Why Diets Work and Then Stop Working
Almost every diet works for the same boring reason — and fails for the same one
Keto, low-fat, paleo, intermittent fasting, carnivore — head-to-head trials repeatedly find that when calories and protein are matched, the differences between diets are small. What actually predicts weight loss is energy balance (eating less than you burn), and what predicts keeping it off is adherence — whether you can actually live this way.
- When calories and protein are matched, diets perform similarly in head-to-head trials
- Most diets work by cutting a food category, which incidentally cuts calories
- Long-term results are driven by adherence, not metabolic magic
Stanford researcher Christopher Gardner ran the DIETFITS trial to settle the low-fat versus low-carb question properly: 609 adults, randomized, followed for a full year with real nutritional support rather than a handout. The result, published in JAMA in 2018, was that both groups lost roughly the same amount — about 5–6 kg on average — with no significant difference between them. Gardner also tested whether genetic markers or insulin response predicted who'd do better…
Why it matters: Rigorous head-to-head trials keep finding that diet composition matters far less than sticking with it. Choose based on what you can sustain, not on which camp has the better marketing.Choose one real claim or decision today and test it against this idea: When calories and protein are matched, diets perform similarly in head-to-head trials
The Minimum Effective Dose of Exercise
The biggest health jump is from zero to something — not from a lot to more
The health returns of exercise are steeply non-linear. Going from completely sedentary to modest activity produces a far larger drop in mortality risk than going from moderately active to athlete.
- Health returns from exercise are steeply non-linear — zero to something matters most
- Guidelines suggest 150 min moderate or 75 min vigorous weekly, plus 2 resistance sessions
- Meaningful mortality benefit appears well below guideline levels
A team led by Chi Pang Wen followed 416,175 adults in Taiwan for an average of eight years, tracking exercise against mortality. Published in The Lancet in 2011, the results were striking: just 15 minutes of moderate exercise a day — well below the standard 30-minute recommendation — was associated with a 14% reduction in all-cause mortality and about three years of additional life expectancy compared to being inactive. Every additional 15 minutes reduced…
Why it matters: The exercise dose-response curve is steepest at the start: 15 minutes a day cut mortality 14% in a 416,000-person cohort. If you do nothing, doing a little is the single highest-leverage health change available.Choose one real claim or decision today and test it against this idea: Health returns from exercise are steeply non-linear — zero to something matters most
What Sleep Deprivation Actually Does to You
The scariest part is that you can't tell how impaired you are
Sleep isn't downtime — it's when memory consolidation, metabolic clearance in the brain, hormone regulation, and immune function happen. Short sleep is causally linked to impaired glucose handling, increased appetite (ghrelin up, leptin down), weakened immune response, and worse emotional regulation, with much of this demonstrated in controlled experiments rather than just correlations.
- Sleep is when memory consolidation, brain clearance, hormone regulation, and immune work happen
- Restriction impairs glucose handling, raises appetite hormones, and weakens immunity
- Subjective sleepiness plateaus while objective performance keeps declining
Hans Van Dongen and David Dinges at the University of Pennsylvania ran a two-week laboratory study restricting participants to 4, 6, or 8 hours in bed, measuring cognitive performance daily. The 6-hour group — a sleep duration millions consider normal — showed cognitive deficits after two weeks equivalent to going 24 hours without sleep at all, roughly the impairment of being legally drunk. The finding that made the study famous was the dissociation: participants…
Why it matters: Six hours a night for two weeks produced impairment equivalent to 24 hours of total sleep deprivation — and participants couldn't perceive it. The feeling of having adapted to short sleep is itself a symptom.Choose one real claim or decision today and test it against this idea: Sleep is when memory consolidation, brain clearance, hormone regulation, and immune work happen
Protein: The One Macro Most People Underrate
It keeps you full, protects muscle, and costs energy just to digest
Protein does three things no other macronutrient does as well. It's the most satiating — gram for gram it keeps you fuller than carbs or fat, so higher-protein diets tend to reduce calorie intake without conscious effort.
- Protein is the most satiating macro — it reduces intake without conscious restriction
- 20–30% of protein calories are burned digesting it, versus 0–10% for fat and carbs
- Adequate protein preserves muscle during weight loss, protecting strength and metabolic rate
Robert Morton and colleagues pooled 49 randomized controlled trials covering 1,863 participants to answer a question the field had argued about for decades: how much protein actually helps when you're resistance training? Their 2018 meta-analysis in the British Journal of Sports Medicine found that protein supplementation significantly increased strength and lean mass — but the benefit flattened out at about 1.6 g per kilogram of bodyweight per day, with a confidence interval extending to…
Why it matters: Protein is uniquely satiating, metabolically expensive to digest, and necessary to hold onto muscle while losing fat. Around 1.6 g/kg/day covers most people who train — well above standard guidelines, but far below what supplement marketing implies.Choose one real claim or decision today and test it against this idea: Protein is the most satiating macro — it reduces intake without conscious restriction
Ultra-Processed Food and the Calorie Trap
The first trial to test it found people ate 500 more calories a day without noticing
'Processed' is a fuzzy word — freezing vegetables is processing. The concept that matters is ultra-processed food: industrial formulations built from refined ingredients plus additives, engineered for shelf life and palatability — sodas, packaged snacks, reconstituted meats, most breakfast cereals.
- 'Ultra-processed' means industrial formulations of refined ingredients plus additives — not just any processing
- These foods are calorie-dense, low-fiber, fast to eat, and weakly satiating for the calories delivered
- The issue is overconsumption by design, not a single toxic ingredient
Kevin Hall at the NIH ran the first randomized controlled trial to test ultra-processed food directly, and he designed it to be hard to argue with. Twenty adults lived in a metabolic ward for a month, spending two weeks on an ultra-processed diet and two weeks on an unprocessed one, in random order. The two diets were carefully matched for calories presented, sugar, fat, fiber, and macronutrients — and participants could eat as much…
Why it matters: The NIH's controlled trial showed ultra-processed food drives ~500 extra calories a day even when nutrients are matched and people eat freely. The effect is invisible to the eater — which is why environment design beats willpower.Choose one real claim or decision today and test it against this idea: 'Ultra-processed' means industrial formulations of refined ingredients plus additives — not just any processing
How to Read a Health Headline Without Getting Fooled
'Doubles your risk' can mean going from 1-in-10,000 to 2-in-10,000
Nutrition headlines mislead through a few repeatable tricks. The biggest is relative versus absolute risk: 'eating bacon daily raises bowel cancer risk 18%' sounds alarming, but that's a relative increase on a small baseline — in absolute terms it shifts lifetime risk by roughly 1 percentage point.
- Relative risk inflates small effects; always ask for the absolute change
- Most nutrition headlines are observational and can't establish causation
- People who eat 'healthy' foods differ in many other ways — confounding is rampant
In 2015 the WHO's International Agency for Research on Cancer classified processed meat as a Group 1 carcinogen — the same category as tobacco and asbestos — and headlines worldwide announced that bacon was as dangerous as smoking. The classification had been badly misread. IARC groups describe how confident scientists are that something causes cancer at all, not how much it raises risk. Cancer Research UK put out the actual numbers: smoking causes about…
Why it matters: Health headlines exploit relative risk, observational data, and single studies. Asking three questions — absolute or relative? observational or randomized? one study or many? — defuses most of them.Choose one real claim or decision today and test it against this idea: Relative risk inflates small effects; always ask for the absolute change