Irregular cycles. Acne. Hair where it was not before. In a fourteen-year-old these are, between them, both the diagnostic criteria for PCOS and an accurate description of ordinary puberty.
That overlap is the entire problem, and the 2023 international guideline answers it in a way that surprises people who expect a guideline to make diagnosis easier. For adolescents it makes it harder, on purpose, and the reasoning is worth following whether you are the teenager or the parent in the room.
What is different under the adolescent rules
| Adults | Adolescents | |
|---|---|---|
| Findings needed | Two of three | Both of the two |
| Irregular cycles | Counts as one criterion | Required |
| Hyperandrogenism | Counts as one criterion | Required |
| Ultrasound | Accepted for the third criterion | Not recommended |
| AMH | Accepted as an alternative to ultrasound | Not recommended |
| If only one feature is present | Not diagnostic | “At risk”, reviewed over time |
The adult criteria are two of three, which means any two of irregular ovulation, hyperandrogenism, and polycystic ovarian morphology. Take the third option away and require both of the remainder, and the standard has moved in one direction only.
Why the third criterion is removed
Multifollicular ovaries are common in healthy adolescents. So is a raised AMH. Neither finding separates PCOS from an ovary that is doing what ovaries do at that age, and a test that returns “abnormal” for a large share of people without the condition is not measuring the condition.
The guideline therefore excludes both from diagnosis until eight years after a first period, the point at which the hypothalamic-pituitary-ovarian axis is treated as mature. AMH inherits the same restriction it was added to relieve in adults — it is an alternative to ultrasound, not an exemption from the age rule.
The practical consequence runs against intuition. A scan is the thing most people assume settles the question, and in this age group it is the thing most likely to produce a finding that means nothing.
Why irregular cycles are not enough on their own
The guideline’s cycle definitions already account for age, and they are more forgiving early.
In the first year after a first period, irregularity is treated as normal — part of the pubertal transition rather than a finding at all. Between one and three years, the window is shorter than 21 days or longer than 45. Only after three years does it narrow to the adult 35.
Applying the adult number to a fifteen-year-old turns a normal transition into a problem, and it is the single most common error in material written about this. A 42-day cycle two years after a first period is inside the definition of regular.
“At risk” is an answer, not a shrug
An adolescent with one of the two required features — irregular cycles but no hyperandrogenism, or the reverse — is placed in an “at risk” category and offered longitudinal evaluation rather than a diagnosis or a dismissal.
This is the part that frustrates families most, and it is the most defensible thing in the whole section. The alternatives are both worse. Diagnosing on one feature labels a large number of girls whose symptoms resolve as puberty finishes, and that label follows them into insurance, into future appointments, and into how they think about their own bodies. Dismissing them means the ones who do have PCOS lose years.
Review over time is the only option that does neither. It costs patience, which is a real cost when a teenager is distressed about her skin or her hair and has been told the answer is “we will see”.
What most articles get wrong
They apply adult criteria to teenagers. Two of three, adult cycle lengths, ultrasound as the settler. Every one of those is wrong under 18, and the error always runs toward over-diagnosis.
They treat a scan as the definitive test. In this age group the guideline specifically excludes it.
They present “at risk” as a failure to diagnose. It is a defined category with a defined action — longitudinal evaluation — not a clinician avoiding a decision.
What makes the waiting productive
If the answer is review over time, then what gets reviewed matters, and almost all of it is information only the household has.
The two required features are both things that reveal themselves across months rather than in a consultation. Cycle dates answer the first one arithmetically — eight cycles a year is a count, not a judgement. The second shows up as a trend in skin and hair that nobody can reconstruct accurately from memory at a six-month review, because the comparison being asked for is with a version of yourself you stopped noticing.
Written down as it happens, that is a record a clinician can read, and the “we will see” appointment becomes a shorter one. It is also the only part of this process anyone outside a clinic can actually do.
A last thing worth saying plainly to anyone reading this about their own daughter: the guideline being strict here is not the system withholding an answer. It is the system declining to hand out a lifelong label on evidence that, at this age, cannot carry it.