By TwinPare Research | Evidence review: 19 August 2026
Standfirst. Fasting is often described as a single switch: insulin falls, fat burning starts and the body enters a special repair mode. Human biology is less tidy. Different fasting patterns produce different responses, several hormones move in different directions, and the same protocol can feel—and measure—very differently from one person to another.
Quick answer
What does the research show in brief?
When food stops arriving, the body shifts from storing incoming energy toward maintaining blood glucose and mobilising stored fuel. Insulin generally falls, glucagon becomes relatively more important, fat mobilisation increases and ketone production may rise as fasting continues. Appetite, stress and metabolic hormones also change, but not in a universal sequence or by the same amount in everyone.
Key takeaways
- By TwinPare Research | Evidence review: 19 August 2026
- Standfirst. Fasting is often described as a single switch: insulin falls, fat burning starts and the body enters a special repair mode. Human biology is less tidy. Different fasting patterns produce different responses, several hormones move in different directions, and the same protocol can feel—and measure—very differently from one person to another.
- When food stops arriving, the body shifts from storing incoming energy toward maintaining blood glucose and mobilising stored fuel. Insulin generally falls, glucagon becomes relatively more important, fat mobilisation increases and ketone production may rise as fasting continues. Appetite, stress and metabolic hormones also change, but not in a universal sequence or by the same amount in everyone.
Fasting is not one intervention
“Fasting” can mean an overnight gap between dinner and breakfast, a daily eating window such as 16:8, a 24-hour fast, several days without energy intake, or a diet that deliberately mimics some features of fasting. These interventions differ in duration, frequency, hydration, calorie intake, meal timing and total energy balance. Results from one cannot simply be assigned to another.
Time matters, but it is not the only variable. Liver glycogen, recent meals, physical activity, body composition, sleep, circadian timing, medication and metabolic health all influence what happens next. Ketones, for example, can begin to rise during an overnight-to-one-day fast, but there is no single hour at which every person crosses the same metabolic threshold.
The first shift: from incoming fuel to stored fuel
After a meal, insulin helps direct glucose and other nutrients into tissues and storage. As the post-meal period ends, insulin usually declines. The liver initially supports blood glucose partly through glycogen breakdown and increasingly through gluconeogenesis. Lower insulin and changing counter-regulatory signals make stored fat more available; the liver can convert fatty acids into ketone bodies when conditions favour it.
This is a coordinated adjustment, not an on/off transition. A person can be oxidising both carbohydrate and fat. “Fat burning” is not identical to losing body fat over time, because long-term fat loss still depends on cumulative energy balance and adherence.
A hormone orchestra—not a single headline
Insulin and glucagon. The balance moves away from nutrient storage and toward fuel mobilisation. The magnitude depends on the previous meal, activity and insulin sensitivity.
Ghrelin. Often called the hunger hormone, ghrelin follows meal-related and circadian patterns. It does not necessarily rise in a straight line for every hour without food. In the 24-hour human study above, the average change was small while individual responses ranged widely.
Cortisol. Cortisol helps maintain energy availability and follows a strong daily rhythm. Longer or more demanding fasts can act as a physiological stressor, but a single cortisol value is difficult to interpret without timing and context.
FGF21. This liver-linked metabolic signal is associated with energy balance and lipid metabolism. A twin study measured fasting FGF21 in 46 monozygotic and 75 dizygotic pairs and estimated heritability at 40% in that sample. It measured fasting levels, not the response to a fasting intervention.
Growth hormone and IGF-1. Growth hormone can rise markedly during fasting, while IGF-1 may stay unchanged in a short fast or fall during longer energy deprivation. This uncoupling is one reason a rise in growth hormone should not be translated into “more muscle.”
Testosterone and reproductive signalling. Severe energy deprivation can suppress the hypothalamic–pituitary–gonadal axis. In a small 83-hour fasting experiment in healthy young men, luteinising hormone and total and free testosterone fell substantially. This does not tell us that every short eating window lowers testosterone, but it shows that fasting does not increase all supposedly anabolic hormones.
What twin research adds
The CREAT twin protocol is unusually relevant because it includes approximately 18 hours of fasting in monozygotic twins discordant for anorexia nervosa and plans measurements including ghrelin, cortisol, thyroid hormones and metabolic markers. It is a protocol designed around negative energy balance; it should not be presented as proof of published fasting-response heritability.
Other twin studies help with pieces of the puzzle. Fasting FGF21 varies partly with inherited factors in a population, yet liver fat and triglycerides were important acquired correlates. In 16 monozygotic female pairs, twins who reported restrained eating had higher ghrelin during a milkshake-preload experiment than their co-twins despite similar intake. And monozygotic twins discordant for obesity differed in fasting ghrelin. Together, these results show why “same genes” does not mean “same current physiology.” None is a direct trial of matched fasting durations across a large twin cohort.
Is fasting better for weight loss?
Fasting can be a practical structure for some people, but the current evidence does not justify calling it uniquely superior. A 2026 Cochrane review of 22 randomised trials involving 1,995 adults with overweight or obesity found that intermittent fasting may make little or no difference to weight loss compared with regular dietary advice. Evidence for quality of life and adverse events was limited or uncertain. The useful question is therefore often not “Is fasting metabolically magical?” but “Is this a safe, sustainable structure that helps this person meet their goals?”
What this research does—and does not—show
hormone systems.
responses.
components.
changes from hunger, ketones or a wearable.
approaches for weight loss or health.
- It shows that fasting coordinates changes across multiple fuel and
- It shows substantial variation between people in some measured
- It shows that several relevant baseline traits have genetic
- It does not identify one ideal fasting duration for everyone.
- It does not prove that a consumer can infer hormone or autophagy
- It does not establish that fasting is superior to other sustainable
TwinPare Perspective
The most useful insight is not that one hormone rises or falls. It is that a fasting response is a pattern: meal timing, sleep, activity, training, perceived hunger, energy, recovery and body-weight trend interact over days and weeks. Twins make this especially visible. Shared DNA can narrow part of the biological difference, while current behaviours, health and environment can still pull outcomes apart.
From research to self-tracking
If fasting is appropriate for you, track outcomes that a consumer tool can actually observe: eating-window consistency, sleep duration and timing, training quality, perceived energy, hunger, recovery and longer-term weight or waist trends. Change one main variable at a time. Do not treat an app pattern as a hormone test, genetic estimate or diagnosis.
Explore in TwinPare
Use TwinPare Health & Fitness to follow your own patterns over time and, where both people actively choose it, compare routines with a twin or partner. The goal is better questions and more consistent observation—not a claim that the app can measure GH, testosterone, insulin or autophagic flux.
Safety note
Fasting is not appropriate for everyone. Extra caution or clinical guidance is important for people who are pregnant or breastfeeding, have diabetes or use glucose-lowering medication, have a current or past eating disorder, are underweight, are adolescents, are frail or older, or have another condition affected by food or medication timing. Longer water-only fasts should not be treated as a casual self-experiment.
Sources and methodology
Ghrelin/GH/IGF-1 Axis in Healthy Humans. J Clin Endocrinol Metab. 2022;107:e3769–e3780. doi:10.1210/clinem/dgac353.
Influence Circulating FGF21 Levels in Healthy Young Adult Twins. J Clin Endocrinol Metab. 2011;96:E351–E355. doi:10.1210/jc.2010-1326.
in Twins (CREAT): a study protocol. BMC Psychiatry. 2020;20:507. doi:10.1186/s12888-020-02903-7.
plasma ghrelin to a milkshake preload in restrained eaters. Physiol Behav. 2014;129:50–56. doi:10.1016/j.physbeh.2014.02.008.
monozygotic twins discordant for obesity. Metabolism. 2009;58:174–179. doi:10.1016/j.metabol.2008.09.010.
fasting-induced hypogonadotropic hypogonadism in men. J Clin Endocrinol Metab. 1997;82:1543–1548. doi:10.1210/jcem.82.5.3947.
or obesity. Cochrane Database Syst Rev. 2026;2:CD015610. <u>doi:10.1002/14651858.CD015610</u>.pub2.
Editorial method: primary studies, protocols and official health guidance were checked; evidence streams were kept separate. This article is educational and is not medical advice.
- Hollstein T, et al. Effects of Short-term Fasting on
- Tyynismaa H, et al. Liver Fat But Not Other Adiposity Measures
- Seidel M, et al. Comprehensive Risk Evaluation for Anorexia nervosa
- Myhre R, et al. A twin study of differences in the response of
- Leskelä M, et al. Fasting plasma total ghrelin concentrations in
- Aloi JA, et al. Pulsatile intravenous GnRH administration averts
- Alruwaili N, et al. Intermittent fasting for adults with overweight
Source notes
The source has been verified and editorially reviewed for this article. The limitations below show which level of conclusion the sources support.
- [source-1] <u>doi:10.1002/14651858.CD015610</u> Linked publication or source record. doi.org, 2026. Evidence type: Primary or supporting source listed in the article review Limitation: Read the source together with the article’s visible limitations; the link does not support every broader interpretation. Open source
Editorial source review
This section shows how the article's key factual claims are linked to the source.
Phrasings that require caution
- Association is not automatically causation.
- Twin findings describe studied groups and do not predict an individual.
- The article is general research information, not medical advice.
| ID | Claim | Source support | Caution |
|---|---|---|---|
| FASTING-HORMONES-1 | What happens to insulin, GH, IGF-1, ghrelin and testosterone during fasting — and what twin research can and cannot tell us. | 2026 | Interpret the finding within the studies’ population, method, and limitations. |