For about thirty years, “luteal phase defect” (LPD) was a routine diagnosis in fertility clinics. The story went: ovulation happens, but the corpus luteum produces too little progesterone, the endometrium doesn’t develop properly, and embryos can’t implant. The fix was supposed to be simple — supplement with progesterone, problem solved.
Then the evidence caught up. In 2015, the American Society for Reproductive Medicine (ASRM) Practice Committee published an opinion that effectively retired LPD as a stand-alone clinical diagnosis. This post walks through what the original idea was, why it didn’t hold up, what a short luteal phase actually predicts, and when it’s still worth investigating.
What “luteal phase defect” originally meant
The luteal phase is the stretch from ovulation to the next period — usually 11 to 14 days. It is named for the corpus luteum, the temporary structure left behind after ovulation that produces progesterone. Progesterone matures the uterine lining so a fertilized egg can implant.
The classical LPD hypothesis was that some women ovulate but don’t make enough progesterone for long enough, and so:
- The luteal phase is short (under 10 days), or
- Progesterone levels are low in mid-luteal phase, or
- The endometrium looks “out of phase” on biopsy.
For decades, endometrial biopsy with histologic dating (the Noyes criteria, from 1950) was the gold-standard test. Patients had a biopsy in late luteal phase, and a pathologist judged whether the endometrium was at the expected developmental stage.
Why the diagnosis fell apart
Three large lines of evidence undermined LPD as a real entity:
1. The biopsy test was unreliable
A landmark NICHD-funded multicenter study (Coutifaris et al., Fertil Steril 2004) compared endometrial biopsies from fertile women and infertile women. “Out of phase” biopsies were as common in fertile women as in infertile women. Inter-observer agreement among pathologists was also poor — the same biopsy could be dated differently by different readers.
If a “diagnostic test” cannot distinguish patients who have the disease from those who do not, it is not a diagnostic test. The biopsy fell out of clinical use after this paper.
2. Short luteal phases are common in fertile women
Schliep et al. (Fertil Steril 2014), using the BioCycle Study, tracked luteal phase length across hundreds of cycles in healthy, regularly cycling women. A meaningful fraction had at least one luteal phase under 10 days — but most still conceived without difficulty. An isolated short luteal phase is a normal variant, not a disease state.
3. Empiric progesterone does not help most natural-conception cycles
Cochrane reviews of luteal-phase progesterone in non-IVF cycles have repeatedly failed to show a live-birth benefit when given empirically for “LPD” in women trying to conceive naturally. Progesterone clearly helps in IVF (where pituitary suppression disrupts the natural luteal phase) and may help in some recurrent miscarriage cases — but those are different clinical scenarios.
The 2015 ASRM Practice Committee Opinion summarized this: there is no reliable test for LPD, no proven empiric treatment, and the diagnosis itself does not predict outcomes well in natural cycles.
What a short luteal phase actually predicts
This is the part most people miss. The 2015 ASRM opinion did not say “short luteal phases are fine.” It said the historical diagnosis (LPD) is not a reliable clinical entity. A persistently short luteal phase (under 9 days, observed across multiple cycles) is still a useful signal — but it is usually a downstream sign, not the root cause.
Causes of recurring short luteal phases include:
- Suboptimal ovulation. A small or weak corpus luteum makes less progesterone for fewer days. This is the most common explanation.
- Approaching perimenopause. Luteal phases shorten in the years before menopause as ovarian reserve declines and follicular phases shift.
- Thyroid dysfunction. Both hypo- and hyperthyroidism affect cycle architecture. (See our post on thyroid and menstrual cycles.)
- Hyperprolactinemia. Elevated prolactin shortens the luteal phase and can suppress ovulation.
- Hypothalamic suppression. Stress, energy deficiency, or excessive exercise can shorten the luteal phase before they cause missed cycles. See hypothalamic amenorrhea.
In other words, a 7-day luteal phase is worth investigating — not because “LPD” is the diagnosis, but because it’s a clue pointing toward something else.
How short is short?
The thresholds most reproductive endocrinologists use:
- 11 to 14 days: typical, normal.
- 10 days: low end of normal. Most clinicians don’t act on this in isolation.
- Under 10 days, occasional: a normal variant — Schliep et al. observed this in fertile women.
- Under 9 days, recurring (2 to 3 cycles): worth investigating for an upstream cause.
- Under 7 days, recurring: definitely worth a workup.
To know your luteal phase length, you need to know when you ovulated. Calendar estimates are not precise enough. The reliable methods are basal body temperature (which shifts at ovulation) and LH-strip-confirmed ovulation. We cover those in tracking BBT for conception and BBT vs LH vs mucus. The Ovulation Calculator gives you a starting estimate, but for luteal-phase questions, BBT charting is the home-friendly gold standard.
When to investigate
If your luteal phase is consistently under 9 days across 2 to 3 charted cycles, talk to a clinician. The standard workup is not “treat for LPD” — it’s:
- TSH and free T4 to rule out thyroid disease.
- Prolactin to rule out hyperprolactinemia.
- AMH and antral follicle count if there is concern for diminished ovarian reserve. See our post on AMH and ovarian reserve.
- Mid-luteal progesterone (day 21 in a 28-day cycle, or 7 days after ovulation) to confirm ovulation occurred. A value above 3 ng/mL confirms ovulation; values above 10 ng/mL suggest a robust corpus luteum.
- Energy availability and stress assessment — especially in athletes, dancers, and people with low body weight or restrictive eating.
Treatment, if any, targets the underlying cause: levothyroxine for hypothyroidism, dopamine agonists for hyperprolactinemia, refeeding and reduced training load for hypothalamic suppression. Empiric progesterone is not first-line.
The IVF and recurrent miscarriage exceptions
Two scenarios where progesterone support is genuinely standard of care:
IVF cycles. GnRH agonist or antagonist protocols suppress the pituitary and can blunt the natural luteal phase. Vaginal or intramuscular progesterone after embryo transfer is well established and improves live-birth rates in IVF. This is not “treating LPD” — it’s replacing hormones the protocol suppressed.
Recurrent first-trimester miscarriage. The PROMISE trial (Coomarasamy et al., NEJM 2015) found no benefit of empiric vaginal progesterone for unexplained recurrent miscarriage. The follow-up PRISM trial (NEJM 2019) found a modest benefit in women with early bleeding and prior miscarriages. Both are nuanced — talk to a reproductive endocrinologist before starting progesterone.
For the typical person trying to conceive naturally with regular cycles, neither situation applies, and empiric progesterone is not indicated.
What this means for your tracking
If you chart your cycles, you do not need to worry about a stray short luteal phase. The pattern matters:
- One cycle with a 9-day luteal phase: normal variation.
- Three out of four cycles with luteal phases under 9 days: bring this to a clinician.
- Luteal phase of 11 to 14 days, no concerns about conception: routine. Keep tracking if it is useful to you, but no investigation needed.
This is one of those areas where medical thinking has changed substantially in the last decade. If you read older blog posts, fertility forums, or even older textbooks, you will see LPD discussed as a primary diagnosis with progesterone as the fix. The 2015 ASRM opinion (still the guiding document in 2026) reframed it: LPD is mostly a label that hides the real question, which is why the luteal phase is short.
The bottom line
The luteal phase is a useful number to know. A persistently short luteal phase is a real signal, worth bringing to a clinician. But “luteal phase defect” as a stand-alone diagnosis with empiric progesterone as treatment is no longer supported by ASRM, and the underlying evidence does not support it. Investigate the cause, treat the cause, and skip the unnecessary progesterone.
If you want to start charting luteal phase length, the Ovulation Calculator is the simplest entry point, and our BBT charting guide covers the home method that gives you the cleanest data.