GLP-1's lesser-known side effects: what the emerging signals actually show

Most of what people feel on a GLP-1 is the familiar gastrointestinal set — nausea, constipation, and the rest, covered on our main side-effects page. But a second tier of rarer and newer signals gets talked about constantly online, and here the honest answer isn’t one answer — it’s a different answer for each. Some are documented risks. One or two are being actively studied and genuinely unsettled. Some looked worrying and then faded under scrutiny. And a few are widely reported but have almost no evidence behind them at all. This page grades each on its own merits.

One framing up front: these are approved medicines with real, well-established benefits for the people they’re right for. This page deliberately zooms in on the rarer risks — that’s one column of a benefit-and-risk ledger, not the whole ledger. The point is to weigh it against the benefits with your clinician, not to scare you off a conversation. It’s education, not a substitute for your prescriber’s advice.

How to read this page

Every signal below carries its own evidence grade, because lumping them together is exactly how a viral worry ends up feeling as solid as a documented risk. A quick key:

Established strong, consistent human evidence · Emerging real data, not yet conclusive · Observational a real-world association, no proof of cause · Anecdotal widely reported, little or no evidence.

The point isn’t to reassure or to alarm — it’s to let you tell these apart.

The eye: NAION and retinopathy

NAION (non-arteritic anterior ischemic optic neuropathy) — a sudden, usually painless loss of vision in one eye from reduced blood flow to the optic nerve. Emerging

This is the signal to understand carefully, because it’s both the newest and the most misreported. A 2024 study from a single eye center found people prescribed semaglutide had a higher rate of NAION than matched patients on other medicines, and a large nationwide Danish cohort found a roughly two-fold higher rate. On the strength of these, the World Health Organization and European regulators flagged it in mid-2025 and the European label was updated to list it as a “very rare” effect — on the order of 1 in 10,000 or fewer. Two things keep this at emerging rather than established: the studies show an association, not proof that the drug causes it, and they have design limitations — as observational data, they can’t fully rule out confounding (for example by underlying diabetes severity). As of mid-2026, the US FDA had not added a NAION warning. So: a real signal, taken seriously by regulators, but unproven and rare in absolute terms.

Diabetic retinopathy worsening is a separate, better-understood issue. Supported but limited In the SUSTAIN-6 cardiovascular trial, people with pre-existing diabetic retinopathy who lowered their blood sugar quickly had more retinopathy complications — a known effect of rapid glucose lowering, not unique to these drugs, and mostly relevant if you already have retinopathy and diabetes.

What this means for you: whatever the cause, sudden vision loss or a sudden change in vision in one eye is a same-day emergency — get seen urgently. If you have diabetes and eye disease, tell your prescriber before starting so your eyes can be monitored.

Gallbladder and biliary problems Established

The best-evidenced entry here. A meta-analysis of 76 randomized trials found GLP-1 medicines modestly raise the risk of gallbladder and bile-duct problems — gallstones and inflammation — with the risk higher at higher doses and, importantly, partly driven by rapid weight loss itself (losing weight quickly is a classic trigger for gallstones, drug or no drug). The absolute risk is still small, but this one is real and documented.

What this means for you: severe, steady pain in the upper-right belly — often after fatty meals, sometimes spreading to the right shoulder or back — needs prompt medical attention. And belly pain with fever or yellowing of the skin or eyes is an emergency: seek urgent care now (that combination can signal a serious infection of the bile ducts).

Mood, depression, and suicidal thoughts Observational — and fading

A worth-knowing example of a signal that shrank under scrutiny. Early adverse-event reports raised a worry about depression and suicidal thoughts, and pharmacovigilance databases did contain such reports. But when the US FDA (starting in 2024) and European regulators reviewed the evidence, they did not find an established causal link — and after a large pooled analysis the FDA formally requested removal of the suicidal-behavior warning from these medicines in early 2026. This is the opposite trajectory to the eye signal: reviewed, and so far not upheld.

Here’s the important distinction, though: what faded is the population-level alarm — the claim that the drug causes suicidality across everyone taking it. That says nothing about you. Rapid weight loss, big changes in eating, and life all affect mood, so a real change in how you feel is worth acting on regardless of what the databases show. Any new or worsening low mood, or thoughts of self-harm, deserves a prompt conversation with your clinician.

If you are having thoughts of harming yourself, this is urgent — do not wait: contact your local emergency number now, or in the US call or text 988 (Suicide & Crisis Lifeline). The faded signal is a reason not to let an old headline scare you off these medicines — never a reason to sit alone with a crisis.

Fertility, pregnancy, and cycle changes Mixed by claim

Three different things get tangled together here:

• Restored fertility is real and well-understood. Established Weight loss can restore ovulation in people whose cycles had stopped (for example with PCOS) — which is why some people conceive unexpectedly. The mechanism (weight loss improving ovulation) is solid; the viral “Ozempic babies” framing is just that mechanism in the wild.
• The pill may work less well — but this is drug-specific. Supported but limited Tirzepatide’s label specifically warns that it can reduce the effectiveness of oral contraceptives (advising a backup method or a non-oral method around starting and dose increases). Semaglutide’s label does not carry this warning — so don’t transfer it between drugs. Check your own medicine’s guidance with a pharmacist.
• Broader menstrual-cycle changes (timing, flow) are widely reported online but barely studied. Anecdotal Some of it plausibly tracks with weight loss and metabolic change; the honest status is that the data isn’t there yet.

What this means for you: these medicines are not for use in pregnancy — if you could become pregnant, use reliable contraception, and if you become pregnant, stop and call your clinician. And if you’re planning a pregnancy, talk to your clinician about stopping well in advance — these drugs linger in the body for weeks, so stopping isn’t something to leave until a positive test. For PCOS specifically, see our PCOS and GLP-1 page.

Feeling cold / temperature changes Anecdotal

The clearest candor case on the page. Feeling colder is reported in online communities but has essentially no clinical or mechanistic evidence behind it as a drug effect. Some of what people notice may simply follow from losing weight (less insulation, lower metabolic rate). We’re listing it precisely because it shows up in the Reddit-scale analysis below — and because the honest grade for it is anecdotal, not established. One caution, though: persistent cold intolerance can also point to something else worth checking — an underactive thyroid, anemia, or (if you also take insulin or a sulfonylurea) a low blood sugar. So if a symptom is new, persistent, or worrying, mention it to your clinician regardless of its grade — “anecdotal” means the drug link is unproven, not that your experience should be ignored.

Muscle loss and hair shedding

Two more that get grouped as “hidden” effects but already have their own detailed pages:

• Muscle (lean-mass) loss Established — some of the weight lost is muscle, and protecting it matters. Full detail, and what actually helps, on GLP-1 and muscle loss.
• Hair shedding Supported but limited — a temporary, self-limited shedding (telogen effluvium) that tends to follow rapid weight loss rather than the drug directly. Full detail on GLP-1 and hair loss.

What people report vs. what’s proven

A 2026 analysis in Nature Health mined over 410,000 posts from GLP-1 online communities to surface what people actually report — and it’s a useful mirror, with one big caveat. It’s excellent at showing what patients experience and talk about (it’s how the reproductive and temperature complaints got noticed), but it cannot show cause — a forum is not a controlled study, and popular ≠ proven. That gap between what’s reported and what’s established is the entire reason this page grades each signal rather than just listing them. Use online communities for solidarity and for questions to raise; use graded evidence — and your clinician — for what to actually do.

For the common side effects and the full red-flag list, see the GLP-1 side-effects page; to look up a specific symptom, try the Side-Effect Helper. As always, whether a GLP-1 is right for you, and how to weigh these risks against the benefits, is a conversation for you and a clinician.