Abstract
Background
A short cervix in the second trimester is known to increase the risk of preterm birth, which can be reduced with the administration of vaginal progesterone. However, some studies have suggested that a significant number of cases still experience preterm birth despite progesterone treatment.
Objective
This study was aimed to investigate the potential value of transvaginal cervical elasticity measured by E-Cervix as a predictor for spontaneous preterm birth (sPTB) in singleton pregnancies receiving progesterone treatment for a short cervix (CL ≤ 2.5 cm) diagnosed at 18 to 24 weeks’ gestation.
Study design
This prospective study was conducted at a single center premature high-risk clinic from January 2020 to July 2022. Singleton pregnancies with a short cervix at 18 to 24 weeks’ gestation were enrolled. Cervical elastography using E-Cervix was performed, and maternal and neonatal demographic characteristics, cervical length (CL), elasticity contrast index (ECI), cervical hardness ratio, mean internal os strain (IOS), and mean external os strain (EOS) were compared before and after progesterone treatment in sPTB and term birth groups. Multivariate logistic regression was used to analyze the association between elasticity parameters and spontaneous preterm birth. The screening performance of CL and optimal cervical elasticity parameters in predicting sPTB was evaluated using receiver-operating characteristic (ROC) curve analysis.
Results
A total of 228 singleton pregnant women were included in the study, among which 26 (11.4%) had sPTB. There were no significant differences in maternal characteristics and gestational age at enrollment between women with and without sPTB. At the start of progesterone treatment, there were no significant differences in cervical elasticity parameters between the two groups. After two weeks of progesterone treatment, women who had sPTB showed significantly higher levels of ECI, IOS, EOS (p = 0.0108, 0.0001, 0.016), and lower hardness ratio (p = 0.011) compared to those who had a full-term birth. Cervical length did not show significant differences between the two groups, regardless of whether progesterone treatment was administered before or after. Among the post-treatment cervical elasticity parameters, IOS and EOS were associated with a 3.38-fold and 2.29-fold increase in the risk of sPTB before 37 weeks (p = 0.032, 0.047, respectively). The AUROC of the combined model including CL, IOS, and EOS (0.761, 95% CI0.589–0.833) was significantly higher than the AUROC of CL alone (0.618, 95% CI 0.359–0.876). At a fixed false-positive of 13%, the addition of IOS and EOS in the CL model increased sensitivity from 34.6% to 57.6%, PPV from 25.7% to 36.5%, and NPV from 91.1% to 94.1%.
Conclusion
When assessing the risk of sPTB in singleton pregnancies with a short cervix receiving progesterone therapy, relying solely on cervical length is insufficient. It is crucial to also evaluate cervical stiffness, particularly the strain of the internal and external os, using cervical elastography.
Introduction
According to the World Health Organization, preterm birth (PTB), defined as birth before 37 weeks’ gestation, is the leading cause of neonatal mortality worldwide [Citation1]. Approximately two thirds of all premature births are classified as spontaneous preterm birth (sPTB), which includes premature onset of labor or preterm pre-labor rupture of membranes (PPROM) [Citation2]. Premature babies are at a high risk of short-term health complications such as respiratory distress syndrome (RDS), bronchopulmonary dysplasia, and long-term neurodevelopmental disabilities including cerebral palsy and cognitive impairment [Citation3,Citation4].
Cervical shortening, defined as a cervical length (CL) of 25 mm or less, as assessed through transvaginal sonography (TVS) between 22 + 0 to 24 + 6 weeks of gestation, is a common precursor to spontaneous PTB [Citation5]. Previous studies have shown that vaginal progesterone can reduce the risk of sPTB by 40% in asymptomatic women with a singleton pregnancy and a short cervix [Citation6]. However, despite treatment with vaginal progesterone, sPTB still occurs in one third of women with a short cervix [Citation7]. This may be because cervical length only evaluates morphological changes and does not provide a high positive predictive value [Citation8]. Therefore, there is a need for innovative approaches to enhance the screening and predictive capabilities for identifying individuals at risk of preterm delivery who are undergoing progesterone treatment. Alternative treatment options, such as cerclage, are considered for individuals with a short cervix who may not benefit from vaginal progesterone administration [Citation9,Citation10].
One promising technique in this regard is transvaginal ultrasound E-Cervix, which utilizes the differences in deformation caused by the pulsing of the cervical artery between cervixes of varying stiffness [Citation11]. This technique incorporates these indices into a multi-parametric quantification of cervical stiffness [Citation11]. The objective of our study was to investigate the potential value of transvaginal ultrasound E-Cervix in assessing the progesterone response in pregnant women with a short cervix (CL ≤ 2.5 cm) diagnosed at 18 to 24 weeks’ gestation. This was achieved by comparing cervical elasticity parameters before and after progesterone treatment.
Materials and methods
Study design and participants
This study was conducted from January 2020 to July 2022 at Shanghai Changning Maternity & Infant Health Hospital in Shanghai, China. It was a single-center prospective cohort study. We enrolled pregnant women who were diagnosed with a short cervix (≤2.5 cm) at 18 to 24 weeks’ gestation and agreed to receive progesterone treatment for the prevention of preterm birth.
We excluded pregnant women with multiple pregnancies, fetal defects detected during system ultrasound scan, cervical length (CL) ≤0.5 cm at enrollment, cerclage placed for failed progesterone treatment, obstetrical complications (hypertensive disorders complicating pregnancy, gestational diabetes mellitus, preterm premature rupture of membranes, and placental abruption), and cervical malformation.
Spontaneous preterm birth (sPTB) was defined as giving birth before 37 weeks’ gestation. Women with a short cervix on vaginal ultrasound, defined as a cervical length ≤25 mm, were treated with vaginal progesterone [Citation6]. Progesterone treatment involving daily vaginal administration of 200 mg micronized natural progesterone at bedtime. Treatment was commenced upon diagnosis of a shortened cervix and continued until 36 weeks’ gestation. Cervical elastography measurements and cervical length assessment were conducted prior to progesterone treatment (pre-progesterone) and two weeks after initiation of progesterone treatment (post-progesterone). We documented demographic, sonographic, elastographic, and obstetric parameters, as well as delivery outcomes. This study received approval from our ethics committee (CNFBLLWJWSH-201905001), and informed consent was obtained from all participants.
Cervical elastography assessment
Cervical elastography assessment was performed using a SAMSUNG WS80A machine equipped with a transvaginal (3–10 MHz) probe and ElastoScan software. The examinations were carried out by an operator certified for measuring cervical length by the Fetal Medicine Foundation and trained on the new software before the research.
Before the examination, participants were instructed to empty their bladder and were placed in the lithotomy position. The probe was gently inserted into the vaginal fornix without applying pressure to obtain a midsagittal section of the cervix. Participants were instructed to breathe normally while the probe was kept steady and switched to elastography mode. A 2-dimensional grayscale image and an elasticity image were simultaneously displayed on the screen, and the elasticity image was assessed after stabilization of the image.
Cervical tissue stiffness was represented by color on the elastogram. In this study, red represented a lower degree of elasticity and harder tissue, while blue represented a higher degree of elasticity and softer tissue. The regions of interest (ROI) for analysis included only the cervical tissue. Once the ROI was fixed, the results appeared in the bottom right corner of the screen. Quantification of cervical strain was calculated using a data analysis system that directly analyzed raw data from the ROI. It was described as the elasticity contrast index (ECI), hardness ratio, mean strain level within 1 cm from the internal os (IOS) and external os (EOS) (). The elasticity contrast index (ECI) is an average contrast index of the pixels within the entire cervical ROI and range from 0 to 81 which represent the homogeneity or heterogeneity of the object within the ROI. The hardness ratio is the percent of the upper 30% of the red (hard) pixel area divided by the total pixel area within the ROI and its measurement value ranges between 0% and 100% (with 100% being the hardest). The internal ostium (IOS) and external ostium (EOS), each shown as a decimal with two significant digits, are the ratio of the total strain to the total number of pixels in the respective ROI and describe the average strain of the cervical IOS and EOS. The values of IOS and EOS range from 0 to 1, and the values closer to 0 corresponding to harder tissues. For each case, three elastographic images were obtained, and their mean value was utilized for analysis.
Figure 1. Cervical elastography image of women with a short cervix at 24 weeks of gestation. A 2-dimensional grayscale image (left) and an elasticity image (right) are simultaneously displayed on the screen. Elastographic parameters and cervical length are shown in the bottom right corner of the screen.
Statistical analysis
Given the 5–11% rate of preterm delivery among pregnant women, the confidence interval of 95%(α = 0.05), and accuracy of 3%(p = 0.03), the required sample size of the present study was at least 70 participants. The normality of the data was evaluated using the Kolmogorov-Smirnov test. Descriptive data were presented as mean ± standard or median (interquartile range [IQR]) for continuous variables and as numbers and percentages for categorical variables. Comparisons between the preterm and term groups were performed using chi-square tests, continuity correction tests, and Fisher exact tests for categorical variables and using Student t tests or Mann-Whitney U tests for continuous variables.
Multivariable logistic regression analysis was performed with spontaneous preterm birth as the binary outcome variable and each of the E-Cervix parameters as continuous predictors. Gestational age and CL were included as confounding covariates. In this logistic regression model, only parameters with P values of <0.05 in the univariate analysis were included. To evaluate the performance of a parameter or combination of parameters for the prediction of spontaneous preterm birth, the area under the curve (AUC) was calculated by using a receiver operating characteristic curve. Statistical significance was defined as a p-value of <0.05. All statistical analyses were performed using the statistical software package SPSS 25.0 (SPSS Inc, Chicago, IL) and Medcalc (Medcalc Software, Mariakerke, Belgium).
Results
Characteristics of study population
Of the 1129 participants from the clinic titled prevention of preterm birth in high-risk pregnancies, a total of 281 singleton pregnancies were diagnosed with short cervix (CL ≤ 25mm) at 18 to 24 weeks of gestation and received progesterone treatment for the prevention of spontaneous PTB. 15 pregnancies ended before 24 weeks’ gestation due to fetal anomalies. 13 had cerclage placed. 15 had prior iatrogenic preterm birth due to preeclampsia and fetal growth restriction. 10 women were lost to follow-up. Thus, 228 pregnancies were enrolled in final analysis, of which spontaneous PTB before 37 weeks’ gestation occurred in 26 women ().
Figure 2. Flowchart of study population. TOP, termination of pregnancy; PTB, preterm birth.
Maternal and neonatal demographic characteristics for women delivered ≥37 weeks and <37 weeks are shown and compared in . There were no marked differences in the maternal characteristics and the gestational age at the time of enrollment between the full-term birth and spontaneous PTB groups. The group of sPTB had lower birth weight and higher rate of neonatal intensive care unit (NICU) admission than the group of full-term birth.
Table 1. Maternal and neonatal characteristics of the full-term birth group versus the sPTB group.
Cervical elasticity parameters and the length of the cervix before and after progesterone treatment
At the onset of progesterone treatment, there were no significant differences in cervical elasticity parameters between the groups of women who had spontaneous preterm birth and those who had term birth. However, after two weeks of receiving progesterone treatment, the sPTB group showed significantly higher levels of ECI, IOS, EOS (p = 0.0108, 0.0001, 0.016), and lower hardness ratio (p = 0.011) compared to the full-term birth group. Interestingly, irrespective of whether progesterone treatment was given before or after, there were no notable variations in cervical length between the groups of women who had spontaneous preterm birth and those who had term birth (). In full-term birth group, the difference in CL between pretreatment and post-treatment was 0.7(p = 0.075), and the respective difference in sPTB group was 1.08(p = 0.055).
Table 2. Pretreatment and post-treatment cervical elasticity parameters and cervical length of the full-term birth group versus the sPTB group.
Association between post-treatment elasticity parameters and sPTB
After adjusted for gestational age and CL, multivariable logistic regression analysis revealed that among the post-treatment cervical elasticity parameters, IOS and EOS were significantly associated with spontaneous PTB before 37 weeks gestation (OR = 3.38,2.19; p = 0.032,0.047, respectively) ().
Table 3. Association between post-treatment cervical elasticity parameters and sPTB before 37 weeks’ gestation.
Prediction of sPTB using post-treatment cervical length and elasticity parameters
The predictive performance by using post-treatment CL and elasticity parameters associated with sPTB were shown in . The ROC curves for spontaneous PTB prediction using post-treatment CL alone and the combination of CL, IOS and EOS are presented in . In the prediction of sPTB, the AUROC of the combination of CL, IOS and EOS was significantly higher (0.761, 95% CI0.589–0.833) than the AUROC of CL alone (0.618, 95% CI 0.359–0.876; the difference between areas was 0.143, p = 0.0021). At a fixed false-positive of 13%, the addition of IOS and EOS in the CL model increased sensitivity from 34.6% to 57.6%, PPV from 25.7% to 36.5%, and NPV from 91.1% to 94.1%.
Figure 3. Receiver operating characteristic (ROC) curve for the predication of sPTB before 37 weeks of gestation by using cervical length alone (----) and the combination of cervical length, ECI and IOS (——) after the treatment with progesterone.
Table 4. Prediction of sPTB using post-treatment cervical length and optimal elasticity parameters.
Discussion
Principal findings
In this study, we investigated the effectiveness of cervical elasticity parameters in detecting changes after progesterone treatment as a means to predict sPTB in singleton pregnancies with a short cervix. Our findings revealed several key points. Firstly, after two weeks of progesterone treatment, the sPTB group exhibited higher ECI, IOS, EOS, and lower hardness ratio compared to the full-term birth group. Secondly, regardless of whether progesterone treatment was administered before or after, there was no significant difference in cervical length (CL) between the full-term birth group and the sPTB group. Thirdly, higher IOS and EOS were associated with a 3.38-fold and 2.19-fold increase in the risk of sPTB occurring before 37 weeks gestation, respectively. Lastly, combining IOS and EOS with CL measurements taken after treatment showed improved predictive performance for sPTB before 37 weeks gestation compared to using post-treatment CL alone.
Results in the context of what is known
Progesterone plays a critical role in maintaining pregnancy and preventing cervical softening by modifying cytokines, regulating macrophages in the stroma, and altering the extracellular matrix in the cervix [Citation12,Citation13]. Vaginal progesterone administration is recommended for patients with a singleton pregnancy and a short cervix in the mid-trimester to prevent sPTB [Citation6]. Currently, transvaginal ultrasound measurement of CL is the primary method to monitor the efficacy of progesterone treatment [Citation14]. Cervical shortening after progesterone treatment may be an indication of cerclage when CL reaches ≤10mm in subsequent CL measurement [Citation9, Citation15]. Hernandez-Andrade et al. reported that a 20% reduction in CL had highest association with sPTB compared to reductions of ≤10% or ≤5mm [Citation16]. Ninkov et al. defined cervical shortening as a decrease in CL measurement by more than 5 mm and found that a reduction in CL between 16 and 19 weeks was more predictive of sPTB <37 weeks than a reduction between 19 and 22 weeks [Citation17]. In our study, we observed a decrease in CL after progesterone treatment in both groups. However, this reduction was minor, less than 1.5 mm, and did not present a significant difference when comparing the pretreatment and post-treatment CL measurement. Our findings may differ from previous study due to mean time interval of 2 weeks between CL measurement. In a placebo-controlled randomized trial (RCT) conducted by O’Brien et al. [Citation18], the group receiving progesterone treatment showed less CL shortening compared to the placebo group, especially in cases where the cervix was initially short. However, the study did not establish a correlation between CL changes with pregnancy outcomes. Based on the reduction in post-treatment CL, we assessed its performance in predicting sPTB and found only moderate predictability (AUROC of 0.618 with PPV 25.7% and NPV91.1%). Therefore, we evaluated the cervical softness as an additional marker for predicting sPTB.
Cervical length and cervical stiffness are known to be the two key characteristics of the cervix that are associated with the risk of preterm birth. The stiffness of the cervix may play a role in how cervical length change in response to intrauterine pressure. Ultrasound elastography has been used to estimate the elastic properties of tissues [Citation19]. According to a study by Hernandez-Andrade et al. cervical hardness was independent of CL and a soft cervix identified by elastography could be an additional marker for predicting sPTB [Citation20]. Similar to the study, we found even in the absence of changes in CL between the full-term group and the sPTB group, there is the possibility of "physiological remodeling" of the cervix following vaginal progesterone administration, which may explain the observed higher ECI, IOS, EOS, and lower Hardness ratio in women who delivered preterm compared to those who carried to full term. However, our study was different from the study mentioned above in that we used transvaginal E-Cervix to evaluate the stiffness of whole cervix, which compression supplied by natural and inherent organ movements (including vessel pulsation, respiration and even fetal movements against the wall of womb) rather than manual compression such as a shear wave, making it less instability in image quality [Citation21]. Furthermore, we observed the changes of cervical hardness before and after the treatment with progesterone rather than at a single time point.
In women with a short cervix treated by progesterone, Jung YJ et al. found IOS before and after treatment and EOS after treatment were associated with 2.92,4.39 and 3.65-fold increase in the risk of sPTB at <32 weeks [Citation22]. Our observations of cervical elasticity parameters after treatment for spontaneous preterm birth are in line with theirs.
Clinical implications
Our study revealed a preterm birth rate of 11.4% in pregnancies with a short cervix that underwent progesterone treatment, surpassing the 7.82% reported by Jung YJ [Citation22]. This discrepancy, however, could be attributed to the different end-points used in each study. While our research used sPTB before 37 weeks gestation as an end-point, Jung YJ’ study used sPTB before 32 weeks of gestation as the primary outcome [Citation22].
Interestingly, no significant differences were discovered in CL between full-term and preterm birth, even when comparing before and after treatment. Meanwhile, we found that cervical IOS, EOS and ECI exhibited elevation, and hardness ratio declined post-treatment in women who underwent preterm birth compared to those with full-term births. The multivariate regression analysis, considering post-treatment cervical elasticity parameters, demonstrated an increased risk of sPTB associated with higher IOS and EOS. Considering this, it’s evident that relying solely on CL measurement isn’t an accurate method to identify high-risk women for sPTB, who are undergoing progesterone treatment. Therefore, incorporating the evaluation of cervical hardness, particularly IOS and EOS post-treatment, with CL measurement, proved to increase prediction efficiency. Consequently, additional assessments of these cervical rigidity parameters may be essential to predict such births more accurately.
In this study, we found that among the parameters of cervical elasticity following progesterone treatment, IOS and EOS were the most predictive of sPTB. This suggests that if there is no improvement in IOS and EOS after two weeks of progesterone treatment, it may indicate a gradual softening of the cervix in those pregnancies.
Strengths and limitations
One notable strength of our study is its longitudinal and prospective design, which specifically targeted on women with a short cervix in high-risk pregnancies. This approach enabled us to consider individual variations and minimize potential confounding factors. However, it is important to acknowledge certain limitations. Firstly, the acquisition of cervical elastograms was performed by only one operator, which may have limited our ability to assess interobserver consistency. Secondly, our study was conducted at a single center and had a relatively small sample size. Therefore, a multicenter study with a larger participant cohort is needed to provide addition for our findings.
Conclusion
In summary, our study demonstrated the association between cervical elasticity parameters and the risk of preterm birth. When assessing the risk of sPTB in singleton pregnancies with a short cervix receiving progesterone therapy, it is insufficient to rely solely on cervical length. It is crucial to also evaluate cervical stiffness, particularly the strain of the internal and external os, through the use of cervical elastography.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Additional information
Funding
References
- World Health Organization. Recommended definitions, terminology and format for statistical tables related to the perinatal period and use of a new certificate for cause of perinatal deaths. Acta Obstet Gynecol Scand. 1977;56:247–253.
- Goldenberg RL. The management of preterm labor. Obstet Gynecol. 2002;100(5 Pt 1):1020–1037. doi:10.1016/s0029-7844(02)02212-3.
- Blencowe H, Cousens S, Chou D, et al. Born too soon: the global epidemiology of 15 million preterm births. Reprod Health. 2013;10(Suppl 1):S2. doi:10.1186/1742-4755-10-S1-S2.
- Campbell S. Prevention of spontaneous preterm birth: universal cervical length assessment and vaginal progesterone in women with a short cervix: time for action!. Am J Obstet Gynecol. 2018;218(2):151–158. doi:10.1016/j.ajog.2017.12.222.
- Di Renzo GC, Gratacos E, Kurtser M, et al. Good clinical practice advice: prediction of preterm labor and preterm premature rupture of membranes. Int J Gynecol Obstet. 2019;144:340–346.
- Romero R, Conde-Agudelo A, Da Fonseca E, et al. Vaginal progesterone for preventing preterm birth and adverse perinatal outcomes in singleton gestations with a short cervix: a meta-analysis of individual patient data. Am J Obstet Gynecol. 2018;218(2):161–180. doi:10.1016/j.ajog.2017.11.576.
- Norwitz ER, Caughey AB. Progesterone supplementation and the prevention of preterm birth. Rev Obstet Gynecol. 2011;4(2):60–72.
- Kim H, Hwang HS. Elastographic measurement of the cervix during pregnancy: current status and future challenges. Obstet Gynecol Sci. 2017;60(1):1–7. doi:10.5468/ogs.2017.60.1.1.
- Makrydimas G, Barmpalia Z, Sotiriadis A. Cervical cerclage for women with shortening cervix while on progesterone. Am J Obstet Gynecol. 2019;220(2):209–210. doi:10.1016/j.ajog.2018.10.020.
- Enakpene CA, DiGiovanni L, Jones TN, et al. Cervical cerclage for singleton pregnant patients on vaginal progesterone with progressive cervical shortening. Am J Obstet Gynecol. 2018;219(4):397.e1–397.e10. doi:10.1016/j.ajog.2018.06.020.
- Patberg ET, Wells M, Vahanian SA, et al. Use of cervical elastography at 18 to 22 weeks’ gestation in the prediction of spontaneous preterm birth. Am J Obstet Gynecol. 2021;225(5):525.e1–525.e9. doi:10.1016/j.ajog.2021.05.017.
- Kirby MA, Heuerman AC, Custer M, et al. Progesterone receptor-mediated actions regulate remodeling of the cervix in preparation for preterm parturition. Reprod Sci. 2016;23(11):1473–1483. doi:10.1177/1933719116650756.
- Weatherborn M, Mesiano S. Rationale for current and future progestin-based therapies to prevent preterm birth. Best pract. Best Pract Res Clin Obstet Gynaecol. 2018;52:114–125. doi:10.1016/j.bpobgyn.2018.03.008.
- Khalifeh A, Berghella V. Universal cervical length screening in singleton gestations without a previous preterm birth: ten reasons why it should be implemented. Am J Obstet Gynecol. 2016;214(5):603.e1–5. doi:10.1016/j.ajog.2015.12.017.
- Daskalakis G, Loutradis D, Antsaklis A, et al. A stepwise approach for the management of short cervix: time to evolve beyond progesterone treatment in the presence of progressive cervical shortening. Am J Obstet Gynecol. 2019;220(4):404–405. doi:10.1016/j.ajog.2019.01.221.
- Hernandez-Andrade E, Huntley ES, Sibai B, et al. A reduction in cervical length after starting vaginal progesterone in women with short cervix is significantly associated with preterm delivery ≤34 weeks and <37 weeks of gestation. Ultrasound Obstet Gynecol. 2023 Published online November 2. doi:10.1002/uog.27527.
- Ninkov T, Nathan EA, Belcher CN, et al. The clinical utility of ongoing sonographic cervix length surveillance in pregnancies prescribed vaginal progesterone therapy. Aust N Z J Obstet Gynaecol. 2023;63(2):198–203. doi:10.1111/ajo.13592.
- O’Brien JM, Defranco EA, Adair CD, et al. Effect of progesterone on cervical shortening in women at risk for preterm birth: secondary analysis from a multinational, randomized, double-blind, placebo-controlled trial. Ultrasound Obstet Gynecol. 2009;34(6):653–659. doi:10.1002/uog.7338.
- Ophir J, Céspedes I, Ponnekanti H, et al. Elastography: a quantitative method for imaging the elasticity of biological tissues. Ultrason Imaging. 1991;13(2):111–134. doi:10.1177/016173469101300201.
- Hernandez-Andrade E, Maymon E, Luewan S, et al. A soft cervix, categorized by shear-wave elastography, in women with short or with normal cervical length at 18-24 weeks is associated with a higher prevalence of spontaneous preterm delivery. J Perinat Med. 2018;46(5):489–501. doi:10.1515/jpm-2018-0062.
- Du L, Zhang LH, Zheng Q, et al. Evaluation of cervical elastography for prediction of spontaneous preterm birth in low-risk women: a prospective study. J Ultrasound Med. 2020;39(4):705–713. doi:10.1002/jum.15149.
- Jung YJ, Kwon H, Shin J, et al. The feasibility of cervical elastography in predicting preterm delivery in singleton pregnancy with short cervix following progesterone treatment. Int J Environ Res Public Health. 2021;18(4):2026. doi:10.3390/ijerph18042026.