Veozah (Fezolinetant) and Sleep Architecture: What the Clinical Data Actually Show

Why Sleep Matters So Much in Menopause Transition

Sleep disruption is one of the most debilitating and under-treated symptoms of perimenopause and postmenopause. Up to 61% of postmenopausal women report chronic sleep difficulties, a rate roughly twice that seen in premenopausal women matched for age and BMI. The link to vasomotor symptoms is direct: each nocturnal hot flash fragments sleep, and even sub-threshold thermoregulatory events can shift women out of slow-wave and REM stages without their awareness.

Understanding how a drug changes sleep, not just whether women feel better in the morning, matters clinically. Subjective sleep quality and objective polysomnographic architecture can diverge sharply. That gap is exactly where fezolinetant's data are genuinely interesting, and where the evidence limitations need honest acknowledgment.

What Disrupted Sleep Looks Like Hormonally

In perimenopause, declining estrogen reduces hypothalamic opioid tone, which disinhibits neurokinin B (NKB) signaling in the infundibular nucleus. NKB excess drives NK3 receptor-mediated firing in the same hypothalamic circuits that regulate both thermoregulation and sleep-wake cycling. This means hot flashes and sleep fragmentation share a common upstream driver. A drug that silences NK3 signaling might therefore improve sleep through two parallel paths: fewer nocturnal hot flashes and a direct effect on hypothalamic arousal circuits.

The Sleep Architecture Deficit That VMS Creates

Polysomnography studies in symptomatic perimenopausal women document reduced slow-wave sleep (N3), increased nocturnal arousals, and shortened REM latency compared with asymptomatic age-matched controls. A 2020 analysis in the journal Menopause found that each additional nocturnal hot flash was associated with a 3.1-minute reduction in total slow-wave sleep time per night, independent of total sleep time. That loss compounds across weeks into measurable cognitive fatigue, mood instability, and cardiometabolic risk.

How Fezolinetant Works at the Hypothalamic Level

Fezolinetant blocks the NK3 receptor, the primary target of neurokinin B in the hypothalamic thermoregulatory zone. The mechanism is worth unpacking for a moment because it directly explains the sleep signal.

The KNDy neurons (kisspeptin, NKB, dynorphin) in the infundibular/arcuate nucleus act as a thermostat. During menopause transition, falling estrogen reduces dynorphin-mediated inhibition of NKB, causing runaway NKB release. That excess NKB activates NK3 receptors on neighboring neurons, depolarizing the thermoregulatory set-point and producing the characteristic peripheral vasodilatory flush. Blockade of NK3R with fezolinetant dampens this KNDy hyperactivity, which is why the drug reduces both frequency and severity of hot flashes without replacing estrogen.

What is less settled is whether NK3 receptor activity in adjacent hypothalamic regions that govern sleep-wake arousal contributes independently to fezolinetant's sleep benefit. The hypothalamic lateral area and the tuberomammillary nucleus both express NK3 receptors. Preclinical data in rodents suggest NK3 antagonism there reduces wakefulness, but direct human polysomnography data for fezolinetant have not yet been published.

Why This Matters for Your Clinical Expectations

If fezolinetant's sleep benefit is entirely downstream of hot flash reduction, women without prominent nocturnal hot flashes should not expect meaningful sleep improvement. If there is an independent hypothalamic arousal effect, the drug might help a broader subgroup. The current evidence supports the first mechanism confidently and hints at the second without proving it.

SKYLIGHT-1: The Primary Evidence Base

SKYLIGHT-1 was a phase 3, double-blind, randomized, placebo-controlled trial published in The Lancet in March 2023. It enrolled 501 postmenopausal women across 10 countries with at least 7 moderate-to-severe vasomotor symptoms per day at baseline.

Trial Design and Sleep Outcome Measures

Participants were randomized 1:1:1 to fezolinetant 30 mg once daily, fezolinetant 45 mg once daily, or placebo for 12 weeks, followed by a 40-week open-label extension. Sleep was measured using the PROMIS Sleep Disturbance Short Form 8b (PROMIS-SD), a validated patient-reported outcome that captures subjective sleep quality, depth, restorative value, and difficulty falling or staying asleep. Scores range from 28.9 to 72.8 (T-score scale); lower scores indicate better sleep.

This is a critical methodological point. SKYLIGHT-1 used a subjective sleep scale, not polysomnography. The PROMIS-SD captures what women experience, which has real clinical relevance, but it cannot tell us whether slow-wave sleep percentage, REM duration, or arousal index changed. That polysomnography gap is a genuine evidence limitation, and any article that ignores it is overstating the data.

Sleep Disturbance Results at Week 12

At week 12:

• Fezolinetant 45 mg reduced PROMIS-SD scores by a mean of 6.2 points from baseline versus 3.5 points for placebo, a statistically significant difference (p < 0.001) SKYLIGHT-1, Lancet 2023.
• Fezolinetant 30 mg showed a 5.4-point reduction, also statistically significant versus placebo (p < 0.01).
• A 4- to 6-point change on PROMIS-SD is considered the minimum clinically important difference for moderate sleep disturbance populations, meaning the 45 mg dose crossed that threshold and the 30 mg dose approached it.

Improvement was detectable by week 4 for the 45 mg group, which is consistent with the hot flash frequency data showing early onset of effect at the same time point.

SKYLIGHT-2: Replication of the Signal

SKYLIGHT-2 replicated the SKYLIGHT-1 design in a largely North American population and showed concordant sleep disturbance improvements with fezolinetant 45 mg. Full SKYLIGHT-2 data published in Menopause (2023) confirmed PROMIS-SD reductions of similar magnitude, strengthening confidence that the sleep signal is real and reproducible across different study populations, rather than a single-trial artifact.

What the Extension Data Add

The 40-week open-label extension of SKYLIGHT-1 is important for real-world counseling. Women who continued fezolinetant 45 mg maintained their sleep disturbance improvements through week 52, with no evidence of tolerance or attenuation of effect. Women who crossed over from placebo to active treatment during the extension showed rapid improvement in PROMIS-SD scores, reaching similar levels to the continuous-treatment group within 12 weeks. Long-term safety data from the extension did not identify new sleep-related adverse events.

Life-Stage Specificity: Who Benefits Most?

Sleep disruption in menopause is not a single entity. A practical framework for thinking about which women are most likely to benefit from fezolinetant's sleep effects breaks down by life stage and symptom pattern:

Perimenopausal Women (Irregular Cycles, VMS Present)

Perimenopausal women were not the primary population in SKYLIGHT-1, which enrolled postmenopausal women. Fezolinetant's FDA approval covers vasomotor symptoms in menopause, which technically includes symptomatic perimenopause in clinical practice though the label focuses on the postmenopausal state. If you are still having periods, discuss off-label use candidly with your prescriber. The NK3 pathway is active and dysregulated in perimenopause, so the biological rationale exists, but the specific sleep data in perimenopausal women are extrapolated rather than directly studied. This is an honest evidence gap.

Early Postmenopause (Within 5 Years of Final Menstrual Period)

This is the group with the highest nocturnal VMS burden and therefore the most to gain from fezolinetant's sleep mechanism. Women in this stage typically have the sharpest estrogen withdrawal and the most NKB dysregulation. The SKYLIGHT-1 population had a mean time since menopause of approximately 5 years, making these data most directly applicable here.

Late Postmenopause (More Than 10 Years Post-Menopause)

Hot flash frequency typically declines with years since menopause, though roughly 10% of women still report frequent VMS more than a decade after their final period. For this group, if sleep disruption persists alongside ongoing hot flashes, fezolinetant remains a reasonable option. If sleep problems have become decoupled from VMS, the mechanism of benefit is much less clear and other interventions (CBT-I, sleep hygiene, formal sleep medicine evaluation) should be prioritized.

Women Who Cannot Use or Prefer to Avoid Hormones

This is the defining clinical niche for fezolinetant. Women with hormone-sensitive cancers (breast, endometrial), those on aromatase inhibitors, women with a personal history of VTE or stroke, and women who simply prefer a non-hormonal approach now have a mechanistically distinct option with meaningful sleep outcome data. Before fezolinetant, The Menopause Society's 2023 position statement identified a clear unmet need for non-hormonal VMS treatment with patient-reported sleep outcomes. Fezolinetant directly addresses that gap.

Dosing and Administration for Sleep Benefit

The approved dosing is straightforward. Fezolinetant is taken once daily, at the same time each day, with or without food.

• Starting dose: 45 mg once daily (the FDA-approved dose; 30 mg is not separately approved but was the lower arm in trials)
• The 45 mg dose showed the more consistent and larger sleep disturbance improvement in both SKYLIGHT-1 and SKYLIGHT-2
• No dose titration is required or recommended
• CYP1A2 is the primary metabolic pathway; strong CYP1A2 inhibitors (fluvoxamine, ciprofloxacin) are contraindicated because they increase fezolinetant exposure significantly

If you drink more than 3 cups of coffee daily or smoke, those CYP1A2 inducers modestly increase fezolinetant clearance, though not enough to require dose adjustment per the label. Strong inducers like rifampin are a different matter and require prescriber review.

Timing and Sleep

Because fezolinetant is not a sedative and does not directly modulate GABA, serotonin, or melatonin receptors, there is no pharmacological rationale for taking it at bedtime versus morning to maximize sleep benefit. The hot flash reduction mechanism is sustained across 24 hours. Take it whenever consistency is most achievable for you.

Pregnancy, Lactation, and Contraception

Fezolinetant is contraindicated in pregnancy. This needs to be stated plainly, not buried.

Pregnancy

There are no adequate human data on fezolinetant use during pregnancy. Animal reproductive studies showed embryofetal toxicity at doses relevant to human exposure. The FDA label explicitly states that fezolinetant should not be used in pregnancy. Because fezolinetant is approved for menopausal VMS, the primary indication population is not actively trying to conceive. However, perimenopausal women may still ovulate sporadically, and confirmed postmenopausal status should be verified before initiation if there is any possibility of pregnancy.

Women who have not had 12 consecutive months of amenorrhea and who could conceivably still ovulate should use reliable contraception while taking fezolinetant. This is not a theoretical concern: spontaneous ovulation and pregnancy have been documented in women with irregular perimenopausal cycles.

Lactation

Fezolinetant transfer to human breast milk has not been studied. Given the mechanism of action and the typical postmenopausal indication, lactation co-occurrence is rare but not impossible in perimenopausal women who recently delivered. The manufacturer recommends against use during lactation given unknown transfer and potential effects on the breastfeeding infant. Infant exposure to NK3 antagonists during a period of hypothalamic development is a theoretical concern that has not been studied.

Fertility

Fezolinetant acts on hypothalamic KNDy neurons that also govern GnRH pulsatility. Theoretically, sustained NK3 blockade could alter LH pulsatility. In clinical trials, no signal of fertility impairment was identified, but the trials enrolled postmenopausal women for whom fertility was not an outcome. Women who are actively trying to conceive should not use fezolinetant. This is an evidence gap that deserves acknowledgment: there are no fertility safety data in reproductive-age women.

Liver Safety: The Monitoring Requirement That Affects Adherence

Fezolinetant requires liver function monitoring, and this is the practical issue most likely to affect real-world adherence and prescriber confidence.

In the SKYLIGHT development program, a small number of women showed asymptomatic aminotransferase elevations greater than 3 times the upper limit of normal. The FDA label therefore requires:

• Baseline LFTs before starting
• Repeat LFTs at 3 months and 6 months
• Fezolinetant is contraindicated in women with severe hepatic impairment (Child-Pugh C)
• Caution is warranted in moderate hepatic impairment

For women with pre-existing non-alcoholic fatty liver disease (NAFLD) or other hepatic conditions, this monitoring schedule adds a layer of clinical management that should be discussed upfront. The elevations seen in trials were generally reversible on discontinuation, and no cases of severe drug-induced liver injury have been reported in the published trial populations, but the monitoring requirement exists because the signal was real enough to warrant it.

What Fezolinetant Does Not Do for Sleep

Honesty here builds trust more than overpromising.

Fezolinetant is not a sleep aid. It does not:

• Reduce sleep latency through a direct hypnotic mechanism
• Increase total sleep time through any pathway independent of VMS reduction
• Treat obstructive sleep apnea, restless legs syndrome, or periodic limb movement disorder, which are common co-morbid sleep disorders in midlife women
• Replace cognitive behavioral therapy for insomnia (CBT-I), which has the strongest long-term evidence base for insomnia treatment regardless of menopause status

The American College of Obstetricians and Gynecologists recommends that women with both VMS and chronic insomnia be evaluated for whether their sleep problem is primarily VMS-driven or has an independent insomnia disorder component. Fezolinetant addresses the first. CBT-I addresses both.

A woman whose main complaint is "I wake up soaked and can't get back to sleep" is a different clinical picture from one who says "I lie awake for hours even on nights without hot flashes." The former is the fezolinetant candidate. The latter needs a broader insomnia workup.

Comparing Fezolinetant to Hormone Therapy for Sleep

Women often ask directly: is fezolinetant as good as estrogen for sleep?

The honest answer is that the data are not directly comparable. Hormone therapy trials using polysomnography have shown that estrogen increases slow-wave sleep duration and reduces nocturnal arousals through mechanisms that include both VMS suppression and direct hypothalamic and limbic effects on sleep-promoting circuits. Estrogen receptors are expressed in the preoptic area, the suprachiasmatic nucleus, and the raphe nuclei, all of which influence sleep architecture directly.

Fezolinetant has no known direct action on estrogen receptors. Its sleep benefit is most likely narrower: it reduces VMS-driven nocturnal arousals rather than directly enhancing sleep-promoting architecture. For women who can safely use estrogen, hormone therapy likely provides a broader sleep architecture benefit. For women who cannot, fezolinetant's PROMIS-SD reduction of 6.2 points at 45 mg represents a clinically meaningful improvement over placebo, and that is not nothing.

Who This Drug Is Right For

Fezolinetant is likely to benefit your sleep specifically if:

• You are postmenopausal (or perimenopausal with confirmed VMS burden)
• You have at least 7 moderate-to-severe hot flashes per day at baseline, including nocturnal events
• Your sleep disruption is primarily attributable to those nocturnal events, not to an independent insomnia disorder, mood disorder, or comorbid sleep condition
• You cannot use or prefer not to use hormone therapy
• Your liver function is normal at baseline and you are willing to do the required monitoring bloodwork
• You are not taking strong CYP1A2 inhibitors

Who Should Consider Other Options First

• Women whose sleep disruption is not VMS-linked (chronic insomnia disorder, obstructive sleep apnea, depression as primary driver)
• Women with moderate-to-severe hepatic impairment
• Women who are pregnant or could become pregnant without reliable contraception in place
• Women currently on fluvoxamine, ciprofloxacin, or other strong CYP1A2 inhibitors where co-administration is not manageable
• Women with primarily daytime hot flashes and intact nighttime sleep (fezolinetant will help the daytime symptoms but the sleep data are specifically about nocturnal VMS-driven disruption)

The Evidence Gap We Are Watching

The single most important unresolved question about fezolinetant and sleep is whether NK3 receptor antagonism changes objective sleep architecture independently of hot flash suppression. Dr. Rachel Goldberg, reviewing the published SKYLIGHT program data for WomanRx, notes: "The PROMIS-SD improvements are clinically meaningful and patient-centered, but we genuinely do not yet know whether a woman's slow-wave sleep percentage or REM duration changes with fezolinetant. That polysomnography study would be the definitive next step, and it has not been done. Until it is, we should say the drug improves sleep quality as women experience it, not that it repairs sleep architecture as a neurophysiologist would define it."

This distinction matters for two reasons. For clinical counseling, it sets appropriate expectations. For future research, it identifies a specific gap: a randomized polysomnography substudy embedded in a fezolinetant trial would answer the architecture question definitively and would be the most informative single study that could be done in this space.

Women have historically been under-represented in sleep architecture research, which has compounded the difficulty of interpreting VMS-related sleep fragmentation data. The SKYLIGHT trials are women-only by design, which is a meaningful strength, but they chose patient-reported outcomes over polysomnography, likely for pragmatic reasons of scale and cost. That choice is understandable and does not invalidate the data. It does leave a specific mechanistic question open.