Interpregnancy Stability of the First-Trimester Uterine Artery Pulsatility Index: A Within-Woman Comparison with PAPP-A and Free β-hCG in 864 Women

Background/Objectives: First-trimester screening combines a haemodynamic marker, the uterine artery pulsatility index (UtA-PI), with two serum markers, pregnancy-associated plasma protein A (PAPP-A) and free β-human chorionic gonadotropin (free β-hCG). For the serum markers, it is known that the value measured in one pregnancy correlates with the value measured in the same woman’s next pregnancy. For UtA-PI, only decreases in the mean value in second pregnancies have been reported. This study quantifies how much of a woman’s standardised UtA-PI persists from one pregnancy to the next and compares this persistence with that of the serum markers measured in the same woman. Methods: We performed a retrospective longitudinal analysis of the screening database of a single foetal medicine unit (August 2010–August 2026). Among 8925 eligible singleton pregnancies, 864 women were screened in two or more pregnancies (1757 pregnancies; 893 pairs of successive pregnancies; median interval 2.9 years, range 0.4–10.7 years). Marker values, expressed as multiples of the median (MoM) and log-transformed, were analysed with random-intercept models that split the total variation into a component between women and a component between the pregnancies of the same woman. The intraclass correlation coefficient (ICC), the fraction of the variation that lies between women, was estimated with bootstrap confidence intervals. Results: The ICC was 0.349 (95% confidence interval 0.280–0.412) for PAPP-A, 0.339 (0.274–0.397) for UtA-PI and 0.392 (0.290–0.478) for free β-hCG; that is, about one-third of the variation in each marker lay between women and about two-thirds between the pregnancies of the same woman. Adjustment for maternal age, body mass index, parity, smoking, gestational age, calendar time, biochemical analyser and examiner did not change the estimates (0.350 and 0.352). No weakening of the between-pregnancy correlation with a longer interval was detected (interaction p = 0.698 and 0.153), although statistical power was limited for the longest intervals. UtA-PI was lower in the next pregnancy by 6.3% on average (95% confidence interval 4.5–8.2%), whereas PAPP-A did not change. A previous PAPP-A below 0.4 MoM recurred in 17.4% of next pregnancies, and a previous UtA-PI above the 95th centile in 6%. Conclusions: Two distinct phenomena coexist in UtA-PI: a systematic fall at the population level in the next pregnancy, and a modest persistence of the individual value. Approximately one-third of the variation in UtA-PI reflects persistent differences between women—no more than for the serum markers—whereas two-thirds arises between the pregnancies of the same woman. A previous value carries some information about the next pregnancy but cannot replace a new measurement.