Who Shouldn't Take a GLP-1

There is a real list of people for whom a GLP-1 medication is a bad idea, and a longer list for whom it needs careful thought before anyone writes a prescription. A provider who does not ask about these things is not being efficient with your time.

Everything below applies to FDA-approved semaglutide and tirzepatide products. It applies equally to compounded preparations of the same molecules, which are not FDA-approved — a compounded product does not carry a different safety profile because it carries a different price, and the screening conversation matters at least as much. Our guide to compounded GLP-1s covers that distinction properly. For the class as a whole, start with our complete guide to GLP-1 medications.

The absolute contraindications

Medullary thyroid carcinoma and MEN2

Both semaglutide and tirzepatide carry a boxed warning — the strongest warning the FDA applies — regarding thyroid C-cell tumours. In rodent studies, these drugs caused dose-dependent and duration-dependent thyroid C-cell tumours. Whether that translates to humans is not established; the relevance of the rodent finding to people is genuinely unknown.

What is not uncertain is the resulting instruction. These medications are contraindicated in anyone with a personal or family history of medullary thyroid carcinoma, and in anyone with multiple endocrine neoplasia syndrome type 2, a hereditary condition that predisposes to it. This is not a caution to weigh against benefits. It is a stop.

"Family history" here means what it says, and it is worth thinking about properly rather than answering from memory. Thyroid cancer in a relative is not always medullary thyroid carcinoma — most thyroid cancer is not — but if you know there is thyroid cancer in your family and you do not know which type, say so rather than guessing. It is a question worth asking your relatives before your appointment.

Known hypersensitivity

A previous serious hypersensitivity reaction to semaglutide, tirzepatide or any component of the formulation rules out re-exposure.

Pregnancy, planning pregnancy and breastfeeding

These medications are not recommended in pregnancy. Weight loss during pregnancy offers no benefit to the fetus, and there is not adequate human safety data. If you become pregnant while taking one, contact your clinician promptly.

If you are planning a pregnancy, this needs discussing well in advance, because these drugs have long half-lives and clear slowly. The labelling for semaglutide advises discontinuing a considerable period — a matter of months, not weeks — before a planned pregnancy. Your prescriber should give you the specific interval for your product.

There is a further point that catches people out. Weight loss and improved insulin sensitivity can restore ovulation in people with PCOS or with cycles disrupted by weight, meaning fertility may return without warning in someone who had assumed conception was unlikely. Contraception is a live issue on these drugs even if it had not been for years.

Tirzepatide adds a specific consideration: because it delays gastric emptying, it can reduce the absorption of oral contraceptives. The labelling advises using a non-oral contraceptive method, or adding a barrier method, for a period after starting and after each dose increase. This is one of the more commonly missed instructions in the whole class.

Breastfeeding is also a reason not to take these medications, as there is insufficient data on transfer into breast milk.

Gastrointestinal conditions

Prior pancreatitis

Pancreatitis has been reported with GLP-1 medications, and a history of it warrants real caution. Whether the class causes pancreatitis at a higher rate than would be expected in this patient population has been debated for years, and the evidence is not clean. What is not debatable is that someone who has already had pancreatitis has demonstrated a vulnerability, and most clinicians consider other options first. Ongoing heavy alcohol use, gallstone disease and severe hypertriglyceridaemia are related risk factors that belong in the same conversation.

Gastroparesis and severe gastrointestinal disease

These drugs work partly by slowing gastric emptying. Giving them to someone whose stomach already empties poorly is asking for trouble, and severe gastroparesis is generally a reason not to prescribe. Significant inflammatory bowel disease, a history of bowel obstruction or major abdominal surgery affecting motility, and severe chronic constipation all require assessment before starting rather than after.

Gallbladder disease

Rapid weight loss from any cause increases the risk of gallstones, and gallbladder events have been reported on these medications. A history of gallstones does not automatically rule out treatment, but it changes the monitoring and the conversation.

Eating disorders

An active eating disorder — anorexia nervosa, bulimia nervosa, or binge eating disorder in an unstable phase — is a reason to address the eating disorder before considering a weight-management medication, not alongside it.

The reasoning is not that people with eating disorder histories are untrustworthy. It is that a medication which markedly suppresses appetite interacts badly with a condition characterised by restriction, and that the treatment can be recruited into the illness. In binge eating disorder specifically the picture is more nuanced and there is genuine clinical interest in this class, but that is a decision for a clinician who knows the eating disorder history in detail, ideally with input from someone treating it.

If you have a history of disordered eating, tell the prescriber. This is not a disclosure that automatically results in refusal, and concealing it does not lead anywhere good.

Drug interactions worth flagging

• Insulin and sulfonylureas. On their own, GLP-1 medications carry a low risk of hypoglycaemia. Combined with insulin or a sulfonylurea such as glipizide or glyburide, the risk rises meaningfully and those doses usually need reducing. This must be managed by a prescriber, not adjusted at home.
• Oral contraceptives. Particularly relevant with tirzepatide, as above.
• Narrow therapeutic index drugs. Warfarin, levothyroxine, some anti-epileptics and immunosuppressants — anything where the blood level has to sit in a narrow band — may be affected by altered absorption. Monitoring may need adjusting.
• Diuretics, ACE inhibitors, ARBs and NSAIDs. Not interactions with the drug so much as with its side effects: if vomiting or diarrhoea causes dehydration, these medications increase the risk to the kidneys.
• Other GLP-1 medications. Taking two agents in this class together, in any combination, is not appropriate.

Situations requiring caution rather than refusal

Several things do not rule out treatment but should change how it is prescribed and monitored: existing diabetic retinopathy, particularly where blood sugar is likely to improve rapidly; significant kidney impairment, because of the dehydration risk; a history of depression or suicidal ideation, which should be monitored during treatment; and age, at both ends — paediatric use is limited to specific approved indications and age ranges, and older patients need more attention to lean mass, hydration and fall risk. Our article on muscle loss during treatment is particularly relevant to that last group.

Anyone scheduled for surgery, endoscopy or any procedure involving sedation needs to tell the team about the medication in advance, because delayed gastric emptying affects fasting requirements and anaesthetic safety.

Why a proper history is not bureaucracy

Every item above is a question that takes under a minute to ask, and every one of them can change the answer. Some — medullary thyroid carcinoma, pregnancy, active eating disorder — change it completely. The purpose of the intake questionnaire and the consultation is to find those things before a prescription exists, which is the only point at which finding them is easy.

This is the clearest practical difference between a clinical service and an order form. If you can obtain one of these medications after answering four questions, none of which concerned your family history, nobody has assessed whether it is safe for you. That is worth recognising for what it is. Our article on how to spot a prescription mill goes through the other markers.

It follows that a clinician may evaluate you and conclude that a GLP-1 is not appropriate. That is a legitimate clinical outcome rather than a door to try elsewhere. If it happens, ask what the reasoning was and what the alternatives are — there usually are some.

If you want to know where you stand, our weight management service begins with an evaluation that covers all of this, and you can book a consultation with a licensed clinician. If you experience severe abdominal pain, persistent vomiting or signs of an allergic reaction while on treatment, the guidance in our side effects article tells you which symptoms need urgent attention.