Saliva and Urine Hormone Testing vs Blood Testing in Perimenopause

A woman in her mid-40s who asks about hormone testing for perimenopause will find three different products marketed as if they answer the same question: a blood panel, a saliva kit, and a urine test like DUTCH. They draw from different biological samples, measure different things, and rest on wildly uneven clinical evidence, yet the marketing around them frequently treats the three as interchangeable paths to the same answer. That is the structural source of the confusion, and it is compounded by a split in medical opinion itself. Functional and alternative-medicine practitioners sometimes present saliva or urine panels as more sophisticated than blood work, while conventional providers dismiss those same panels outright, and a woman caught between the two gets no stable ground to stand on.

The biology underneath makes this worse before it gets better, because hormones rise and fall through the day and the month, so the value a test captures depends heavily on when and how the sample was taken and which hormone the lab is chasing. A woman can have her estradiol drawn on a Tuesday morning and get a very different number than if the same blood had been drawn Thursday afternoon, and that variability exists before anyone has even picked a testing method. Dr. Brooke's 2025 clinical commentary documents just how inconsistent the responses women get from their own providers tend to be when they raise the subject of testing, with some dismissed outright and others sent for panels that do not match their symptoms or their stage of transition. Even women who show up prepared to advocate for themselves often leave with more questions than they started with.

That confusion deepens because there are few plain-language resources that lay out what each testing method actually measures and where the clinical evidence for it stands. A perimenopause-focused testing service like Intermission Health builds its approach around the clinical validation that blood testing has accumulated, and that grounding is useful precisely because it gives women something to measure marketing claims against. Understanding the underlying biology, and the specific limits of each sample type, lets a woman tell whether a test is genuinely informative or simply well-marketed. That is the work this piece sets out to do.

What perimenopause does to hormones

Perimenopause is a period of instability in which estrogen can spike one month and fall sharply the next, while progesterone drops first, often years before any shift in estrogen becomes apparent. That distinction matters for anyone thinking about testing, because a single reading, taken at a single point in time, only captures one frame of a process that keeps moving. Any method of testing, blood, urine, or saliva, inherits this problem. None of them freezes a hormone system that is itself refusing to sit still.

FSH is a useful illustration of how misleading a single number can be during this transition. Intermittent bursts of estrogen production can suppress FSH back into a range that looks entirely normal, so a woman might show a menopausal-range FSH reading one fortnight and a perfectly ordinary one two weeks later. Neither reading is wrong. Both are just incomplete snapshots of a system whose output is oscillating rather than trending smoothly in one direction. One might argue that this alone should make women skeptical of any single lab value presented as definitive, and that skepticism would be well placed.

Progesterone follows a more predictable pattern, which is part of why it carries real diagnostic weight when captured correctly. It tends to be the first hormone to decline, often starting from the mid-30s onward, well before most women associate any symptoms with perimenopause. A day-21 progesterone level below 16 nmol/L indicates inconsistent ovulation that cycle, and that signal is real and measurable, but only if the blood draw is taken on the right day relative to ovulation. Taken at a random point in the cycle, that same signal disappears into noise. This means timing governs almost every comparison the rest of this piece makes: the value of a measurement depends on when it was taken almost as much as on what it measured.

What blood testing measures

Blood testing holds its place as the clinical standard because it directly quantifies circulating hormone concentrations, both the free fraction and the portion bound to carrier proteins, in a way that neither saliva nor urine can replicate for sex hormones. A single serum draw can return estradiol, progesterone, FSH, LH, testosterone, DHEA-S, and AMH together, giving a clinician a multi-hormone picture that reflects both ovarian function and adrenal contribution in one sitting. That breadth is part of what makes blood the medium clinical guidelines are actually built around.

Androgens offer a clear example of where blood testing earns its reputation. Testosterone and DHEA-S hold relatively steady across the month compared with the sharp swings seen in estrogen and progesterone, which makes blood a dependable medium for tracking them. Dr. Brooke's commentary notes that urine markers for estrogen and progesterone can correlate reasonably well with serum levels, but that correlation weakens noticeably for testosterone, which is exactly the hormone blood testing handles with the most consistency.

None of this works without attention to timing. Estradiol and FSH are best drawn on day 3 or day 5 of the cycle for women who still have one, while progesterone is best captured on day 19 or seven days after confirmed ovulation, when a true peak can be assessed rather than guessed at. Morning draws between 8 and 10 a.m. catch hormones closer to their daily peak, and biotin supplements should be paused beforehand, since biotin interferes with many of the assays labs use to measure these hormones. Reviewing progesterone alongside estradiol on day 19, rather than looking at either number in isolation, lets a clinician assess the ratio between the two. That ratio matters for a woman whose estrogen level looks unremarkable on paper but who reports symptoms consistent with estrogen dominance relative to a progesterone level that has already started to fall.

Blood testing's ability to combine seven hormones in a single draw, and to be timed deliberately around the points in a cycle where each hormone is most informative, is a large part of why it anchors sound perimenopause assessment. The curated approach some services take to blood-based hormone panels, paired with clinician review and the ability to track a woman's results over multiple cycles, is what turns a single multi-hormone snapshot into something closer to a trend line. Intermission Health builds its panel offering around that principle: a snapshot gains meaning when it is timed well and reviewed against what came before it. Insurance coverage for these panels varies by provider and plan, but the clinical case for blood testing rests on accuracy and timing, not on cost. It is also the foundation major guidelines on menopause and perimenopause management are built around, a point the diagnosis section later in this piece returns to directly.

Saliva testing and sex hormones

Saliva testing has an intuitive appeal. It is non-invasive, can be done at home without a clinic visit, and its proponents argue that it measures something more biologically relevant than a blood draw does: the free, unbound hormone fraction that is actually available to interact with tissue, rather than the total pool that includes hormone still attached to carrier proteins. Hormones diffuse from blood capillaries into the salivary glands, so in theory, a saliva sample reflects that active fraction directly. The argument sounds reasonable on its face, and it is worth taking seriously rather than waving off.

The appeal runs out when it meets the practical reality of how low sex hormone concentrations in saliva actually are. Measuring them requires assays sensitive enough to detect very small quantities, and that sensitivity comes at the cost of reliability: diet, hydration, the time of day, and even oral health can all introduce noise into the result. A medical policy on salivary hormone testing for aging and menopause, effective May 1, 2026 and due for its next review in January 2027, codifies where the clinical consensus has landed. Salivary levels do not consistently provide a reasonable representation of circulating serum hormones, and the variability within a single patient's own results, tested under similar conditions, can be large enough to undercut any conclusion drawn from them. The American College of Obstetricians and Gynecologists and the American Society for Reproductive Medicine have reviewed the evidence and concluded there is nothing establishing that hormone levels measured in saliva carry biological meaning for diagnosing or managing menopause. No published studies have established sensitivity, specificity, or predictive values for salivary hormone testing applied to menopause, which leaves the method without the basic evidentiary scaffolding clinical guidelines require.

Cortisol is the genuine exception, and it strengthens rather than weakens this analysis to say so. A clinical endocrinology service evaluation found that waking salivary cortisone gave screening information comparable to a Short Synacthen Test in a meaningful share of cases, and in the Debono study the evaluation cited, 83% of patients preferred collecting the sample at home over a clinic visit. That is a real, validated use case, and it shows the problem with saliva testing is not the sample type itself but the specific hormones it fails to measure reliably. Sex hormones in perimenopause are measured unreliably by saliva testing because the method fails for them specifically.

What the DUTCH urine test measures

The DUTCH urine test (a dried urine panel for hormone metabolites) occupies a different position than saliva testing in this comparison, because it does something blood testing genuinely cannot: it tracks how the body metabolizes and clears estrogen, progesterone, and testosterone after they've been used, including which metabolic pathways dominate that clearance process. It also captures adrenal hormone output and sleep-related markers across a full collection day. That is a real capability, and the liver-clearance picture it produces is interesting on its own scientific terms. Whether interesting translates into actionable for a woman trying to manage perimenopause symptoms is where the case for DUTCH starts to thin out.

Urinary metabolites reflect what the body has already processed and eliminated. They do not necessarily reflect how actively a hormone is signaling in tissue at the moment of the test, or how engaged a given receptor is. A woman can produce what looks like an optimal metabolite pattern on a DUTCH panel and still be dealing with significant, disruptive symptoms, because the test is describing clearance, not current signaling. Future Woman's 2026 comparison guide makes the strongest version of the case for urine testing anyway, arguing that because DUTCH averages multiple samples across a full day, it accounts for hormonal pulsing more effectively than a single blood draw does, which its proponents say makes it more accurate for tracking progesterone and estrogen in cycling women than a timed serum snapshot.

That argument runs into an evidence problem that is hard to get around. The primary research supporting the DUTCH test's validity was conducted by the test's own founder, and no substantial independent, peer-reviewed body of research has established its clinical utility for diagnosing or managing perimenopause specifically. Winona's clinician-reviewed guide states that research has not shown the DUTCH test helps diagnose perimenopause, guide treatment decisions, or manage its symptoms, and that no medical society recommends it for that purpose. And there is a practical dosing issue that undercuts the test's usefulness regardless of how sophisticated its metabolite picture is: most clinicians prescribing hormone replacement therapy adjust estrogen doses based on how a patient feels, not based on a lab value. The elaborate metabolic detail a DUTCH panel provides currently has no clear pathway into that decision-making process in mainstream practice. The science behind it may be worth continued study. It simply is not, at this point, clinically actionable for perimenopause management in the way its marketing often implies.

How perimenopause is diagnosed

If no single test perfectly captures a hormone system that will not hold still, and even the gold-standard method (blood) only produces a snapshot, what is testing actually for? Perimenopause is diagnosed mostly through symptoms rather than a lab value, and the diagnostic process starts there. The British Menopause Society holds that for women over 45, perimenopause is better diagnosed by symptoms than by blood tests. UK NICE guidelines (NG23) back that position, recommending clinical diagnosis based on symptoms for women over 45 and reserving FSH testing for women between 40 and 45, or for investigating premature ovarian insufficiency under 40, where two FSH tests are run four to six weeks apart.

That does not make testing pointless. It gives testing a specific, narrower job: ruling out conditions that mimic perimenopause closely enough to cause real diagnostic confusion. Thyroid dysfunction is the clearest example, producing fatigue, weight changes, brain fog, mood shifts, hair thinning, and temperature dysregulation, nearly the full symptom list associated with the menopause transition. A TSH test can eliminate or confirm thyroid involvement in a way that a symptom inventory alone cannot, and that distinction has real consequences for how a woman gets treated.

Testing also supplies context that gives symptoms a clearer shape. A day-21 progesterone reading below 16 nmol/L signals that ovulation did not happen consistently that cycle, and an FSH reading that comes back elevated on one occasion, even if a later reading looks normal, can still confirm that the transition is underway when read alongside a symptomatic picture. None of this functions as a standalone verdict. The clinical marker for early perimenopause is a persistent variation of seven or more days across consecutive cycle lengths, not a lab value, and late perimenopause is marked by a stretch of amenorrhea, short of 12 full months, within the past year. A single set of lab results captures one moment in that pattern. It does not replace tracking the pattern itself.

What happens when that distinction gets lost matters beyond the testing conversation itself. Dr. Brooke's commentary and the broader clinical record both point to a real cost: perimenopausal women are frequently diagnosed and treated for anxiety, depression, or other mood disorders when the underlying driver is hormonal, a pattern that inadequate menopause training in medical education has helped sustain. Testing will not fix that gap by itself, but a thoughtfully timed panel, read in the context of a woman's symptom history rather than in isolation, gives a clinician one more way to catch what a symptom checklist alone might miss. That is the role testing actually plays, and it sets up the practical question of how to get a blood panel that delivers on that role.

Getting useful results from blood testing

A blood panel only produces useful information when the right markers are requested and the draw lands at the right point in the cycle. Most standard panels ordered through a general practitioner are too narrow to capture the full picture a perimenopausal woman actually needs, often checking a hormone or two in isolation rather than the fuller set that reveals how ovarian and adrenal function relate to each other.

Estradiol and FSH are best measured on day 3 or day 5 of the cycle, when estradiol sits at its lowest baseline and gives the clearest read on whether FSH is elevated. Progesterone should be checked on day 19, or seven days after confirmed ovulation, since that is the window where a true peak, rather than a value depressed simply by poor timing, can be assessed. For women whose cycles have become too erratic to hit a specific day with any confidence, scheduling the draw after a prolonged gap without a period offers a workable alternative, since in those cases a reasonable window still allows ovulation and cycle timing to be assessed.

A few practical details change how reliable the results turn out to be. Testing in the early morning, between roughly 8 and 10 a.m., catches hormones closer to their daily peak. Biotin supplements should be paused before the draw, since biotin interferes with many of the assays used to measure these hormones and can distort results without the patient realizing it. Women on hormonal contraceptives or hormone replacement therapy should also know that both suppress FSH and LH, making those two markers unreliable indicators of underlying ovarian function while the medication is active. A woman who wants to understand what her own ovaries are doing, independent of what a contraceptive is doing to her cycle, faces a genuine limitation there, and that tradeoff is worth raising directly with a clinician before ordering a panel rather than discovering it after results come back and do not make sense.

None of these details are complicated on their own. A woman who asks for the full multi-hormone panel, times the draw deliberately, and brings her cycle history into the conversation is giving blood testing the conditions it needs to do what it does best.