What Actually Happens When You Crash Diet

A patient told me she had lost eighteen pounds in five weeks on around 900 calories a day, and that she was proud of it right up until the point where her hair started coming out and she stopped being able to get through a workout. Her TSH was normal. Her free T3 was at the bottom of the range. She had done exactly what she had been told to do, and her body had responded exactly as bodies do.

Severe restriction works. That is the problem. It works quickly enough to look like success and it triggers a set of adaptations that make the next attempt harder than the last one.

What your body reads a crash diet as

Your body has no way of knowing that the deficit is voluntary. It cannot distinguish a chosen 900 calories from a famine, and it does not try. What it detects is a sustained energy shortfall, and it responds with a coordinated set of changes designed to keep you alive through it.

Energy expenditure falls. Reproductive function is deprioritised. Thyroid hormone conversion shifts. Appetite signalling changes. None of this is a malfunction. It is the system working correctly, in a context it was not designed for.

Metabolic adaptation, which is the real name for starvation mode

"Starvation mode" is the phrase people use, and it is treated as a myth by one half of the internet and as an excuse by the other. The measurable phenomenon underneath it is called metabolic adaptation, or adaptive thermogenesis, and it is real.

Here is what it means precisely. When you lose weight, your resting metabolic rate falls, because a smaller body costs less to run. That part is arithmetic and everyone agrees on it. Metabolic adaptation is the observation that the rate often falls by more than body size alone predicts. The body is not just smaller, it has become more economical.

The most cited evidence is unusually extreme. In a follow up of participants from a televised weight loss competition, resting metabolic rate remained substantially below predicted values six years later, in the region of several hundred calories a day, and most participants had regained a large proportion of the weight. That study is worth reading and worth caveating. Fourteen participants, an intervention far outside anything clinical practice would recommend, and considerable debate about how far it generalises. It is a signal, not a settled law.

The older and in some ways more instructive evidence is the Minnesota Starvation Experiment, conducted in the 1940s, in which healthy men lost roughly a quarter of their body weight on a semi-starvation diet. Their metabolic rate fell dramatically. So did their concentration, their mood and their interest in anything other than food. On refeeding, many ate far past comfort for months. The psychological half of that result is the half people forget, and it is the half I see most often in clinic.

The thyroid part, which is my specialty and gets missed

During significant caloric restriction, the conversion of T4 to active T3 decreases and reverse T3 tends to rise. This is a normal adaptive response, not thyroid disease, and it is one reason people feel cold, flat and slow on a crash diet while their TSH looks entirely unremarkable.

It also means a thyroid panel drawn during an aggressive diet can be genuinely misleading. I have seen people started on thyroid medication on the strength of labs that were describing their diet rather than their thyroid. If you are being tested, it is worth your clinician knowing what you have been eating. This is part of the wider problem we cover in why TSH alone is not a thyroid panel.

The muscle question, stated accurately

You will read that muscle is broken down for fuel first. That is a simplification, and the accurate version is more useful.

The body does not select muscle preferentially ahead of fat. Under an energy deficit it draws on stored glycogen, on fat, and on protein together, and the proportions shift depending on how large the deficit is, how much protein you are eating, and whether your muscles are being given any reason to stay.

What makes severe restriction so costly is that all three of those levers are usually pulled the wrong way at once. The deficit is large, protein intake is typically low because total intake is low, and there is rarely any resistance training. Under those conditions a substantial share of the weight lost is lean tissue rather than fat. Estimates vary widely across studies and populations, which is why I am not going to give you a single percentage, but the direction is not in dispute.

And this is the part that makes it self-reinforcing. Lean tissue is metabolically active. Lose it, and your resting metabolic rate falls further, on top of the adaptive drop that was already happening. The diet that was meant to raise your metabolism has lowered it twice.

You cannot see any of this on a scale, which reports one number falling and calls it progress. It takes a body composition measurement to separate the two, and we explain how that works in DEXA and body composition scans.

Hunger does not go back to normal when the diet ends

Leptin, which signals energy sufficiency, falls with fat loss and falls disproportionately during rapid loss. Ghrelin, which drives hunger, rises. Both changes persist after the diet stops, and in at least one well conducted study these hormonal shifts were still measurable twelve months after the weight loss.

This is why the period after a crash diet feels like a failure of willpower and is not one. You are not eating more because you have become undisciplined. You are eating more because a set of hormonal signals is telling you to, considerably more loudly than before you started. We go through the different kinds of hunger and what each one actually is in the three kinds of hunger.

What I would do instead

A moderate deficit rather than a severe one. It is slower and it holds, and holding is the only thing that has ever mattered.

Adequate protein throughout, because it is the single most effective lever for preserving lean tissue in a deficit.

Resistance training, two or three times a week, which is not there to burn calories. It is there to give your body a reason to keep the muscle it would otherwise let go.

Enough sleep, because short sleep worsens insulin sensitivity and appetite regulation at exactly the moment you need both working.

And a measurement that can tell fat loss from muscle loss, so that you find out which one is happening while there is still time to change it.

If the reason you are considering something drastic is that ordinary approaches have stopped working, that is worth investigating rather than overriding. It is frequently the sign of something underneath, and the first place I look is insulin resistance.

Sources: Fothergill et al., Persistent metabolic adaptation, Obesity 2016 · Sumithran et al., Long-term persistence of hormonal adaptations, NEJM 2011 · NIDDK, Weight Management

If any of this sounds like your experience, you can book a virtual consultation with Dr. Sater to review your history and testing in full.