News Center
The timing question has a clearer answer than most online discussion suggests. Professional guidelines from the American Society for Reproductive Medicine and ACOG recommend evaluation after six months of trying at age 35 or older, and prompt baseline testing at 40 and beyond. The reason is arithmetic: ovarian reserve declines on its own schedule, so the signals worth acting on are the ones that appear in your labs and cycle patterns, not the ones appearing in someone else's success story.
Two documents anchor the timing conversation in the United States. The American Society for Reproductive Medicine (ASRM) advises women under 35 to seek evaluation after twelve months of unprotected intercourse without conception, women 35 to 39 after six months, and women 40 and older to consider evaluation right away. ACOG's clinical guidance on infertility workup mirrors this age-graded structure. These timelines exist because the probability of spontaneous conception per month drops across the late thirties, and because diagnostic testing takes time to arrange and interpret.
IVF enters the picture as a treatment option that also obeys the same age-graded logic. Data published in the CDC's national assisted reproductive technology surveillance reports show that live birth rates per cycle decline steadily with the age of the egg, from roughly the low forties in percentage terms for women under 35 down to single digits by the early forties, with wide variation between individuals and clinics. The exact figures vary by year and by how outcomes are measured, and readers should consult the CDC ART reports and individual clinic outcomes published through SART for current numbers. What the aggregate data make plain is the direction of the curve. Every birthday moves the expected yield per cycle downward, and the move is not linear in a gentle way. The slope steepens in the late thirties and steepens again around 40.
Guidelines, of course, describe populations. Your own situation may argue for earlier or later action, which is exactly why the two decision traps described in this article matter. Most women who delay IVF longer than their data justify are not making a careless choice. They are responding to information that looks credible and feels reassuring, and the two most common sources of that reassurance deserve close examination.
Ask a room of women in their late thirties when they think IVF should begin, and the answers cluster around a feeling of readiness rather than a number. Readiness is a reasonable thing to want. The problem is that the modern information environment systematically inflates the sense of how much time is available, through a mechanism that psychologists have documented for decades and that social platforms have amplified to an unusual degree.
Survivorship bias describes the error of drawing conclusions from the cases that succeeded while the cases that failed remain invisible. Fertility content online is a textbook environment for it. A woman who conceived naturally at 44 has a story worth posting, a pregnancy announcement worth sharing, and a comment section full of people asking for her routine. A woman who did four cycles at 43 without a viable pregnancy usually posts nothing at all. The silence is not evidence that failure is rare. It is evidence that failure is quiet.
The practical consequence is a distorted mental sample. Scroll through a fertility hashtag for twenty minutes and you will see a stream of late-age successes, each one presented with an implied lesson, whether it is a supplement stack, a diet, a mindset shift, or simply faith. Your brain, doing what brains do, updates its estimate of how common these outcomes are. The estimate moves in exactly the wrong direction, because the denominator, the number of women who tried and did not succeed at the same age, never appears on screen.
Clinical data fills in the missing denominator. Published studies of preimplantation genetic testing report that the proportion of embryos that are chromosomally normal falls from roughly sixty percent in the mid-thirties to around forty percent near 39 and closer to twenty percent by 42, with variation across studies and laboratories. The CDC's aggregate cycle data show the same downward drift at the level of live births. Neither body of data appears in a celebration post. Both appear in the decision you actually have to make.
Content platforms optimize for engagement, and engagement follows emotion. A 25-year-old's straightforward IVF cycle generates fewer clicks than a 45-year-old's miracle. The recommendation systems learn this quickly. A single view of one late-age success story leads to more of them, and within a week a user's feed can look like statistical proof that age matters far less than people claim.
There is a second, subtler distortion. Stories that involve reversal of a poor prognosis, a low AMH that rebounded, a failed clinic elsewhere followed by success at a new one, travel for treatment against a doctor's caution, carry an extra charge of drama, and drama travels furthest. The content that survives the algorithmic filter is disproportionately the content describing exceptions. A woman building her sense of the possible from this feed is building it from the tail of the distribution while believing she is seeing the middle.
A useful habit: whenever a story moves you, ask what the person's starting age was, what their diagnostic numbers were, and how many attempts the journey took. Success posts rarely volunteer the denominator. The absence itself is information.
Natural conceptions at 44 and beyond do happen, and they are real events in the lives of the women who experience them. The benchmark question is whether one woman's outcome should update your planning. Population-level data on spontaneous pregnancy at 44 show monthly probabilities in the low single digits, and the pregnancies that do occur carry elevated risks of chromosomal abnormality and miscarriage. A single celebrated case tells you the event is possible. It tells you nothing about the probability, and probability is the only quantity a decision can use.
The comparison becomes more misleading still when the story is used to argue against acting. "My neighbor conceived at 45 without any help" is offered as a reason to wait, when it functions just as easily as a reason to test. If natural conception at an advanced age is plausible enough to cite, the ovarian reserve that makes it possible is measurable, and measuring it costs little. The anecdote argues for information, and information is precisely what waiting defers.
While the feed supplies stories of other people's bodies, your own body publishes a quieter report. Several signals correlate with declining ovarian reserve or with a narrowing window, and each of them is easy to dismiss precisely because none of them is painful. These are the signals that should trigger a lab draw and an ultrasound rather than another month of observation.
A regular cycle that has drifted from 30 days to 26 or 25 days over the course of a year is often described casually as "still regular," and in a narrow sense it is. Shortening cycles can reflect a shortening follicular phase, which in turn can reflect a diminishing pool of follicles competing for the same hormonal signals. The change is gradual enough to normalize. Many women notice it only in retrospect, when they look back at a period tracker and see the trend line. A cycle that has shortened by several days and stayed short is worth a conversation with a gynecologist, ideally one who will order an antral follicle count and anti-Müllerian hormone level rather than simply noting that the cycle is still within a normal range.
Menstrual volume varies naturally, and a single light period means little. A pattern of progressively lighter flow over several cycles, particularly when combined with shortening cycle length, belongs to the same cluster of changes associated with declining ovarian function. It is not diagnostic on its own. It is a prompt. The cost of checking is an appointment and two tests. The cost of assuming it is nothing is measured in months, and months are the currency that matters here.
Women who track basal body temperature often learn to read their charts the way sailors read weather. A typical ovulatory chart shows a pre-ovulation plateau, a clear thermal shift, and a sustained post-ovulation elevation. Charts that become erratic, with a shift that arrives later each month, a luteal phase that shortens below ten days, or temperatures that wander without a readable pattern, can reflect ovulatory function under strain. Charting is imperfect science and no one should diagnose themselves from a thermometer. A chart that used to be legible and has stopped being legible is, at minimum, a reason to bring data to a physician instead of waiting for the chart to fix itself.
Few phrases in reproductive medicine carry more hidden weight than a physician's suggestion to keep observing. Sometimes it is exactly the right advice, as when all testing is normal and the timeline is long. Sometimes it functions as permission to postpone a decision the patient already senses is approaching, and the postponement quietly becomes a year.
The way to use this sentence well is to convert it into a dated checkpoint. "Keep observing" should come with three specifics: which values are being observed, what change would end the observation, and when the follow-up visit happens. A woman told at 38 to keep observing who books a repeat AMH and ultrasound for six months later has turned a vague instruction into a monitoring plan. A woman who hears the same sentence and simply goes home has often accepted an open-ended delay without noticing that she did. Observation without a checkpoint is indistinguishable from waiting, and waiting has a price that the next section makes explicit.
Among the reasons women give for postponing IVF, the most sympathetic is also the most expensive: the plan to arrive at treatment in better condition. Three months of herbal regulation, six months of weight loss, a job change to reduce stress. Each plan is individually reasonable, and each one consumes the same resource it is meant to protect.
The clearest way to see the cost is to follow one variable: the number of usable embryos a cycle can produce. IVF outcomes depend heavily on having enough embryos to survive the losses at each stage, retrieved eggs that fertilize, embryos that develop to blastocyst, blastocysts that test as chromosomally normal. Each stage loses a fraction, and the fraction lost grows with age.
Consider a simplified illustration using the broad patterns reported in published PGT-A studies and CDC aggregate data, with all figures approximate and individual variation substantial. A woman of 37 undergoing a typical retrieval might expect a certain number of blastocysts, of which a majority test normal, leaving her with a reasonable pool from which to select a transfer. The same retrieval at 39 yields somewhat fewer eggs, a lower blastocyst rate, and a normal-embryo fraction closer to forty percent. At 42, the normal fraction approaches twenty percent, which means a woman may need several retrievals to accumulate the same selection pool she would have had at 37 in one or two.
| Age at retrieval | Approximate share of embryos testing chromosomally normal | What it means for planning |
|---|---|---|
| Mid-30s | Roughly 50 to 60 percent in many published series | A single retrieval often produces a workable selection pool |
| Around 39 | Roughly 35 to 45 percent | One to two retrievals commonly needed for the same pool |
| Around 42 | Roughly 15 to 25 percent | Multiple retrievals often required; selection space narrows sharply |
Figures are approximate ranges drawn from published preimplantation genetic testing literature; individual results vary widely and current clinic-specific data should be confirmed with each center and with the CDC's national ART reports.
The table explains why delay is expensive in a way that feels invisible while it happens. A woman who waits two years does not merely arrive two years older. She arrives needing more cycles to reach the same position, which means more stimulation medications, more procedures, more cost, and more of the emotional bandwidth that every cycle consumes. The waiting that was supposed to conserve resources ends up spending them.
Every delay strategy competes against a baseline it rarely acknowledges: the option of starting now. Starting now does not mean transferring embryos now. It means retrieving information and, where appropriate, banking embryos at the current age, then proceeding with life improvements in parallel. The two paths can be compared directly.
Path one is sequential: improve first, treat later. Its total duration is the improvement period plus the treatment period, and the treatment happens at an older biological age. Path two is parallel: begin evaluation and, if indicated, cycle now while the improvement plan runs alongside. Its treatment begins at the current age, and the improvement work is not lost, because egg quality interventions, where they help at all, and general health optimization still apply to the cycles that follow. The sequential path preserves the feeling of order. The parallel path preserves the ovarian biology. Only one of these is a renewable resource.
There is a financial dimension as well, stated carefully and without specific figures. IVF costs in the United States vary by clinic, region, medication protocol, and the number of cycles required, and readers should rely on the published price schedules of individual centers. The structural point holds across all of them: the number of cycles a woman needs tends to rise with age at treatment, and each additional cycle carries its own full cost. Delay can therefore convert a one-cycle plan into a multi-cycle plan, and the improvement plan that justified the delay rarely offsets that difference.
Delay scripts deserve individual treatment, because some contain legitimate medicine and some contain only avoidance wearing medicine's clothes.
The distinction matters because the two feel identical from the inside. Both involve effort, discipline, and a story about doing things properly. Three tests separate them reliably.
Much of the resistance to starting comes from a misunderstanding of what starting involves. The first step in the United States is a consultation and a panel of tests: anti-Müllerian hormone, antral follicle count on ultrasound, cycle-day hormone levels, and standard preconception screening. None of this commits anyone to stimulation, retrieval, or transfer. It produces a snapshot.
The snapshot has decision value on its own. A woman who learns at 38 that her reserve is robust has bought herself genuine flexibility, and her plan to optimize health for a year can proceed with data behind it. A woman who learns her reserve is already diminished has learned the single most valuable fact in her situation, early enough for the information to change what she does. Either result is progress. The only losing outcome is remaining at zero information while the months pass.
A repeat AMH and ultrasound six to twelve months after the first panel adds the dimension a single snapshot lacks: trajectory. Reserve declines at different rates in different women, and two women with identical current numbers can face very different windows. A trend line, even a short one, converts the vague anxiety of "how much time do I have" into a measurable question with a physician's interpretation attached.
| Common belief | What the evidence supports |
|---|---|
| "IVF works whenever I decide to do it, so I can treat it as a backup." | IVF success depends strongly on the age of the eggs being retrieved. It extends the window; it does not remove the window. |
| "My mother conceived naturally at 43, so my timeline is fine." | Ovarian reserve timing varies substantially even within families. Her timeline is anecdote; yours is measurable with AMH and antral follicle count. |
| "One more year of preparation won't matter." | In the late thirties, one year moves expected embryo normalcy and per-cycle success by a meaningful margin, per published age-stratified data. |
| "I'm not pregnant because I'm too stressed." | Stress levels in infertile women largely match those of fertile peers before diagnosis. Attribution errors here delay evaluation of physical causes. |
| "My periods are regular, so my eggs are fine." | Regular cycles can persist while reserve declines. Cycle regularity and ovarian reserve are related but distinct measures. |
The general thresholds, drawn from ASRM and ACOG guidance and adjusted for common situations, look like this:
One more entry belongs on the list. If you have spent months consuming fertility content and feeling simultaneously reassured and anxious, that state of mind is itself a signal. It means the question is live, the information you are gathering is unverified, and the cheapest way to resolve both problems is a single appointment that replaces the feed with your own numbers.
Once the decision is made, the path forward is concrete: a consultation, a baseline panel, and a treatment plan built around your specific numbers rather than around anyone else's story. The two decision traps described here, the distorted urgency created by survivorship-heavy feeds and the quiet expense of open-ended preparation, lose most of their power the moment real data replaces estimated odds. The women who fare best are rarely the ones who waited until they felt ready. They are the ones who found out, early, what they were working with.
