In 2012, Herman Pontzer and his colleagues published a paper measuring the daily energy expenditure of the Hadza, a hunter-gatherer population in Tanzania. The headline finding was simple and counterintuitive: despite walking miles every day to forage and hunt, the Hadza burned about the same number of calories per day as sedentary office workers in the United States, after correcting for body size.
That paper, and the dozen-plus follow-ups that have come since, gave us the constrained energy expenditure model — one of the most-cited and most-misunderstood findings in modern nutrition science.
It’s worth getting right, because the popular interpretation has hardened into a slogan (“exercise doesn’t matter for weight loss”) that overshoots what the evidence supports.
The additive model and why it broke
The classical view of total daily energy expenditure (TDEE) was additive. Your body had a basal metabolic rate (BMR), the energy cost of just being alive. On top of that, you added the thermic effect of food (TEF, ~10% of intake), and on top of that, your activity expenditure. Move more, burn more, all of it linear. A 200-calorie jog added 200 calories to your TDEE.
The Pontzer findings broke that picture. The Hadza, despite far higher activity levels, had TDEE figures that — once you controlled for body composition — looked similar to sedentary populations. The cleaner interpretation was that the body compensates. When activity goes up, other components of energy expenditure quietly come down: resting metabolism, fidgeting, immune system activity, reproductive function, possibly inflammation regulation. The total stays in a relatively narrow band.
This is the constrained energy expenditure model. TDEE behaves less like a sum and more like a regulated quantity with a partial ceiling.
What the model does and doesn’t claim
Three claims are well-supported.
TDEE does not scale linearly with physical activity in the long run. Compensatory adaptations partially offset increases in activity expenditure, especially as activity becomes habitual.
The size of the compensation varies by individual and context. Some people compensate heavily; some barely. This explains a lot of the heterogeneity in exercise-for-weight-loss interventions.
The classical TDEE equations (Mifflin-St Jeor, Harris-Benedict times an activity factor) systematically over-predict TDEE for highly active people. They assume additivity that doesn’t fully hold.
Three claims are not what Pontzer showed.
Exercise does not “fail to burn calories.” The acute energy cost of activity is real and well-measured. The compensation happens over weeks and months, not in the moment.
Exercise is not useless for weight management. The evidence that exercise helps with weight maintenance after weight loss, independent of any acute calorie effect, is strong. The mechanism may be partly about appetite regulation and partly about preserving lean mass during weight loss.
The first law of thermodynamics has not been overturned. Energy in still equals energy out plus or minus storage. What has been refined is our understanding of how flexibly the “energy out” side responds to perturbations.
Why this matters for calorie tracking
If you’re using a calorie tracking app, the constrained-energy-expenditure model has practical implications.
Static TDEE estimates go stale. This is the problem MacroFactor’s adaptive algorithm tries to solve directly: by reading your actual weight trajectory and back-calculating effective TDEE, the app can update the target as your body’s compensation patterns settle in. Static plans built from Mifflin-St Jeor and an activity multiplier will tend to over-predict TDEE for active users, which means they’ll set calorie targets too high for the actual goal.
Plateau is not a moral failing. When weight loss stalls after months of consistent intake, the most common explanation isn’t “you stopped trying.” It’s that compensatory adaptations have closed the gap between intake and expenditure. The intervention is to acknowledge the new equilibrium and either accept it as the new maintenance, or recalibrate intake downward by a modest amount.
Movement still matters, just not for the reason the cardio-equipment-display says. A 45-minute treadmill session displays “burned 480 calories.” The real net effect on your weekly TDEE may be a fraction of that, after compensation. But the cardiovascular, metabolic, mood, sleep, and lean-mass-preservation benefits of regular activity remain robust and substantial. Move because of those.
What good practice looks like in 2026
Three principles synthesize the literature for a thoughtful tracker.
First, treat your TDEE estimate as a starting hypothesis, not a fact. Adjust it from your weight data over 6–8 weeks, not from your wearable’s calorie display.
Second, weight intake-side interventions more heavily than activity-side interventions when the goal is body weight change. The compensatory adaptation literature is unambiguous on which side does more of the work.
Third, do not stop moving. The benefits of activity that are not mediated by calorie expenditure are large, durable, and well-measured. Constrained energy expenditure does not undermine the case for exercise; it just refines what exercise is for.
Pontzer’s own popular book frames the takeaway well: the body is a system that adapts, not a furnace that just burns whatever you put in. Calorie counting still works as a tool, but only when you read it as a feedback signal rather than a deterministic plan.
That’s what the research actually says. The TikTok version is louder and less useful.
