Waking Up at 3am in Perimenopause: Two Sleeps, One Diet

Movement and time outdoors: not diet, but just as non-negotiable.

Same Diet, Two Different Kinds of Sleep

For five years, I ate strict keto — very low carb, no fruit, most of my calories from animal protein and fat. And for most of that time, I slept less than I used to. Most nights I got only six hours of sleep and it never bothered me. I woke up clear-headed, moved through my day without a hint of afternoon fog, and never once thought of it as a problem.

Then perimenopause arrived, and within what felt like a few months, sleep stopped being something I could take for granted. It wasn't a gradual shift — it was a different experience altogether. Waking up at 3am in perimenopause became the pattern: roughly every hour through the night, then a hard, final wake-up by 3:30 or 4am, wide awake, unable to fall back asleep no matter what I did

Something had changed. And it wasn't the diet — I hadn't touched what I was eating in years. So what was it?

Ruling Out the Obvious Suspects waking up at 3 am perimenopause

Ask anyone about perimenopausal sleep problems and you'll hear the same two explanations on repeat: hot flashes, or sleep apnea.

I have neither.

Zero hot flashes. Zero night sweats. Of the whole roster of classic menopausal symptoms, the only one I have is irregular periods — no vasomotor symptoms at all, at any time of day or night.

Sleep apnea took more checking. My Apple Watch has been tracking breathing disturbances and it has never flagged anything. I want to be precise about what that actually means, rather than overstate it: It isn't built to catch mild cases — it's a screening signal, not a diagnosis. But clean data from a tool with that kind of specificity, combined with none of the usual apnea risk factors, is real supporting evidence, even if it isn't certainty.

So: no hot flashes, no apnea red flags. The two textbook explanations didn't fit. Something else was going on.

Why 3:30am, Specifically

The clue turned out to be timing. My hardest, most consistent wake-up wasn't randomly scattered through the night — it clustered right around 3:30-4am, and it was a hard stop. Once awake at that hour, I couldn't get back to sleep.

That window matters. Cortisol — the hormone that mobilizes energy and, not coincidentally, helps you wake up — naturally starts rising in the 2-4am range as part of the normal circadian rhythm. That's not a malfunction; it happens every night, in everyone.

What's supposed to happen alongside it: your liver holds a reserve of glycogen — stored glucose, typically 80-120g on a normal diet — that it draws on smoothly overnight to keep blood sugar steady while you're not eating. When cortisol rises in the early morning, a well-stocked liver just releases a bit more glucose to match it, and you sleep straight through.

Strict keto keeps that glycogen reserve chronically low. So when cortisol does its normal early-morning rise, the liver can't lean on a smooth glycogen release — it falls back on gluconeogenesis, manufacturing glucose from amino acids and glycerol instead. That pathway is less smooth and more sensitive to cortisol's dose, producing a more volatile blood glucose swing — one that's been proposed as enough to pull the body out of sleep entirely.

That's a plausible, specific explanation for the hard 3:30-4am stop — not just "perimenopause causes insomnia," but a mechanistic reason for that particular hour.

Why the Hours Before That, Too

But that mechanism doesn't fully explain the hourly waking that started before the terminal wake-up. That's a different kind of disruption — not one clean cutoff, but general fragility through the whole night. That points to a second, independent system: progesterone.

Progesterone breaks down into a metabolite called allopregnanolone, which acts on GABA receptors in the brain — the same calming, sedative pathway benzodiazepines work on. It's part of what keeps sleep consolidated and resistant to minor disruptions. As progesterone declines through perimenopause, that buffering effect weakens, and sleep becomes measurably lighter and easier to fragment — by noise, movement, minor temperature shifts, the kind of small nightly disturbances that wouldn't have registered before.

So, most likely, two things were happening on top of each other: perimenopause's effect on progesterone and GABA making my sleep more fragile in general, and the glycogen/cortisol mechanism creating one specific, harder failure point in the early morning. Same diet as before, but a body that could no longer absorb its effects the way it used to.

The Fix — and Why It Worked

The fix wasn't a return to a high-carb diet. I'm still low carb, generally somewhere around 100-150g a day, sometimes less, sometimes more. What changed was making sure carbohydrate — mostly from fruit, vegetables, and starchy roots — actually showed up somewhere in the day, and reliably by dinner.

My days usually look something like this: breakfast is mostly protein and fat, with a small piece of fruit afterward. I usually have more fruit after lunch. Dinner is always meat with roots and vegetables — no fruit at that meal. I eat two or three meals depending on my schedule — I've deliberately kept this flexible, and I only eat when I'm actually hungry, not because a clock says it's time.

Breakfast: mostly protein and fat, some fruit after.

The pattern I noticed, consistently: on days I didn't get enough carbohydrate-containing food in by dinner, that night's sleep was noticeably worse. That's not proof of the mechanism above — it's an N=1 observation — but it lines up closely enough that I trust it.

Dinner: meat, roots, vegetables

Diet Isn't the Whole Story

None of this means diet is the whole story, though — and I want to be clear about that, because it would be easy to read this as "just eat more carbs and you're fixed." It's one piece.

Nutrient quality still matters enormously to me: animal protein and fat as the base of most meals, real food over processed substitutes, the kind of nutrient density I've written about before. But sleep, movement, and stress management are their own independent levers, not just diet's supporting cast.

My version of stress management doesn't look like a meditation app. It's outdoor exercise — hiking, walking, stair HIIT — daylight and time in the forest, slow breathing practiced consistently through the Buteyko method, healthy relationships, and deliberately balancing time alone with time with the people I love. None of that shows up on a nutrition label, but all of it is doing real work.

A Framework, Not a Prescription

If there's one thing I'd want a reader to take from this, it's not "eat 120g of carbs a day" — it's the process. I noticed a pattern, looked for a mechanism that would explain it, tested it against my own data, and adjusted based on what actually happened rather than what a diet camp told me should happen.

Perimenopause changes the rules for all of us differently. The framework that's worked for me — quality nutrition with carbohydrate placed where my body seems to need it, real sleep, real movement, real stress management — is a starting point for figuring out your own version, not a protocol to copy exactly.

If you're in a similar situation and feel stuck, reach out — I'm happy to help.

Schedule Your Free Metabolic Reset Call

15min

Book Now