Ghrelin: the hormone that makes you hungry

Ghrelin was not discovered by anyone looking for a hunger signal. It was identified in 1999 by Kojima and colleagues, who were hunting for the natural ligand of the growth hormone secretagogue receptor — a receptor known to exist, but whose partner molecule was unknown. They purified it from rat stomach: a 28-amino-acid peptide carrying an unusual octanoyl (fatty acid) group attached to serine-3. That acyl group is not decoration. Without it, the peptide does not activate the receptor at all. They named the molecule ghrelin, from the Proto-Indo-European root ghre, "to grow", because it releases growth hormone.^[1]

The appetite role came second. Only once the peptide was in hand did it become clear that the same molecule released before meals and made animals — and then people — eat. That dual identity is still the most interesting thing about it: a single stomach-derived peptide that speaks to the pituitary about growth and to the hypothalamus about food, and that rises in states of energy shortage when both messages are useful.

There are two circulating forms, and the distinction matters when reading any study. Acyl ghrelin is the active, appetite-driving form. Des-acyl ghrelin — the unacylated form — is far more abundant in blood, and its physiological role is still being worked out.^[14] A paper reporting "total ghrelin" is therefore measuring mostly the form that does not directly drive eating, which is one reason results across the literature are not always as tidy as summaries suggest.

In 2001, Cummings and colleagues did the experiment that made ghrelin credible as a meal-initiation signal. They sampled plasma 38 times over 24 hours in 10 healthy people eating on a fixed schedule. Ghrelin rose nearly twofold immediately before every meal and fell to trough levels within an hour of eating — a pattern almost exactly reciprocal to insulin. Between meals it followed a diurnal rhythm, climbing through the evening to a zenith at 1 a.m. and falling to a nadir at 9 a.m. Ghrelin also correlated positively with age.^[2]

The detail worth holding onto is that this anticipatory rise tracks expected meal times rather than an empty stomach as such. This is why someone who habitually eats at 12:30 becomes hungry at 12:30, and why the first few days of shifting a meal schedule feel harder than the second week. Part of what you call appetite is a learned, scheduled hormonal event.

Correlation with meal times is suggestive. Giving the hormone and watching what happens is decisive. Wren and colleagues infused intravenous ghrelin (5.0 pmol/kg/min) or saline into nine healthy volunteers in a randomised, double-blind crossover study. Every single participant ate more on ghrelin — a mean increase of 28 ± 3.9% in energy consumed from a free-choice buffet (p<0.001), with higher appetite scores and no change in gastric emptying. It was the first circulating hormone shown to stimulate food intake in humans.^[3]

The clinical mirror image is just as instructive. In cancer patients with anorexia, a ghrelin infusion increased energy intake by 31 ± 7% and improved meal appreciation by 28% — again, every patient ate more.^[7] That is why ghrelin biology is being explored as a treatment for wasting and cachexia, not only as a target in obesity. A hormone that reliably makes people eat is a problem in one clinic and a therapy in another.

This is the section that changes how people think about their own weight history. In 2002, Cummings and colleagues followed 13 people with obesity through a six-month diet. A 17% loss of body weight was accompanied by a 24% increase in the area under the 24-hour ghrelin curve.^[4] Losing weight did not quieten the hunger signal. It amplified it.

Sumithran and colleagues then showed how long that lasts. In the New England Journal of Medicine in 2011, 50 overweight or obese adults completed a 10-week very-low-energy diet and lost 13.5 ± 0.5 kg. Ghrelin rose significantly; leptin, peptide YY, cholecystokinin, insulin and amylin fell; subjective appetite increased. The critical finding came later: at 62 weeks — more than a year after the initial weight loss — ghrelin, leptin, peptide YY, cholecystokinin, insulin and hunger were all still significantly different from baseline. The hormonal push toward regain does not fade.^[9]

I want to be plain about what this means and what it does not. It is a biological explanation for why weight regain is the norm rather than the exception, and it is not a character failing on the part of the person regaining. It is also not a reason to give up on weight loss. It is an argument for two specific things: losing weight at a rate and by a method the body will tolerate, and planning maintenance as a permanent strategy rather than an afterthought bolted on when the diet ends. A plan that has no answer for month fourteen is not finished.

Spiegel and colleagues ran a randomised two-period crossover in 12 healthy young men — mean age 22, mean BMI 23.6 — with caloric intake and physical activity held constant. Two days of sleep restriction, compared with two days of sleep extension, produced an 18% fall in leptin (p=0.04), a 28% rise in ghrelin (p<0.04), a 24% rise in hunger (p<0.01) and a 23% rise in appetite (p=0.01). The largest increases — 33 to 45% — were for calorie-dense, high-carbohydrate foods.^[5]

The same signal appears at population scale. Taheri and colleagues examined 1,024 people in the Wisconsin Sleep Cohort and found that habitual 5-hour sleepers had a predicted 15.5% lower leptin and 14.9% higher ghrelin than 8-hour sleepers, independent of BMI, with a U-shaped relationship between sleep duration and BMI.^[6]

Two nights. Food and activity unchanged. A 28% rise in the hormone that makes you eat, and a specific pull toward the foods that are easiest to overeat. If you are working on appetite and sleeping six hours, sleep is the first lever, not the last — our guide to sleep optimization covers what actually moves it.

Blom and colleagues fed 15 healthy men two isocaloric breakfasts in a single-blind crossover — one with 58% of energy from protein, one with 19%. The high-protein breakfast suppressed postprandial ghrelin significantly more than the high-carbohydrate breakfast (p<0.01), increased glucagon and cholecystokinin, and slowed gastric emptying.^[8]

The honest caveat sits in the same paper. Appetite ratings did not differ significantly between the two breakfasts, and ad libitum energy intake was not significantly affected in this short study. So protein is a defensible lever with a clear mechanism — it is not a magic switch, and the hormone moving did not translate into measurably less eating over the hours that followed.

Fibre deserves the same scepticism. Oat beta-glucan added to an energy-restricted diet for three months in 66 overweight women did not enhance weight loss or change ghrelin meaningfully.^[12] Twelve weeks of 45 g/day resistant starch in adults with prediabetes had no effect on appetite, food intake, or total or active ghrelin.^[13] A pattern runs through this literature: the acute meal-test studies are considerably more encouraging than the chronic trials, and supplements sold on the strength of the former rarely deliver the latter.

A common worry is that training simply makes you hungrier and cancels itself out. The acute hormonal data do not support that. A meta-analysis of acute exercise in overweight and obese individuals — six studies, 73 participants, mean BMI 30.6 — found that exercise moderately suppressed acylated ghrelin AUC (pooled SMD −0.34, 95% CI −0.53 to −0.15), with a larger effect at higher BMI and only trivial effects on PYY and GLP-1.^[10]

Intensity appears to matter. A 2023 meta-analysis of 12 studies comparing high-intensity interval and sprint interval training with moderate-intensity continuous training found that interval protocols — sprint intervals especially — suppressed acylated ghrelin more at every analysed time point.^[11]

The practical read is modest but useful: exercise does not make you acutely hungrier, and harder, shorter efforts suppress ghrelin more than steady-state work. These are acute effects measured over hours in controlled settings, not proof of long-term appetite control, and they say nothing about what you choose to eat that evening.

Most ghrelin is made in the fundus of the stomach, which is why bariatric surgery changes the picture so sharply. In the 2002 Cummings study, people who had lost 36% of body weight after gastric bypass had a 24-hour ghrelin AUC 77% lower than normal-weight controls and 72% lower than matched obese controls — and the normal meal-related fluctuations and diurnal rhythm were absent entirely.^[4] A 2025 narrative review reports that fasting acyl ghrelin falls markedly after sleeve gastrectomy — more so than after Roux-en-Y gastric bypass — consistent with removal of the ghrelin-producing fundus.^[14]

The newest development is LEAP2, liver-expressed antimicrobial peptide 2: an endogenous antagonist at the ghrelin receptor that blocks ghrelin-induced feeding. It is produced mainly by the liver, is regulated by nutrients and body weight, and tends to be elevated in obesity — the mirror image of ghrelin.^[15][14] I mention it because it is where the interesting work is, not because it is available to anyone: LEAP2 is an active research target, not a treatment, and the human data remain limited.

A word on GLP-1 receptor agonists, since it is the question I am asked most. These drugs reduce appetite primarily through GLP-1 receptor signalling and slowed gastric emptying, rather than through a well-established direct ghrelin-lowering effect, and the ghrelin data in people taking them are inconsistent. If someone tells you semaglutide works by lowering your hunger hormone, they are compressing a messier story than the evidence supports.

Ghrelin is a signal, not an enemy. It is the message that energy is needed, and it is the same message that keeps people with cancer cachexia eating enough to stay alive. You cannot abolish hunger, and a plan built on the assumption that you should is a plan that will fail in a way that feels personal.

What you can do is avoid the modifiable things that amplify it unnecessarily. Sleep enough that you are not paying a 28% ghrelin tax for no return. Lead meals with protein, which suppresses the post-meal rise more than carbohydrate does. Train, including some harder efforts, knowing that exercise does not acutely make you hungrier. Lose weight slowly enough that you can live with the method, and treat maintenance as the real intervention. And be sceptical of anything sold as an appetite fix on the strength of a single acute meal test.

If you want to see where your own hormonal picture sits rather than reason from population averages, the relevant markers — including the hormonal panel we track — are worth reviewing with your own physician alongside your history, not in isolation.