EDS and pregnancy

In the largest series ever assembled on pregnancy in vascular Ehlers-Danlos syndrome, Murray, Pepin, Peterson, and Byers reported 30 pregnancy-related deaths across 565 deliveries. The death rate of 5.3 percent per delivery, published in Genetics in Medicine in 2014, came from women with protein-altering COL3A1 variants. Among 39 women interviewed about their deliveries, 14.5 percent had a life-threatening complication: arterial dissection or rupture in 9.2 percent, uterine rupture in 2.6 percent, surgical complications in 2.6 percent.

A standard obstetric textbook chapter on connective tissue disease in pregnancy will not contain those numbers. A standard prenatal visit will not include a question about joint hypermobility, easy bruising, family history of arterial dissection, or skin texture. Most pregnant women with any subtype of Ehlers-Danlos syndrome go through prenatal care with no clinician on the team who has been formally trained on the condition.

This is the information desert. The clinical risks are documented in genetics literature, in single-center obstetric series, and in anesthesia case reports. They are largely absent from the textbooks, guidelines, and prenatal-care pathways that determine what an OB/GYN actually does in a fifteen-minute appointment.

What the literature shows

The vEDS numbers are the loudest signal. The Pepin et al. 2014 paper in Genetics in Medicine, drawn from the University of Washington COL3A1 cohort, established that survival in vEDS depends on mutation type. Median survival was 51 years across the cohort. Pregnancy is one of the high-risk windows because the gravid uterus, the major arteries, and the bowel all depend on type III collagen, the protein COL3A1 encodes. Older series put pregnancy mortality at 5 to 12 percent per delivery. A 2026 retrospective multicenter cohort published in BJOG by Bersselaar and colleagues reported lower rates with modern preconception management, planned cesarean in the late preterm window, and care at experienced centers, although events still occurred.

For hypermobile, classical, and other non-vascular subtypes the numbers are smaller and less consistent. The Lind and Wallenburg 2002 paper in Acta Obstetricia et Gynecologica Scandinavica, drawn from 128 pregnancies in 46 women in the Dutch EDS Association, concluded that pregnancy was generally well tolerated outside vEDS, with higher rates of unusually rapid delivery in the hypermobility type and severe maternal complications concentrated in the vascular type. A larger picture has emerged from registry and claims analyses since. Women with EDS carry roughly three times the odds of cervical insufficiency, with adjusted odds ratios in the range of 2.17 to 3.11 across published cohorts. Reported preterm delivery rates in hEDS pregnancies range from 6 percent to 25 percent depending on cohort definition, with premature rupture of membranes a recurrent driver. The 2024 PLOS One scoping review and consensus guidelines led by Pezaro and colleagues, published as the first international expert co-creation effort on hEDS and hypermobility spectrum disorder pregnancy, found no evidence to support prophylactic cerclage and recommended individualized assessment.

The connective-tissue mechanism explains the cluster. The cervix, the fetal membranes, and the pelvic floor are all collagen-rich. When collagen behaves with more mechanical give than typical, the cervix is more likely to shorten and dilate prematurely, the membranes are more likely to rupture early, and the pelvic floor is more likely to stretch beyond functional recovery.

Pelvic floor, prolapse, and the postpartum body

Pelvic organ prolapse rates in EDS cohorts run far above the general-population baseline. The McIntosh et al. 1995 series in the American Journal of Obstetrics and Gynecology reported 75 percent of women with Marfan or Ehlers-Danlos syndrome had a history of pelvic organ prolapse and 50 percent had urinary incontinence, in a small cohort but with a signal that subsequent work has reproduced. A 2023 case series in Case Reports in Urology placed POP prevalence in EDS in the range of 13 to 75 percent depending on subtype and severity. A 2025 comparative study in Discover Medicine found women with EDS scored a median 117 of 300 on the Pelvic Floor Distress Inventory, almost twice the median of 60 in non-EDS controls.

Pelvic floor injury during vaginal delivery is a recognized risk in EDS pregnancies. The Murray vEDS series reported third- and fourth-degree perineal lacerations in 20 percent of vaginal deliveries. In hEDS the postpartum picture often includes pelvic girdle pain, sacroiliac instability, and prolapse symptoms that emerge months after delivery rather than immediately, when the relaxin-mediated ligament laxity of pregnancy compounds an already lax baseline.

The cesarean is not the safer default

The standard mental model that cesarean delivery avoids the pelvic floor and is therefore safer in EDS does not hold. Wound healing after cesarean in classical EDS and vEDS is poor. Skin and fascia handle sutures unpredictably. Dehiscence and incisional hernia rates are elevated. The orphananesthesia recommendations published by Wiesmann and colleagues in 2014 in the Orphanet Journal of Rare Diseases describe the surgical-tissue-handling adjustments experienced centers use: avoiding tension, supporting tissue with deeper suture lines, longer absorbable suture choice, slower return to activity.

Anesthesia adds another layer. People with EDS, especially the hypermobile type, report rapid-onset, short-duration, or absent local anesthetic effect at rates well above background. A study of dental procedures in people with EDS by Schubart and colleagues found 88 percent reported partial, complete, or rapidly waning resistance to local anesthetics. The mechanism is unsettled. Hypotheses include altered drug diffusion through abnormal connective tissue, sodium channel variants, and shorter functional binding. The clinical implication for labor and surgery is concrete: an epidural or spinal that works to standard duration in a typical labor may not. Documented Tarlov cysts, craniocervical and atlantoaxial instability, and dural ectasia in some people with EDS also affect neuraxial planning. MRI or ultrasound spine imaging before placement is recommended when symptoms suggest those findings. Post-dural puncture headache rates appear higher.

Postpartum joint instability and pelvic girdle pain

The postpartum window in hEDS is its own clinical entity. Relaxin, the hormone that increases ligament laxity to allow pelvic widening for delivery, persists for months after birth. In a connective tissue that is already lax, the result is a period of compounded instability across the spine, the pelvis, the shoulders, and the hands. Symphysis pubis dysfunction and sacroiliac joint pain are common during pregnancy in hEDS and frequently fail to resolve at the conventional six-week postpartum mark. Subluxations may increase. Carrying an infant on one hip can dislocate it. The 2024 Pezaro consensus guidelines call out postpartum physical therapy with a clinician familiar with hypermobility as a specific element of care.

Genetic counseling and what is testable before conception

Most subtypes of EDS in the 2017 international classification by Malfait and colleagues are autosomal dominant. That includes vascular (COL3A1), classical (COL5A1, COL5A2), classical-like (rare COL1A1 variants and TNXB in some forms), arthrochalasia (COL1A1, COL1A2), and several others. Inheritance risk for each pregnancy when one parent is affected is 50 percent. A subset of subtypes is autosomal recessive, including kyphoscoliotic (PLOD1, FKBP14), dermatosparaxis (ADAMTS2), and spondylodysplastic forms. Recessive inheritance changes the counseling picture: an affected child requires both parents to be carriers, and partner carrier testing for known-gene subtypes is indicated.

For known-gene subtypes, prenatal testing is available once the family-specific variant is identified. Chorionic villus sampling at 10 to 13 weeks and amniocentesis from 15 weeks onward both allow direct sequencing of the fetal sample for the family variant. Preimplantation genetic testing for monogenic conditions, performed during in vitro fertilization before embryo transfer, is also an option for known-gene subtypes. Confirmation of PGT-M results by CVS or amniocentesis is recommended by the American Society for Reproductive Medicine 2023 committee opinion.

Hypermobile EDS sits outside this framework. No gene has been confirmed for hEDS, so no prenatal genetic test exists. The Norris Lab at the Medical University of South Carolina runs the largest active gene-discovery effort. Until a gene is identified and validated, prenatal counseling for hEDS is empirical: a parent with hEDS will tell their child's clinicians the family history; the child may or may not develop hEDS, and diagnosis if it occurs will be clinical.

For vEDS specifically, preconception counseling is the highest-value intervention available. The French national reference center pathway, organized through AP-HP, is the most developed: COL3A1 confirmation before conception, celiprolol initiation when not contraindicated (Ong et al., The Lancet, 2010, BBEST trial), arterial imaging at baseline, planned delivery in a level-three maternity unit with vascular surgical and interventional radiology backup, and planned cesarean in the 35 to 37 week window for many cases. The 2026 BJOG cohort by Bersselaar and colleagues found event rates with modern coordinated care below the historic Murray figures, although still elevated relative to the general obstetric population.

What the guidelines do not say

ACOG has no condition-specific practice bulletin for Ehlers-Danlos syndrome in pregnancy. The Royal College of Obstetricians and Gynaecologists has no Green-top guideline for EDS. The French Society of Anesthesia and Resuscitation and the French connective tissue reference network have produced anesthesia and vascular-surgery guidance for vEDS, but no comprehensive obstetric protocol. The Williams Obstetrics chapter on connective tissue disease in pregnancy treats EDS in roughly a paragraph. The 2024 Pezaro consensus guidelines in PLOS One are currently the most comprehensive expert document on hEDS and HSD pregnancy management, and they exist precisely because the standard obstetric guideline literature does not cover the population.

A pregnant woman with EDS who searches for prenatal information finds patient association documents, individual case reports, single-center protocols, and the consensus paper. She does not find a NICE guideline, an ACOG bulletin, or a Williams chapter that organizes the clinical risks her delivery team needs to know. The information desert is not metaphorical. It is a documented gap between what the genetics and connective-tissue literature describes and what the obstetric guideline infrastructure has codified.

What the data still does not show

The reason the guidelines do not cover EDS in pregnancy is partly that the evidence base is thin. The largest vEDS pregnancy series, the Murray 2014 paper, drew on 565 deliveries through retrospective pedigree review and interview, with the inherent biases of a registry assembled through specialist referral. The Lind and Wallenburg 2002 series covered 128 pregnancies in one Dutch advocacy population. The 2024 Pezaro scoping review identified a small set of single-center hEDS pregnancy series with cohort sizes in the dozens to low hundreds. No large prospective study has linked EDS phenotype, genotype where available, prenatal management, and pregnancy outcomes across centers and countries with consistent definitions.

Evidence-based obstetric management of EDS pregnancy requires three things the field largely lacks: prospective outcome data linked to phenotype and genotype, multi-center cohorts large enough to detect real effects on rare events such as uterine rupture and arterial dissection, and longitudinal follow-up that captures postpartum prolapse, joint instability, and wound complications that emerge months after delivery. The current literature is single-center series and registry analyses. The clinical risks are real, the management practices at experienced centers are reasonable, and the population-scale data needed to turn those practices into guidelines does not exist outside a handful of national reference networks.