Fetal Growth Restriction (FGR)

Definition and Epidemiology

Fetal growth restriction (FGR; formerly known as intrauterine growth restriction) is the condition in which the fetus fails to reach its genetic growth potential due to a maternal, fetal or placental pathological factor. It affects approximately 10% of pregnancies and is a leading cause of perinatal morbidity and mortality (ACOG PB 227, 2021; SMFM Consult Series #52, 2020). Antenatal detection is an integral component of prenatal care; Because although treatment options are limited, fetal monitoring is initiated after diagnosis, aiming to reduce the risk of stillbirth and optimize the timing of birth (UpToDate, "Screening and diagnosis", 2026).

The antenatal diagnosis rate of FGR is low: approximately 75–80% of small-for-gestational-age (SGA) newborns are undetected before birth. In undetected cases, the risk of stillbirth increases eightfold; therefore, a systematic screening approach is critical (UpToDate, 2026). FGR is also associated with long-term consequences such as neurodevelopmental delay in childhood and increased risk of hypertension, dyslipidemia, type 2 diabetes mellitus, and coronary heart disease in adulthood.

Distinguishing the definitions: FGR or SGA?

The clinical definition of SGA (birth weight or estimated fetal weight <10th percentile) does not distinguish constitutionally healthy fetuses from those with pathological growth restriction. FGR, on the other hand, refers to the failure to realize growth potential as a result of an underlying pathological process (most commonly placental insufficiency). Not every fetus diagnosed with antenatal FGR may be SGA at birth; Similarly, not every newborn born SGA may be detected as FGR during pregnancy (UpToDate, 2026). This distinction is important to avoid unnecessary interventions (constitutionally small fetus) and to ensure appropriate timing of follow-up and birth (true FGR).

Etiology and Pathophysiology

FGR is not a homogeneous entity; The causes are grouped under the headings of placental, maternal and fetal, and they are not independent of each other (for example, maternal vascular disease causes placental changes, causing placental insufficiency).

• Placental (most common): Placental insufficiency is the most common risk factor. Velamentous cord insertion, circumvallate placenta, and single umbilical artery are also associated with FGR. Placental insufficiency cannot be measured directly antenatally and is a diagnosis of exclusion; EFW <10th percentile, when other causes (genetics, structural anomaly, infection, drug/substance) are excluded — especially if umbilical or uterine artery Doppler is abnormal — it is attributed to placental insufficiency (UpToDate, 2026).
• Maternal: Chronic hypertension and other maternal vascular diseases, kidney disease, pregestational diabetes (with vasculopathy), collagen tissue diseases and antiphospholipid syndrome. According to NICHD data, preeclampsia, particularly severe with onset at <34 weeks, is associated with consistent impairment in fetal growth and can begin as early as 22–23 weeks, before maternal symptoms appear (UpToDate, "Evaluation"). Modifiable factors: tobacco, alcohol, cocaine, opioid use; warfarin and some antiepileptic/chemotherapeutic drug exposures.
• Fetal: Genetic anomalies, syndromes and congenital anomalies. The risk of chromosomal anomaly in FGR that appears isolated was found to be 6.4% in a systematic review (most commonly trisomy 21 and 18). Congenital infections (cytomegalovirus, syphilis, rubella, varicella, toxoplasmosis, malaria) are responsible for approximately 5-10% of cases (UpToDate, 2026).

Screening and Diagnosis

Evaluation of growth

Normal/abnormal growth distinction requires accurate determination of gestational age and fetal size and comparison with a growth standard. Pregnancy dating is ideally based on first trimester crown-rump distance (CRL). Ultrasound is the standard method for fetal size: estimated fetal weight (EFW) is calculated by a formula that includes biparietal diameter, head circumference, abdominal circumference (AC) and femur length (most commonly Hadlock in the USA). The error between EFW and actual birth weight is greater than 14% across studies at the 95% confidence interval; this limitation contributes to both over- and underdiagnosis (UpToDate, 2026).

Scanning strategy

A risk-based approach, consistent with UpToDate and various international guidelines, is recommended:

• Low-risk pregnancies: Routine fundal height measurement from the 24th week onwards. If there is a difference of ≥4 cm from the expected value or if the fundal height cannot be adequately evaluated, a growth ultrasound is performed. Fundal height alone is an inadequate screening tool (FGR detection sensitivity 17-86%) (UpToDate, 2026).
• Pregnancies with risk factors: EFW and amniotic fluid assessment starting at week 28 at the earliest, repeated every 3-4 weeks according to the risk profile. Scans are not performed ≤2 weeks apart because interobserver measurement variability can cause spurious shifts in biometric percentiles (an interval of at least 3 weeks is recommended) (SMFM #52; UpToDate, 2026).

Routine third trimester ultrasound screening in the low-risk population is not recommended by international guidelines; because it has not been proven to reduce perinatal mortality. It has been stated that if a single scan is to be performed, it should be performed closer to the 36th week (it detects SGA better instead of the 32nd week), and if two scans are to be performed, 32 and 36 weeks are reasonable (UpToDate, 2026).

Diagnostic criteria

In the USA (SMFM #52, ACOG PB 227, AIUM), the diagnosis of FGR is based on EFW <10th percentile or AC <10th percentile on a population-based or customized growth chart. AC performs similarly to EFW in predicting SGA. In well-dated fetuses without anomalies, EFW or AC <3rd percentile indicates a high likelihood of FGR regardless of gestational age (expert consensus). When EFW/AC is between the 3rd–10th percentiles, abnormal Doppler parameters and a declining growth trajectory support a diagnosis of pathological placental FGR and reduce the likelihood of constitutional smallness (UpToDate, 2026).

ISUOG (2020) and Delphi consensus (Gordijn 2016) also include Doppler parameters in the definition (Table 1). In a comparative study, US (SMFM-ACOG-AIUM) criteria showed higher sensitivity (55% versus 29%) but lower specificity (93% versus 98%) than ISUOG criteria in predicting SGA; both approaches were poor at predicting composite adverse neonatal outcome (sensitivity 10-15%) (UpToDate, 2026). This limitation emphasizes that no set of diagnoses is the “gold standard” and that the clinical context is decisive.

Source: Gordijn SJ et al. Ultrasound Obstet Gynecol 2016;48:333-339; The ISUOG Practice Guidelines (Lees 2020) adopt this framework. US guidelines (SMFM/ACOG) continue to use EFW/AC <10th percentile as a stand-alone threshold.

Supporting features

EFW/AC <3rd percentile identifies a group at increased risk of stillbirth even if amniotic fluid and Doppler are normal (approximately three times higher than the 3rd to 5th percentile, four to seven times higher than the 5th to 10th percentile). Oligohydramnios may be due to fetal hypoxemia reducing renal blood flow and EFW <3rd percentile is associated with an adverse outcome. Polyhydramnios, when accompanied by FGR, increases the risk of aneuploidy (especially trisomy 18) or genetic syndrome; Genetic counseling and chromosomal microarray are recommended in these cases (UpToDate, 2026).

Doppler Evaluation: Multivessel Approach

The hemodynamic physiology of FGR forms the basis for the use of Doppler ultrasound in monitoring and management. When 30% of the placental villous vascularization is obliterated, umbilical artery (UA) indices are impaired; At 60-70% obliteration, flow may disappear or reverse at the end of the cardiac cycle (UpToDate, "Evaluation").

• Umbilical artery (UA): It is the primary Doppler parameter in FGR monitoring. Pulsatility index (PI) is preferred; Above the 95th percentile for gestational age is abnormal. Order of decay: high PI → end-diastolic flow loss (AEDF) → reverse diastolic flow (REDF). Reverse flow is associated with five times higher perinatal mortality compared to loss of flow. UA Doppler monitoring significantly reduces perinatal death by triggering preterm birth when abnormal (review of 19 studies: RR 0.71, 95% CI 0.52-0.98) (UpToDate, 2026).
• Middle cerebral artery (MCA): In progressive hypoxemia, cerebral blood flow increases (brain-protective — "brain-sparing" effect) and MCA PI decreases (<5th percentile). Subsequent normalization of MCA PI is a bad sign indicating loss of brain protective mechanism. While SMFM/ACOG/AIUM recommend against using MCA Doppler in routine management, ISUOG and RCOG support its use; RCOG recommends the use of abnormal MCA Doppler (PI <5th percentile) in timing birth in a term small fetus with normal UA Doppler (UpToDate, 2026).
• Cerebroplacental ratio (CPR): MCA PI / UA PI. It reflects both placental status and fetal response. In late-onset FGR, the only Doppler change present may be a decrease in CPR. The inverse ratio, the umbilicocerebral ratio (UCR = UA PI / MCA PI), can also be used (UpToDate, 2026).
• Ductus venosus (DV): With advancing UA resistance, fetal cardiac performance deteriorates, atrial pressure increases and diastolic flow decreases in the DV a-wave. Late in the disease, the a-wave disappears or reverses; this is a sign of impending acidemia and fetal death (sensitivity 65%, specificity 95% for fetal pH <7.20). SMFM/ACOG/AIUM recommend not using DV in routine management; Outside the USA, DV is used in birth timing (UpToDate, 2026).

Fetal Monitoring and Frequency

Monitoring begins at the gestational age at which delivery would be considered for perinatal benefit. Key parameters include EFW, Doppler, and fetal behavior (biophysical profile [BPP] and NST/cardiotocography [CTG]). Because cardiovascular (Doppler) and behavioral (BPP/NST) deterioration can occur largely independently of one another, the fetus is monitored using multiple modalities (UpToDate, "Management," 2026).

• Growth: EFW typically assessed every ~3 weeks; It is not done more often than 2 weeks due to measurement error. A decrease in the EFW percentile is associated with increased perinatal morbidity-mortality.
• Doppler-based frequency: Strategies that use Doppler to determine behavioral testing frequency are associated with lower stillbirth rates compared to empirical strategies (UpToDate, 2026). Frequency is summarized in Table 2 according to UA/MCA findings.

*For patients with reactive NST and BPP 8/8, 10/10 or 8/10 with normal amniotic fluid. 3rd–10th in the presence of comorbidity (oligohydramnios, preeclampsia). In percentile cases, timing is individualized; Most patients are delivered between 34+0 – 37+6 weeks. Birth should not exceed 39+0 weeks (risk of term stillbirth increases significantly). If NST is nonreactive or BPP <8/10, management is individualized according to findings and gestational age; In case of recurrent late decelerations, immediate delivery is indicated. Source: UpToDate "Management" (2026), consistent with SMFM #52.

Birth Timing: Evidence and Threshold Logic

There is no consensus on the optimal time of birth in FGR; The timing balances the risk of iatrogenic preterm birth outcomes and stillbirth with ongoing follow-up.

• Reverse diastolic flow (REDF): Since the risk of fetal death exceeds neonatal morbidity-mortality rates at least from the 30th week, delivery at 30+0 – 32+0 weeks is recommended.
• Absent end-diastolic flow (AEDF): because the risk of fetal death exceeds neonatal risks from 33–34 weeks onward, delivery is recommended at 33+0–34+0 weeks.
• Least complicated FGR (UA normal): The upper threshold of 37+0 weeks is based mainly on expert opinion and limited data; The risk of stillbirth increases with advancing gestational age (37th week 28/10,000, 38th week 41/10,000, 39th week 77/10,000, 40th week 194/10,000 — EFA <10th percentile) (UpToDate, 2026).

Indicative randomized trials

GRIT (24-36 weeks, immediate vs delayed birth): Fewer stillbirths (2 vs. 9) but more neonatal/infant deaths (27 vs. 18) in the immediate birth group; No difference in neurodevelopmental outcomes through age 13. DIGITAT (>36 weeks, induction vs waiting): Composite adverse outcome and cesarean section rates are similar; Both approaches are reasonable, the choice depends on patient preference. TRUFFLE (26-32 weeks, very early FGR; Lancet 2015): DV-based birth with late changes tends to increase neurodevelopmental impairment-free survival compared to short-term variability (STV)-based birth (late DV 95% vs STV 85%) but at the expense of a small increase in unexpected fetal death. Post-hoc analysis highlighted that both DV and computerized CTG evaluation are necessary together (UpToDate, 2026).

Corticosteroid and Magnesium Sulfate

When preterm birth is anticipated, antenatal corticosteroids are administered at <34+0 weeks. Although evidence for effectiveness in the preterm neonate with FGR is controversial, one cycle is recommended until definitive information is available (UpToDate, 2026). Application of 34+0 – 36+6 weeks does not reduce the need for respiratory support and increases the rate of neonatal hypoglycemia; It is recommended by some guidelines in selected late-preterm cases.

Magnesium sulfate is administered for fetal neuroprotection at expected birth at <32 weeks; When examined specifically in fetuses with FGR, a reduction in significant neurodevelopmental impairment and mortality was observed (UpToDate, 2026). Dosage and administration details follow the institutional neuroprotection protocol.

Mode of Delivery and Intrapartum Management

Cesarean delivery is performed for standard obstetric indications; otherwise, a trial of vaginal delivery is acceptable, and an unfavorable cervix is not a reason to avoid induction. However, when reversed UA end-diastolic flow is the indication for delivery, planned cesarean delivery is offered because most of these fetuses cannot tolerate labor—especially if cervical conditions are unfavorable. For cervical ripening, a mechanical method (balloon catheter) may be safer than prostaglandins in this setting (UpToDate, 2026). Continuous intrapartum fetal heart rate monitoring is indicated.

Controversial Points and Current Changes

• Diagnostic criteria: US (SMFM/ACOG, <10th percentile alone) and ISUOG/Delphi (including Doppler) approaches are different; US criteria are more sensitive, ISUOG is more specific, and both are poor predictors of adverse neonatal outcome (UpToDate, 2026).
• Role of MCA and DV: SMFM/ACOG/AIUM do not recommend these in routine management, while ISUOG/RCOG supports them. DV reverse flow is not considered an indication for delivery in the United States; Used in birth timing outside the USA.
• Maternal interventions: There is no intervention proven to improve FGR in a healthy pregnant woman. Sildenafil (phosphodiesterase-5 inhibitor) was found to be futile in early-onset FGR, and a Dutch study was stopped early due to increased neonatal pulmonary hypertension (UpToDate, 2026). Pravastatin pilot data is promising but not yet proven. Anticoagulation (UFH/LMWH) does not reduce recurrent placenta-mediated late pregnancy complications; obstetric antiphospholipid syndrome is an exception.
• Subsequent pregnancy: SGA births tend to recur (~23% vs. ~3% in the second pregnancy in those with SGA in the first pregnancy). Low-dose aspirin may be effective in FGR secondary to preeclampsia; It is not recommended in the absence of risk factors.

Clinical Summary

• FGR, EFA or AC <10. percentile (severe: <3rd percentile) and affects ~10% of pregnancies; Differentiation from SGA is critical in terms of pathological vs constitutional smallness.
• The most common cause is placental insufficiency; In isolated early FGR (<32 weeks, especially <3rd percentile) genetic testing (microarray) and, if abnormalities are present, fetal evaluation are recommended.
• UA Doppler is the primary monitoring tool. Frequency of monitoring and timing of birth are determined according to the finding of UA: REDF → 30+0–32+0 weeks, AEDF → 33+0–34+0 weeks, high UA PI → 37+0 weeks; Cases with normal UA are delivered at 37–39 weeks; delivery should not be later than 39+0 weeks.
• Antenatal corticosteroid (betamethasone 12 mg IM ×2 / dexamethasone 6 mg IM ×4) when delivery is anticipated at <34+0 weeks; MgSO4 neuroprotection is applied at <32 weeks.
• In case of UA reverse flow, the option of planned cesarean section is offered; Otherwise, an unfit cervix does not prevent a vaginal birth attempt.

🔄 Current Developments (2026)

This section adds new and changing data from the last two years, on top of ACOG PB 227 (2021), SMFM #52 (2020) and ISUOG 2020 resources, which form the backbone of the guide. Diagnostic thresholds (EFA/AC <10% / <3%) and umbilical artery (UA) Doppler-based birth timing algorithm (REDF ~30-32 weeks, AEDF ~33-34 weeks, normal UA 37-39 weeks) did not change during 2024-2026; The following developments either update or add nuance to the current approach.

RCOG GTG 31, 3rd edition — emphases affecting practice

• SGA–FGR distinction sharpened: Interventions are targeted at FGR cases at real risk; Unnecessary early-term induction in constitutionally small (only SGA) fetuses is avoided. NHS "Saving Babies' Lives" v1's imposition of 37+0–37+6 births in SGA has been criticized as unnecessarily increasing iatrogenic preterm birth and interventions; the threshold is shifted to individualization.
• Early risk assessment: Risk stratification and assignment to an agreed monitoring pathway at the booking visit (≤14 weeks) is recommended.
• Personalized growth chart: plot the symphysis–fundal height (SFH) measurement on a customized chart rather than a population chart; refer for ultrasound if measurements cross centiles or growth is static. This approach increases detection while reducing false-positive SGA diagnoses.

Adding abdominal circumference (AC) to diagnosis: new alert

SMFM included AC <10th percentile in addition to EFW <10th percentile in its definition of FGR. A 2025 secondary analysis of NICHD Fetal Growth Study–Singletons data (n=2400) evaluated this expansion: the broader definition increased FGR prevalence from 7.5% to 9.0% (181→215 cases) without improving prediction of composite neonatal morbidity (AUC 0.52 → 0.51; both poor), while SGA prediction remained similar (AUC 0.70 → 0.71). The authors emphasize adopting the expanded SMFM definition cautiously, particularly in low-risk populations, given the potential for unnecessary surveillance and iatrogenic preterm birth (Shea A et al., J Ultrasound Med 2025;44(7):1261-1271; PMID 40099719).

SOGC No. 442 (2023) — late FGR definition and CPR emphasis

This GRADE-based Canadian guideline (Kingdom J et al., J Obstet Gynaecol Can 2023;45(10):102154; PMID 37730302), which is not included in the current guideline, provides clear criteria for late FGR (≥32 weeks): AC or EFW <3rd percentile alone; or at least two of the following three—AC/EFW <10th percentile, two-quartile drop in percentile, abnormal Doppler (UA PI >95th percentile or CPR <5th percentile). MCA is preferably interpreted as the cerebroplacental ratio (CPR = MCA PI / UA PI); CPR <5th percentile is abnormal and is an indication for either intensive monitoring or delivery, depending on the gestational age. This is where sources diverge: SOGC and RCOG incorporate CPR into monitoring/scheduling, while SMFM #52 does not recommend routine clinical use of MCA/CPR due to insufficient evidence—both approaches are valid. SOGC also includes PlGF/sFlt-1 angiogenic markers and aspirin prophylaxis in screening for placenta-derived FGR.

Frequency of monitoring and prevention

• Frequency of UA Doppler: RCT (n=292) comparing UA Doppler every week versus every two weeks in high-risk 3rd trimester pregnancies found no difference in NICU admission in the first 24 hours (risk difference −2.63; 95% CI −12.92 – 7.65) and concluded that monitoring every two weeks may be more practical and cost-effective (Salama MH et al., Obstet Gynecol Sci 2023;66(3):161-168; PMID 36593702). These data suggest that more frequent screening does not improve outcome for normal Doppler findings—a nuance to the “once a week” frequency in Table 2.
• Prevention with aspirin: The effect of low-dose aspirin in preventing FGR depends on the dose and timing of initiation. Initiation at ≤16 weeks is associated with a significant reduction and a dose–response relationship (RR 0.56; 95% CI 0.44–0.70); the effect is weaker when started after 16 weeks (Roberge S et al., Am J Obstet Gynecol 2017; PMID 27640943). Practical implication: in a high-risk pregnant patient, initiation preferably at 12–16 weeks at ≥100–150 mg/day. This strengthens the current guide’s statement that aspirin may be effective in FGR secondary to preeclampsia by specifying a dose/timing window.

Note: The contents of SMFM Consult Series #52 (2020) have been reaffirmed and are in effect in 2024; The current guide's reliance on this source remains current. All of the above numerical claims depend on the relevant guide number/year or PMID/DOI; sources have been verified with PubMed metadata and none have been retracted. (Information taken from PubMed.)

← All scientific guides