Evidence-based articles exploring the science of nutrition, metabolism, and health technology.
A landmark CVPR 2021 study introducing a 5,006-dish dataset with video, depth, and nutritional annotations — enabling computer vision models that outperform professional nutritionists at predicting calories and macronutrients from food images.
How the USDA Healthy Eating Index and Harvard Alternative Healthy Eating Index score diet quality across 13 components — and what decades of validation research reveal about their link to chronic disease and mortality.
The NOVA system classifies all foods into four groups by degree of processing. Large cohort studies now link ultra-processed food consumption to obesity, cardiovascular disease, and all-cause mortality.
EAR, RDA, AI, UL — the NASEM Dietary Reference Intake framework underpins every nutrition label and dietary guideline. Understanding these concepts reveals why nutrient flagging thresholds matter.
A comparative review of dietary limits from the WHO, American Heart Association, and Dietary Guidelines for Americans — where they converge, where they differ, and the evidence behind the numbers.
Åkerstedt and Folkard's 1987 model describes how homeostatic sleep drive, circadian rhythm, and sleep inertia interact to shape daily alertness — validated across 35+ years of military and aviation research.
The Sleep, Activity, Fatigue, and Task Effectiveness model — developed at Walter Reed Army Institute of Research — is the gold standard for biomathematical fatigue prediction. But it has zero nutrition inputs.
A systematic meta-analysis finds a small but highly significant effect (ES = -0.21, P < 0.0001) of dehydration on attention, executive function, and motor coordination — even at mild levels of 1-2% body mass loss.
The largest personalized nutrition study ever conducted found that identical twins show different metabolic responses to the same foods — gut microbiome, meal timing, and sleep matter more than genetics alone.
Time-restricted eating research shows that a 10-12 hour eating window improves insulin sensitivity and metabolic markers — independent of calorie intake. When you eat matters as much as what you eat.
Glucose spikes and crashes impair attention, memory, and executive function with a striking r=0.83 correlation between glycemic variability (MAGE) and cognitive scores — even in non-diabetic adults.
Caffeine blocks adenosine receptors within 20 minutes, peaks at 45 minutes, and has a half-life of 4.5-6.4 hours. Its dose-response follows an inverted-U curve — 200mg is optimal, 400mg+ shows diminishing returns.
