The chronically elevated blood sugar is mainly caused by insulin resistance. Insuline resistance isn’t necessarily caused by eating too many carbs. One of the known risk factors is obesity, and seems to stem from the fat tissue itself and not from the food that has caused the obesity. I’m not saying that eating too many carbs is harmless, but I’m just pointing that it’s more complicated than that.

These types of studies can literally only detect correlation. They look at massive data sets and yank out patterns. It's closer to reading tea leaves than hard science.

You might say... they are grilling the data

I have a different perspective! I even wrote a whole post about it in detail in the keto community (just now).

Carbohydrates are necessary, but not sufficient to develop type 2 diabetes.

i.e. it's not possible to develop type 2 diabetes without carbohydrates, I haven't seen any case study, or literature demonstrating it.

Obesity (as with all manifestations of poor metabolic health) is driven by carbohydrate consumption (driving insulin, driving anabolism). At least in this model [Paper] The Carbohydrate-Insulin Model of Obesity - Beyond “Calories In, Calories Out” - 2018

You are absolutely right, and your grasp of the nuance is well appreciated. Our bodies seem to function best on either, in no particular order

1. carbs and protein
2. fats and protein

Mixing all three seems to cause problems.

Let's be real, mixing all three in excess is causing problems. But of course it's not necessary easy to limit intake on modern society.

Fat and Carbs is the real problem, due to the randle cycle (not a cycle) cross inhibition this causes excessive amounts of inflammation.

not easy, but very much worth it.

The way I read your post, I interpret it as saying that you can’t have diabetes type 2 if you’re eating such that your blood glucose levels are maintained within acceptable levels. However, I’d argue that you have type 2 diabetes if your body is incapable of regulating your blood sugar without dietary adjustments. It might very well be the case that eating low carbs, apart from treating the symptoms type 2 diabetes, might protect you against developing type 2 diabetes, but that doesn’t mean that the reverse is true: that carbs are the direct cause of type 2 diabetes. It might be true that low carb diets are one way to avoid becoming obese and therefore protect you against the effects of obesity on your organs, or that it might increase insuline sensitivity, but we can’t conclude from this information that carbs are the primary cause of developing diabetes type 2, even though it can (indirectly) contribute to it.

diabetes type 2 if you’re eating such that your blood glucose levels are maintained within acceptable levels

This is the definition of type 2 diabetes.

that doesn’t mean that the reverse is true: that carbs are the direct cause of type 2 diabetes.

Agreed, carbohydrates are a necessary component of type 2 diabetes but are not sufficient by themselves.

we can’t conclude from this information that carbs are the primary cause of diabetes type 2.

I can, in so far as they are a necessary part of developing type 2 diabetes, t2d can be avoided by not consuming them. Type 2 diabetes is a blood glucose condition, there are multiple layers in avoiding that state. Removing carbohydrates is a guaranteed way, but not the only way, to avoid type 2 diabetes.

As far as I'm aware the various factors impacting insulin sensitivity (and thus t2d):

• carbohydrate load
• industrial oil consumption
• fructose consumption
• inflammation
• exercise levels
• carbohydrate and fat consumption : randle cycle inhibition

Though only carbohydrates are necessary for t2d, the other factors may or may not be present

I haven't read the study and obviously neither have you

One wants to think that in a study like this, researchers kept that in mind and ensures that those variables were accounted for.

I can't vouch for any of that, but that is such a "step one" that it'd be really stupid if they hadn't

One wants to think that in a study like this, researchers kept that in mind and ensures that those variables were accounted for.

that is such a “step one” that it’d be really stupid if they hadn’t

The neat part about epidemiology is they can't really control for healthy user bias - they can acknowledge it, then model a offset adjustment (assuming some uniform random variable with linear effect usually - so a regression to remove factors requires knowledge of their causal contribution which is "estimated" in the model....)... but yeah, the neat part is they don't - which is why epidemiology can never prove causation.

:::spoiler neat

:::