Do GLP-1 drugs cause nutrient deficiencies in teenagers?

---

The short answer

A 2026 analysis of US insurance claims followed 2,031 adolescents aged 10–17 for a year after starting a GLP-1 receptor agonist. Within that year, 16.9% picked up at least one nutritional deficiency diagnosis, most commonly vitamin D deficiency at 12.4% (PMID 42751742).

The number that should bother you more is the other one: only 23.3% had any nutrition therapy or counselling visit within 180 days, and among those who did, the average wait was 149 days.

So roughly one in eight of these teenagers accumulated a deficiency diagnosis, and roughly three in four never saw anyone about food.

That second finding is the solid one. The first comes with caveats serious enough that we are going to spend most of this page on them, because the version circulating this week — "1 in 8 children on GLP-1s become vitamin D deficient" — is not quite what the study can support.

Evidence tier: 3. Retrospective analysis of administrative insurance claims, read in full. Not a trial, no control group reported, and deficiencies identified from billing codes rather than measured blood levels.

Key points:

• 16.9% of adolescent GLP-1 users had ≥1 nutritional deficiency diagnosis within a year.
• 12.4% had vitamin D deficiency specifically; nutritional anaemia 1.55%, iron-deficiency anaemia 1.44%.
• Only 23.3% received nutrition counselling within 180 days. Mean time to first visit: 149 days.
• 78.6% of the cohort was on liraglutide, not semaglutide or tirzepatide. This says little about the drugs most teenagers start today.
• No control group. Adolescents with obesity already have high baseline rates of vitamin D deficiency.

What the study actually did

Researchers queried the Inovalon administrative claims database — covering over 100 million patients — for the years 2017 to 2022. They pulled 2,031 patients aged 10 to 17 who started a GLP-1 receptor agonist, met continuous-enrolment criteria, and had no prior diagnosis of nutritional deficiency. That last filter matters: these are newly recorded diagnoses, not pre-existing ones.

The cohort was 61.5% female, mean age 15, and 62.6% carried an obesity diagnosis.

Deficiencies were then counted using ICD-10 codes recorded in the following 12 months.

Three reasons not to read this as "GLP-1s cause vitamin D deficiency"

One: these are billing codes, not blood tests. A diagnosis appears in claims data when a clinician tested for something and coded the result. It does not capture teenagers who were deficient and never tested — and, going the other way, a cohort under active medical supervision gets tested more than the general population does. Some of this 12.4% is detection, not incidence.

Two: there is no control group. The study reports what happened to adolescents on GLP-1s. It does not report what happened to comparable adolescents who were not. That distinction is the whole ballgame here, because adolescents with obesity already carry elevated rates of vitamin D deficiency independent of any medication — it is one of the better-established associations in paediatric nutrition. Without a comparator, there is no way to separate the drug's contribution from the population's baseline.

Three: this is mostly a liraglutide study. The breakdown was liraglutide 78.6%, dulaglutide 10.4%, semaglutide 9.1%. Tirzepatide does not appear. The data runs 2017–2022, which predates the current prescribing pattern almost entirely. Anyone reading this as a finding about semaglutide or tirzepatide in teenagers is extrapolating well past what was measured.

None of this makes the study worthless. It makes it a signal worth testing properly rather than a conclusion.

The finding that doesn't need caveats

The care gap does not depend on causation at all.

Fewer than a quarter of these adolescents saw anyone about nutrition within six months of starting a drug whose primary mechanism is eating less. For those who did, the average wait was almost five months.

That is a straightforward description of how care was delivered, and it holds regardless of whether the drug causes deficiency. A medication that works by suppressing appetite, given to a growing adolescent, with no dietary support in three cases out of four, is a gap you can see without any statistics at all.

If there is one thing to take from this paper, it is that one — not the vitamin D number.

Why this is showing up now

GLP-1 prescribing in adolescents has expanded quickly, and the research base has not kept pace. Our own community signal this week was dominated by GLP-1 dosing and non-response threads — 56 dosing posts and 61 experience posts across Reddit in seven days, with tirzepatide mentioned 151 times, far ahead of anything else. The conversation is overwhelmingly about adults, which is part of why paediatric questions get so little attention.

The claim also travelled this week via a widely-shared post from a science communicator, accurately quoting the 1-in-8 figure. The study it points to is real and the numbers are right. What got lost in transit is that it is claims data without a comparator, on a drug mix almost nobody uses now.

What we could not determine

• Whether the deficiency rate exceeds baseline for adolescents with obesity who are not on these drugs. The study reports no comparator.
• Whether vitamin D levels were actually low, as opposed to tested and coded. Claims data cannot distinguish these.
• Whether semaglutide or tirzepatide behave the same way. They were 9.1% and 0% of this cohort.
• Whether supplementation was prescribed. The analysis covers diagnoses and counselling visits, not prescriptions.

Limitations

This is educational content, not medical advice, and nothing here is a reason to start or stop any medication.

• Retrospective claims analysis, tier 3. One database, one country, 2017–2022.
• We have not independently re-analysed the data, only read the published report.
• Paediatric prescribing decisions belong with a paediatrician or paediatric endocrinologist, who can measure rather than infer.

The bottom line

The headline version — one in eight teenagers on a GLP-1 becomes vitamin D deficient — overstates a study that cannot establish causation, measured billing codes rather than blood, and ran mostly on a drug that has largely been superseded.

The quieter finding is the one worth acting on. Three out of four adolescents started an appetite-suppressing drug without seeing anyone about nutrition, and those who did waited an average of five months. That is not a question about pharmacology. It is a question about how the care is organised, and it has an obvious answer.

If a teenager in your life is starting one of these, the useful question for the prescriber is not "will this cause a deficiency" — it is "who is checking, and when."