[iamt2]

> Interpretations vary, but giving up on fruit or grain and starch is obviously not part of it.

For typical genetics, this is true. However, I've yet to find any sources that show prevention and/or reversal of Type 2 that holds over 5+ years that includes a significant (>20g) of net daily carbs from fruit, grain and/or starches. I'm not alone, but in a definite minority of people who control their Type 2 through only diet and exercise, not even metformin, but all of our diets, while varied, have one characteristic in common: we restrict net carbs to various degrees, many at 20-50g net daily.

If you know how Type 2 diabetics like myself can keep our condition reversed while adding grains, starches and fruits in greater quantities, then please shoot a pointer my way. My plants intake these days is limited to very low carb green leafy vegetables like Romaine hearts.

[jerf]

As I've been trying to reconcile both the behavior of my own body and all the data that has been coming at us over the past couple of decades, one of the things that I've wondered is whether a crap, sugar-filled diet can do permanent damage that can never be undone, making the body especially prone to take carbs and behave badly with them.

My idea here is that perhaps some of the issue here is that people like you (and to a non-clinical extent, me) who may have had a very bad diet in the past kick our bodies into this "mode", for lack of a better word, we find we have to watch our carbs relatively closely. Or perhaps for some people, their genetics simply start them there. Meanwhile, people who never got to that point are saying "It's no big deal, I eat a lot more carbs a day and never get fat." It may still be the case that when you get down to it, the sugar and the white bread aren't doing them any favors, but they never overwhelmed their body's ability to deal with it.

I think one of the perils of medical studies is that it's really easy to run a study like "How Important Is It To Eat Beets 5 Times A Day" and get back a statistical result that it's a 3% detriment or something, without noticing that the 3% detriment is that two people had a horrible reaction and everybody else had no reaction. As an individual, the 3% isn't really interesting, because that number doesn't correspond to the result that anybody had; the question is, are you in the set of people who had an extreme reaction or not? I wish I had time to study this question concretely; how many studies are presenting their results using Gaussian-based statistics when the underlying data is fundamentally bi-modal ("it worked really well for a few people and didn't do anything to most"), thus causing potentially useful treatments to get statistically fuzzed out of existence?

[iamt2]

> My idea here is that perhaps some of the issue here is that people like you (and to a non-clinical extent, me) who may have had a very bad diet in the past kick our bodies into this "mode", for lack of a better word, we find we have to watch our carbs relatively closely. Or perhaps for some people, their genetics simply start them there.

For me there is a strong genetic component. Practically every member of my extended family on one side of my parents is either pre-Type 2, or is full-blown Type 2. Did not know this growing up, and dutifully ate a diet laden with starches, which came to bite me in the ass in a big way as an adult. The rest of the immediate and extended family is now aware, and knows that I'm living proof that if they catch the same subtype, it can be prevented and (if too late) reversed without resorting to medication, and hopefully I stop it dead in its tracks with my generation, and it never gets even a toehold in future generations of my family.

> As an individual, the 3% isn't really interesting,...

This is why open source and open data is so critical to scientific research moving forward. My hope is either JupyterLab or some project like it with massive collaboration and scaling becomes the standard way to present scientific findings in the future, and all raw data and software tooling used becomes accessible by anyone. Then we could definitively answer questions like yours.