Hair started shedding at month three. Does the timing tell me whether it is the drug or the weight loss?

Started tirzepatide in February. Nothing unusual until early June, when the shedding started, and it is now substantial: handfuls in the shower, hair on the pillow, and a visible thinning at the temples and along the parting. I am not imagining it and neither is my hairdresser.

What I want to know is whether this is the drug or the weight loss, because the two lead to different decisions. If it is a direct drug effect I have a choice to make about continuing. If it is the rate of loss, I can slow down and keep the drug.

The specific thing I keep reading that I do not understand is that the timing itself is supposed to be informative. Several sources say a two-to-four-month lag points to one mechanism and immediate onset points to another, but none of them explain why, and I would like to understand the reasoning rather than take the conclusion on trust.

Timeline, as precisely as I can reconstruct it: first dose 8 February, weight loss started immediately and has been fast, about 14 kg by the end of May. Shedding noticed first week of June. Protein intake honestly probably poor for the first two months, better since.

On the nutritional contributors, since this is the part that is actually modifiable and the part where the advice is usually vague.

Protein

Hair shaft is roughly 90% keratin, a protein with a high cysteine content. Follicles are among the most metabolically active tissues in the body and they are also expendable: in a state of restricted amino acid availability, the body downregulates hair growth long before it compromises anything important. That is a feature, not a malfunction, and it is why protein-energy restriction produces telogen effluvium reliably.

The arithmetic that matters on this class: appetite suppression reduces protein intake disproportionately, because protein-dense foods are the most volume-expensive and the most likely to be aversive. Meat, eggs and fish are the classic aversion categories during the nauseated period, and the drift is toward crackers, toast and yoghurt.

Worked example of what "probably poor for the first two months" typically means. Someone who was eating 2,400 kcal with 100 g of protein drops to 1,200 kcal and, without deliberate effort, keeps protein at roughly the same share of energy:

• Before: 100 g protein, 400 kcal, 16.7% of 2,400
• After, same share: 0.167 × 1,200 = 200 kcal = 50 g of protein

Fifty grams a day for someone at, say, 90 kg is 0.56 g/kg, which is below even the minimal reference intake and far below what is appropriate during a large deficit. Common recommendations for preserving lean tissue on a substantial deficit are in the region of 1.2-1.6 g/kg, and evidence in energy-restricted populations supports the upper part of that range. So the gap is not marginal; it is threefold. That is entirely sufficient to produce effluvium without invoking any drug effect.

Iron

The best-evidenced nutritional association with hair shedding after protein-energy restriction. Mechanism is plausible: the follicle matrix is a rapidly proliferating tissue and iron is required for ribonucleotide reductase. Points worth knowing:

• Ferritin, not haemoglobin. You can have iron-deficient hair loss with a completely normal full blood count. Iron stores deplete long before anaemia appears.
• The threshold is contested. Various authors have proposed ferritin targets for hair growth around 30, 40 or 70 µg/L, all above the level at which a laboratory flags a result as low. The evidence for a specific threshold is weak, but the association between low-normal ferritin and shedding is reported consistently enough to be worth acting on.
• CRP alongside, always, because inflammation raises ferritin and can hide deficiency.
• Reduced red meat intake is near-universal on this class, from aversion and from volume constraints, and red meat is the main dietary source of haem iron.

Zinc

Genuine zinc deficiency causes hair loss, and it also causes changes in taste, poor wound healing and skin lesions. It is uncommon in isolation in the absence of malabsorption. Serum zinc is a mediocre marker, and it falls with low albumin and with inflammation. Reasonable to check in someone with markedly reduced intake; not reasonable to supplement indefinitely on suspicion, because high-dose zinc induces copper deficiency, which itself causes anaemia and neurological problems.

Others, briefly

• Vitamin D. Association reported; causal evidence weak. Cheap to check, commonly low anyway.
• B12 and folate. Worth checking as part of the same panel; deficiency causes shedding among other things.
• Selenium. Both deficiency and excess cause hair loss. An argument against enthusiastic supplementation.
• Vitamin A. Excess causes hair loss. Another argument for the same.
• Biotin. Deficiency causes hair loss and is very rare in people who eat food. Supplementation in the absence of deficiency has no good evidence for hair, and it interferes with a number of immunoassays including thyroid and troponin tests, which is a genuine clinical nuisance. Discussed further below.

The practical summary: protein and iron are where the evidence is, both are commonly deficient in exactly this population for structural reasons, and both are measurable or countable. The rest is worth checking once and not worth chasing.

Adding the trial incidence, since the question is partly whether this is drug-attributable and the safety tables have something to say.

Alopecia appears as a reported adverse event in the obesity trials of this class at rates in the low single figures, higher than placebo but not dramatically so. In SURMOUNT-1 alopecia was reported at roughly 4-6% across the tirzepatide arms against about 1% on placebo [1]. STEP 1 reported it at a lower rate, in the region of 3% against about 1% on placebo [2]. Reported rates in the higher-dose multi-agonist programmes appear higher again, which is consistent with the pattern below.

Three observations that matter more than the absolute numbers.

1. The rate tracks weight loss, not the molecule. Within and across these programmes, alopecia reporting is higher in the arms that lost more weight faster. Tirzepatide arms lost more than semaglutide arms in their respective trials and reported more alopecia; the aggressive multi-agonist arms lost the most and reported the most. That gradient is what you would expect if the mechanism is telogen effluvium secondary to rapid loss, and it is not what you would expect from a receptor-mediated effect, since the agents have quite different receptor profiles and the ordering follows weight change rather than pharmacology.

2. The comparison that settles it is bariatric surgery. Post-surgical telogen effluvium is very well described, with reported frequencies commonly in the range of 40-70% in the first year, peaking at three to four months post-operatively. No GLP-1 receptor is involved anywhere in that. A single mechanism, rapid loss with reduced protein and micronutrient intake, produces the same syndrome with the same timing at ten times the frequency. That is close to a natural experiment and it argues strongly that the drug is a route to the insult rather than the insult.

3. Trial ascertainment badly understates it. "Alopecia" as a spontaneously reported coded adverse event in a trial where nobody was asked about hair is not a prevalence estimate. Participants raising it depends on whether they thought it worth mentioning. Real-world reporting is far higher than 3-6%, and the gap is ascertainment rather than a difference in populations. So do not read the low trial figures as evidence that this is rare; read them as evidence that it was not asked about.

The synthesis: hair shedding on this class is common, is dose-related only insofar as dose relates to rate of loss, has a well-understood mechanism that has nothing specifically to do with GLP-1 receptors, and recovers. The variable to manage is the rate of loss and the adequacy of intake, not the choice of agent.