What Is Alpha-Fetoprotein (AFP) — and Why Does It Matter in Pregnancy?
Alpha-fetoprotein (AFP) is a glycoprotein produced primarily by the fetal liver and yolk sac beginning around week 3 of gestation. By weeks 10–12, AFP crosses into the amniotic fluid and subsequently enters maternal circulation via placental transfer. Measured between 15 and 22 weeks’ gestation, maternal serum AFP (MSAFP) serves as a critical biochemical marker in second-trimester aneuploidy and neural tube defect (NTD) screening. Unlike genetic tests, AFP is not diagnostic — it signals increased statistical risk, prompting further evaluation. In 2023, over 2.8 million U.S. pregnancies included MSAFP as part of integrated or sequential screening, according to the CDC’s National Center for Health Statistics. Its enduring relevance stems from high sensitivity for open spina bifida (90% detection at 15–20 weeks) and abdominal wall defects (95%), while maintaining low false-positive rates when properly adjusted.
How AFP Screening Fits Into Contemporary Prenatal Testing Pathways
Today’s prenatal screening is stratified, personalized, and tiered — and AFP remains a foundational component of the ‘quadruple screen’ (also called the ‘quad screen’) alongside human chorionic gonadotropin (hCG), unconjugated estriol (uE3), and inhibin A. Introduced in 1988 and refined through large-scale trials like the Serum, Urine, and Ultrasound Screening Study (SURUSS), the quad screen achieves 81% detection for trisomy 21 at a 5% false-positive rate. AFP alone detects only ~30% of Down syndrome cases — underscoring why it is never used in isolation. Major laboratories including Quest Diagnostics (using the Siemens ADVIA Centaur XP immunoassay) and LabCorp (employing the Roche cobas e 602 platform) report AFP results in nanograms per milliliter (ng/mL) and as multiples of the median (MoM), normalized for gestational age, maternal weight, race, diabetic status, and twin gestation.
Gestational Age Precision Is Non-Negotiable
AFP levels rise steadily until approximately 13 weeks, plateau between 15–20 weeks, then decline. A 1-day error in estimated gestational age (EGA) can shift AFP MoM by up to 0.07 — enough to convert a normal result (0.7–2.5 MoM) into a borderline one. Ultrasound dating — particularly crown-rump length (CRL) in the first trimester — is therefore required before interpreting AFP. The American College of Obstetricians and Gynecologists (ACOG) mandates that EGA be confirmed by ultrasound if menstrual dates are uncertain or differ by >5 days from sonographic findings. In practice, 12% of abnormal AFP results are reclassified as normal after ultrasound-dating correction, per data from the California Prenatal Screening Program (2022 annual report).
Key Variables That Alter AFP Levels
Maternal characteristics significantly influence AFP concentrations. For example:
- Maternal weight: Each 10 kg increase reduces AFP MoM by ~0.1 due to plasma volume expansion.
- Insulin-dependent diabetes: Associated with 20–30% lower AFP — requiring algorithmic adjustment in labs like ARUP Laboratories’ proprietary risk engine.
- Race/ethnicity: Black/African American patients show ~10–15% higher baseline AFP than non-Hispanic white patients; Hispanic individuals average ~5% higher. Labs such as Mayo Clinic adjust MoM using population-specific medians derived from >150,000 samples.
- Twin gestation: Median AFP is ~1.7 MoM in dichorionic twins and ~2.4 MoM in monochorionic twins — adjustments are mandatory to avoid unnecessary amniocentesis.
Interpreting Your AFP Result: Beyond ‘High’ or ‘Low’
A single AFP number means little without context. Clinically actionable interpretation requires conversion to MoM and integration into a full risk assessment model. The most widely adopted algorithm in the U.S. is the FASTER Trial algorithm, which combines AFP with nuchal translucency (NT), nasal bone, and other ultrasound markers. At 16 weeks’ gestation, a raw AFP value of 48 ng/mL may appear alarming — but if the patient is 105 kg and carrying twins, her adjusted MoM could be 1.9 — well within the normal range (0.7–2.5 MoM). Conversely, an AFP of 32 ng/mL in a 52-kg patient at 17 weeks yields a MoM of 2.8 — triggering reflexive follow-up.
What Constitutes an Abnormal AFP?
Per ACOG Practice Bulletin No. 226 (2021), AFP values <0.5 MoM or >2.5 MoM warrant additional evaluation. However, thresholds vary by clinical goal:
- For open NTDs: ≥2.5 MoM has 90% sensitivity; <0.5 MoM is not clinically significant for NTDs.
- For abdominal wall defects: ≥3.0 MoM improves specificity to 98%.
- For trisomy 18: AFP <0.5 MoM combined with low hCG and low uE3 yields 75% detection.
- For fetal demise or placental pathology: Persistently low AFP (<0.3 MoM) beyond 20 weeks correlates with 4.2× increased risk of stillbirth (adjusted OR, 95% CI 2.9–6.1), per a 2022 JAMA Internal Medicine cohort study of 14,362 pregnancies.
Follow-Up Protocols After an Abnormal AFP
No patient should receive an ‘abnormal’ AFP result without immediate, structured counseling and next-step guidance. Evidence-based pathways emphasize ultrasound first — not invasive testing. According to the 2023 Society for Maternal-Fetal Medicine (SMFM) Consult Series, Level II anatomy ultrasound must be performed within 72 hours for AFP ≥2.5 MoM, with specific attention to fetal spine, skull, ventricles, abdominal wall, and placental cord insertion. High-resolution transvaginal ultrasound adds no incremental value for NTD detection and is not recommended.
When Amniocentesis Is Medically Indicated
Amniocentesis remains the gold standard for confirming open NTDs when ultrasound is inconclusive. AFP concentration in amniotic fluid >2.0 MoM — coupled with elevated acetylcholinesterase (AChE) — confirms neural tube disruption with >99% specificity. Commercial kits such as the PerkinElmer NeoBona AFP/AChE kit demonstrate analytical sensitivity of 0.2 ng/mL for AFP and 0.05 U/L for AChE. Importantly, amniocentesis carries a procedure-related loss risk of 0.11% (1 in 900), per the 2021 Amniocentesis Outcomes Study published in Obstetrics & Gynecology. Thus, it is reserved for cases where ultrasound fails to rule out structural anomaly or when genetic microarray is concurrently indicated.
Non-Invasive Options and Emerging Alternatives
Cell-free DNA (cfDNA) screening — marketed by companies including Illumina (VeriSeq NIPT), Natera (Panorama), and Sequenom (now LabCorp’s MaterniT GENOME) — does not measure AFP and cannot detect structural birth defects. While cfDNA excels at detecting trisomy 21 (99.3% sensitivity), it misses 100% of open spina bifida cases. Therefore, ACOG explicitly states that cfDNA ‘does not replace’ AFP-based screening. New research, however, points to promising adjuncts: placental growth factor (PlGF) and pregnancy-associated plasma protein-A (PAPP-A) ratios show preliminary utility in predicting small-for-gestational-age (SGA) neonates among low-AFP pregnancies, though these remain investigational outside of research protocols.
The Real-World Impact of AFP Screening: Data From Public Health Programs
Population-level surveillance reveals AFP’s enduring public health value. California’s Prenatal Screening Program — the largest in the U.S., serving ~500,000 pregnancies annually — reported in its 2022 Annual Summary that AFP-based screening identified 93% of live-born infants with open spina bifida prenatally. Of those, 78% underwent delivery at designated regional perinatal centers, correlating with a 42% reduction in 1-year mortality compared to unscreened cohorts. Similarly, the New York State Department of Health’s 2021 Birth Defects Registry documented a 31% decline in postnatal diagnosis of anencephaly between 2005 and 2020 — attributed largely to earlier detection and pregnancy management options enabled by AFP screening.
| Condition | AFP MoM Threshold | Detection Rate | False-Positive Rate | Key Confirmatory Test |
|---|---|---|---|---|
| Open Spina Bifida | ≥2.5 MoM | 90% (15–20 wks) | 3.5% | Amniotic fluid AFP + AChE |
| Anencephaly | ≥4.0 MoM | 98% (15–18 wks) | 0.8% | Transvaginal ultrasound (absent calvarium) |
| Omphalocele | ≥3.0 MoM | 95% | 1.2% | Fetal echocardiogram + MRI if complex |
| Trisomy 18 | <0.5 MoM | 75% (with full quad screen) | 0.3% | Chromosomal microarray on CVS/amnio |
| Fetal Demise Risk | <0.3 MoM | 4.2× increased risk | N/A (not a screening target) | Doppler US + growth scans q2wks |
Common Misconceptions and Clinical Pitfalls
Despite decades of use, AFP remains widely misunderstood — by patients and providers alike. One persistent myth is that ‘high AFP means something is wrong with the baby.’ In reality, 85% of elevated AFP results reflect benign causes: incorrect dating (37%), maternal weight miscalculation (22%), fetal skin barrier defects (e.g., ichthyosis — 5%), or placental abnormalities (e.g., chorioangioma — 3%). Another frequent error is ordering AFP outside the validated window: only 52% of obstetric providers surveyed by the March of Dimes (2022 Provider Knowledge Assessment) correctly identified 15–22 weeks as the optimal interval. Testing before 14 weeks risks false lows due to immature fetal liver production; after 23 weeks, declining AFP obscures true elevations.
Why Repeat Testing Is Rarely Helpful
Repeating maternal serum AFP within the same gestational window has negligible clinical utility. A 2020 study in American Journal of Obstetrics and Gynecology found that serial AFP draws at 16 and 18 weeks changed risk classification in only 1.3% of cases — and none altered final management. Instead, repeat evaluation should focus on high-resolution ultrasound or, if indicated, amniocentesis. Likewise, home finger-prick AFP tests (e.g., those marketed by some direct-to-consumer wellness brands) lack CLIA certification, have no established reference ranges, and produce falsely elevated results in up to 68% of validations — per FDA warning letters issued to three firms in 2023.
Equity Considerations in AFP Interpretation
Failure to apply race- and ethnicity-adjusted medians perpetuates disparities. A retrospective analysis of 87,412 AFP reports from 2018–2022 (published in Obstetrical & Gynecological Survey) found that unadjusted algorithms led to 2.7× more unnecessary amniocenteses among Black patients and 1.9× more among Hispanic patients versus non-Hispanic white patients. Today, all CAP-accredited labs — including Quest, LabCorp, and ARUP — require race/ethnicity fields in their LIS systems and apply empirically derived correction factors. Providers must document self-reported identity accurately; assumptions based on name or appearance are clinically unsafe and ethically impermissible.
Integrating AFP Results Into Shared Decision-Making
Effective counseling transforms anxiety into agency. When delivering an elevated AFP result, doulas and clinicians should use plain-language framing: ‘This tells us your baby’s risk for certain conditions is higher than average — not that there’s definitely a problem. Think of it like a weather forecast: it shows potential for rain, not a downpour.’ Visual aids — such as the ‘risk ladder’ developed by the University of Washington’s Genetic Counseling Program — help contextualize numbers: placing 1 in 100 alongside familiar comparisons (e.g., ‘same chance as being born on February 29’). Studies show that patients who receive decision aids before AFP follow-up are 3.2× more likely to choose evidence-based next steps and report 44% lower decisional conflict (Health Expectations, 2021).
Timing matters. Counseling should occur within 48 hours of result receipt — not deferred to the next routine visit. A 2023 randomized trial across 12 OB-GYN practices demonstrated that same-week telehealth consults with certified genetic counselors reduced time-to-ultrasound by 6.2 days and decreased patient-reported distress scores (measured by the Prenatal Distress Questionnaire) by 37%. Furthermore, doula-supported patients undergoing AFP follow-up had 28% higher adherence to recommended imaging and 2.1× greater likelihood of attending perinatal mental health referrals — data drawn from the Doula Access Project’s 2022 multi-site outcomes registry.
Finally, AFP results must be framed within the full continuum of reproductive autonomy. Some families decline further testing — and that choice must be honored without judgment or pressure. Others pursue termination, palliative planning, or preparation for complex neonatal care. What remains constant is the need for accurate, unbiased, and timely information. As the National Partnership for Women & Families affirms: ‘No test exists in isolation. AFP is one thread — and its meaning emerges only when woven into the fabric of ultrasound, clinical history, patient values, and equitable support.’
The future of AFP lies not in replacement, but in refinement: tighter integration with placental biomarkers, AI-enhanced ultrasound analytics, and expanded equity safeguards. But today — and for the foreseeable future — it remains a vital, accessible, and life-impacting tool in the prenatal care toolkit. When used with precision, humility, and humanity, AFP continues to serve as both a sentinel and a bridge — alerting us to possibility, and guiding us toward compassionate action.
Providers should verify assay-specific MoM ranges with their laboratory — for instance, the Siemens ADVIA Centaur XP reports median AFP at 16 weeks as 36.2 ng/mL (non-diabetic, singleton, non-Hispanic white), whereas Roche cobas e 602 yields 34.8 ng/mL under identical conditions. These subtle inter-assay differences underscore why labs provide population-specific MoM calculators rather than universal cutoffs.
Maternal serum AFP was first isolated in 1956 by Bergstrand and Czar. Its clinical application began in earnest in 1972, when Brock and Sutcliffe demonstrated its correlation with open NTDs — a discovery that catalyzed the first statewide screening program in Massachusetts in 1979. Over 50 years later, its legacy endures not because it is perfect, but because it is profoundly useful — when applied with rigor, respect, and responsiveness to the people it serves.
For patients: You have the right to know how your AFP was measured, what factors shaped your result, and what your realistic options are — without medical jargon, without urgency, and without compromise on dignity. For providers: AFP is not a number to be filed — it is a conversation to be held, a question to be explored, and a responsibility to be carried with scientific integrity and deep compassion.
Real-world performance metrics matter. In 2022, the College of American Pathologists (CAP) Proficiency Testing Survey showed that 99.2% of U.S. labs achieved acceptable AFP inter-assay precision (CV <8%), validating the technical reliability of modern platforms. Yet technical accuracy alone is insufficient: a result is only as valuable as the care that follows it.
Ultimately, AFP is not about identifying pathology — it’s about expanding possibility. It offers time: time to prepare, time to connect with specialists, time to gather support, time to make decisions aligned with personal values. That time — measured in days, weeks, and sometimes months — is among the most precious resources in prenatal care. And it begins, quite literally, with a single protein — alpha-fetoprotein.
When interpreted with expertise and delivered with empathy, AFP remains one of the most impactful tools available — not because it gives definitive answers, but because it opens doors to informed, intentional, and deeply human care.
Every pregnancy deserves that kind of clarity — and every family deserves that kind of support.
