GLP-1s, Cycles, and Perimenopause

I practice in Los Angeles, which means a large fraction of my patients are women between 35 and 55 who have been told, by someone with a large audience and no medical license, something confidently wrong about their hormones.

So let me take the four questions I get most, in the order I get them, and answer each one at the level of confidence the evidence actually supports. Where that level is low, I am going to say so rather than fill the gap with something that sounds authoritative.

1. PCOS: does this help?

Yes, and the mechanism makes sense.

Polycystic ovary syndrome is, for most women who have it, fundamentally a disorder of insulin resistance and androgen excess, with the two feeding each other. Hyperinsulinemia drives ovarian androgen production and suppresses sex hormone-binding globulin, which raises free testosterone further.

Anything that improves insulin sensitivity tends to improve the whole cascade. Weight loss does this. Metformin does this. GLP-1 receptor agonists do it, and in the trials and cohort data available they improve insulin sensitivity, reduce androgen levels, and increase the rate of ovulatory cycles — often more than metformin does, though head-to-head data remains limited.

The international PCOS guideline now includes anti-obesity pharmacotherapy as an option in appropriate patients. It is not a first-line replacement for the full management of PCOS, which is broader than weight, but it is a legitimate tool.

The consequence nobody prepares you for: if you have not been ovulating, and you start ovulating, you can become pregnant. Women who have spent a decade being told they would struggle to conceive are conceiving, quickly, on medications that are contraindicated in pregnancy. Which brings me to the next question.

2. Contraception: is there an interaction?

This one has a specific, actionable answer that too few people receive.

Tirzepatide reduces the absorption of oral contraceptives. Delayed gastric emptying changes the pharmacokinetics of oral drugs, and for combined oral contraceptives the effect was significant enough that it is in the label. The manufacturer’s guidance: use a non-oral contraceptive method, or add a barrier method, for 4 weeks after starting tirzepatide and for 4 weeks after each dose increase.

Semaglutide’s labelling does not carry the equivalent warning; the studied effect on oral contraceptive exposure was not clinically significant. I still mention it, because if someone is having significant vomiting from any of these drugs, absorption of anything oral becomes unreliable.

If you take a pill and you are on tirzepatide and nobody told you this, you are not alone, and it is worth a conversation this week rather than next month.

3. Fertility and pregnancy

These medications are not for use during pregnancy. Animal reproductive toxicity data showed adverse effects, human data is limited, and the intervention itself — a substantial energy deficit — is not what a pregnancy needs.

For planning purposes, the half-life matters. Semaglutide’s labelling recommends discontinuing at least 2 months before a planned pregnancy, because the drug clears slowly. Tirzepatide’s washout is shorter but the same principle applies. If pregnancy is on your horizon at all, that timeline should be part of the conversation before you start, not after.

If you conceive unexpectedly while taking one: stop the medication and call your obstetric clinician. This is a known situation, it is managed, and the reaction you deserve is a plan, not a lecture.

What I cannot tell you, because the data does not exist yet, is the long-term picture for children conceived shortly after exposure, or the effects of pre-conception GLP-1 use on pregnancy outcomes at scale. Registries are being built. We will know more in 5 years than we do now.

4. Perimenopause: the question I care most about

Here is the setup. Somewhere in the late thirties to late forties, estradiol begins its long, non-linear decline. Independent of any change in diet or activity, most women experience a redistribution of fat toward the abdomen, a decline in lean muscle mass, and an acceleration of bone loss. The scale may barely move while the body composition underneath it changes substantially.

So a woman arrives in my office at 47, frustrated, doing everything she did at 37, and it is not working — because the underlying substrate changed.

Do GLP-1s work in this population? Yes. They are effective across the age range studied, and the metabolic problems of midlife — visceral adiposity, insulin resistance, rising blood pressure and lipids — are exactly what these drugs address.

But two things change about the plan:

Muscle and bone become the central concern, not a footnote. You are now stacking a pharmacologic energy deficit on top of an estrogen-driven decline in lean mass and bone mineral density. This is the population in whom I am most insistent about protein targets and resistance training, and the population in whom I most want a baseline DXA. I have written about why this is the whole game — in perimenopause, it is not an exaggeration.

Menopausal hormone therapy is a separate conversation, and often a parallel one. A GLP-1 is not a substitute for hormone therapy and hormone therapy is not a substitute for a GLP-1. They address different problems. For many of my patients the right answer involves both, and that requires a clinician willing to think about them together rather than treating one and referring out the other.

A practical note on tracking

Everything in this piece is easier to act on if you are recording it rather than reconstructing it.

Two situations in particular. If you are on tirzepatide and taking an oral contraceptive, the 4 week backup window resets after every dose escalation — which is exactly the kind of rolling, dose-linked date that people lose track of, and the consequence of losing track is not trivial. And if you are perimenopausal, cycle changes, symptoms and body-composition changes are all happening at once from at least two causes, and separating them later without a record is guesswork.

I use the Zenday App for this, and the reason it is the right tool rather than a generic health app is that it treats the injection as the organising event — so a dose increase and everything that follows from it sit on the same timeline, rather than in different places. It is also worth saying that this is sensitive information: it runs in a browser rather than as an app on your phone, and for a category of medication that people are routinely judged for taking, that is a reasonable thing to want.

Where the honest answer is “we don’t know”

I would rather be useful than impressive, so:

• We have very little trial data on GLP-1 use specifically stratified by menopausal status. Almost everything I have said about perimenopause is inference from general physiology plus clinical experience, not from a trial designed to answer it.
• The effect of these drugs on bone mineral density over multi-year horizons is not well characterised, and it is the emerging question I watch most closely.
• Interactions with menopausal hormone therapy are not systematically studied.
• The effect on breastfeeding is not established; the drugs are not recommended during lactation.

If someone tells you they have confident answers to these, ask them what they are reading. Then read it yourself.

Men have their own version of this, and it is the mirror image: adipose tissue converts testosterone to estradiol, which suppresses the signal to make more. GLP-1s and men’s hormones covers it, including why testosterone replacement is usually the wrong first move.

One label instruction worth acting on

Tirzepatide labelling advises that oral contraceptives may be less effective, and recommends a non-oral method or an added barrier method for 4 weeks after starting and after each dose increase. That instruction lands precisely when fertility is returning, which is the combination worth taking seriously — the detail is in GLP-1s and your other medications.

What I would want a friend to do

Ask the four questions above out loud, at your first appointment, before the prescription is written. If your prescriber cannot answer the tirzepatide-and-the-pill question, that tells you something useful about the quality of supervision you are about to receive — and given how many of these prescriptions now come from a ten-minute online intake, it is worth finding out early.

Bone density deserves a mention alongside all of this, because the loss that comes with rapid weight change lands hardest in exactly the group this article is about — the bone health rankings cover what is worth doing and what is worth measuring.