GLP-1 and Sleep: What the Labels and the Studies Actually Say

Sleep is one of the few things here you can change. One GLP-1 — liraglutide — lists insomnia as a common side effect, mainly in the first three months. The others do not list it. No study proves cause, so the usable half is the routine: a fixed wake time, a wind-down hour, an earlier caffeine cut-off.

An evidence article, not medical advice. For your dose, your side effects or a sleep problem that will not shift, ask your prescriber.

By Matt Cole. Editor: Sezen Soykut.

---

Key takeaways

• Insomnia is listed on one GLP-1 label — liraglutide — at up to 1 in 10 people, mainly in the first three months. Not for semaglutide or tirzepatide.
• Tiredness is listed for all of them: 5–7% against 3% on placebo in the Zepbound trials.
• Two small studies found sleep improved on treatment, but neither had a comparison group. The behavioural levers — wake time, wind-down, caffeine timing — are the documented part, and yours.

---

What the labels actually say

The product information is the only place anyone has counted. Liraglutide is the exception. Saxenda's European product information lists insomnia under psychiatric disorders at the frequency "common", defined there as up to 1 in 10 people; the leaflet adds it "usually occurs during the first 3 months of treatment" [1]. The frequencies come from five placebo-controlled trials in 5,813 adults [1].

For the weekly medicines the word is not there. Insomnia, somnolence and sleep disorder appear nowhere in the adverse-reaction text for Mounjaro, Wegovy, Ozempic or Zepbound. Absence from a label is not proof it does not happen — only that it was not seen often enough in the trials to be listed.

Tiredness is listed for all of them, and the small print is the story. Mounjaro's table puts fatigue in the "very common" column, but footnote 8 says that frequency came from a cardiovascular-outcomes trial in type 2 diabetes; in the weight-management trials it was "common", up to 1 in 10 [2]. If you take it for weight, the footnote describes you and the column does not. Wegovy lists fatigue as very common, "Mainly seen in the dose-escalation period" [3]; Ozempic, the same molecule at diabetes doses, sits a tier lower [4]. The US Zepbound label prints numbers instead: 5–7% across its three doses, against 3% among the 958 people on placebo [5]. Neither label says the effect ends — both say where it was mainly seen.

Our titration guide and what the first month looks like cover those weeks.

What the studies show, and what they cannot

In a prospective study of 78 people with obesity in Turkey, median Pittsburgh Sleep Quality Index scores fell from 6.0 to 3.0 over three months, and the share above the poor-sleep threshold fell from 55.1% to 16.7% [6]. The improvement did not track how much weight people lost, and did not differ by medicine [6]. The authors chose their words: treatment "was followed by" it [6]. No control group, no randomisation.

In 164 people on tirzepatide in Greece, followed for a year, overnight polysomnography showed sleep-disordered breathing improve, the apnoea–hypopnoea index falling by 6.64 events an hour [7] — breathing, not sleep quality, not insomnia. Those authors note "the absence of a control group limits causal inference" [7].

Nothing here establishes that a GLP-1 causes or cures a sleep problem. The studies are small, most have no comparison group, and they describe groups, not you.

The evidence runs the other way

The strongest study here is about sleep changing appetite, not medicine changing sleep. In a randomised trial of 80 adults aged 21 to 40 with overweight who habitually slept under 6.5 hours a night, one sleep-counselling session added about 1.2 hours a night and cut objectively measured daily energy intake by about 270 calories against the control group [8]. Two limits travel with that number: nobody in it was on a GLP-1, and people with insomnia or sleep apnoea were excluded. It does not answer a reader asking about their insomnia.

What you can change this week

No asterisks here. The NHS insomnia guidance is specific, free and loggable [9].

• Keep the wake time fixed — "wake up and get out of bed at the same time every day", and "do not sleep in after a bad night's sleep" [9]. A lie-in borrows from tomorrow night.
• Take the hour before bed: "relax at least 1 hour before bed – for example, take a bath or read a book", and no television or phone right before bed [9].
• Move the caffeine earlier. The NHS line is nothing within six hours of bed [9]; a meta-analysis of 24 studies puts coffee at "at least 8.8 h prior to bedtime" and found caffeine cut total sleep time by 45 minutes [10]. For most people, lunch is the last cup.
• Only go to bed when you are sleepy, in a room kept "dark and quiet" [9]. Skip the daytime nap, exercise during the day rather than within four hours of bed, and skip the big late meal [9].

Pick two; all four at once is how a routine collapses by Thursday, and the habit readiness assessment tells you which you can hold. Then log it — hours slept, wake time, last caffeine, whether it was a dose week. Four weeks of that is a pattern you can show someone.

The part that is not yours to manage

Sleep apnoea is a diagnosis a clinician makes. If you snore heavily, stop breathing in your sleep, or are very sleepy in the day, that is a conversation with your doctor. And if your sleep changed when your dose did, when and how you take your medicine is a question for whoever prescribes it.

---

The Gila pilot is free and currently on iPhone: sign up here. Our weekly newsletter is here.

---

Sources

1. European Medicines Agency, Saxenda (liraglutide) — Summary of Product Characteristics, Annex I §4.8 and package leaflet. https://www.ema.europa.eu/en/documents/product-information/saxenda-epar-product-information_en.pdf
2. European Medicines Agency, Mounjaro (tirzepatide) — Summary of Product Characteristics, Annex I §4.8 (Table 1 and footnote 8). https://www.ema.europa.eu/en/documents/product-information/mounjaro-epar-product-information_en.pdf
3. European Medicines Agency, Wegovy (semaglutide) — Summary of Product Characteristics, Annex I §4.8, Table 3. https://www.ema.europa.eu/en/documents/product-information/wegovy-epar-product-information_en.pdf
4. European Medicines Agency, Ozempic (semaglutide) — Summary of Product Characteristics, Annex I §4.8, Table 1. https://www.ema.europa.eu/en/documents/product-information/ozempic-epar-product-information_en.pdf
5. US FDA / Eli Lilly, ZEPBOUND (tirzepatide) injection — Prescribing Information, §6.1 "Clinical Trials Experience", via DailyMed (NIH). https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=487cd7e7-434c-4925-99fa-aa80b1cc776b
6. Kır Y, Haymana C, Toros B, et al. "Prospective changes in sleep quality and depressive symptom burden following initiation of GLP-1-based therapies in patients with obesity: a real-world study." Journal of Endocrinological Investigation, published online 18 August 2026. https://doi.org/10.1007/s40618-026-03024-w
7. Adamidis N, Margariti T, Georgakopoulou VE, et al. "Effects of Tirzepatide on Body Composition, Metabolic Parameters, and Sleep Outcomes: A Real-World One-Year Prospective Study." Cureus 2026. https://doi.org/10.7759/cureus.107081
8. Tasali E, Wroblewski K, Kahn E, Kilkus J, Schoeller DA. "Effect of Sleep Extension on Objectively Assessed Energy Intake Among Adults With Overweight in Real-life Settings: A Randomized Clinical Trial." JAMA Internal Medicine 2022;182(4):365–374. https://doi.org/10.1001/jamainternmed.2021.8098
9. NHS, Insomnia. Page last reviewed 19 March 2024. https://www.nhs.uk/conditions/insomnia/
10. Gardiner C, Weakley J, Burke LM, et al. "The effect of caffeine on subsequent sleep: A systematic review and meta-analysis." Sleep Medicine Reviews 2023;69:101764. https://doi.org/10.1016/j.smrv.2023.101764