions could be drawn. So annoying headline maybe, but honestly reported. And in this topic the no-story, no-effect is the story. Perhaps artisanal sourdough is mostly similar to white bread, and the range of different glycemic responses gets subsumed by differences in individual responses to any generic bread. Artisanal bread and wonderbread, equally bad, equally good.

Microbiome is becoming the scapegoat for some studies when they can't find a root cause, is it REALLY that important or are we just seeing things our models failed at pushed to some little understood part of dietary science?

I've seen the microbiome argument used as an explanation for weight loss, carbs processing, fat processing, protein absorption and now this, it's just feeling too much of our failed model relying on this unknown territory.

Outside of this not-really-helpful study, I can say that sourdoughs do tend to have a longer shelf life without preservatives; they resist staling and mold better than white bread, all things being equal. Commercial bread usually has preservatives added to balance this out, but it won't work out the same.

I've found sourdoughs, especially big miches, will actually taste better after 2 or 3 days. Commercial packaged breads are usually a straight downhill slope.

Source: direct experience, various baking books, historical info.

It would be great to see a larger study, but it's interesting nonetheless.

Something I thought about fermented foods is that they tend to have more bioavailable nutrients. So, sourdough contains less phytic acid than quick breads, making it so the nutrition in the bread can be digested more easily. Amongst other things. If this is true, that seems like a good enough reason to choose it. Flavor and texture is good enough too, really.

The point about individuals discovering what works for them is critical. One of the best things you can do for yourself is just pay attention to your diet and how it makes you feel. Maybe the crappiest white bread is just fine for you. Maybe it's totally not. I believe there aren't many absolute truths about nutrition, so often the onus is on individuals to try and discover these things about themselves.

Just a warning, as noted in the article, this is nothing to base a diet on. But I wouldn't necessarily base a diet on the epidemiological studies alone either.

A click-bait headline on nutrition scientific publication is a way for the media to get ad revenue, not a way to inform.

That said, the article did a pretty good job at informing on some really interesting science. It puts some good upper bounds one could expect in the population this was drawn from. Other populations, or sub populations could have larger possible effects.

The critiques are basically that this study should not be the end-all view. Of course some one should do an n=20 on this type of profiling before expanding to a larger cohort. And it seems that there's a lot to make a larger study really interesting.

So people on average are average. That's not surprising at all. It bothers me that so few studies are built to identify which individuals will see a difference and which will not. Rather they assume all individuals are identical and will have identical reactions.

To properly detect subgroups you need much larger sample sizes, in thousands...

The trial only lasted for 1 week each way and contained a sample of 20. I wouldn't be making any strong conclusions based on this.

Reply on news.ycombinator.com