Nutrition Research, Read Properly
Every statistic here links to the study it came from. When the evidence is preliminary or contested, we say so in the article rather than in a footnote.
Food Nutrition Apps: What That Score Out of 100 Is Actually Made Of
The Nutrition Facts panel is required to carry four micronutrients out of the 27 the same regulation calls essential in human nutrition. Everything a barcode scanner tells you is built on that panel and an ingredient list — and Yuka's score, the one method in this category published in full, is 60% nutrition, 30% additives and 10% an organic bonus. What that single number can and cannot carry.
- Nutrition apps
Nutrition Apps Compared: What Each One Can Actually Count
MyFitnessPal counts six micronutrients — the pre-2016 US Nutrition Facts label set, vitamin D not among them. Cronometer counts up to 95. An honest comparison of the main nutrition apps on what they can count, where their numbers come from, and what sits behind the subscription, including where our own app is the wrong choice.
- GLP-1
GLP-1 Hair Loss: What the Drug Labels Say, and Why Biotin Is the Wrong Answer
Hair loss is a counted adverse reaction on the Zepbound and Wegovy labels — 5.8% at Wegovy's maximum dose, 8.4% of women in that trial — and both labels associate it with weight reduction rather than a nutrient shortfall. Why biotin is the wrong answer, and what is worth doing instead.
- GLP-1
The GLP-1 Nutrition Gap: What 461,382 Patient Records Actually Show
In a peer-reviewed analysis of 461,382 adults, 22.4% had a nutritional deficiency diagnosed within a year of starting a GLP-1. Vitamin D led, protein is the bigger gap, and the hair-loss evidence points somewhere other than vitamins.
- Calorie tracking
How Accurate Are AI Calorie Apps? What the NIH Found When It Weighed the Food
NIH researchers weighed 102 meals to a tenth of a gram and compared them against four popular calorie apps. Measured one by one, the apps ran between roughly 250 and 345 calories short per meal. Here's why photo tracking misses, and how to correct for it.