Sleep Before IVF: What the Evidence Actually Shows — For Both of You

Evidence-based · Written by Dr. Lejla Fazlicic, D.Ac, L.Ac · All key claims cited to peer-reviewed research

The short answer: Sleep-disordered breathing — snoring and sleep apnea — carries the strongest signal, associated in pooled IVF cohorts with lower clinical pregnancy and live birth. Sleep quality shows a moderate association. Sleep duration in women shows almost none. And no randomized trial has ever tested whether improving sleep improves an IVF outcome.

Two things follow from that. First, the largest cohort to test female sleep duration — nearly 7,000 women trying to conceive — found nothing, while in the same cohort family the male partner's short sleep was associated with reduced fecundability. That is the finding almost nobody mentions. Second, I could not identify an ASRM, NICE or ESHRE guideline recommending a specific sleep target as an evidence-based fertility treatment, because the trial evidence to support one has not been generated. You deserve to know that before anyone sells you a sleep protocol.

Working out which of these actually applies to the two of you is the hard part to do alone.

Book a free 15-minute call · See what working together looks like

If you are reading this after a cycle that did not work, start here: your sleep did not fail your cycle. Nobody has ever demonstrated that it could. If nobody told you any of this before your last retrieval, that is not your failure — many patients are not given this nuance before treatment.

But if you have roughly three months before the next retrieval, sleep is worth a serious look — for reasons more specific and more interesting than "get eight hours." This is useful time to identify persistent sleep difficulty or possible sleep-disordered breathing and discuss it with your medical team. It is not a deadline, a guarantee, or evidence that a fixed sleep-improvement window changes IVF outcomes.

Does sleep affect IVF success?

Let us be precise about what has been measured, because this is a field where the headline and the data have drifted apart.

The most rigorous summary available is a 2026 systematic review and meta-analysis in BMC Pregnancy and Childbirth. It screened in 14 prospective IVF and ICSI cohorts totalling 9,902 women, of which seven contributed data to the pooled analyses (Huang et al., 2026). It found three different things about three different aspects of sleep, and the differences matter:

• Sleep-disordered breathing — the strongest result. Clinical pregnancy odds ratio 0.52, live birth odds ratio 0.47, with essentially no heterogeneity between studies.
• Poor subjective sleep quality (a Pittsburgh Sleep Quality Index score above 5) — clinical pregnancy odds ratio 0.78 under one statistical model, but imprecise and compatible with no association under the more conservative one.
• Sleep duration and timing — a possible U-shape, but the studies disagreed too much to pool at all.

The review authors then did something admirable: they concluded that their own findings "do not support routine targeted sleep screening as a standard component of infertility evaluation, but they indicate that sleep-disordered breathing may be relevant to consider when clinically indicated." That is a research team declining to oversell its own result — and the second half of that sentence is exactly where the useful action sits.

The largest single IVF cohort points the same way. In 3,183 women assessed before embryo transfer, good sleep quality was associated with a live birth rate of 50.5% versus 45.7% for poor sleep — adjusted relative risk 1.12 — while sleep duration showed no significant association with any outcome (Liu et al., 2023). A separate prospective cohort of 1,276 women found fewer mature oocytes, fewer normally fertilized oocytes and fewer good-quality embryos among women reporting trouble falling asleep more than three times a week (Yao et al., 2022) — though sleep in that study was recorded on the day of retrieval, after stimulation, which leaves the causal arrow genuinely uncertain.

The study that gets miscited

You will find "more sleep means more eggs" attributed to a 2017 pilot study that used objective actigraphy in 24 IVF cycles. What that study actually found was an association of about 1.5 additional oocytes per extra hour of sleep at p = 0.09 — not statistically significant. The significant model it is quoted from was driven by AMH and day-3 FSH, not by sleep (Goldstein et al., 2017). It is a reasonable pilot. It is not evidence for a sleep-and-eggs claim, and this is a plausible hypothesis, not demonstrated evidence: the study was too small and the sleep association was not statistically significant.

What about sleep and getting pregnant without IVF?

Here the picture inverts in a way that is genuinely useful, and almost never reported.

PRESTO is a large North American preconception cohort that follows couples actively trying to conceive. Among 6,873 women, fecundability — the per-cycle probability of conception — was flat across every sleep duration category. Under 6 hours: fecundability ratio 0.89. Six hours: 0.95. Seven: 0.99. Nine or more: 0.96. Every estimate crossed the null. The only significant finding was trouble sleeping more than half the time, at 0.87. Shift work showed no association at all (Willis et al., 2019).

In the same cohort family, among 1,176 male partners, men sleeping fewer than 6 hours a night had a fecundability ratio of 0.62 compared with men sleeping 8 hours — and the association held among men not doing night or rotating shifts, and among men with no infertility history (Wise et al., 2018).

Read those two together. In the population where both were measured, his short sleep tracked with reduced conception and hers did not. This is worth sitting with, because the entire cultural conversation about sleep and fertility is addressed to women. Same window. Same clock. Same team — and on this particular input, the evidence leans toward him.

An honest caveat on that finding: the pattern was not tidy. Six hours looked fine (ratio 1.06) while under six looked markedly worse, which is biologically odd and suggests the very-short-sleep group differs in ways nobody measured. And a separate cohort of 1,228 women found no association between preconception sleep and live birth at all (Freeman et al., 2023).

What does sleep actually change in your biology?

This is the part worth understanding, because it explains why sleep keeps appearing in fertility conversations even where the outcome data is thin. Sleep is not a proven fertility intervention. It supports several systems relevant to general reproductive and metabolic health.

1. Insulin sensitivity

In a controlled inpatient study, 20 healthy men restricted to 5 hours a night for a week showed insulin sensitivity reduced by roughly 20% on intravenous glucose tolerance testing and 11% on the euglycemic clamp — the reference standard — with salivary cortisol up 51% (Buxton et al., 2010). Your maturing eggs sit in follicular fluid that reflects your metabolic state, which is why insulin keeps surfacing in the pre-IVF conversation. If insulin is your particular battleground, we go further in our guide to PCOS, inositol, sleep and insulin before IVF.

2. Cortisol recovery

Sleep loss does not raise cortisol during the night. It raises it the following evening — 37% higher after partial sleep deprivation and 45% after total deprivation, with the quiescent period of cortisol secretion delayed by at least an hour (Leproult et al., 1997). The interpretation the authors offer is the useful one: sleep loss degrades the recovery half of the stress cycle rather than the activation half. Recovery is a skill — and skills can be built. We take that apart properly in what nobody can tell you yet about preparing for IVF.

3. Testosterone — with an honest asterisk

One week of sleep restricted to 5 hours a night was associated with a 10–15% decline in daytime testosterone in young healthy men, in a short JAMA research letter (Leproult & Van Cauter, 2011). It is a small, extreme, widely cited study — and the free-living evidence complicates it. A cohort of 796 men found no association between sleep duration and any reproductive hormone (Chen et al., 2016), and a 2025 meta-analysis of night-shift work found testosterone nominally higher in shift workers (Viramgami et al., 2025). The mechanism is plausible. The chain from a 10% testosterone dip to a fertility outcome has not been measured in anyone.

How much sleep should you get before IVF?

The defensible framing is avoid the extremes, not optimize. Some observational studies have associated very short or long sleep with less favorable reproductive outcomes. Seven to eight hours is a reasonable general-health reference range, but fertility-specific evidence does not establish an optimal sleep duration or show that changing your duration improves IVF outcomes.

Does his sleep affect fertility?

In 796 men followed longitudinally, semen volume and total sperm number formed an inverted U with a peak at 7 to 7.5 hours. Above 9 hours, total sperm number was 39% lower; the short-sleep estimates crossed the null. Notably, men who shifted toward 7 to 7.5 hours between study years improved — the closest thing to a causal hint the literature offers (Chen et al., 2016).

The best-designed male study is more equivocal than its own summary suggests. Among 690 men in a population-based preconception cohort with 1,247 semen samples, every point estimate for sleep duration pointed the same way — but every confidence interval crossed the null. Only poor global sleep quality reached significance, and barely, at a 19.2% lower total sperm count. The authors nonetheless summarise their result as supporting a relationship between poor sleep health and semen quality, which is worth knowing before quoting the study in either direction (Coleman et al., 2026).

Sleep factor	Strength of evidence	What was actually found
Sleep-disordered breathing / apnea	Strongest	Pooled from the seven cohorts contributing data to a 14-cohort review of 9,902 women: clinical pregnancy OR 0.52, live birth OR 0.47, no heterogeneity
Sleep quality / trouble falling asleep	Moderate, repeatable	Live birth RR 1.12 for good vs poor sleep quality; fewer mature oocytes and good-quality embryos with frequent difficulty falling asleep
His sleep duration	Moderate, one strong cohort	Under 6 hours: fecundability ratio 0.62 vs 8 hours
Her sleep duration	Weak to absent	Flat across all categories in 6,873 women; no live birth association in 1,228 women
Shift work	Narrow	Across 16 cohorts and 123,403 women (Stocker et al., 2014): menstrual disruption adjusted OR 1.15; night-shift early pregnancy loss adjusted OR 1.41; the infertility signal disappeared after adjustment
Melatonin supplements	Low certainty, no live birth benefit	Live birth RR 1.23 (not significant); pregnancy effect vanished in every subgroup

Does sleep apnea affect IVF success?

This is the one part of the sleep literature with a large, consistent signal — and if there is one thing in this article to act on, it is not a bedtime. It is breathing.

Sleep-disordered breathing produced the only large, consistent, low-heterogeneity signal in the entire literature. And it is common in exactly the population reading this: in a prospective cohort of 156 women with PCOS undergoing their first IVF cycle, 37.2% screened positive for symptoms consistent with obstructive sleep apnea. Those who screened positive had lower AMH, lower antral follicle counts, needed more gonadotropin and retrieved fewer oocytes; among those who reached transfer, the clinical pregnancy rate was 42.3% versus 60.2% (Zhang et al., 2024). If PCOS is part of your picture, this belongs alongside the rest of the metabolic conversation — see know your PCOS phenotype before IVF.

Two honest limits. That study is PCOS-only, so obesity and insulin resistance are tangled into it. And no trial has shown that treating apnea improves an IVF outcome. But snoring, waking unrefreshed, witnessed pauses in breathing, or morning headaches are things worth mentioning to your own physician regardless of fertility — and this is a testable, treatable condition rather than a habit to feel guilty about. That conversation belongs with your medical team, not with an article.

Does melatonin help?

Melatonin is genuinely present in follicular fluid and behaves as an antioxidant there; intrafollicular oxidative damage markers run inversely to melatonin concentrations (Tamura et al., 2008). That is real biology and it is why the supplement is marketed hard.

The trial evidence does not follow. A 2025 meta-analysis of 11 trials found clinical pregnancy relative risk 1.24 — but live birth relative risk 1.23 with a confidence interval from 0.85 to 1.80, which is not significant, and the pregnancy effect disappeared in every single subgroup analysis (Tang et al., 2025). An earlier IVF-specific review of 7 randomized trials found only one improved outcome — mature oocyte count — with two of the seven trials at high risk of bias (Mejlhede et al., 2021). The Cochrane review covering antioxidants in female subfertility grades the whole class low to very low certainty, and carries an editorial note that seven of its included studies have since been retracted (Showell et al., 2020).

None of that makes melatonin harmful. It makes it a question for your clinician rather than a purchase.

What the evidence does not show

Modify the modifiable. Measure the measurable. Part of that discipline is naming what is not measurable yet:

• No randomized trial has tested a sleep intervention with a fertility endpoint. Not one. Every association above is observational.
• Sleep has not been shown to change egg quality or sperm DNA integrity. The study most often cited for the sperm DNA claim found no association with DNA fragmentation index or comet assay; its only significant result was on a different measure and ran in an unexpected direction at both sleep extremes (Wang et al., 2018). What does have strong evidence behind it for fragmentation is covered in lowering sperm DNA fragmentation.
• Reverse causation is barely addressed. Anxiety about a cycle disturbs sleep. A poor response disturbs sleep. Untangling that from sleep disturbing the cycle has essentially not been done.
• One study tested a commonly assumed explanation and did not find support for it. The single study that formally tested whether perceived stress explained the observed sleep–outcome association did not find evidence that it did (Bariya et al., 2025).
• The largest US cohort of infertile women found nothing. Across 1,603 women, short sleep and clinical symptoms of apnea were not associated with conception or live birth (Eisenberg et al., 2021).
• Blue light and screens have essentially no human ovarian data. The handful of human studies on light at night rely on satellite-measured outdoor brightness at a residential address — a proxy for living in a city, not for light reaching anyone's eyes — and none measures a fertility outcome.
• I could not identify an ASRM, NICE, or ESHRE guideline that recommends a specific sleep target as an evidence-based fertility treatment. The trial evidence needed to support one has not been generated.

What we do with this

Sleep in the pre-IVF window is worth attention, and it is worth attention for reasons that are honest rather than inflated. A reasonable, evidence-anchored position looks like this:

1. Discuss possible sleep-disordered breathing with your physician if there is loud habitual snoring, unrefreshing sleep, witnessed breathing pauses, morning headaches, or substantial daytime sleepiness. It was the most consistent signal in the available IVF/ICSI cohort evidence, but no trial has shown that treating it improves IVF outcomes.
2. Target sleep difficulty, not bedtime arithmetic. Across the most independent datasets, quality and trouble falling asleep survive; hours mostly do not.
3. Use roughly 7 to 8 hours as a practical general-health reference if it fits your life, without treating it as an IVF prescription. The fertility data do not establish a precise optimum, and they do not show that moving into any particular range improves outcomes.
4. Include him deliberately. His sleep is not the afterthought here. In the one cohort that measured both, it was the half that showed a signal.
5. Do not add melatonin on the strength of a mechanism. Ask your clinician, and know that live birth has not moved in the pooled trials.

And if your cycle is closer than three and a half months away, use the time you have. Sleep habits may shift quickly, though evaluation and treatment of persistent insomnia or sleep-disordered breathing can take longer. No window, used or unused, has ever been shown to decide an outcome on its own.

Related reading

• Egg quality in the 90 days before retrieval
• Lowering sperm DNA fragmentation in 74 days
• PCOS, inositol, sleep and insulin before IVF
• Exercise in the 90 days before IVF
• What nobody can tell you yet about preparing for IVF
• The research library

Frequently asked questions

Does sleep affect IVF success?

Partly, and less than most articles claim. In a 2026 systematic review that screened in 14 prospective cohorts totalling 9,902 women, sleep-disordered breathing was associated with substantially lower clinical pregnancy and live birth odds, and poor sleep quality with a modest reduction. Sleep duration showed no consistent association. No randomized trial has tested whether improving sleep improves an IVF outcome, so every finding here is an association rather than proof of cause.

How many hours of sleep should I get before IVF?

Roughly 7 to 8 hours is where the curves peak, and the honest framing is to avoid both extremes rather than to optimize. Nine or more hours has been associated with fewer oocytes and lower clinical pregnancy odds in IVF cohorts, and with lower total sperm number in men; under 6 hours has been associated with reduced fecundability in male partners. In women trying to conceive, the largest cohort — 6,873 participants — found no association between sleep duration and per-cycle conception at any level.

Does my partner's sleep matter for IVF?

The best evidence concerns natural conception rather than IVF, and it is the finding most often missed. In a North American preconception cohort, men sleeping under 6 hours a night had a fecundability ratio of 0.62 compared with men sleeping 8 hours, and the association held among men not working night shifts. In the same cohort family, female sleep duration showed no association. No study has tested male sleep against an IVF outcome. Sperm are produced over roughly 74 days (isotope studies measure the core production cycle itself at about 64, before transit), so his preparation window and hers overlap almost completely.

Should I take melatonin before IVF?

That is a question for your clinician, and the trial evidence is weaker than the marketing. A 2025 meta-analysis of 11 trials found no significant improvement in live birth, and the clinical pregnancy benefit disappeared in every subgroup analysis. An IVF-specific review of 7 randomized trials found only mature oocyte count improved. The Cochrane review of antioxidants in female subfertility grades this class of evidence low to very low certainty.

Can poor sleep damage sperm DNA?

This claim outruns its evidence. The study usually cited for it measured sperm chromatin in 796 young men and found no association between sleep duration and DNA fragmentation index or any comet assay parameter. Its only significant result concerned a different measure and moved in an unexpected direction at both short and long sleep. Smoking, by contrast, has a well-documented association with sperm DNA fragmentation.

Does shift work reduce fertility?

Less than commonly stated. A meta-analysis of 16 cohorts and 123,403 women found menstrual disruption with an adjusted odds ratio of 1.15 and night-shift early pregnancy loss with an adjusted odds ratio of 1.41, but the infertility association disappeared after adjustment for confounders. The authors concluded there is insufficient evidence to advise restricting shift work. In men, a 2025 meta-analysis found no significant association between night-shift work and sperm count or testosterone.

---

This article is educational and reflects associations reported in the scientific literature. It is not medical advice, not a diagnosis, and not a way to interpret your own results. As a Doctor of Acupuncture I offer lifestyle guidance that works alongside your physician and fertility clinic — I do not diagnose, interpret labs, or prescribe. Always make decisions about testing and treatment with your medical team.