Is fast weight loss bad? Trials challenge the go-slow rule

The standard advice to lose weight slowly rests on a belief that trials have not confirmed: that weight shed quickly comes back quickly. In the study built to test it, people who lost weight rapidly and people who lost it gradually had regained almost exactly the same proportion three years later [s1]. The rate of loss, on the evidence, matters far less to lasting success than the popular rule assumes.

Guidelines have long recommended gradual weight loss, roughly half a kilogram to a kilogram a week, on the reasoning that a slower pace is easier to keep off [s1]. An Australian randomised trial put that to the test. Researchers assigned 200 adults with obesity to either a 12-week rapid programme or a 36-week gradual one, both aiming for the same 15% weight loss [s1]. More of the rapid group hit the target — 81% reached at least 12.5% loss versus 50% of the gradual group — and those who did moved on to a maintenance phase [s1]. By 144 weeks, both groups had regained most of what they lost, and by nearly identical amounts: 70.5% regain in the rapid group against 71.2% in the gradual group among completers, with intention-to-treat analysis showing 76.3% in both [s1]. The authors concluded the rate of loss does not affect how much is regained — a result the guidelines' logic did not predict [s1].

An older study agrees, and goes further

A separate analysis found fast losers did better, not worse. Drawing on the TOURS lifestyle trial, researchers sorted middle-aged women by how quickly they lost weight in the first month, then tracked them to 18 months [s2]. The fast group had lost far more by then — 10.9 kg versus 3.7 kg in the slow group — and were no more prone to regaining it [s2]. Fast losers were 5.1 times more likely than slow losers to have achieved a 10% weight loss at 18 months [s2]. Early, rapid progress predicted better long-term outcomes, not a rebound [s2].

Where "fast" does carry a cost

None of this makes crash dieting harmless, and the caveat is about body composition, not regain. A randomised trial that produced the same amount of weight loss over 5 weeks versus 15 weeks found the tissue lost differed: the slow group lost more fat and preserved a more favourable body composition, while the rapid group lost more lean body mass and had a larger drop in resting metabolic rate [s3]. Both approaches improved cholesterol and blood-sugar measures, and some metabolic markers actually improved more with rapid loss [s3] — but losing muscle and slowing metabolism are real downsides of going too fast.

There are safety limits too. In the Australian trial, one participant in the rapid-loss group developed gallbladder inflammation requiring surgery [s1] — a known hazard of very fast weight loss, which is why aggressive very-low-calorie diets are normally run under medical supervision.

It is also worth being clear about what these trials did and did not compare. In the Australian study the two groups were built to reach the same 15% target, and the rapid arm simply got there sooner [s1]; in the body-composition trial the amount of weight lost was deliberately matched, only the duration differing [s3]. So the comparison is between reaching a given loss quickly or slowly, not between losing more and losing less. That is the honest frame for the "go slow" rule: it is a claim about pacing, and on pacing the trials find little long-term penalty for speed [s1][s2], with the qualification that the tissue lost is more favourable when the pace is gentler [s3].

What it means

The evidence undercuts the idea that losing weight fast dooms you to regaining it: across these trials the rate of loss had little bearing on long-term maintenance, and early rapid loss sometimes predicted better results [s1][s2]. What does the damage is not speed itself but how the weight is lost — extreme crash diets cost lean mass, blunt metabolic rate, and carry risks such as gallstones [s1][s3]. The durable predictor of success remains the same whatever the pace: whether the change is one a person can sustain.