Doppler Ratio Shows Promise for Predicting Perinatal Risk in Preeclampsia

A single Doppler ultrasound ratio that blends three fetal measurements—amniotic fluid depth, umbilical artery blood flow resistance, and middle cerebral artery blood flow resistance—may help clinicians spot which pregnancies complicated by preeclampsia are heading toward a difficult birth. In a retrospective cohort of 151 women, the Amniotic-Umbilical-to-Cerebral Ratio, or AUCR, achieved an area under the receiver operating characteristic curve of 0.87 for predicting composite adverse perinatal outcomes among the 54 women with preeclampsia, outperforming the cerebroplacental ratio, middle cerebral artery pulsatility index, and umbilical artery pulsatility index in the same group. Yet the authors themselves caution that the difference between the ratio and the established Doppler indices did not reach statistical significance, and that the finding requires prospective validation before it can change clinical practice.

The study, published in Archives of Gynecology and Obstetrics, was conducted at University Hospital Münster in Germany. Researchers led by L. M. Müller of the Department of Obstetrics and Gynecology analysed data from women who underwent ultrasound examinations at the hospital between January 2017 and December 2023. To be included, pregnancies had to be singleton, have complete measurements of umbilical artery pulsatility index, middle cerebral artery pulsatility index, and single deepest pocket of amniotic fluid, have fetal biometry available, and have the ultrasound performed no more than 14 days before delivery. Gestational age had to fall between 23 and 42 weeks. Pregnancies with multiple gestations, congenital anomalies, incomplete data, or stillbirth were excluded. After applying these criteria, the final cohort comprised 54 women with preeclampsia and 97 healthy controls.

Preeclampsia, defined according to the International Society for the Study of Hypertension in Pregnancy guidelines as de novo gestational hypertension combined with proteinuria, other maternal organ dysfunction, or uteroplacental dysfunction, affects 2 to 5 percent of pregnancies worldwide and remains a leading cause of maternal and perinatal mortality. Infants born to mothers with the condition are more often born prematurely, weigh less, and face higher risks of neonatal morbidity and death. Because preeclampsia and fetal growth restriction share the underlying mechanism of placental dysfunction, international guidance recommends applying the ISUOG fetal-growth-restriction surveillance protocol to all pregnancies affected by the hypertensive disorder.

The AUCR itself is a relatively new construct. It was first proposed by Stumpfe and colleagues, who combined the single deepest pocket measurement of amniotic fluid with the ratio of umbilical artery pulsatility index to middle cerebral artery pulsatility index—effectively folding the cerebroplacental ratio and the amniotic fluid assessment into a single number. In that earlier retrospective study of 165 singleton pregnancies with small-for-gestational-age fetuses at term, AUCR emerged as the strongest predictor of adverse perinatal outcomes, outperforming the individual parameters. The formula was adopted unchanged in the Münster study, with the amniotic fluid depth recorded in centimetres.

In the new cohort, the clinical picture differed sharply between the two groups. Women with preeclampsia had a significantly higher body mass index, 27.8 versus 24.2 kilograms per square metre, and a higher rate of overweight and obesity, 65 versus 42.3 percent. Mean gestational age at Doppler measurement was significantly lower in the preeclampsia group, 31.74 weeks versus 37.86 weeks, and the interval between measurement and delivery was shorter, 4.50 versus 6.92 days. Fetal growth restriction occurred in 33 of the 54 preeclamptic pregnancies, or 61 percent, with 23 of those cases classified as early-onset. Delivery occurred at a median of 34 weeks in the preeclampsia group versus 39 weeks in controls. Preterm labour affected 85 percent of the preeclamptic women versus 3 percent of controls, neonatal intensive care unit admission 89 versus 21 percent, primary caesarean section 74 versus 28 percent, and birth weight below the 10th percentile 33 versus 12 percent.

The Doppler measurements themselves told a consistent story. Median umbilical artery pulsatility index was significantly higher in the preeclampsia group, 1.24 versus 0.89, while the cerebroplacental ratio was significantly lower, 1.21 versus 1.76. Although absolute middle cerebral artery pulsatility index values did not differ significantly between groups, gestational-age-adjusted percentiles revealed significantly lower values in the preeclamptic pregnancies. Median AUCR was significantly lower in the preeclampsia group, 4.92 versus 7.82. Notably, a subgroup analysis within the preeclampsia cohort found no significant difference in AUCR between pregnancies with and without fetal growth restriction, 4.5 versus 5.04, suggesting that the ratio’s performance was not simply a proxy for growth restriction.

To compare predictive power directly, the researchers multiplied the cerebroplacental ratio and AUCR values by minus one before analysis, so that higher values consistently indicated higher risk across all four parameters. Among women with preeclampsia, AUCR yielded the largest area under the curve for composite adverse perinatal outcomes at 0.872, followed by the cerebroplacental ratio at 0.811, middle cerebral artery pulsatility index at 0.72, and umbilical artery pulsatility index at 0.707. Composite adverse perinatal outcomes were defined as the presence of at least one of three neonatal events: a five-minute Apgar score below 7, umbilical cord arterial pH below 7.1, or admission to the neonatal intensive care unit. Pathological values of all three individual parameters were also significantly associated with adverse outcomes, with odds ratios of 12 for umbilical artery pulsatility index, 6 for middle cerebral artery pulsatility index, and 13 for the cerebroplacental ratio with respect to the composite outcome.

The contrast between risk groups was striking. AUCR had a significantly higher area under the curve in women with preeclampsia than in healthy controls, 0.872 versus 0.589, with a p-value below 0.001. In the low-risk control group, the ratio showed little predictive accuracy. The areas under the curve for the other three parameters were also higher in the preeclampsia group, but those differences were not statistically significant: 0.811 versus 0.637 for the cerebroplacental ratio, 0.720 versus 0.601 for middle cerebral artery pulsatility index, and 0.707 versus 0.543 for umbilical artery pulsatility index. Pairwise comparisons of the areas under the curve among the four Doppler parameters within the preeclampsia group did not reach statistical significance either, a limitation the authors attribute to the small cohort size.

The authors are careful about what their comparison can and cannot show. Doppler measurements in the preeclampsia group were obtained at a significantly lower gestational age than in the controls, and established indices such as the middle cerebral artery and umbilical artery pulsatility indices are known to vary with gestational age. Gestational age-specific reference ranges or percentiles for AUCR do not currently exist, so the observed difference in diagnostic performance between the two groups may partly reflect differences in gestational age at assessment rather than disease effects alone. The researchers describe the ROC comparison between groups as exploratory and call for interpretation with caution. They also note that recruiting a truly gestational-age-matched healthy control group proved infeasible, because AUCR measurements would need to be taken shortly before birth in women who then delivered preterm without any relevant disease—a combination that is difficult to assemble ethically and practically.

Several further limitations shape the conclusions. The retrospective single-centre design may introduce bias, the sample size of 151 pregnancies limits representativeness, and no subdivision of the preeclampsia cohort by disease severity or therapeutic intervention was possible. Confounding factors cannot be ruled out, and no long-term outcome data were available. No standardized cut-off values for interpreting AUCR exist to date. The authors conclude that in this retrospective cohort AUCR demonstrated good discriminative ability for adverse perinatal outcomes in preeclamptic patients but did not outperform established Doppler indices, and that given the limited sample size and the absence of a validated threshold, the ratio should be regarded as a promising adjunct pending confirmation in larger, prospective studies. To their knowledge, only one prior study by Karabay and colleagues had specifically examined AUCR in preeclampsia, and it excluded fetuses with suspected growth restriction; the Münster findings, showing lower AUCR values regardless of growth-restriction status, are consistent with that work and suggest the ratio’s value in this population is not driven solely by the high prevalence of fetal growth restriction.

Subject of Research: Diagnostic performance of the Amniotic-Umbilical-to-Cerebral Ratio for predicting adverse perinatal outcomes in preeclampsia

Article Title: Amniotic-Umbilical-to-Cerebral Ratio (AUCR) as a diagnostic tool for predicting adverse perinatal outcome in pregnancies complicated by preeclampsia

Article References: Müller, L. M., Oelmeier, K., Schmitz, R., Möllers, M., Oberste, K., & Willy, D. (2026). Amniotic-Umbilical-to-Cerebral Ratio (AUCR) as a diagnostic tool for predicting adverse perinatal outcome in pregnancies complicated by preeclampsia. Archives of Gynecology and Obstetrics, 313(1), Article 274. https://doi.org/10.1007/s00404-026-08531-y

Image Credits: AI Generated

DOI: 10.1007/s00404-026-08531-y

Keywords: preeclampsia, AUCR, Doppler ultrasound, perinatal outcome, cerebroplacental ratio, fetal growth restriction, umbilical artery, middle cerebral artery, amniotic fluid, placental dysfunction, ROC analysis, retrospective cohort