Multiple good studies agree. Treat as settled.
Yes, it causes real weight loss. In a 72-week trial of 3,127 adults without diabetes, the top dose of the GLP-1 pill took off 11.2% of body weight, against 2.1% on a dummy pill. It has not been tested head to head against injected semaglutide or tirzepatide; compared across trials, it does less than tirzepatide.
In a 2023 pooling of fiber-supplement trials, the gel-forming soluble kinds lowered LDL cholesterol about 9.8 mg/dL and the soluble kinds that do not gel about 3.7. For constipation the best-tested fiber is psyllium, a soluble one; wheat bran, the classic insoluble fiber, had one small trial in the newest pooling we could read. Most foods carry both kinds.
220 minus age is close to the average of a mixed group of healthy adults and off by more than 10 beats a minute for many of them: by our count of one lab's published data on 230 adults, mostly recreational athletes, it missed 70, and Tanaka's formula, counted the same way, missed 60. Age explains less than half of why one person's maximum differs from another's. The only way to know yours is to measure it, and some people should not try that alone.
If you eat no animal foods, you need vitamin B12 from a supplement or a fortified food: in a British study, 52% of vegan men had levels in the deficient range. Over 50, US advice is to get most of it the same way. For energy in people whose levels are normal, small placebo-controlled trials of B12 on its own found no clear benefit.
Oats, barley, psyllium, plant sterols, nuts, soy and unsaturated fats each lower LDL cholesterol by a few percent in pooled trials, sterols by up to about 12%. Combined as the Portfolio diet, with the food delivered, they lowered LDL 28.6% against 30.9% for 20 mg of lovastatin in the first month of a trial of 46 people; completed, with 34 people trying each treatment for a month, the trial found the statin significantly better, 33.3% against 29.6%. Advised to eat that way at home, people's LDL fell about 13%, against 3% on a standard low-fat diet.
In a randomised trial, losing 5% of body weight raised insulin sensitivity by about a quarter. Of three trials here that matched the weight loss on different diets, none found a clear difference in the muscle; in the largest, a liver measure improved two to three times as much on keto. In type 2 diabetes, low-carb pulled ahead on blood sugar for about six months, and by a year that pooled lead had roughly halved. “Once and for all” did not hold where it was tested: after the low-carb diets ended, fasting insulin did not stay down.
Gram for gram, maltodextrin raises blood sugar about as much as pure glucose and more than table sugar, so that part of the viral claim is true. Worse than sugar for insulin resistance, belly fat or a fatty liver is another matter: trials of glucose against fructose and against table sugar found glucose no worse, and on some measures better. And the 65 to 250 pounds a year traces to nothing we could find.
Six of seven kinds of exercise cut visceral fat across 84 randomised trials, and in one eight-month trial it fell 6.9% while weight fell 2.6%. The green tea sold to shrink it was tested mostly by companies that sell tea. Of three trials we read with no company among their authors, two found no change and one, a small trial alongside exercise, found less.
Yes, for the fat: in a year-long randomised trial, 64% of people on a lifestyle programme had liver fat back under the line that defines the disease, against 20% on usual care, and liver fat starts falling within days of eating less. The scarring, which is what predicts how the disease ends, is a slower and smaller story.
Statins do raise blood sugar: by about 0.7 mg/dL on average in the blinded trials that measured it, enough to add roughly one diabetes diagnosis per 1,000 people a year on a standard dose, most of them in people already close to the line. Here is that number beside the heart attacks and strokes statins prevent, measured in the same people where trials counted both, and who sits where.