Thread by @Dmozaffarian

A DETAILED PRIMER on NUTRITION and DIABETES:

One of the most common questions I’m asked is how various foods influence risk of developing T2 diabetes (DM). This (long) thread will go through the science

(Evidence on foods to treat DM is broadly similar but won’t be covered here

First, a common mistake is to only consider acute (post-eating) effects on blood glucose (BG). The thinking goes: If a food acutely raises BG, it increases risk; and if it does not, it must be ok.

This is not correct.

Foods influence DM through many paths: BG/insulin, but also liver fat synthesis, inflammation, microbial dysbiosis, more.

Fructose is a prime example. Fructose (which with glucose forms sucrose, or table sugar) has a low glycemic index (GI, 25), with small effects on BG/insulin.

But, fructose in excess clearly raises diabetes risk, through other paths. In the liver, fructose is converted to fat (a process termed de novo lipogenesis [DNL]). In excess, these fats can’t be burned and lead to abnormal fat storage in the liver (nonalcoholic fatty liver) and…

… around other abdominal organs (visceral fat). Over weeks to months (not considering a single meal), this is a prime driver of insulin resistance and diabetes.

So, fructose makes clear that GI and acute BG responses are not the whole story. We’ll come back to this.

Second, still, a major overall driver is of course foods that rapidly raise BG. Prime offenders: refined starch (essentially 100% glucose), added sugars (e.g., sucrose, which is 50% glucose and 50% fructose). These “fast carbs” are digested quickly, so when eaten at high doses…

…flood the liver. High BG spikes insulin, leading to multiple negative effects for FM. Fructose is converted to fat by DNL, and excess glucose (beyond what’s taken up by muscles or fills glycogen stores) is also converted to fat by DNL. (as is excess protein… stay tuned)

So, foods that drive either higher BG/insulin or DNL increase diabetes risk. (at high doses, glucose does both)

That’s why lowering refined starch and sugar has such profound benefits for diabetes - both these harmful pathways are shut down.

Third, excess protein intake also drives higher insulin and DNL! What’s excess? Anything not needed for building new muscle. If you’re doing regular strength training, dietary protein will build muscle. But if not, that excess protein spikes insulin levels and generates new fat.

If you’re just measuring BG, these harms will be missed.

Take-home message: (a) you don’t build muscle just by eating protein; (b) if you eat more than your muscles use, then effects on insulin, DNL, fatty liver, and visceral fat are not much different from eating refined carbs.

This is one reason why higher protein intakes are consistently linked to HIGHER onset of diabetes. This is under recognized, and so many people (and start-up companies) are gushing about “more protein” for health. Despite breathless marketing, Americans are not protein deficient.

Protein also alters the microbiome. There’s growing evidence these effects vary based on the protein source and processing. Like for refined carbs, processed proteins might have some different health effects than protein in intact foods.
ncbi.nlm.nih.gov/pmc/articles/P…

Bottom line: if you’re not actively strength training, you don’t need more protein. And beware processed protein-rich foods and powders. Likely not doing your body any good, and maybe causing harms.

Fourth, for starches and sugars, both dose (amount) and rate of digestion (GI) matter. At low doses and rates (e.g., fruit, veggies, minimally processed whole grains, beans), the (low) flux of glucose and fructose into the liver matches the body’s metabolic needs…

…Glucose is used by brain and muscles, and fructose is converted to fat which is used by the heart and some other tissues. We evolved to eat low, slow doses of glucose and fructose - all the tissues are happy. So all carbs are not the problem, just high doses of processed carbs.

Of note, the human body has no need for dietary carb: the liver can synthesize all it needs from fats and proteins. But other compounds in these foods, like prebiotic fibers and diverse phytonutrients, create health benefits. And don’t forget taste and joy of eating.

If eating refined starches or sugars, their flux can be slowed by mixing with other foods. That’s one reason ice cream has a low GI (50) compared to many foods. Douse bread in olive oil, load pasta or rice with veggies, nuts, beans, eggs, chicken, fish. A bit of refined starch…

… or added sugar in an otherwise healthy food or meal is not that big of a deal. Think “low refined starch and added sugar” rather than “low-carb” for prevention of diabetes. Fruits, veggies, beans, and minimally processed whole grains have many health benefits.

Fifth, minimally processed whole grains are good for diabetes, from both observational studies and trials. Benefits arise from prebiotic fibers (nurturing gut bacteria) and diverse phytonutrients. The starch also has slower digestion due to partly intact plant structure. …

… One can slow the digestion further by mixing in other healthy, low GI foods: nut butters, seeds, nuts, berries, yogurt, etc.

Health effects of highly processed whole grains (many industrial breads, cereals) are more controversial. They contain all the fiber and minerals, but higher GI and fewer trace phytonutrients. Observational studies and two small trials in diabetic patients suggest benefit.

Overall, processed whole grains are likely healthier than their refined grain counterparts, but worse than minimally processed whole grains. Slowing their GI by adding other foods to the meal is probably even more helpful.

Sixth, foods can damage the pancreas and microbiome to increase diabetes risk, separate from acute BG effects. Heme iron is a classic example. Iron is a highly reactive, oxidant compound.

Excess iron causes diabetes in animal models. In humans, it’s well known that conditions of iron overload damage the pancreas and insulin-producing beta cells, and that reducing iron overload reverses these effects.

Iron also has important microbiome effects. Both iron deficiency and high iron intakes cause microbial dysbiosis and gut inflammation. These pathways are strongly related to diabetes risk. ncbi.nlm.nih.gov/pmc/articles/P…

The prime source of heme iron is red meat (heme iron is what makes red meat “red”).

Like animal models and conditions of iron overload, women with higher body iron stores have higher risk of gestational DM. In general populations, higher dietary heme iron intake, higher iron body stores, and higher red meat intake are each associated with higher incidence of DM.

Like many bioactive natural compounds, dietary heme iron has a natural “sweet spot.” Too little, or too much, and bad things happen.

The microbiome and diabetes effects of many other food compounds are still emerging. In one trial in humans, four different artificial and natural low-calorie sweeteners adversely altered the microbiome, and two of these also (over many months) led to worse glucose tolerance.

Similar early findings for microbiome harms have been seen in animal models for other common additives, like stabilizers, emulsifiers, and thickeners.

Putting it all together:
1. Avoid refined starches and added sugars (especially liquid added sugars).
2. When you do enjoy these, mix other foods to slow digestion and metabolic harms. (Think ice cream, honey coated nuts, or 70% cocoa over candy or pastries …

… and mix starches with plenty of fruits, nuts, veggies, beans, plant oils, yogurt, or fish, or modest eggs & poultry).
3. Enjoy minimally processed whole grains, especially mixed with healthy low GI foods. Processed whole grains are better than their refined grain counterparts.

4. Don’t seek out “protein,” unless you’re actively strength training and building muscle. Americans get plenty of protein from a regular diet. Avoid refined proteins and powers in ultraprocessed foods and shakes.

5. If you eat red meat, stick to modest amounts. It has some important nutrients, but the heme iron and protein, in excess, will increase risk of diabetes in the long-term.

6. As possible, avoid ultraprocessed foods with lots of additives. We’re just scratching the surface on their effects, and better safe than sorry.

p.s. I know there are high red meaters who will disagree with some of these points, sometimes passionately. Hopefully people will be civil…. we’re all trying to fix the food system.