Genetics, metabolism, willpower, discipline. If you've ever wondered what separates the people who keep weight off from the people who regain it, that's probably your shortlist.
A systematic review tested that shortlist directly. Forty-nine studies. Over 31,000 participants. A hundred and twenty-four different predictors of weight maintenance, each classified by strength of evidence. The results demolished every trait on the list.
What Actually Predicts Keeping Weight Off Long Term
Weight maintenance is predicted by specific behaviors, not identity traits. A systematic review of 49 studies and 31,741 participants found that self-monitoring, increasing physical activity, and dietary changes carry strong evidence. Age, gender, stress, and weight history carry none. What you change matters. Who you are when you start does not.
— Varkevisser et al. 2019 · Obesity Reviews · n=31,741
Age was non-predictive. Gender was non-predictive. Ethnicity, socioeconomic status, education level: none of them predicted whether someone maintained their weight loss.
Stress carried the same verdict, backed by strong evidence. And the finding that might hit hardest: weight history was non-predictive. The number of past diet attempts, the total amount previously lost and regained, the years spent cycling through programs. None of it predicted whether this time would be different.
Your past failures are not evidence against your future.
What did predict it were specific, trackable behaviors. Self-monitoring (regularly tracking weight and food intake) showed the strongest evidence across the entire review. Eight out of ten studies that measured it found consistent prediction of maintenance success.
Increasing physical activity predicted maintenance in three-quarters of studies that measured it. But being active when you started the diet showed no predictive power. Starting fit was irrelevant. Becoming more active was what counted.
The same pattern held for dietary changes: cutting junk food, eating more fruits and vegetables, reducing sugary drinks, controlling portions. Each behavior independently predicted success.
Underneath all of it sat a single unifying pattern: baseline traits did not predict outcomes. Behavioral changes from baseline did. Where you started had nothing to do with where you ended up. What you changed about your daily habits had everything to do with it.
One psychological factor did carry strong negative evidence. Not willpower, not discipline as a personality trait, but something more specific: internal disinhibition, the tendency to lose control of eating in response to emotional triggers. A pattern most people recognize instantly, and one that has nothing to do with how strong-willed you are.
The mechanism behind this matters. Weight loss plateaus are not primarily caused by metabolic adaptation. They are driven by adherence decay: a measurable erosion of behavioral consistency that follows a predictable curve. Women dropped from 80% adherence in month one to 40% by month three. Men from 80% to 70% by month five. The behaviors that predict maintenance are the exact same behaviors that erode over time. Tracking them is how you spot the drift before the scale does.
If most diets end the same way, and your identity never predicted the outcome, the question was never whether you're the type of person who can keep weight off. It was always about what you measure, and whether those measurements track what the evidence says actually matters.
All 124 predictors. Every evidence level. The complete breakdown.