Gabapentin for hot flashes and sleep

Gabapentin is an inexpensive generic drug, approved for seizures and nerve pain, that also reduces hot flashes and may improve sleep. It is not FDA-approved for menopause, but The Menopause Society’s 2023 position statement rates the evidence Level I and recommends it. Its main appeal is for women whose night sweats wreck their sleep, since the drowsiness it causes can be put to use at bedtime.

What it is and how it works

Gabapentin is structurally related to GABA, the brain’s main calming neurotransmitter. How it reduces hot flashes is not fully understood; it is thought to steady the hypothalamic temperature center that misfires as estrogen falls.

Unlike SSRIs and SNRIs, gabapentin is not a serotonin drug, and the tamoxifen concern the position statement raises for some antidepressants does not apply to it. Unlike fezolinetant, it requires no routine lab monitoring. Its costs are sedation, dizziness, and the need to titrate.

What the evidence shows

900 mg a day versus placebo. The foundational trial randomized 59 postmenopausal women with at least seven hot flashes a day to gabapentin 300 mg three times daily or placebo for 12 weeks. Gabapentin reduced hot flash frequency 45% and the composite score (frequency times severity) 54%, versus 29% and 31% for placebo. Half the gabapentin group reported at least one adverse event versus 28% on placebo, and 13% withdrew for side effects versus 3%. In the open-label extension, women who increased to as much as 2,700 mg a day had 54% and 67% reductions.

2,400 mg a day versus estrogen. A 60-woman trial compared gabapentin titrated to 2,400 mg, conjugated estrogens 0.625 mg, and placebo over 12 weeks. The hot flash composite score fell 71% with gabapentin, 72% with estrogen, and 54% with placebo; gabapentin and estrogen did not differ. The cost was a cluster of headache, dizziness, and disorientation that the authors estimated would affect about one in four women treated at this dose.

Pooled analysis. An individual-patient pooled analysis of three gabapentin trials found hot flash reductions 35% to 38% greater than the corresponding placebo arms at 4 weeks, similar to or better than paroxetine (41%) and venlafaxine (33%) in the same analysis.

Sleep. A secondary analysis of the 900 mg trial found better sleep quality scores on the Pittsburgh Sleep Quality Index at weeks 4 and 12 and better overall PSQI score at week 4, compared with placebo.

How strong is this? The direction of effect is consistent across trials and the position statement rates it Level I. But the trials are small (59 and 60 women), mostly 12 weeks, and the sleep data are secondary. Large trials of the kind run for fezolinetant do not exist for gabapentin.

Dose, how it is taken, time to effect

The position statement’s suggested range is 900 to 2,400 mg a day in divided doses:

• Start with 100 to 300 mg at night (100 mg if you are sensitive to medications).
• Add a second 300 mg at night.
• Then add a separate 300 mg in the morning.
• Continue adjusting with your clinician toward 900 mg, or higher if needed and tolerated.

Because the half-life is short and drowsiness is the main side effect, the statement specifically suggests bedtime dosing for women with sleep disruption from night sweats. Some women take a single bedtime dose only. Side effects of dizziness, unsteadiness, and drowsiness typically appear in the first week, improve in the second, and resolve by week 4. The position statement notes that non-hormonal prescription options typically start working within 2 weeks, and the pooled trials measured benefit at 4 weeks.

Ask your clinician before stopping after regular use, rather than stopping abruptly.

Side effects and who should be cautious

The common side effects are drowsiness, dizziness, unsteadiness, and impaired balance or coordination. At higher doses, headache and disorientation become more frequent.

The position statement notes a warning for all antiepileptic drugs, including gabapentin, about uncommon suicidal thoughts or behaviors; report mood changes.

Be cautious if you drive or operate machinery until you know how it affects you, are at risk of falls, drink alcohol regularly, or take opioids, benzodiazepines, or other sedatives, all of which add to its sedative effect. Pregabalin, its relative, was tested in one hot flash trial (59% to 61% reduction versus 35% for placebo) but is not recommended by the position statement because of side effects and its Schedule V controlled-substance status.

Cost and insurance in the US

Gabapentin is available as a generic, which generally keeps cost low, though the sources reviewed here do not list US prices. Because the hot flash use is off-label, check how your plan handles it. There are no required lab tests.

How it compares with other options

Option	Reduction in hot flashes	Sleep effect	Lab monitoring	Cost
Gabapentin 900 to 2,400 mg	45% vs 29% placebo at 900 mg; matched estrogen at 2,400 mg	Sedating; improved sleep quality scores	None	Generic
SSRIs/SNRIs	About 1 to 2 fewer/day vs placebo	Variable; can cause insomnia	None	Generic
Fezolinetant 45 mg	About 2.5 fewer/day vs placebo	Insomnia reported in 3.9% vs 1.8%	Liver tests	Brand, about $780/month
CBT for insomnia	No change in flash count	70% vs 24% reached no-insomnia range	None	Varies
Low-dose estradiol	52.9% vs 28.6% placebo at 8 weeks	Indirect, via fewer night sweats	None routine	Generic

If your nights are the main problem, read why perimenopause wakes you at 3 am; a combination of bedtime gabapentin and CBT for insomnia addresses both the sweats and the conditioned wakefulness.

Questions for your clinician

• Would a single bedtime dose be enough for my pattern of symptoms, or do I need daytime doses too?
• How fast should I increase the dose, and what is the target?
• What should I avoid combining with it (alcohol, sleep aids, pain medications)?
• If it helps my sleep but not my daytime flashes, what would we add?

Our symptom check can help you log how many flashes happen at night versus during the day before the visit, which makes this decision easier.