Provocative Testing

Basal hormone levels capture the HPG axis at a single, static moment. Provocative tests instead push the axis — deliver an artificial dose of a stimulating signal, then watch how it responds — to reveal its capacity to react rather than merely its resting state. This is the tool that separates an axis that is simply quiet from one that is genuinely deficient.

The distinction matters because so many of the deviations look identical at rest. A boy with constitutional delay, a boy with partial congenital hypogonadotropic hypogonadism (CHH), and a boy whose axis is transiently suppressed by illness or training can all show the same low LH, low FSH, and low testosterone on a single morning draw. What separates them is not the number at rest but how the machinery behaves when provoked.

Why a static number under-reads the axis

Two features of the axis make single measurements unreliable, and both are fixed by provocation:

• LH is pulsatile. GnRH — and therefore LH — is released in bursts every 1–2 hours (see pulsatility). A single blood draw can land in an inter-pulse trough and read falsely low. FSH, with its longer half-life, is smoother and less affected. A provocative test forces a response on demand instead of waiting to catch a spontaneous pulse.
• The axis normally sits idle for years. Through the long childhood quiet the same cells are dormant by design. A low resting value does not distinguish “dormant but intact” from “damaged or missing” — but a stimulus does: intact machinery responds, absent machinery does not.

The inventory of provocative tests

Each test injects the stimulus at a different tier of the axis, so each interrogates a different part of the machinery.

GnRH stimulation test (native GnRH / gonadorelin)

The classic test. An intravenous bolus of GnRH (gonadorelin), with LH and FSH measured at 0, 30, and 60 minutes; LH normally peaks around 30 minutes. It asks the narrowest question: can the pituitary gonadotropes release LH and FSH when handed GnRH directly?

Its limitation is priming. The gonadotropes need recent exposure to endogenous GnRH pulses to respond briskly. A perfectly normal but not-yet-activated axis (constitutional delay) can under-respond to a single brief bolus and mimic true deficiency — a false positive. This is why the short native-GnRH test does not cleanly separate constitutional delay from CHH on its own.

GnRH-agonist stimulation test (leuprolide / leuprorelin)

The more discriminating version, and the one usually meant today. A single subcutaneous dose of a longer-acting GnRH agonist delivers sustained receptor stimulation over hours rather than a brief pulse. LH and FSH are sampled across the first few hours (the agonist “flare”), and — the valuable addition — testosterone is measured at 24 hours, reporting whether the Leydig cells downstream actually respond to the LH that was provoked.

The sustained stimulus pulls a response even from a relatively unprimed-but-intact axis, so it discriminates constitutional delay from CHH better than the short bolus. (Note the apparent paradox with therapy: the same agonist, given continuously, eventually suppresses the axis — that is how leuprolide treats precocious puberty and prostate cancer. The diagnostic test uses a single dose and reads the initial flare, not the later shutdown. Continuous versus pulsed is the whole difference — see pulsatility.)

hCG stimulation test

Where the GnRH tests probe the brain-to-pituitary tiers, the hCG stimulation test probes the gonad. Human chorionic gonadotropin acts as an LH analog at the Leydig cell, so giving it (as a single injection or over several days) and measuring the testosterone rise asks: can the testes make testosterone when driven directly?

This localizes the problem below the pituitary. A good testosterone response means the testes are present and capable, pointing any deficit upstream (central/secondary hypogonadism). A flat response means the gonad itself cannot answer — it confirms primary testicular failure, or, in a child with non-palpable testes, helps distinguish functioning undescended testes from anorchia (absent testes).

Clomiphene (SERM) challenge

A test of the feedback loop rather than a single tier. Clomiphene, a selective estrogen-receptor modulator, blocks estrogen’s negative feedback at the hypothalamus and pituitary (see feedback architecture). If the whole hypothalamic–pituitary unit is intact, removing that brake should drive LH and FSH up. A rise indicates an intact, suppressible-and-releasable axis; a flat response points to a hypothalamic or pituitary defect.

Pulsatile GnRH priming (localizing hypothalamus vs. pituitary)

A single blunted GnRH stimulation is ambiguous: it cannot tell a pituitary that can’t respond from one that is merely de-primed by chronically absent GnRH. Administering GnRH in physiologic pulses over hours to days re-primes the gonadotropes; if responsiveness then recovers, the pituitary was fine and the lesion is hypothalamic (GnRH deficiency). If it stays flat despite priming, the defect is at the pituitary itself. This is the most direct way to separate those two central levels.

What the tests tell you — capacity vs. drive

The whole point is to answer a different question than basal labs do:

Question	Test	A response means
Can the pituitary make LH/FSH?	GnRH / GnRH-agonist	Gonadotropes are intact; deficit (if any) is upstream in the drive
Can the testes make testosterone?	hCG	Gonad is present and functional; deficit is central
Is the feedback loop intact and releasable?	Clomiphene	Hypothalamic–pituitary unit works
Hypothalamus or pituitary?	Pulsatile GnRH priming	Recovery with priming → hypothalamic; no recovery → pituitary

The single most useful application is separating constitutional delay of growth and puberty (CDGP) from partial CHH — two conditions that look identical at rest (see deviations). A robust stimulated LH rise, and especially a good 24-hour testosterone rise on the agonist test, favors CDGP: the machinery works and is poised to activate. A blunted response favors CHH.

The interpretation is asymmetric

A crucial subtlety governs when the test is worth doing:

• A robust response is conclusive on its own. Functional suppression, illness, or a missed pulse can only blunt a response — none of them can fake a strong one. So a brisk stimulated rise proves the machinery works no matter the surrounding conditions, and it does so without depending on catching a good moment.
• A blunted response is only trustworthy once confounders are removed. A weak response during acute illness, an energy deficit, or heavy training cannot be told apart from a true deficit. To make a blunted result mean something, the functional suppressors must be cleared first.

The practical corollary: never run a stimulation test during an acute illness (a blunted result would be uninterpretable), and read a blunted result cautiously if the person is still under a known suppressor.

Limitations

• CDGP and partial CHH lie on a continuum. No single provocative test cleanly splits them; the result shifts probability rather than delivering certainty. Inhibin B, AMH, and — often most decisively — time remain essential companions.
• Priming effects blunt the short native-GnRH test in genuinely normal but unactivated axes; the agonist test and pulsatile priming mitigate but do not fully erase this.
• The FSH-first mimic of early puberty. At the very onset of gonadarche the awakening pulse generator fires slow pulses, so FSH leads and LH lags — a picture that can resemble a partial deficit. Repeating over months, not just provoking once, is what resolves it.
• Exogenous androgen invalidates the pituitary-level tests. Testosterone (or an anabolic steroid) suppresses the axis by negative feedback, so the GnRH and clomiphene tests must be done before starting, or during a washout from, testosterone therapy. The hCG test, which acts below the pituitary, is less affected but still best interpreted off exogenous androgen.

Key idea

Provocative tests convert a static snapshot into a functional probe: they inject a stimulus at a chosen tier of the axis and read whether that tier can respond. Because a strong response cannot be faked by suppression, provocation is the cleanest way to prove the machinery is intact — while a weak response is only as trustworthy as the conditions it was measured under. Used well, a stimulation test answers the one question basal labs cannot: not what the axis is doing, but what it is able to do.