← Back

Fertility

Everything in the library studied for it, medical and holistic together. Within each kind, what was shown to help comes first, strongest evidence first. What was tested and not shown to help stays on the list.

Strong several reviews agree · Moderate one review or consistent trials · Limited one or a few small studies · Untested no study in people, listed with what it rests on

Medication14 Hormone therapy6 Supplements10 Therapies6 Lab markers14

Medication

Letrozole and clomipheneStrongStudied in womenVery common

Live birth in PCOS: letrozole compared with clomiphene

Shown to raise live births in women with PCOS against clomiphene, 27.5% against 19.1% in the largest trial, about one extra birth per 10 women treated.

Compared with clomiphene or other selective estrogen receptor modulators [3]; clomiphene [4][5]

The numbers and sources

How common: Very common: more than 1 in 10 people.
Effect: Across 11 trials (2,060 women), live births after ovulation induction and timed intercourse were higher with letrozole than with selective estrogen receptor modulators such as clomiphene (OR 1.72, 95% CI 1.40 to 2.11, I² 0%), with a number needed to treat of 10. Where 20% of women would have a live birth on clomiphene, 27% to 35% would on letrozole [3]. A second meta-analysis, with 8 trials reporting live birth, found a risk ratio of 1.49 (95% CI 1.27 to 1.74) [4]. In the PPCOS II trial, 103 of 374 women (27.5%) on letrozole and 72 of 376 (19.1%) on clomiphene had a live birth over up to five cycles (rate ratio 1.44, 95% CI 1.10 to 1.87) [5].
Population: 2,060 women with anovulatory PCOS in 11 trials, within a review of 41 trials and 6,522 women [3]. 3,952 women with PCOS in 29 trials [4]. 750 women aged 18 to 40 with PCOS, double-blind, at US centers [5].
Sources: [3], [4], [5]
Caveats: The Cochrane review set aside six earlier trials over concerns about the validity of their data, and a funnel plot showed mild asymmetry suggesting some trials favoring the comparison drugs may be missing [3]. The authors of the second review describe the quality of the evidence as mixed [4].

Letrozole and clomipheneStrongStudied in womenVery common

Ovulation and pregnancy in PCOS: letrozole compared with clomiphene

Shown to raise ovulation and pregnancy rates in women with PCOS against clomiphene. In the largest trial, 61.7% of letrozole cycles and 48.3% of clomiphene cycles were ovulatory.

Compared with clomiphene [4][5][6]; clomiphene or other selective estrogen receptor modulators [3]

The numbers and sources

How common: Very common: more than 1 in 10 people.
Effect: Ovulation was more frequent with letrozole across 22 trials (RR 1.14, 95% CI 1.06 to 1.21) [4] and across 50 trials (RR 1.20, 95% CI 1.13 to 1.26), and pregnancy across 50 trials (RR 1.44, 95% CI 1.28 to 1.62). The endometrium was thicker on letrozole (SMD 0.89) and fewer follicles developed (SMD -0.56) [6]. In PPCOS II, 834 of 1,352 letrozole cycles (61.7%) and 688 of 1,425 clomiphene cycles (48.3%) were ovulatory [5]. Clinical pregnancy was higher with letrozole across 23 trials (OR 1.69, 95% CI 1.45 to 1.98): where 24% would conceive on clomiphene, 32% to 39% would on letrozole [3].
Population: Women with PCOS in 22 trials [4] and 50 trials [6]. 750 women with PCOS [5]. 3,321 women in 23 trials [3].
Sources: [3], [4], [5], [6]
Caveats: Heterogeneity was moderate to high in the 50-trial analysis (I² 54% for ovulation, 66% for pregnancy, 98% for endometrial thickness) [6]. Ovulation and pregnancy are steps toward a live birth, which is reported in the block above.

Hormonal birth controlModerateStudied in women

Fertility after stopping

Shown that 83% of women became pregnant within 12 months of stopping contraception, with no difference between hormonal methods and IUDs or by length of pill use.

Compared with other contraceptive methods, including IUDs [10]

The numbers and sources

Effect: The pooled pregnancy rate within 12 months of stopping was 83.1% (95% CI 78.2 to 88). It did not differ significantly between hormonal methods and IUDs, and neither progestin type nor length of pill use changed it [10]. Pregnancies occurred as early as 7 to 14 days after implant removal [11].
Population: 14,884 women in 22 studies [10]; implant trial participants [11].
Sources: [10], [11]
Caveats: Studies varied, and the review included the contraceptive injection, which is not covered on this page.

Ulipristal acetateModerateStudied in women

Emergency contraception (ella)

Shown to lower pregnancies after unprotected sex compared with levonorgestrel emergency pills, 1.8% against 2.6% within 72 hours in the largest trial.

Compared with levonorgestrel emergency contraception [7][8]

The numbers and sources

Effect: Pooled, ulipristal was linked to fewer pregnancies than levonorgestrel (RR 0.59, 95% CI 0.35 to 0.99, 2 RCTs, 3,448 women) [7]. In the larger trial, pregnancies occurred in 1.8% on ulipristal and 2.6% on levonorgestrel when taken within 72 hours (OR 0.68, 95% CI 0.35 to 1.31). Between 72 and 120 hours, all 3 pregnancies were in the levonorgestrel group [8]. In women with BMI over 30, the pregnancy rate on ulipristal was 3.1%, not significantly below the 4.5% expected without emergency contraception [9].
Population: 2,221 women with regular cycles seeking emergency contraception within 5 days [8]; 115 emergency contraception trials with 60,479 women, 2 of them for this comparison [7]; women in the label trials [9].
Sources: [7], [8], [9]
Caveats: The single trial on its own did not reach significance [8]. This is a single 30 mg dose, not the 5 mg daily fibroid dose.

Hormonal birth controlLimitedStudied in women

Preventing pregnancy

Linked to 4.55 unintended pregnancies per 100 women a year on pills, patch or ring, against 0.27 on IUDs and implants, in a large US cohort.

Compared with IUDs and implants [1]; no comparison group in the label trials [2][3]

The numbers and sources

Effect: The failure rate was 4.55 per 100 participant-years on pills, patch or ring and 0.27 on long-acting methods (adjusted HR 21.8, 95% CI 13.7 to 34.9). Women under 21 on pills, patch or ring had almost twice the risk of older women [1]. In the label trial of drospirenone 3 mg with ethinyl estradiol 0.02 mg, the Pearl Index was 1.41 pregnancies per 100 woman-years [2]. In Mirena's extension, the cumulative pregnancy rate over years 6 to 8 was 0.68% [3].
Population: 7,486 women given their choice of method at no cost in St. Louis [1]; 1,027 women aged 17 to 36 [2]; Mirena extension users [3].
Sources: [1], [2], [3]
Caveats: An observational cohort in which women chose their method [1]. Label trials have no control group.

Letrozole and clomipheneLimitedStudied in women

Pregnancy in PCOS: clomiphene compared with placebo

Limited evidence: clomiphene raised clinical pregnancy against placebo in women with PCOS, in 3 small trials of 133 women. No placebo-controlled trial reported live birth.

Compared with placebo [7]; gonadotropin injections [7]

The numbers and sources

Effect: Clomiphene was linked to more clinical pregnancies than placebo (OR 5.91, 95% CI 1.77 to 19.68). If 5% conceived on placebo, between 8% and 50% would on clomiphene. No live birth, miscarriage, multiple pregnancy or ovarian hyperstimulation data were reported against placebo. Against a gonadotropin, clomiphene was linked to fewer pregnancies, ongoing pregnancies and live births, in limited evidence [7].
Population: 133 women in 3 trials, within a review of 28 trials and 3,377 women with anovulatory subfertility [7].
Sources: [7]
Caveats: The reviewers rated the evidence low to very low in quality, mainly for imprecision and poor reporting, and advise caution because of low event rates [7].

Letrozole and clomipheneModerateStudied in womenCommon

Multiple pregnancy

The clomiphene label reports 8% of pregnancies in its trials were multiple. Pooled PCOS trials found multiple pregnancy in 2.2% of women on clomiphene and 1.6% on letrozole.

The numbers and sources

How common: Common: between 1 in 100 and 1 in 10 people.
Effect: Of 2,369 pregnancies with known outcome in the clomiphene trials, 7.98% were multiple: 6.9% twins, 0.5% triplets, 0.3% quadruplets and 0.1% quintuplets [1]. In pooled PCOS trials, multiple pregnancy occurred in 2.2% on selective estrogen receptor modulators and 1.6% on letrozole (OR 0.74, 95% CI 0.42 to 1.32) [3]. In PPCOS II, 7.4% of clomiphene pregnancies and 3.4% of letrozole pregnancies were twins, not a significant difference [5]. In AMIGOS, multiple gestations made up 9% of ongoing clomiphene pregnancies, 13% of letrozole pregnancies and 32% of gonadotropin pregnancies. All those on the oral drugs were twins, while gonadotropins produced 24 twin and 10 triplet pregnancies [12]. Twin pregnancy did not differ between letrozole and clomiphene in unexplained infertility (RR 0.81, 95% CI 0.39 to 1.68) [10].
Population: 2,369 pregnancies in uncontrolled label trials [1]. 2,247 women in 14 trials [3]. 750 women with PCOS [5]. 900 women with unexplained infertility [12].
Sources: [1], [3], [5], [12], [10]
Caveats: None of these trials had an untreated group, so the rates show how often multiple pregnancy occurred on each drug, not how much each drug raised it.

Letrozole and clomipheneModerateCommon

Birth defects and the letrozole pregnancy warning

The letrozole label states it can cause fetal harm and contraindicates use in pregnancy. Pooled fertility trials found no rise in birth defects against clomiphene, at very low certainty.

The numbers and sources

How common: Common: between 1 in 100 and 1 in 10 people.
Effect: The letrozole label cites postmarketing reports of miscarriage and birth defects after use during pregnancy, and malformations in rats and rabbits at doses below the human dose, while stating the data are insufficient to give a drug-associated risk. It puts the US background rate of major birth defects at 2% to 4% [2]. Across 20 fertility trials, birth defects did not differ between letrozole and clomiphene (RR 1.69, 95% CI 0.51 to 5.58, very low certainty) [16]. In 911 newborns from Canadian fertility centers, anomalies occurred in 2.4% after letrozole and 4.8% after clomiphene, major ones in 1.2% and 3.0%, and heart anomalies in 0.2% and 1.8% [17]. PPCOS II recorded four major anomalies on letrozole and one on clomiphene (P = 0.65) [5]. AMIGOS found no significant differences in anomalies between letrozole, clomiphene and gonadotropins [12]. The clomiphene label states human data do not suggest a rise above the background rate, while animal studies showed malformations, and it contraindicates use in pregnancy [1].
Population: 6,679 patients in 20 trials [16]. 911 newborns [17]. 750 women [5]. 900 women [12].
Sources: [2], [16], [17], [5], [12], [1]
Caveats: The newborn study was retrospective [17]. In fertility use, letrozole is taken on cycle days 3 to 7, before ovulation [15], while the label warning concerns exposure during pregnancy [2]. Trials of a few hundred pregnancies cannot rule out a rise in a rare defect.

LevothyroxineModerate

Pregnancy outcomes with subclinical hypothyroidism or thyroid antibodies

Mixed results: fewer preterm and low birth weight births in one review, lower miscarriage in women with thyroid antibodies, and no difference in an NIH trial.

The numbers and sources

Effect: In randomized trials of women with subclinical hypothyroidism in pregnancy, levothyroxine was associated with less preterm birth (RR 0.56, 95% CI 0.41 to 0.77) and fewer low birth weight infants (RR 0.56, 95% CI 0.35 to 0.89). Small for gestational age and macrosomia did not differ [5]. In euthyroid pregnant women with thyroid peroxidase antibodies, miscarriage was lower (RR 0.78, 95% CI 0.63 to 0.98), while preterm birth (RR 0.78, 95% CI 0.55 to 1.13) and live birth (RR 1.05, 95% CI 0.99 to 1.12) did not differ [6]. In subclinical hypothyroidism, gestational hypertension (OR 1.03, 95% CI 0.85 to 1.25) and pre-eclampsia (OR 1.02, 95% CI 0.66 to 1.58) did not differ from controls [7].
Population: 30 RCTs and cohort studies with 18,568 pregnant women with subclinical hypothyroidism [5]; 8 RCTs with 1,645 women [6]; trials and observational studies in pregnant women with subclinical or overt hypothyroidism or thyroid autoimmunity [7].
Sources: [5], [6], [7]
Caveats: [5] pooled RCTs with cohort studies, and the figures above are its RCT estimates. Preterm birth fell in [5] but not in [6], which studied a different group. The NIH trial of treatment from mid-pregnancy found no difference in pregnancy outcomes [8].

Ulipristal acetateLimitedStudied in women

Pregnancy after ulipristal for fibroids

Case reports: of 71 pregnancies after ulipristal for fibroids, 50 ended in live births and 19 in miscarriage, with no comparison group.

Compared with no comparison group in the case series [10]

The numbers and sources

Effect: A systematic review found 24 pregnancies after ulipristal (19 live births, 6 miscarriages), and a case series added 47 (31 live births, 13 miscarriages, 1 fetal death, 2 terminations, 1 ongoing). Of the 71, 44 had myomectomy after ulipristal. Five women who did not have surgery had delivery complications related to their fibroids [10].
Population: 71 pregnancies from 6 case reports, 1 retrospective series and a multicenter chart review [10].
Sources: [10]
Caveats: No comparison group, so these counts cannot show whether ulipristal changed the chance of pregnancy or miscarriage.

Letrozole and clomipheneStrongStudied in women

Unexplained infertility: letrozole compared with clomiphene

Not shown to differ from clomiphene in women with unexplained infertility, in two meta-analyses. One large trial found live birth rates of 18.7% on letrozole and 23.3% on clomiphene.

Compared with clomiphene [10][11][12]; gonadotropin injections [12]

The numbers and sources

Effect: Across 8 trials, a positive pregnancy test occurred in 24% on letrozole and 23% on clomiphene (RR 1.08, 95% CI 0.85 to 1.36). Clinical pregnancy (RR 1.15, 95% CI 0.71 to 1.85) and live birth (RR 0.94, 95% CI 0.83 to 1.08) did not differ [10]. In insemination cycles, total pregnancy did not differ (RR 1.06), and miscarriage was lower with letrozole (RR 0.61) [11]. In the AMIGOS trial of stimulation with insemination for up to four cycles, live birth rates were 32.2% with gonadotropins, 23.3% with clomiphene and 18.7% with letrozole. Letrozole's pregnancy rate was lower than gonadotropins' (P < 0.001) and did not differ significantly from clomiphene's (P = 0.10) [12].
Population: 2,647 women with unexplained infertility in 8 trials [10]. Women with unexplained infertility having insemination, in trials the abstract does not count [11]. 900 ovulatory women aged 18 to 40 at US centers [12].
Sources: [10], [11], [12]
Caveats: Heterogeneity between trials was significant (I² 60.8%) [10]. Most of these trials paired the drug with insemination, so they do not test the drug alone.

Letrozole and clomipheneStrong

Miscarriage

Not shown to differ between letrozole and clomiphene. About one pregnancy in four ended in miscarriage on either drug in pooled PCOS trials.

The numbers and sources

How common: Very common: more than 1 in 10 people.
Effect: Miscarriage per pregnancy was 25% on selective estrogen receptor modulators and 24% on letrozole (OR 0.94, 95% CI 0.66 to 1.32, high certainty) [3]. Pregnancy loss did not differ between letrozole and clomiphene across fertility trials (RR 1.17, 95% CI 0.91 to 1.50) [16] or in unexplained infertility (RR 0.92, 95% CI 0.61 to 1.38) [10]. In PPCOS II, 31.8% of letrozole pregnancies and 29.1% of clomiphene pregnancies were lost [5]. The clomiphene label reports a miscarriage rate of 20.4% and a stillbirth rate of 1.0% in its trials [1], and the letrozole label gives a US background miscarriage rate of 15% to 20% of recognized pregnancies [2].
Population: 736 pregnancies in 15 trials [3]. 20 trials [16]. 8 trials [10]. 257 pregnancies [5]. 2,369 pregnancies in label trials [1].
Sources: [3], [16], [10], [5], [1], [2]
Caveats: These trials compare the two drugs with each other, so they cannot show whether either changes miscarriage risk against no drug.

Letrozole and clomipheneModerateStudied in women

Live birth in unexplained infertility: clomiphene compared with no treatment

Not shown to raise live births in women with unexplained infertility against no treatment or placebo. In one trial, 14% on clomiphene and 17% with no treatment had a baby.

Compared with no treatment or placebo [8]; expectant management, meaning no treatment for six months [9]

The numbers and sources

Effect: Clomiphene did not raise live births against no treatment or placebo (OR 0.79, 95% CI 0.45 to 1.38). Clinical pregnancy did not differ with insemination (OR 2.40, 95% CI 0.70 to 8.19), without it (OR 1.03, 95% CI 0.64 to 1.66), or with an hCG trigger (OR 1.66, 95% CI 0.56 to 4.80) [8]. In a Scottish trial, live births were 26 of 192 (14%) with clomiphene, 32 of 193 (17%) with expectant management and 43 of 191 (23%) with unstimulated insemination. Against expectant management, the odds ratio was 0.79 (95% CI 0.45 to 1.38) for clomiphene and 1.46 (95% CI 0.88 to 2.43) for insemination [9].
Population: 1,159 women in 7 trials, at clomiphene doses of 50 to 250 mg for up to 10 days [8]. 580 women in couples with over two years of infertility, confirmed ovulation, open tubes and motile sperm, in five Scottish hospitals [9].
Sources: [8], [9]
Caveats: The review's live birth estimate is identical to the Scottish trial's, so it may rest on that one trial. Heterogeneity between trials ranged from 34% to 58% [8].

Letrozole and clomipheneLimitedStudied in women

PCOS: letrozole compared with ovarian drilling surgery

Limited evidence: in one trial of 141 women with PCOS, live births did not differ significantly between letrozole and laparoscopic ovarian drilling.

Compared with laparoscopic ovarian drilling [3]

The numbers and sources

Effect: Live birth OR 2.07 (95% CI 0.99 to 4.32). Where 22% would have a live birth after drilling, 24% to 47% would on letrozole. Clinical pregnancy did not differ significantly across 3 trials (OR 1.47, 95% CI 0.95 to 2.28) [3].
Population: 141 women in 1 trial for live birth, 367 women in 3 trials for pregnancy [3].
Sources: [3]
Caveats: The reviewers rated the live birth evidence very low in certainty and the pregnancy evidence low [3].

Hormone therapy

DHEAModerateStudied in women

IVF and ICSI outcomes

Shown to raise clinical pregnancy rates and eggs retrieved in women having IVF or ICSI, on low-quality evidence. Live birth data were insufficient.

The numbers and sources

Effect: In women undergoing IVF or ICSI, DHEA increased endometrial thickness by a mean difference of 0.93 (95% CI 0.27 to 1.60) and clinical pregnancy rate with an OR of 1.34 (95% CI 1.08 to 1.67). Oocytes retrieved rose by 0.73 (95% CI 0.36 to 1.10). Evidence was insufficient to conclude on live birth, ongoing pregnancy or miscarriage [11].
Population: 1,973 women in 16 RCTs [11].
Sources: [11]
Caveats: Quality of evidence was rated low to very low, and heterogeneity was linked to disease type, stimulation protocol and when DHEA was started [11].

hCGModerateStudied in men

Sperm production in hypogonadotropic hypogonadism

Shown to produce sperm in about 75% of men with hypogonadotropic hypogonadism, mostly with hCG combined with FSH.

The numbers and sources

Effect: The overall success rate for producing sperm was 75% (95% CI 69 to 81), with a mean sperm concentration of 5.92 million/mL (95% CI 4.72 to 7.13). Success was higher with hCG combined with FSH, and in men with lower baseline gonadotropins. Earlier testosterone therapy did not change the results. Studies of men with pre-pubertal onset had a lower success rate than mixed groups (68% vs 84%) [2].
Population: Men with hypogonadotropic hypogonadism and azoospermia, 44 gonadotropin studies [2].
Sources: [2]
Caveats: The pooled studies combine hCG with FSH preparations, so hCG alone is not isolated. The abstract does not say how many were controlled.

hCGLimitedStudied in men

Pregnancy rates in male infertility

Limited evidence: hCG-combined therapy was linked to higher pregnancy rates in men with low sperm counts or after varicocele surgery, with wide intervals.

The numbers and sources

Effect: In men with oligospermia, 3 months of hCG-combined therapy was linked to higher pregnancy rates (OR 2.91, 95% CI 1.57 to 5.38). After varicocele surgery, hCG for 3 months was linked to higher pregnancy rates (OR 34.49, 95% CI 1.88 to 632.45). In men with severe testiculopathy there was no difference (OR 5.13, 95% CI 0.27 to 98.56) [4].
Population: 926 men in 7 RCTs [4].
Sources: [4]
Caveats: The search ended in December 2019. Intervals were wide, no pregnancy rate was reported for hypogonadotropic hypogonadism, and live births and adverse events were rarely reported [4].

hCGLimitedStudied in men

Intratesticular testosterone during testosterone use

One small trial found testosterone alone cut testicular testosterone by 94%, while adding hCG kept it near baseline at 250 to 500 IU in healthy men.

The numbers and sources

Effect: Testosterone enanthate alone suppressed intratesticular testosterone by 94%. With hCG, it rose linearly with dose: 25% below baseline at 125 IU, 7% below at 250 IU and 26% above at 500 IU [5].
Population: 29 healthy men, randomized to 3 weeks of testosterone with placebo or one of three hCG doses [5].
Sources: [5]
Caveats: One small 3-week trial. It measured hormone levels in testicular fluid, not sperm or fertility.

hCGLimitedStudied in men

Semen during testosterone therapy

Limited evidence: in a review without controls, semen did not change in 26 men on testosterone plus hCG, and none lost sperm entirely.

The numbers and sources

Effect: Semen parameters did not change over more than a year, and no man became azoospermic. Nine of 26 men contributed to a pregnancy during follow-up [6].
Population: 26 men with low testosterone, mean age 35.9, on testosterone plus hCG, mean follow-up 6.2 months [6].
Sources: [6]
Caveats: A retrospective review with no control group.

hCGModerate

IVF pregnancy with intrauterine hCG before embryo transfer

Not shown to change live birth or pregnancy rates with intrauterine hCG before embryo transfer in people having IVF.

The numbers and sources

Effect: In an individual participant data meta-analysis, intrauterine hCG before embryo transfer did not change live birth rates (OR 0.99, 95% CI 0.83 to 1.19) or clinical pregnancy rates (OR 1.04, 95% CI 0.83 to 1.31) compared with placebo or no intervention [3].
Population: 2,244 people undergoing IVF in 7 RCTs with individual data, out of 28 RCTs located [3].
Sources: [3]
Caveats: The 21 trials without individual data all failed the review's trustworthiness checks, and those trials had shown higher pregnancy rates [3]. This is hCG placed in the uterus, not the injected trigger used for ovulation.

Supplements

Coenzyme Q10StrongStudied in men

Male fertility and semen quality

Shown to raise sperm count and motility in men with unexplained infertility. Reviews disagree on sperm concentration and on the size of the effect.

The numbers and sources

Effect: In 9 studies (781 men) with idiopathic infertility, CoQ10 raised sperm concentration by 10.22 million/mL (95% CI 3.51 to 16.93) and total motility by 4.95 percentage points (95% CI 1.15 to 8.75), and raised the odds of clinical pregnancy (OR 6.02, 95% CI 1.97 to 18.41) [21]. In 8 RCTs (877 men), total sperm count, total and progressive motility and normal morphology were higher on CoQ10, and sperm concentration and semen volume did not change [22]. A network meta-analysis of 16 studies in men with unexplained infertility found CoQ10 improved sperm concentration over placebo [23].
Population: Men with idiopathic or unexplained infertility [21][23]; men in RCTs of semen quality [22].
Sources: [21], [22], [23]
Caveats: One review reported standardized mean differences as large as 13.38, several times the 2.98 reported by the network meta-analysis [23], with significant heterogeneity between studies [22]. The pregnancy interval is wide [21]. The reviews disagree on sperm concentration [21][22].

FolateModerate

Miscarriage and pregnancy rate (high dose against standard dose)

Shown to raise pregnancy rates slightly with 4 to 5 mg a day before conception against 0.4 mg. Fewer miscarriages appeared only among women who conceived.

The numbers and sources

Effect: Across 3 RCTs (5,144 women trying to conceive), 4 to 5 mg a day of folic acid before conception, against 0.4 mg, was linked to fewer miscarriages among the 1,488 who became pregnant (RR 0.73, 95% CI 0.56 to 0.94, I² 0%). Counting all randomized women, the reduction was not significant (RR 0.79, 95% CI 0.61 to 1.04). Pregnancy rate was higher on the high dose (RR 1.10, 95% CI 1.02 to 1.18) [15].
Population: Nonpregnant women of reproductive age [15].
Sources: [15]
Caveats: Certainty was rated low to moderate. Restricting to women who became pregnant breaks randomization [15].

InositolModerateStudied in women

Menstrual cycles and ovulation in PCOS

Shown to make a regular cycle about 1.8 times as likely as placebo in women with PCOS. The guideline review called the evidence limited.

The numbers and sources

Effect: A regular cycle was 1.79 times as likely on inositol as on placebo (95% CI 1.13 to 2.85), and results did not differ from metformin [1]. In an earlier analysis, ovulation rate (RR 2.3, 95% CI 1.1 to 4.7) and cycle frequency (RR 6.8, 95% CI 2.8 to 16.6) were higher than placebo [2]. A review for the 2023 international PCOS guideline found possible ovulation benefits for D-chiro-inositol and called the overall evidence limited and inconclusive [3].
Population: 1,691 women in 26 trials [1]; 601 women in 10 trials [2]; 2,230 women in 30 trials [3].
Sources: [1], [2], [3]
Caveats: Sources disagree on certainty. Pregnancy rates came from single trials, and no trial reported live births or miscarriage [2]. Ovulation heterogeneity was high (I² 75%) [2].

N-acetylcysteineModerateStudied in women

Polycystic ovary syndrome

Shown to raise progesterone and endometrial thickness in women with PCOS. These are hormone and ultrasound measures, not pregnancy rates.

The numbers and sources

Effect: Across 22 studies (2,515 women), NAC was associated with higher progesterone (SMD 0.95, 95% CI 0.13 to 1.77) and thicker endometrium (SMD 0.58, 95% CI 0.10 to 1.06) than placebo and other drugs. Estradiol, SHBG and FSH did not differ [7].
Population: Women with PCOS, compared with placebo, metformin or clomiphene [7].
Sources: [7]
Caveats: Hormone and ultrasound measures, not pregnancy or live birth rates. Comparators were mixed.

N-acetylcysteineModerateStudied in men

Male fertility and semen quality

Shown to improve sperm concentration, motility, morphology and semen volume in men with infertility, measured as semen quality rather than pregnancy.

The numbers and sources

Effect: Across 5 RCTs in men with infertility or impaired semen, NAC raised semen volume (MD 0.69 mL), sperm concentration (MD 4.43), total motility (MD 9.69) and normal morphology (MD 1.36) against control [8]. In 3 placebo-controlled trials in idiopathic male infertility, NAC improved sperm concentration, morphology, motility and ejaculation volume [9].
Population: Men with infertility or impaired semen parameters [8][9].
Sources: [8], [9]
Caveats: Few trials, likely overlapping between the reviews. Semen measures, not pregnancy or live birth. The abstract of [9] reports p-values only.

TribulusModerateStudied in men

Sperm quality

Shown to improve some sperm measures in infertile men in 6 of 7 studies, though results were not pooled and some studies had no control group.

The numbers and sources

Effect: A systematic review found 7 studies of tribulus in men with idiopathic infertility. Six reported improvement in some or all of sperm count, motility and morphology, and one uncontrolled study found no effect [10].
Population: Infertile men in clinical trials and quasi-experimental studies [10].
Sources: [10]
Caveats: The review did not pool results, and some included studies had no control arm [10].

D-aspartic acidLimitedStudied in men

Sperm quality in men with infertility

One trial found a product combining D-aspartic acid, ubiquinol and zinc raised sperm motility in infertile men. D-aspartic acid's own share is unknown.

The numbers and sources

Effect: In infertile men given 2,660 mg of D-aspartic acid with 200 mg of ubiquinol and 10 mg of zinc daily for 3 months, progressive sperm motility rose from 10.63 to 15.21 (p = 0.047) and total testosterone from 5.06 to 5.89 (p = 0.009). A placebo group took starch granules [7].
Population: 24 men per group with idiopathic infertility in Egypt [7].
Sources: [7]
Caveats: A three-ingredient product, so the effect of D-aspartic acid alone cannot be separated. The abstract reports the active group's change and does not give the between-group comparison [7].

InositolModerateStudied in women

Fertility treatment in PCOS

Mixed results: one review found higher clinical pregnancy rates in women with PCOS in fertility treatment, and a larger review found no effect.

The numbers and sources

Effect: In 17 intervention studies of women with PCOS undergoing assisted reproduction, myo-inositol or D-chiro-inositol raised the clinical pregnancy rate (RR 1.64, 95% CI 1.25 to 2.15) and top-grade embryos (RR 1.12), and lowered antral follicle count and anti-Mullerian hormone [8]. In 11 studies (981 women undergoing IVF), myo-inositol raised the mature (MII) oocyte rate (OR 1.55, 95% CI 1.04 to 2.31) and the fertilization rate (OR 1.62, 95% CI 1.21 to 2.16). The MII oocyte rate did not rise in poor responders [9]. A 2026 meta-analysis of 39 studies found fewer irregular periods on myo-inositol (RR 0.78, 95% CI 0.68 to 0.89) and no significant effect on clinical pregnancy rate [10].
Population: Women with PCOS in assisted reproduction [8][10]. Women undergoing IVF, with and without PCOS [9].
Sources: [8], [9], [10]
Caveats: The reviews disagree on clinical pregnancy [8][10]. None of the abstracts reports live births for inositol. Heterogeneity was high for several outcomes [8][10].

Coenzyme Q10LimitedStudied in women

IVF outcomes in women with poor ovarian response

One trial found CoQ10 with folic acid raised egg counts in 100 women with poor ovarian response, but pregnancy rates did not differ.

The numbers and sources

Effect: In 100 women with poor ovarian response, CoQ10 plus folic acid before and during an ICSI cycle raised oocyte count and antral follicle count against folic acid alone. Chemical and clinical pregnancy rates did not differ [24].
Population: 100 women classified as poor ovarian responders [24].
Sources: [24]
Caveats: One trial without a placebo. Pregnancy rates did not differ [24].

MacaModerateStudied in men

Semen quality

Not shown to raise sperm concentration against placebo, with mixed effects on other semen measures in infertile and healthy men.

The numbers and sources

Effect: In a meta-analysis of 5 RCTs, maca did not raise sperm concentration compared with placebo (mean difference 2.22, 95% CI -2.94 to 7.37, p = 0.4). Effects on other semen measures were mixed in infertile and healthy men [6].
Population: Infertile men and healthy men [6].
Sources: [6]
Caveats: Five small trials [6].

Therapies

Endometriosis surgeryStrongStudied in womenVery common

Pregnancy in mild endometriosis, compared with diagnostic laparoscopy

Shown to raise pregnancy rates in women with minimal or mild endometriosis against diagnostic laparoscopy alone: 30.7% against 17.7% reached 20 weeks in the largest trial.

Compared with diagnostic laparoscopy alone [4][8]; placebo, as the network analysis names its comparison [9]

The numbers and sources

How common: Very common: more than 1 in 10 people.
Effect: Surgery raised viable intrauterine pregnancy against diagnostic laparoscopy (OR 1.89, 95% CI 1.25 to 2.86, 3 trials, 528 women, moderate certainty). No trial in the review measured live birth [4]. In a Canadian trial, 50 of 172 women after resection or ablation and 29 of 169 after diagnostic laparoscopy had pregnancies lasting 20 weeks or more within 36 weeks (cumulative 30.7% against 17.7%, rate ratio 1.9, 95% CI 1.2 to 3.1). Fetal loss was 20.6% and 21.6% [8]. A network meta-analysis found higher odds of pregnancy after surgical laparoscopy than placebo (OR 1.63, 95% CI 1.13 to 2.35) [9].
Population: 528 women in 3 trials [4]. 341 infertile women aged 20 to 39 with minimal or mild endometriosis [8]. 2,245 women with endometriosis-related infertility in 26 trials [9].
Sources: [4], [8], [9]
Caveats: Live birth was not measured in the pooled trials [4]. These trials covered minimal or mild disease only.

Endometriosis surgeryStrongStudied in womenHow often is not known

Lower anti-Mullerian hormone after endometrioma surgery

Shown to lower anti-Mullerian hormone, a marker of egg supply, after endometrioma surgery in women, by about 35% to 54% in pooled studies.

The numbers and sources

How common: How often is not known: the sources give no rate.
Effect: AMH fell by 1.77 ng/mL at 1 to 6 weeks, 1.17 ng/mL at 2 to 6 months and 2.13 ng/mL at 9 to 18 months after cystectomy, mean falls of 44.4%, 35.1% and 54.2%. Antral follicle counts did not change significantly [15]. A second review found falls of 1.39, 1.13 and 2.12 ng/mL over the same periods, with no difference between removing a cyst from one ovary or both [16].
Population: 650 women in 14 prospective studies [15]. Women in 30 studies [16].
Sources: [15], [16]
Caveats: These are before-and-after measures in the same women, and heterogeneity was high (I² 92% to 94%) [15]. A lower AMH is not the same as lower fertility, which these reviews did not measure.

Uterine fibroid proceduresLimitedStudied in women

Pregnancy after hysteroscopic removal of fibroids inside the uterus

One small trial found more pregnancies after hysteroscopic removal of submucous fibroids than with timed intercourse alone, but the difference could be chance.

Compared with expectant management with regular fertility-oriented intercourse for 12 months [10]

The numbers and sources

Effect: In women with otherwise unexplained subfertility and submucous fibroids, clinical pregnancy odds were higher after hysteroscopic myomectomy (OR 2.44, 95% CI 0.97 to 6.17, P = 0.06) and miscarriage did not clearly differ (OR 1.54, 95% CI 0.47 to 5.00). No data on live birth or complications were found [10].
Population: 94 women with unexplained subfertility and submucous fibroids [10].
Sources: [10]
Caveats: Very low-quality evidence from one small trial [10].

Uterine fibroid proceduresLimited

Pregnancy after embolization compared with myomectomy

Limited evidence: one small trial found fewer live births after embolization than after myomectomy, while a larger trial found similar small numbers of pregnancies.

Compared with myomectomy [5][6][7]

The numbers and sources

Effect: In a subgroup of one trial, live birth (OR 0.26, 95% CI 0.08 to 0.84) and pregnancy (OR 0.29, 95% CI 0.10 to 0.85) were less likely after embolization than myomectomy [5]. In FEMME, 12 women after embolization and 6 after myomectomy reported pregnancies over 4 years, with 7 and 5 live births (hazard ratio 0.48, 95% CI 0.18 to 1.28) [6]. A meta-analysis reported comparable pregnancy rates [7].
Population: 66 women in one trial subgroup [5]; 254 women in FEMME [6]; 13 studies [7].
Sources: [5], [6], [7]
Caveats: The Cochrane authors rated the live birth evidence very low quality, from a subgroup of a single study, and urged extreme caution in reading it [5]. Pregnancy was not the primary outcome of FEMME [6].

AcupunctureLimitedStudied in women

Live birth after IVF

One trial of 848 women found live births did not differ between acupuncture and sham acupuncture during IVF, 18.3% against 17.8%.

Compared with sham acupuncture with a non-penetrating needle [18]

The numbers and sources

Effect: Live births occurred in 74 of 405 women given acupuncture (18.3%) and 72 of 404 given sham (17.8%) (risk difference 0.5%, 95% CI -4.9% to 5.8%, relative risk 1.02) [18].
Population: 848 women having a fresh IVF cycle at 16 centers in Australia and New Zealand [18].
Sources: [18]
Caveats: One trial, with treatment only during ovarian stimulation and around embryo transfer. No no-treatment group was included.

Uterine fibroid proceduresLimited

Pregnancy after focused ultrasound

Linked to lower pregnancy and live birth rates than myomectomy in pooled studies, live births 17% against 44%, with no head-to-head trials.

Compared with women who had myomectomy, in separate studies [13]

The numbers and sources

Effect: After high-intensity focused ultrasound, the pooled pregnancy rate was 23.3% (95% CI 11.5 to 37.6) and live birth rate 17.3% (95% CI 7.8 to 29.3), against 56.9% (95% CI 45.6 to 67.9) and 44.1% (95% CI 34.9 to 53.4) after myomectomy. Studies of younger women and of women trying to conceive reported better outcomes [13].
Population: Published studies of women with fibroids up to March 2024 [13].
Sources: [13]
Caveats: The groups come from different studies, which differ in age and in whether women were trying to conceive. The authors call for comparative trials [13].

Lab markers

Thyroid peroxidase antibodies (TPO antibodies)Strong

Miscarriage and preterm birth

Linked to more miscarriage and preterm birth in women with normal thyroid function, with preterm birth in 6.6% of antibody-positive women against 4.9%.

The numbers and sources

Effect: A meta-analysis of women with normal thyroid function found an odds ratio for miscarriage with thyroid autoantibodies of 3.90 in 19 cohort studies (95% CI 2.48 to 6.12) and 1.80 in 12 case-control studies (1.25 to 2.60). The odds ratio for preterm birth was 2.07 (1.17 to 3.68) [10]. In pooled individual data on 47,045 pregnant women, preterm birth occurred in 6.6% of TPO antibody-positive women and 4.9% of antibody-negative women (odds ratio 1.33, 95% CI 1.15 to 1.56) [11]. A pooled analysis of 11,212 pregnant women found the association with premature delivery rose with the antibody concentration [6].
Population: 31 studies with 12,126 women for miscarriage and 5 studies with 12,566 women for preterm birth [10]. 47,045 pregnant women in 19 cohorts, of whom 7.5% were TPO antibody positive [11]. 11,212 pregnant women in Dutch birth cohorts [6].
Sources: [10], [11], [6]
Caveats: Observational. Each study applied its own assay and cutoff, so a positive result does not mean the same level across studies.

Anti-Müllerian hormone (AMH)ModerateStudied in women

Response to ovarian stimulation in IVF

Linked to the response to IVF stimulation: AMH alone predicted a poor response with an area under the curve of 0.78, against 0.61 for age.

The numbers and sources

Effect: In pooled individual data, age alone predicted a poor response to ovarian stimulation with an area under the curve of 0.61. AMH alone reached 0.78 and the antral follicle count 0.76, and combining the two did not improve on either (0.80, P = 0.19) [13]. The American Society for Reproductive Medicine states that the ability of AMH and the follicle count to predict oocyte yield, and poor and excessive responses, has been well demonstrated, and that extremely low AMH values are not a reason to refuse IVF treatment [12].
Population: 28 study databases with 5,705 women undergoing IVF [13].
Sources: [13], [12]
Caveats: The finding is for women having IVF. It concerns the number of eggs collected, not whether a pregnancy follows (see the next block).

ProgesteroneModerate

Early pregnancy that does not continue, in women with pain or bleeding

Linked to non-viable pregnancy at low levels: with pain or bleeding and an inconclusive scan, progesterone under 3.2 to 6 ng/mL predicted it with 98.4% specificity.

The numbers and sources

Effect: In women with symptoms and an inconclusive ultrasound, a single progesterone below cut-offs of 3.2 to 6 ng/mL predicted a non-viable pregnancy with a pooled sensitivity of 74.6% (95% CI 50.6% to 89.4%) and specificity of 98.4% (90.9% to 99.7%). A low result raised the probability of a non-viable pregnancy from a median of 73.2% to 99.2%. In women with symptoms alone, a threshold of 10 ng/mL gave a sensitivity of 66.5% (53.6% to 77.4%) and a specificity of 96.3% (91.1% to 98.5%), raising the probability from 62.9% to 96.8% [7].
Population: 26 cohort studies with 9,436 women in the first 14 weeks of a spontaneous pregnancy, all with pain or bleeding [7].
Sources: [7], [2]
Caveats: Observational studies of test accuracy in women who already had symptoms, most of whose pregnancies were non-viable, so the figures do not describe pregnancy in general. A low level may mark a failing pregnancy and not cause it. MedlinePlus describes this use of the test as largely superseded by other tests [2].

ProgesteroneModerateStudied in women

Pregnancy in IVF cycles

Linked to fewer pregnancies in IVF at both ends: raised progesterone before egg collection in fresh cycles, and luteal levels under 10 ng/mL in frozen transfers.

The numbers and sources

Effect: In fresh IVF cycles, progesterone at or above 0.8 ng/mL on the day of the trigger injection went with a lower probability of pregnancy: odds ratios of 0.79 at 0.8 to 1.1 ng/mL, 0.67 at 1.2 to 1.4, 0.64 at 1.5 to 1.75 and 0.68 at 1.9 to 3.0. No such effect was seen in frozen-thawed or donor cycles [9]. Around frozen embryo transfer, among studies that used thresholds under 10 ng/mL, a level above the threshold went with more ongoing pregnancies or live births (relative risk 1.47, 95% CI 1.28 to 1.70) and fewer miscarriages (0.62, 0.50 to 0.77) in cycles using only vaginal progesterone and blastocysts [10].
Population: 63 studies of 55,199 fresh cycles, 9 studies of 7,229 frozen-thawed cycles and 8 studies of 1,330 donor cycles [9]. Cohort studies of women having frozen embryo transfer, searched to March 2021 [10].
Sources: [9], [10]
Caveats: Observational. These findings are for women in IVF treatment, most of them given hormones, and do not carry over to natural cycles. The second review found it uncertain whether thresholds of 10 ng/mL or above mattered, and calls for trials of whether raising a low level changes outcomes [10]. Assays disagree most at the low levels the first finding rests on (see Accuracy).

ProlactinModerate

Periods and ovulation in trials that lowered high prolactin

Shown in a trial of women with high prolactin and no periods: the drug that normalized prolactin more often also restored ovulation more often, 72% against 52%.

The numbers and sources

Effect: In women with high prolactin and absent periods, prolactin became stably normal in 83% on cabergoline and 59% on bromocriptine, and 72% against 52% had ovulatory cycles or became pregnant. Absent periods persisted in 7% and 16%. Adverse effects were recorded in 68% and 78% [4]. A systematic review found that dopamine agonists, against no treatment, reduced the prolactin level and the likelihood of persistently high prolactin (relative risk 0.90, 95% CI 0.81 to 0.99), and that cabergoline did more than bromocriptine to reduce persistently high prolactin, absent or infrequent periods and milky discharge [5].
Population: 459 women, most with a small prolactin-secreting pituitary tumor or no cause found, treated for 24 weeks [4]. 8 randomized and 178 nonrandomized studies with more than 3,000 patients [5].
Sources: [4], [5]
Caveats: The trial compared two drugs and had no placebo group. It enrolled women with high prolactin and absent periods, so it does not describe mildly raised results found without symptoms [4].

Follicle-stimulating hormone (FSH)LimitedStudied in men

Sperm production in men seen for infertility

Limited evidence: in men seen for infertility, higher FSH went with abnormal sperm counts, and FSH above 7.6 mIU/mL went with azoospermia caused by failed sperm production.

The numbers and sources

Effect: Among 153 men with no sperm in the ejaculate, 89% of those whose testicles were not producing sperm had FSH greater than 7.6 mIU/mL or a testicular long axis of 4.6 cm or less. Of the men with a blockage, 96% had FSH of 7.6 or less or a long axis greater than 4.6 cm. FSH alone separated the two groups with an area under the curve of 0.87 [10]. In men seen at one infertility clinic, FSH above 7.5 IU/L carried a five to thirteen times higher risk of abnormal semen quality than FSH below 2.8 IU/L, depending on the semen measure. The authors report a raised risk of abnormal sperm concentration and shape above 4.5 IU/L, a level inside their laboratory's reference range of 1.4 to 18.1 IU/L [11].
Population: 153 azoospermic men evaluated from 1995 to 2000 [10]. 610 men at one urology infertility clinic from 2004 to 2008, of whom 153 were excluded [11].
Sources: [10], [11]
Caveats: Both studies are of men already being seen for infertility at single centers, so the cut points may not hold for men in general. Neither sets a reference interval.

Luteinizing hormone (LH)LimitedStudied in women

Polycystic ovary syndrome

One study found 75% of women with polycystic ovary syndrome had LH above the 95th percentile of controls when sampled every 10 minutes for 8 to 12 hours.

The numbers and sources

Effect: Excluding women who had ovulated in the previous 21 days, 75.0% of women with the syndrome had a pooled LH level above the 95th percentile of control women, and 94% had a raised ratio of LH to FSH. Pooled LH was within the normal range in the 9 women studied soon after an ovulation. Pooled LH fell as body mass index rose (R = -0.59) [8].
Population: 61 women with polycystic ovary syndrome defined by infrequent periods and raised androgens, and 24 control women in the early follicular phase [8].
Sources: [8]
Caveats: One study of 85 women that pooled samples taken every 10 minutes, which is not how a routine single blood draw is done. The ratio of LH to FSH is a different quantity from the LH level on this page.

ProgesteroneLimitedStudied in women

Ovulation

One study found a single progesterone of 5 ng/mL or above confirmed ovulation with 98.4% specificity and 89.6% sensitivity against ultrasound.

The numbers and sources

Effect: In a study that tracked ovulation by ultrasound, a single serum progesterone of 5 ng/mL or above had a specificity of 98.4% (95% CI 96.0 to 99.5) and a sensitivity of 89.6% (85.2 to 92.9) for ovulation having occurred [6]. The American Society for Reproductive Medicine uses a lower figure, above 3 ng/mL about a week before the expected period, as presumptive and sufficient evidence of recent ovulation [5].
Population: 102 women over 260 menstrual cycles in a European multicenter study [6].
Sources: [6], [5]
Caveats: One study, a secondary analysis. The two sources use different cut points and neither is a reference interval. The committee opinion states that a single value does not assess the quality of the luteal phase [5].

Anti-Müllerian hormone (AMH)Moderate

Natural conception in women without infertility

Mixed results: three cohorts found low AMH did not lower the chance of natural conception, one larger cohort did, and a pooled analysis found poor predictive value.

The numbers and sources

Effect: In women aged 30 to 44 trying to conceive for 3 months or less, those with AMH below 0.7 ng/mL had a predicted probability of conceiving of 65% by 6 cycles (95% CI 50% to 75%) against 62% (57% to 66%) with normal values, and 84% against 75% by 12 cycles [8]. In women with one or two earlier losses, low AMH (below 1.00 ng/mL) and high AMH (above 3.5) were not associated with the chance of conceiving in a cycle (odds ratios 1.13, 95% CI 0.85 to 1.49, and 1.04, 0.87 to 1.24) [7]. In 102 women planning pregnancy, AMH did not predict time to ongoing pregnancy once age was accounted for [10]. The American Society for Reproductive Medicine's 2026 opinion describes the studies as conflicting: it cites a cohort of 3,150 women in which AMH below 1 ng/mL went with a lower chance of pregnancy, and a meta-analysis of 11 studies with 4,388 women in which AMH had poor predictive value for spontaneous pregnancy (area under the curve 0.59) [11].
Population: 750 women aged 30 to 44 [8]. 1,202 women aged 18 to 40 at four US centers [7]. 102 women planning pregnancy [10].
Sources: [8], [7], [10], [11], [3], [12]
Caveats: Observational. The American College of Obstetricians and Gynecologists states that using AMH for fertility counseling in women without a diagnosis of infertility is not supported by data from high-quality sources, and that a single level in women with presumed fertility does not appear useful in predicting time to pregnancy and is not for use in counseling on it [3]. The American Society for Reproductive Medicine states that markers of ovarian reserve are not to be used as a fertility test for women who are not infertile or whose fertility is untested, or to promote planned egg freezing [12].

Anti-Müllerian hormone (AMH)ModerateStudied in women

Live birth after IVF

Mixed results: AMH had some association with live birth after IVF in one meta-analysis, which called its predictive accuracy poor, and added nothing to age in another.

The numbers and sources

Effect: In a meta-analysis of 13 studies, the diagnostic odds ratio of AMH for live birth was 2.39 (95% CI 1.85 to 3.08) in women of unknown ovarian reserve and 4.63 (2.75 to 7.81) in women expected to have low reserve. The authors conclude that AMH has some association with live birth and that its predictive accuracy is poor [14]. In pooled individual data, age predicted ongoing pregnancy after IVF with an area under the curve of 0.57 and no ovarian reserve test added to it [13]. Across four assays in one clinic, the accuracy of pregnancy prediction was uniformly low (0.62 to 0.63) [15].
Population: 5,764 women of unknown ovarian reserve and 542 expected to have low reserve [14]. 5,705 women [13]. 83 patients [15].
Sources: [14], [13], [15], [12]
Caveats: Observational. The American Society for Reproductive Medicine states that AMH has only a weak association with oocyte quality, clinical pregnancy rates and live birth rates [12].

Follicle-stimulating hormone (FSH)ModerateStudied in women

Response to ovarian stimulation in IVF

Mixed results: raised early-cycle FSH is described as specific but not sensitive for low ovarian reserve, and no ovarian reserve test predicted ongoing pregnancy after IVF.

The numbers and sources

Effect: The American Society for Reproductive Medicine describes an elevated basal FSH, measured on day 2, 3 or 4 of the cycle, as a specific but not sensitive test for diminished ovarian reserve [8]. A meta-analysis of individual data from studies of FSH, anti-Müllerian hormone and antral follicle count before IVF found that age alone predicted a poor response with an area under the curve of 0.61, and that anti-Müllerian hormone or the follicle count alone reached 0.78 and 0.76. For ongoing pregnancy, age was the single predictor (0.57) and none of the tests added to it [9].
Population: 28 study databases with 5,705 women undergoing IVF [9].
Sources: [8], [9]
Caveats: The pooled analysis does not report in its abstract how FSH alone performed for a poor response. The committee opinion gives no FSH cut point [8].

Thyroid peroxidase antibodies (TPO antibodies)Moderate

Live birth and miscarriage in trials of levothyroxine for antibody-positive women

Mixed results: two large trials of levothyroxine in antibody-positive women with normal thyroid function found no change in live births or miscarriage, after small earlier trials reported fewer miscarriages.

The numbers and sources

Effect: In a placebo-controlled trial of 952 women with TPO antibodies, normal thyroid function and a history of miscarriage or infertility, the live-birth rate was 37.4% with levothyroxine 50 micrograms a day and 37.9% with placebo (relative risk 0.97, 95% CI 0.83 to 1.14), with no significant differences in pregnancy loss or preterm birth [12]. In an open-label trial of 600 antibody-positive women having in vitro fertilization, miscarriage rates were 10.3% with levothyroxine and 10.6% without, and live-birth rates 31.7% and 32.3% [13]. Earlier, a pooled estimate from two small randomized studies gave a relative risk of miscarriage of 0.48 (0.25 to 0.92) [10]. In one earlier study, 115 antibody-positive pregnant women were divided into a levothyroxine group and an untreated group, with miscarriage in 3.5% and 13.8% [14].
Population: 952 women at 49 UK hospitals, drawn from 19,585 tested [12]. 600 women at one hospital in China [13]. 115 antibody-positive pregnant women in Italy [14].
Sources: [12], [13], [10], [14]
Caveats: These trials treated women who had the antibodies and did not set out to lower the antibody level. The largest trial started treatment before conception and the Chinese trial was open label [12][13].

Follicle-stimulating hormone (FSH)Limited

Natural conception in women without infertility

Limited evidence: in two cohorts early-cycle FSH did not predict natural conception, including FSH above 10 mIU/mL in women aged 30 to 44.

The numbers and sources

Effect: In women aged 30 to 44 who had been trying to conceive for 3 months or less, those with early follicular FSH above 10 mIU/mL had a predicted probability of conceiving of 63% by 6 cycles (95% CI 50% to 73%) against 62% (57% to 66%) in women with normal values, and 82% against 75% by 12 cycles. Neither difference was significant [6]. In 102 women planning pregnancy and followed for a year, basal FSH did not predict time to ongoing pregnancy once age was accounted for (hazard ratio 0.96, 95% CI 0.86 to 1.06) [7].
Population: 750 women aged 30 to 44 without a history of infertility, 83 of them with FSH above 10 mIU/mL [6]. 102 women planning pregnancy [7].
Sources: [6], [7], [8]
Caveats: Observational. Both cohorts enrolled women with no known infertility, so they do not describe women already having trouble conceiving. The American Society for Reproductive Medicine states that markers of ovarian reserve are poor predictors of reproductive potential independently of age [8].

ProgesteroneLimitedStudied in pregnancy

Live birth in a trial of progesterone for bleeding in early pregnancy

One trial found vaginal progesterone for bleeding in early pregnancy did not significantly raise live births: 75% against 72% on placebo.

The numbers and sources

Effect: Women with vaginal bleeding in early pregnancy were given vaginal progesterone 400 mg twice daily or placebo until 16 weeks. A live birth after at least 34 weeks occurred in 75% (1,513 of 2,025) on progesterone and 72% (1,459 of 2,013) on placebo, a relative rate of 1.03 (95% CI 1.00 to 1.07, P = 0.08). Adverse events did not differ significantly [8].
Population: 4,153 women at 48 hospitals in the United Kingdom [8].
Sources: [8]
Caveats: A marker that tracks an outcome is not always one whose correction changes it. The trial enrolled women by bleeding and did not select them by blood progesterone or treat to a blood level, so it does not test raising a low result [8].
Why limited: This rests on one trial, of 4,153 people. Strength counts independent studies that agree, so one trial is limited at any size.