Longitudinal evaluation of cervical length for the prediction of mid-trimester cervical length in low-risk population Authors: Bilge Cetinkaya Demir, Murat Deniz Celenk, Gokhan Ocakoglu Published: 2026 ABSTRACT BackgroundPreterm birth (PTB) remains a leading cause of perinatal mortality and long-term morbidity. While screening strategies primarily target high-risk populations, a substantial proportion of PTBs occur in low-risk women. Mid-trimester cervical length (CL) assessment may be implemented too late for effective prevention. Earlier evaluation in the first trimester may improve risk stratification.ObjectiveTo longitudinally evaluate cervical length from the first to the second trimester and assess its ability to predict mid-trimester CL and PTB in a low-risk population.MethodsIn this prospective cohort study, 292 low-risk pregnant women underwent transvaginal ultrasound at 11–14, 16–20, and 20–24 weeks of gestation. Cervical measurements included single-line CL (A), two-line components (B: internal os to cervical angulation; C: angulation to external os), and anterior (D) and posterior (E) cervical lip widths. Longitudinal changes were analyzed, and receiver operating characteristic (ROC) analysis was used to evaluate predictive performance for PTB <35 weeks.ResultsMedian single-line CL (A) increased from 37.7 mm (29–58) at 11–14 weeks to 41.8 mm (28–66) at 20–24 weeks. This increase was primarily driven by elongation of the B component (13% increase, p<0.001), while the C component remained stable. First-trimester CL demonstrated limited predictive performance for PTB, with an area under the curve (AUC) of 0.61.ConclusionCervical length increases from the first to the second trimester in low-risk pregnancies, mainly due to changes in the proximal cervical segment. First-trimester CL measurement has limited value in predicting PTB, supporting the continued use of mid-trimester assessment for risk stratification. Introduction Preterm birth (PTB) affects nearly 15 million newborns worldwide each year [1, 2]. As it is associated with perinatal mortality and long-term morbidity, the prediction and prevention of PTB remain major public health priorities. Although a history of PTB is the strongest predictor, most women who deliver preterm have no identifiable risk factors [3]. In light of this, various screening strategies—including biochemical markers, uterine activity monitoring, and risk-scoring systems—have been investigated to identify women at increased risk of PTB [3]. In current clinical practice, transvaginal ultrasonographic cervical length (TVU CL) measurement is considered the most reliable method for predicting PTB risk, and both the American College of Obstetricians and Gynecologists (ACOG) and the Society for Maternal-Fetal Medicine (SMFM) have proposed diagnostic protocols [3, 4]. Based on current evidence, guidelines recommend mid-trimester TVU CL measurement for this purpose [4, 5]. However, despite progesterone treatment initiated in the second trimester, up to 30% of women with a short cervix still deliver preterm [6]. Therefore, earlier prediction of PTB—prior to the mid-trimester—and the establishment of timely treatment strategies may be crucial for reducing perinatal morbidity and mortality [7]. In the literature, only a limited number of studies have focused on first-trimester endocervical canal length (ECL) measurements as an alternative to mid-trimester TVU CL assessment for predicting PTB [8–16]. Elastography, which assesses cervical tissue softening, has also been explored as a predictive tool for preterm delivery [11]. Additionally, two-line method has been reported to be more sensitive than single-line measurement in the first trimester for predicting preterm delivery before 34 weeks of gestation [13]. Earlier detection and intervention may improve the effectiveness of PTB prevention. However, evidence regarding the ability of first-trimester TVU CL to predict second-trimester cervical length remains inconsistent [11–16]. The aim of this prospective cohort study was to evaluate the predictive value of endocervical canal length measured between 11 and 14 weeks of gestation in low-risk singleton pregnancies for subsequent cervical length at 20–24 weeks and for preterm birth before 35 weeks. We hypothesized that early identification of cervical changes using these parameters may improve the management of a short cervix and contribute to the existing literature. Materials and methods This prospective cohort study included low-risk pregnant women for preterm birth who were examined and followed until delivery at the maternal–fetal medicine center of a university hospital between May 2022 and March 2023. The study was approved by the Institutional Ethics Committee of the Medical Faculty and conducted in accordance with the Declaration of Helsinki (approval number: 2022-10/41). Written informed consent was obtained from all participants prior to enrollment. Exclusion criteria were: multiple pregnancy, history of cervical insufficiency, previous preterm birth, presence of abdominal cerclage, uterine anomalies, use of progesterone therapy, history of cervical conization, and withdrawal of consent during the study period. All cervical measurements were performed using the same ultrasound system (Voluson S6, GE Healthcare, Milwaukee, WI). Transvaginal ultrasonography and cervical length (CL) measurements were conducted by a single experienced maternal–fetal medicine specialist more than 15 years of experience to minimize operator-dependent variability. Measurements were performed according to the International Society of Ultrasound in Obstetrics and Gynecology (ISUOG) guidelines [5]. At least three measurements were obtained over a 5-minute period, and the shortest value was recorded. At 11–14 weeks, the internal os was evaluated separately, excluding the inferior uterine segment and isthmus, and this measurement was defined as endocervical length (ECL). During second-trimester assessments, CL was measured in a sagittal plane where both the internal and external os were clearly visualized. Cervical measurements were defined as follows: (A) single-line measurement from the internal to external os (Fig. 1a);(B + C) 2-line measurement, where (B) represents the distance from the internal os to the point of cervical angulation and (C) from the angulation point to the external os (Fig. 1b);(D) anterior cervical lip length and (E) posterior cervical lip length (Fig. 1c). (A) single-line measurement from the internal to external os (Fig. 1a); (A) single-line measurement from the internal to external os (Fig. 1a); (B + C) 2-line measurement, where (B) represents the distance from the internal os to the point of cervical angulation and (C) from the angulation point to the external os (Fig. 1b); (B + C) 2-line measurement, where (B) represents the distance from the internal os to the point of cervical angulation and (C) from the angulation point to the external os (Fig. 1b); (D) anterior cervical lip length and (E) posterior cervical lip length (Fig. 1c). (D) anterior cervical lip length and (E) posterior cervical lip length (Fig. 1c). Fig. 1Cervical measurements. A: Single line cervical measurement extending from internal to external os (ECL and CL). B: Cervical measurement from internal os to cervical angulation. C: Cervical measurement from cervical angulation to external os. d: The measurements of anterior cervical lip. e: The measurements of posterior lip Cervical measurements. A: Single line cervical measurement extending from internal to external os (ECL and CL). B: Cervical measurement from internal os to cervical angulation. C: Cervical measurement from cervical angulation to external os. d: The measurements of anterior cervical lip. e: The measurements of posterior lip Maternal characteristics and obstetric history were recorded for all participants. Cervical measurements were performed prospectively at 11–14, 16–20, and 20–24 weeks of gestation. The 16-week assessment was included to detect early cervical changes and avoid delay in potential prophylactic interventions prior to the routine 20–24-week scan. A short cervix was defined as CL < 25 mm. The primary objective was to assess the correlation between first-trimester ECL and second-trimester CL (20–24 weeks). The secondary objective was to evaluate whether these measurements could predict spontaneous preterm birth longitudinally in an asymptomatic singleton population. Normality of continuous variables was assessed using the Shapiro–Wilk test. Normally distributed variables are presented as mean ± standard deviation, and non-normally distributed variables as median (min–max). Within-subject comparisons across gestational time points (11–14, 16–20, and 20–24 weeks) were performed using paired t-tests or Wilcoxon signed-rank tests, as appropriate. Correlations between measurements were evaluated using correlation analysis, and Spearman’s rank correlation coefficient (ρ) was reported when distributional assumptions were not met. An a priori sample size calculation was performed for the primary outcome, assuming a correlation coefficient of r = 0.30 between first- and second-trimester cervical measurements, with a two-sided α level of 0.05 and 80% power. The estimated minimum sample size was 85 subjects. The final study population (n = 113) was sufficient to ensure adequate statistical power. To ensure the reproducibility of our findings, we implemented a rigorous intra-examination consistency protocol: All evaluations followed the standardized ISUOG guidelines. For every participant at each time point, a minimum of three independent measurements were obtained over a 5-minute period. The shortest consistent measurement was recorded to prevent bias from transient uterine contractions. By maintaining the same operator and a strict internal measurement protocol across all 1,550 images, we ensured that the longitudinal “trends” observed in our Linear Mixed-Effects Model reflect true biological remodeling rather than technical variability. Receiver operating characteristic (ROC) curve analysis was used to determine the optimal cervical length cut-off at 20–24 weeks for predicting preterm birth. The optimal threshold was identified using Youden’s index, and sensitivity and specificity were reported with 95% confidence intervals. Longitudinal changes in cervical length were analyzed using a linear mixed-effects model, with cervical length (mm) as the dependent variable and time (11–14, 16–20, and 20–24 weeks) as a categorical factor. Time-varying covariates included B, C, D, and E measurements. A subject-specific random intercept was included to account for within-subject correlation. Interaction terms (B×time, C×time, D×time, and E×time) were included in the primary model to assess time-dependent effects; a secondary model without interaction terms was also constructed. The combined variable BC was excluded a priori due to multicollinearity, as it represents a linear combination of B and C. Estimated marginal means (EMMs) were reported for cervical length across time points, with Bonferroni-adjusted pairwise comparisons. Fixed-effect estimates (β), standard errors, and 95% confidence intervals were presented. Model fit was evaluated using − 2 log-likelihood, Akaike information criterion (AIC), and Bayesian information criterion (BIC). A two-sided p-value < 0.05 was considered statistically significant. Statistical analyses were performed using IBM SPSS Statistics for Windows, Version 23.0 (IBM Corp., Armonk, NY). Results A total of 292 singleton pregnant women presenting for first-trimester fetal aneuploidy screening were initially assessed for eligibility during the study period. Of these, 18 were excluded due to predefined exclusion criteria and 78 declined participation. An additional 80 women were lost to follow-up and 3 experienced miscarriage during the study period. Ultimately, 113 women who completed all scheduled evaluations at 11–14, 16–20, and 20–24 weeks of gestation were included in the final analysis (Fig. 2). Participants diagnosed with a short cervix were managed according to standard clinical protocols. In total, 1,550 cervical length measurements were analyzed to evaluate longitudinal changes in the cervical canal. The demographic and obstetric characteristics of the study population are presented in Table 1. The mean maternal age was 28.7 years (range: 18–42), and 27.4% (n = 31) had at least one systemic condition (hypothyroidism, hypertension, epilepsy, or asthma). Table 1Demographic values of patient populationValueAge (year)a28.70 ± 5.82 (18–42)Systemic diseaseb31(27.40%) Hypotyroidism12(10.60%) Hypertension2(1.80%) Epilepsy3(2.7%) Asthma3(2.7%)Nulliparityb47 (41.5%)Multipartyb66 (58.5%)History of 1st trimester abortionb24 (21.2%)History of 2nd trimester abortionb5 (4.40%)CL changeb Increased by weeks78 (69%) Decreased by weeks35 (31%)Short cervix at 20–24 weeksb0PPROM0Preterm birth < 34 weeks8 (7%)PPROM Preterm premature rupture of membranesavalues are given median (min-max), bValues are given number (percent) Table 1Demographic values of patient populationValueAge (year)a28.70 ± 5.82 (18–42)Systemic diseaseb31(27.40%) Hypotyroidism12(10.60%) Hypertension2(1.80%) Epilepsy3(2.7%) Asthma3(2.7%)Nulliparityb47 (41.5%)Multipartyb66 (58.5%)History of 1st trimester abortionb24 (21.2%)History of 2nd trimester abortionb5 (4.40%)CL changeb Increased by weeks78 (69%) Decreased by weeks35 (31%)Short cervix at 20–24 weeksb0PPROM0Preterm birth < 34 weeks8 (7%)PPROM Preterm premature rupture of membranesavalues are given median (min-max), bValues are given number (percent) Demographic values of patient population PPROM Preterm premature rupture of membranes avalues are given median (min-max), bValues are given number (percent) Fig. 2Study cohort diagram Study cohort diagram Cervical length measurements across gestational time points are summarized in Tables 2 and 3, with subgroup analyses for multiparous and nulliparous women presented in Tables 4 and 5 and Fig. 3. Table 2All measurements of cervix (mm) at three time periodsGestationalWeeksA(mm)B(mm)C(mm)BC(mm)Anterior lipD (mm)Posterior lipE (mm)11–14 wk37.70(29–58)20.20(11.40–39.60)20.70(1–34)40.90(27.30–69.90)12(7.20–19.60)12.56(7.30–21.90)16–20 wk39.65(21.20–59.20)21.90(11.10–38)21.10(1-33.60)42.40(22.60–59.80)12.40(6.40–19)13.75(7.40–20.80)20–24wk41.80(28–66)23.50(7.90–49.30)19.90(1-40.70)43.60(25.60–72)12.60(6.10–19.90)15.12(8.20–24)A, B, C, BC, D, E: defined at the materials & methods section and Fig. 1 (a to c). Values are given as median (min-max) Table 2All measurements of cervix (mm) at three time periodsGestationalWeeksA(mm)B(mm)C(mm)BC(mm)Anterior lipD (mm)Posterior lipE (mm)11–14 wk37.70(29–58)20.20(11.40–39.60)20.70(1–34)40.90(27.30–69.90)12(7.20–19.60)12.56(7.30–21.90)16–20 wk39.65(21.20–59.20)21.90(11.10–38)21.10(1-33.60)42.40(22.60–59.80)12.40(6.40–19)13.75(7.40–20.80)20–24wk41.80(28–66)23.50(7.90–49.30)19.90(1-40.70)43.60(25.60–72)12.60(6.10–19.90)15.12(8.20–24)A, B, C, BC, D, E: defined at the materials & methods section and Fig. 1 (a to c). Values are given as median (min-max) All measurements of cervix (mm) at three time periods A, B, C, BC, D, E: defined at the materials & methods section and Fig. 1 (a to c). Values are given as median (min-max) Table 3Percent change in cervical measurements between each time interval%∆ : percent change in cervical measurement two time periods (last minus first)+ : positive percent change, - : negative percent change values are given as median (min-max) Table 3Percent change in cervical measurements between each time interval%∆ : percent change in cervical measurement two time periods (last minus first)+ : positive percent change, - : negative percent change values are given as median (min-max) Percent change in cervical measurements between each time interval %∆ : percent change in cervical measurement two time periods (last minus first) + : positive percent change, - : negative percent change values are given as median (min-max) Fig. 3Cervical measurement changes on each time period for nulliparous and multiparous patients Cervical measurement changes on each time period for nulliparous and multiparous patients At 11–14 weeks, the median single-line CL (A) was 37.7 mm (IQR: 29–58), and two-line CL (BC) was 40.9 mm (IQR: 27.3–69.9) (Table 2). Both measurements increased over time, reaching 41.8 mm (IQR: 28–66) for CL-A and 43.6 mm for BC at 20–24 weeks, corresponding to increases of 9.29% (p < 0.01) and 5.12% (p = 0.005), respectively (Tables 2 and 3). The median posterior cervical lip (E) length increased from 12.56 mm (IQR: 7.3–21.9) at 11–14 weeks to 15.12 mm (IQR: 8.2–20.2) at 20–24 weeks, representing a 27.2% increase (p < 0.001) (Tables 2 and 3). No participant had a short cervix (< 25 mm) at the final measurement. The change in B measurement between 11 and 14 and 20–24 weeks was statistically significant (p < 0.001), whereas no significant change was observed for C measurements across gestation. Similarly, no significant differences were found in D measurements (Table 3). Subgroup analyses showed that in multiparous women (n = 66), the median ECL-A increased from 37.7 mm to 41.8 mm (10% increase, p < 0.001), while BC measurements increased from 41.2 mm to 43.9 mm (7.15% increase, p = 0.002). Posterior lip (E) measurements also increased significantly across gestation in this group (p = 0.013, p < 0.001, and p = 0.029) (Tables 4 and 5). Table 4All measurements of cervix (mm) at three time periods for multiparous and nulliparous patientsParityGestationalWeeksA (mm)B (mm)C (mm)BC (mm)Anterior lip D (mm)Posterior lip E (mm)Multiparous11–1437,35(29–58)21.65(12-39.60)20.90(1-30.30)41.20(30-69.90)12,60 ± 2,75(7,20 − 19,10)12,90 ± 2,41(8–18,70)16–2040,60(28–59,20)23(15.20–30.70)22.10(1-33.60)43.75(29-59.80)12,66 ± 2,50(9–19)14,25 ± 3,25(8–19,70)20–2442,25(28–54)24.30(15-37.40)19.95(1–29)43.90(29–55)12,98 ± 2,57(8–19,90)15,40 ± 2,87(9,90 − 24)Nulliparous11–1438(31,30–50,90)20,30(13–34)19,90(11,80 − 33,60)40,80(27,30–60,30)11,80(8,50 − 19,60)12,10 ± 3,13(7,30 − 21,90)16–2037,65(21,20–52,10)19,50(2,67 − 32,70)20,70(11,10–38)40,10(22,60 − 49,30)12,40(6,40 − 19)13,14 ± 2,82(7,40 − 20,80)20–2440,60(29,90 − 66)19,70(12,20–40,70)22,10(7,90 − 49,30)41,80(25,60–72)12,40(6,10–17,30)14,73 ± 3,04(8,20–23,10)A, B, C, BC, D, E: defined at the materials methods section values are given as median (min-max) and mean Table 4All measurements of cervix (mm) at three time periods for multiparous and nulliparous patientsParityGestationalWeeksA (mm)B (mm)C (mm)BC (mm)Anterior lip D (mm)Posterior lip E (mm)Multiparous11–1437,35(29–58)21.65(12-39.60)20.90(1-30.30)41.20(30-69.90)12,60 ± 2,75(7,20 − 19,10)12,90 ± 2,41(8–18,70)16–2040,60(28–59,20)23(15.20–30.70)22.10(1-33.60)43.75(29-59.80)12,66 ± 2,50(9–19)14,25 ± 3,25(8–19,70)20–2442,25(28–54)24.30(15-37.40)19.95(1–29)43.90(29–55)12,98 ± 2,57(8–19,90)15,40 ± 2,87(9,90 − 24)Nulliparous11–1438(31,30–50,90)20,30(13–34)19,90(11,80 − 33,60)40,80(27,30–60,30)11,80(8,50 − 19,60)12,10 ± 3,13(7,30 − 21,90)16–2037,65(21,20–52,10)19,50(2,67 − 32,70)20,70(11,10–38)40,10(22,60 − 49,30)12,40(6,40 − 19)13,14 ± 2,82(7,40 − 20,80)20–2440,60(29,90 − 66)19,70(12,20–40,70)22,10(7,90 − 49,30)41,80(25,60–72)12,40(6,10–17,30)14,73 ± 3,04(8,20–23,10)A, B, C, BC, D, E: defined at the materials methods section values are given as median (min-max) and mean All measurements of cervix (mm) at three time periods for multiparous and nulliparous patients Multi parous Nulli parous A, B, C, BC, D, E: defined at the materials methods section values are given as median (min-max) and mean In nulliparous women (n = 47), median A values at 20–24 weeks increased by 4.51% and 9.17% compared with first- and second-trimester measurements, respectively (p = 0.045 and p = 0.007). Posterior lip (E) measurements increased by 16.59% compared to 16–20 weeks and by 28.14% compared to 11–14 weeks (p = 0.007 and p < 0.001, respectively) (Table 5). Eight patients delivered preterm before 34 weeks of gestation. The median maternal age was 28.6 years. None of these patients had a cervical length of < 25 mm at 24 weeks of gestation. At 11–14 weeks, the median cervical length (CL) was 38 mm (range: 29–51 mm), which decreased slightly to 36.5 mm (range: 28–43 mm). Preterm premature rupture of membranes was not observed in any case. All preterm births occurred in the context of threatened preterm labor unresponsive to treatment. Receiver operating characteristic (ROC) curve analyses were performed to evaluate the ability of cervical length (A, mm) to predict preterm birth at ≤ 35 weeks (Fig. 4). At 11–14 weeks, the area under the curve (AUC) was 0.61 (SE = 0.09; 95% CI: 0.50–0.70; p = 0.224), with an exploratory threshold of ≤ 34.8 mm (sensitivity 50.0%, specificity 77.38%). At 16–20 weeks, the AUC was 0.58 (SE = 0.09; 95% CI: 0.46–0.70; p = 0.343), with a threshold of ≤ 31.5 mm (sensitivity 21.43%, specificity 96.72%). At 20–24 weeks, the AUC was 0.66 (SE = 0.09; 95% CI: 0.56–0.75; p = 0.060), with a threshold of ≤ 38.6 mm (sensitivity 56.25%, specificity 79.76%). None of these results reached statistical significance. Table 5Percent change in cervical measurements between each time interval for multiparous and nulliparous patients%∆ : percent change in cervical measurement between two time periods (last minus first)+ : positive percent change, - : negative percent change Values are given as median (min-max) Table 5Percent change in cervical measurements between each time interval for multiparous and nulliparous patients%∆ : percent change in cervical measurement between two time periods (last minus first)+ : positive percent change, - : negative percent change Values are given as median (min-max) Percent change in cervical measurements between each time interval for multiparous and nulliparous patients %∆ : percent change in cervical measurement between two time periods (last minus first) + : positive percent change, - : negative percent change Values are given as median (min-max) Fig. 4Receiver operating characteristic (ROC) analyses to evaluate the ability of CL (A) to predict preterm birth at 35th week Receiver operating characteristic (ROC) analyses to evaluate the ability of CL (A) to predict preterm birth at 35th week Longitudinal analysis using a linear mixed-effects model with time-by-measure interactions showed that the overall main effect of time was not significant after accounting for interactions (p = 0.484). Adjusted mean cervical length (A) values were 39.72 mm at 11–14 weeks, 40.26 mm at 16–20 weeks, and 40.97 mm at 20–24 weeks. Pairwise comparisons demonstrated a significant increase between 11 and 14 and 20–24 weeks (mean difference: 1.25 mm; p = 0.010), while other comparisons were not significant. Significant time-by-measure interactions were observed for B (p = 0.002) and C (p = 0.016), but not for D (p = 0.669) or E (p = 0.950). In the interaction model, B and C showed the strongest associations with cervical length (both p < 0.001), followed by E (p = 0.001), while D was not significant (p = 0.097). Effect estimates varied by gestational age. At 11–14 weeks, the change in A per 1-mm increase was 0.69 for B, 0.58 for C, 0.23 for E, and − 0.12 for D. At 16–20 weeks, the corresponding values were 1.01, 0.51, 0.20, and − 0.19, respectively. At 20–24 weeks, they were 0.71 (B), 0.74 (C), 0.18 (E), and − 0.03 (D) (Table 6). Table 6Linear mixed-effects model coefficients for CL (A (mm))βSE95% CIp-valueIntercept8.162.293.66 : 12.66 < 0.001 Time = 11–14 w. vs. 20–24 w.2.763.02-3.18 : 8.700.361Time = 16–20 w. vs. 20–24 w.-0.853.53-7.80 : 6.100.811B (mm)0.710.060.60 : 0.82 < 0.001 C (mm)0.740.060.62 : 0.85 < 0.001 D (mm)-0.030.12-0.26 : 0.200.775E (mm)0.180.10-0.01 : 0.380.066(Time = 11–14) × B (mm)-0.010.08-0.18 : 0.150.871(Time = 16–20) × B (mm)0.300.090.12 : 0.48 0.001 (Time = 11–14) × C (mm)-0.160.08-0.32 : -0.01 0.048 (Time = 16–20) × C (mm)-0.220.08-0.38 : -0.07 0.005 (Time = 11–14) × D (mm)-0.090.16-0.40 : 0.230.590(Time = 16–20) × D (mm)-0.160.18-0.50 : 0.190.375(Time = 11–14) × E (mm)0.050.15-0.25 : 0.340.751(Time = 16–20) × E (mm)0.010.15-0.28 : 0.300.923Time was modeled as categorical with 20–24 weeks as the reference. CL: cervical length. BC (mm) was excluded a priori due to collinearity with B (mm) and C (mm). Model fit: −2LL (REML) = 1549.57; AIC = 1553.57; BIC = 1560.92. Estimated marginal means were evaluated at the sample means of covariates; pairwise comparisons used Bonferroni adjustment Table 6Linear mixed-effects model coefficients for CL (A (mm))βSE95% CIp-valueIntercept8.162.293.66 : 12.66 < 0.001 Time = 11–14 w. vs. 20–24 w.2.763.02-3.18 : 8.700.361Time = 16–20 w. vs. 20–24 w.-0.853.53-7.80 : 6.100.811B (mm)0.710.060.60 : 0.82 < 0.001 C (mm)0.740.060.62 : 0.85 < 0.001 D (mm)-0.030.12-0.26 : 0.200.775E (mm)0.180.10-0.01 : 0.380.066(Time = 11–14) × B (mm)-0.010.08-0.18 : 0.150.871(Time = 16–20) × B (mm)0.300.090.12 : 0.48 0.001 (Time = 11–14) × C (mm)-0.160.08-0.32 : -0.01 0.048 (Time = 16–20) × C (mm)-0.220.08-0.38 : -0.07 0.005 (Time = 11–14) × D (mm)-0.090.16-0.40 : 0.230.590(Time = 16–20) × D (mm)-0.160.18-0.50 : 0.190.375(Time = 11–14) × E (mm)0.050.15-0.25 : 0.340.751(Time = 16–20) × E (mm)0.010.15-0.28 : 0.300.923Time was modeled as categorical with 20–24 weeks as the reference. CL: cervical length. BC (mm) was excluded a priori due to collinearity with B (mm) and C (mm). Model fit: −2LL (REML) = 1549.57; AIC = 1553.57; BIC = 1560.92. Estimated marginal means were evaluated at the sample means of covariates; pairwise comparisons used Bonferroni adjustment Linear mixed-effects model coefficients for CL (A (mm)) Time was modeled as categorical with 20–24 weeks as the reference. CL: cervical length. BC (mm) was excluded a priori due to collinearity with B (mm) and C (mm). Model fit: −2LL (REML) = 1549.57; AIC = 1553.57; BIC = 1560.92. Estimated marginal means were evaluated at the sample means of covariates; pairwise comparisons used Bonferroni adjustment Sensitivity analysis using a main-effects-only model showed a significant overall effect of time (p = 0.035), with higher cervical length at 20–24 weeks compared to 11–14 weeks (p = 0.030). Model fit indices (− 2LL, AIC, BIC) indicated better fit for the interaction model. Discussion In contrast to most previous studies that relied on a single cervical length measurement [8, 10–12], the present study adopted a longitudinal design, enabling evaluation of cervical changes from the first to the second trimester in an asymptomatic, low-risk population. Notably, CL increased in the majority of women (69%) at 20–24 weeks compared with first-trimester measurements, suggesting that early gestational measurements may not accurately reflect subsequent cervical dynamics. Feng et al. reported that two-line CL measurement was superior to single-line assessment in predicting preterm delivery [13]. However, our findings did not demonstrate a significant difference in predictive performance between these two methods (A vs. BC). Both single- and two-line measurements showed a significant increase between 11 and 14 and 20–24 weeks of gestation. This observation contrasts with prior reports indicating that CL begins to decrease after 16 weeks, highlighting potential variability in cervical remodeling patterns across populations. Only a limited number of studies have evaluated both single- and two-line measurement techniques concurrently [11, 13]. To our knowledge, this is the first study to separately analyze the B and C components of the two-line method. The B distance exhibited a significant increase across gestational weeks, independent of parity, and appeared to be the primary contributor to the overall elongation observed in both measurement techniques. In contrast, the C distance remained stable. These findings support the hypothesis that cervical remodeling during pregnancy may originate in the endocervical region, reflecting physiological rather than pathological changes in this low-risk cohort. We also assessed anterior and posterior cervical lip widths as additional markers of cervical remodeling. Previous studies employing shear-wave elastography have demonstrated the relevance of cervical softening in predicting preterm birth risk [11]. In our study, the anterior lip (D value) was significantly associated with CL at 20–24 weeks but showed no correlation with gestational age. Conversely, posterior lip width (E value) increased linearly with advancing gestation and was not influenced by parity. Similar to our approach, Becerra-Mojica et al. evaluated cervical consistency indices in the first trimester and reported predictive value for preterm birth, particularly in high-risk populations [17]. The absence of high-risk patients in our cohort may explain the limited predictive associations observed. Our findings are partially consistent with those of Gudicha et al., who reported longer CL in multiparous women compared with primiparous women [14]. Likewise, both single- and two-line measurements were greater in multiparous women in our study. However, when all measured parameters were analyzed collectively, parity did not significantly influence the overall pattern of cervical change, suggesting that its effect may be limited to absolute values rather than dynamic remodeling. Using a cut-off value of 25 mm for short cervix, no cases of midtrimester CL < 25 mm were identified in our cohort. While Souka et al. proposed a first-trimester threshold of 15 mm [15], and midtrimester CL < 25 mm is widely accepted as a strong predictor of preterm delivery, the absence of such cases in our study reflects the low-risk nature of the population. This, together with the relatively small sample size, limits the ability to draw definitive conclusions regarding the relationship between early CL and subsequent shortening. Previous studies have demonstrated that women with a history of preterm delivery exhibit greater cervical shortening over time [18]. Additionally, methodological differences—particularly inclusion of the uterine isthmus—may significantly influence CL measurements and their predictive value. Accurate delineation of the internal os remains especially challenging in the first trimester, when the lower uterine segment has not yet fully developed [8]. This anatomical limitation may lead to overestimation of CL in early pregnancy. Consistent with earlier reports, first-trimester CL measurements in low-risk populations appear to have limited predictive value for preterm birth. Studies by Rosenberg et al. and Zalar et al. demonstrated no significant association between early CL and gestational age at delivery [19, 20]. Similarly, Weitzner et al. found that neither early second-trimester CL nor interval changes predicted preterm birth reliably [21]. Collectively, these findings support the notion that cervical shortening detected in the midtrimester is more clinically relevant for risk stratification than early measurements. Early identification of cervical changes may also facilitate timely intervention, enabling patients who are most likely to benefit from therapies aimed at preventing or treating cervical shortening to access treatment at an earlier stage [22, 23]. Although a CL < 25 mm before 14 weeks has been associated with increased risk of preterm delivery, its predictive strength is substantially lower than when identified between 14 and 24 weeks [8]. Therefore, early screening may be justified primarily in selected high-risk populations, such as women with prior cervical surgery or second-trimester loss, rather than as a routine approach. A key strength of this study is the use of a single experienced operator and a standardized transvaginal ultrasound protocol, minimizing interobserver variability. Furthermore, this is the first longitudinal study to evaluate both single- and two-line CL measurements alongside anterior and posterior cervical lip dimensions, offering a more comprehensive assessment of cervical remodeling. However, the predictive value of first-trimester CL assessment remains controversial. Current clinical practice continues to rely predominantly on midtrimester screening due to its superior sensitivity and predictive accuracy. Our findings underscore the need for novel, combined approaches to improve early prediction models, while emphasizing the importance of strict adherence to standardized measurement techniques. However, several limitations should be acknowledged. The study population consisted exclusively of low-risk women, and the sample size was relatively small, limiting the ability to assess predictive performance for preterm birth. In addition, the absence of cases with short cervix restricted subgroup analyses. In conclusion, first-trimester cervical length measurements appear to have limited value in predicting midtrimester shortening in low-risk populations. Midtrimester assessment remains the most reliable approach for identifying women at risk of preterm birth. Given the significant perinatal and long-term consequences of preterm birth, further large-scale, multicenter studies are warranted to clarify the role of first-trimester cervical assessment and its integration into predictive strategies for preterm delivery.