US claims data link HMG protocols to higher live birth odds

The study identified 14,174 women aged 18 to 42 who received gonadotropins after an infertility diagnosis. It sorted them by the first gonadotropin regimen they received: HMG alone, follitropin alfa alone, follitropin beta alone, or a mixed protocol that combined HMG with one of the follitropins ¹. The researchers used logistic regression adjusting for baseline characteristics, which means the comparison accounts for recorded differences between groups ¹. The researchers then looked at live births linked to that first regimen.

From 2016 to 2021, the share of first regimens using HMG in a mixed protocol rose from 60.1% to 71.2%, and HMG monotherapy rose from 4.3% to 5.7%. Over the same period, follitropin alfa monotherapy fell from 21.7% to 11.4%, and follitropin beta monotherapy fell from 13.9% to 11.6%. PubMed

Compared with follitropin alfa alone, the adjusted odds of live birth were 1.31 for the mixed protocol, with a confidence interval (the range likely to contain the true effect) from 1.20 to 1.45. For HMG monotherapy, the odds ratio was 1.28, with a confidence interval from 1.08 to 1.53. Both lower bounds sat above 1.0, which means the live birth odds were higher in the HMG arms. The estimate for follitropin beta versus follitropin alfa was 0.96, with a confidence interval from 0.84 to 1.10, which does not show a meaningful difference. PubMed

Ovarian hyperstimulation syndrome (OHSS), a serious complication where the ovaries over-respond and fluid can collect, was uncommon. It appeared in 1.2% to 2.9% of cycles across the groups, and severe cases were rare at 0.1%. PubMed

A randomized evidence base already pointed in the same direction. The 2026 Cochrane review combined 59 randomized trials covering 18,119 women and compared recombinant FSH with HMG or highly purified HMG ². It found that recombinant FSH had 17% lower odds of live birth, with an odds ratio of 0.83. In other words, the HMG arms had higher live birth odds in the randomized comparisons ². The same review found recombinant FSH carried 42% higher odds of ovarian hyperstimulation syndrome, with a confidence interval from 1.12 to 1.80. Wiley

Not every comparison favors HMG. A systematic review of trials comparing a 2:1 mix of recombinant FSH and recombinant LH against HMG alone found the mix retrieved about 1.4 more oocytes per cycle and had 18% higher clinical pregnancy ³. That endpoint was clinical pregnancy, not live birth, so the measurements do not fully overlap.

Short term outcomes measure how many follicles grow and how many eggs are retrieved. One study of 4,525 antagonist cycles, a common IVF stimulation method that uses a medicine to stop an early ovulation, looked at how gonadotropin type and antral follicle count, a baseline ultrasound measure of a woman's egg supply, together related to follicle recruitment and egg yield. Elsevier

For egg number, the 2:1 recombinant mix outperformed HMG alone by about 1.4 oocytes per cycle in a systematic review ³. For OHSS risk, HMG-containing regimens look lower-risk than recombinant FSH alone in the randomized evidence ², and the new claims data recorded a low overall rate of 1.2% to 2.9% across all groups. PubMed

Mid term outcomes include clinical pregnancy and live birth. The claims data found HMG protocols had higher live birth odds than follitropin alfa alone ¹. The randomized review found the same direction for HMG versus recombinant FSH ². However, one 2024 study compared HMG with low-dose hCG, a related hormone that acts like luteinizing hormone, as the add-on during ovarian stimulation and found no difference in live birth rates after fresh and frozen embryo transfer ⁶. HMG products also carry hCG activity; commercial highly purified HMG contains beta-hCG making up 18 to 47 percent of gonadotropin content ⁷. You can read more about that hormone in the HCG research summary.

In frozen embryo transfer preparation, letrozole plus HMG matched hormone-replacement cycles for live birth and reduced cesarean section and gestational diabetes rates ⁸. Real-world data combining recombinant FSH with highly purified HMG reached pregnancy rates comparable to recombinant FSH alone, with lower OHSS in frozen-transfer cycles, 6.9% versus 13.2%. Nature

Long term outcomes are harder to study in claims data. The new analysis covered live births linked to the index regimen but did not follow all later cycles or losses in detail ¹. Earlier observational work raised a question about epithelial ovarian cancer after gonadotropin exposure, with residual confounding likely and no modern confirmation ⁹. Batch variability and the roughly 20 to 30 percent non-gonadotropin protein in urinary products keep immunogenicity questions open ⁷. The new claims study was not designed to test cancer risk or other long term safety outcomes. PubMed

Claims data cannot show why doctors chose one protocol over another, and the authors adjusted for measured baseline differences but not unmeasured ones ¹. The study was not designed to compare cumulative live birth across multiple cycles or to separate fresh and frozen transfers ¹. It also reported severe OHSS as rare, but rare events need larger samples to estimate precisely. The confidence interval for HMG monotherapy live birth was wider than for the mixed protocol ¹. We still do not know whether the pattern holds in patients with different ovarian reserve profiles or in letrozole-based frozen transfer protocols, which come from a separate study design ⁸. Future research can test these real-world patterns in prospective cohorts.