Objective
We aimed to identify genetic factors that influence the rate of the first stage of labor.
Study Design
We prospectively enrolled 233 laboring nulliparous parturients. Demographic, clinical, and genetic data were collected. We evaluated the influence of population and individual variability using a nonlinear mixed effects model.
Results
Parturients who were homozygous for “G” at oxytocin receptor gene rs53576 transitioned to active labor later and thus had slower labor. Catechol- O -methyltransferase rs4633 genotype TT was associated with slower latent phase labor. Labor induction with prostaglandin was associated with faster labor, and request for meperidine was associated with slower labor. Birthweight was related inversely to the rate of the active phase.
Conclusion
There are demographic, clinical, and genetic factors that influence an individual’s rate of labor progress. This information could be used in automated form to improve the prediction of the length of the first stage of labor.
The time course of the first stage of labor varies tremendously among individual parturients. The reason for this variability is multifactorial. Demographic, clinical, and genetic factors have been shown to predict significant differences in the length of time that is required to progress to full cervical dilation. Friedman was the first tofull-text-online describe 2 phases of the first stage of labor mathematically; he developed a population-based labor curve that continues to be used today. Friedman’s model considered transition between active and latent labor to occur at the same cervical dilation in all parturients. The cervical dilation at which transition occurs has been debated in the intervening years; however, there is no evidence that it occurs at the same cervical dilation in all cases.
More recently, efforts have been made to identify factors that influence the first stage of labor on an individual basis. In the past several years, there has been a large increase in the proportion of women who deliver by cesarean section. Many of these procedures are performed for the arrest of cervical dilation. However, it may be that the rate of labor progress has not arrested but is simply slower than expected, based on individual factors that remain to be identified. Given the multitude of risks that are associated with cesarean section during delivery in the postpartum period and during subsequent pregnancies, it would be beneficial to identify characteristics of individual women that would allow more accurate prediction of the length of the first stage of labor and thus potentially prevent unnecessary cesarean section delivery.
Using mathematic modeling techniques that can address individual variability, Debiec et al found that Asian ethnicity, maternal weight, and neuraxial analgesia influenced the length of the first stage of labor. Also using these methods, Reitman et al found that a genetic polymorphism in the β2 adrenergic receptor, which is known to be important in asthma and hypertension, is associated with significantly slower progress of labor that may be responsible for the effect of Asian ethnicity described by Debiec et al.
We prospectively enrolled a cohort of 233 nulliparous parturients and evaluated both population and individual labor progress using the nonlinear mixed effects models described by Debiec et al and Reitman et al. We tested the hypothesis that the relatively common polymorphisms in the β2 adrenergic and oxytocin receptors (OXTRs) that previously have been shown to affect the time course of the first stage of labor and polymorphisms in the catechol- O -methyltransferase (COMT) gene that have been implicated in labor progress in preclinical studies would predict variability in labor progress. These findings were corrected for the effects of demographic- and treatment-related variables that significantly altered labor progress.
Materials and Methods
Data collection
With institutional review board approval from King Fahad Medical City (Riyadh, Saudi Arabia) and written informed consent, 233 nulliparous parturients were enrolled from March-September 2010. Parturients who came to the labor and delivery rooms in Maternity Hospital in King Fahad Medical City were interviewed by a physician and offered enrollment if they met inclusion criteria. Healthy nulliparous parturients who anticipated vaginal delivery of a singleton infant of estimated gestational age of 37-42 weeks were included. Patients were excluded from the study if they had any significant ongoing health problem that included preeclampsia or gestational hypertension. Patients with preeclampsia and gestational hypertension were excluded because of the disease or its treatment with magnesium and/or beta blockade might impact the progress of labor.
Demographic data that were collected on enrollment included gravidity, parity, maternal age, weight, height, and estimated gestational age. Clinical data that were recorded with respect to the time included cervical dilation, level of pain (recorded on an 11-point numerical rating scale), onset of contractions, rupture of membranes, induction, oxytocin administration, mode of delivery, type of analgesia, and birthweight.
Genomic DNA was obtained from a blood sample that was taken at the time of enrollment in the study. Blood samples were processed at Columbia University; DNA was extracted with the PureGene DNA Extraction Kit (Gentra, Minneapolis, MN). The genetic material was sent to Sequenom (Danvers, MA) for analysis at 2 functional β2 adrenergic receptor at amino acid 27 single nucleotide polymorphisms (SNPs; rs1042713 and rs1042714), 3 OXTR SNPs (rs2228485, rs2254298, and rs53576), and 3 COMT SNPs (rs4633, rs4646312, and rs6269). The data were combined into a database with demographic and clinical variables with the use of R ( http://www.r-project.org/ ). The data were then analyzed with nonlinear mixed effects models with PLTTools as a graphic user interface as described by Debiec et al and Reitman et al.
Construction and analysis of progress model
Labor progress was assessed with a biexponential equation ( Figure 1 ). Important elements of the equation include cervical dilation, L 1 (the active labor rate constant), L 2 (the latent labor rate constant), and C 1 (equal to 10 minus the cervical dilation at which transition between active and latent labor occurs). Because all patients in this analysis reached full cervical dilation, they are aligned at the time of complete cervical dilation. Time is run in reverse such that time is equal to 0 at 10-cm cervical dilation; all other times represent the time before full cervical dilation. This convention has been used in other time series models of labor progress.
Each variable was tested for an independent effect on each of the 3 parameters that describe labor progress (L 1 , L 2 , C 1 ). The –2 log likelihood values were compared in nested models with and without the variable of interest. The effect was considered statistically significant if chi-square analysis that compared the value of the objective function of the new model to the base model reached a probability value of < .05 (reduction of 3.84 for 1 degree of freedom). Variables that had significant effects on a model parameter were placed in rank order according to significance.
The final model was constructed by adding the significant univariate effects together in a stepwise fashion beginning with the most significant. Those effects that improved the model after correction for other effects were retained. Those effects that did not improve the model to statistical significance by chi-square distribution were discarded. Factors were then excluded from the final model in reverse order because those that did not significantly worsen the model on removal were discarded. Model accuracy was assessed with calculation of median predicted error. Model precision was assessed with median absolute predicted error. The 95% confidence limits were derived with log-likelihood analysis as previously described. To assess whether the effect size was clinically relevant, the predicted labor course for parturients with extreme values of significant variables was constructed with a simulation of labor progress that moved forward in time from a fixed cervical dilation with Microsoft Excel (Microsoft Corporation, Redmond, WA). Changes in latent labor were simulated from a small cervical dilation; simulation of changes in active labor rates were begun at 4 cm to capture the transition point in most patients.
The sample size was chosen to have 90% power to detect the effect of β2 adrenergic receptor at amino acid 27 polymorphism described by Reitman et al at a probability value of < .05. This would require 200 patients who had vaginal delivery; therefore, 250 patients were targeted to allow for cesarean deliveries during labor and other dropout. The sample size was chosen a priori on the basis of the β2AR polymorphism because we had only a good idea of effect size variability and the frequency of the minor allele for this polymorphism.
Results
Data collection
A total of 233 patients were enrolled according to preplanned inclusion criteria between March 2010 and September 2010. Of the initial cohort, 206 parturients reached a cervical dilation of 10 cm; 11% of them required cesarean delivery during the first stage of labor. The 206 women who completed the first stage of labor are the subject of the following analysis. Demographic and clinical data are shown in Table 1 . Population frequencies for each single nucleotide polymorphism are shown in Table 2 .
| Characteristic | Measure a |
|---|---|
| Demographic | |
| Age, y | 24 (16–37) |
| Height, cm | 158 (142–175) |
| Weight, kg | 71 (45–135) |
| Body mass index, kg/m 2 | 29 (19–58) |
| Gestational age, d | 278 (259–297) |
| Birthweight, g | 3100 (2030–4330) |
| Clinical, n (%) | |
| Intervention | |
| Artificial rupture of membranes | 120 (58) |
| Induction | 31 (15) |
| Oxytocin | 96 (47) |
| Type of analgesia | |
| None | 9 (4) |
| Meperidine | 195 (95) |
| Neuraxial | 2 (1) |
| Type of delivery | |
| Normal spontaneous vaginal | 157 (76) |
| Instrument vaginal | 31 (15) |
| Cesarean | 18 (9) |
| Single nucleotide polymorphism (total population) | Genotype | n (%) |
|---|---|---|
| β2 adrenergic receptor | ||
| rs1042713 (202) | AA | 31 (15.3) |
| AG | 94 (46.5) | |
| GG | 77 (38.1) | |
| rs1042714 (200) | CC | 115 (57.5) |
| CG | 72 (36) | |
| GG | 13 (6.5) | |
| Oxytocin receptor | ||
| rs2228485 (187) | CC | 8 (4.3) |
| CT | 51 (27.2) | |
| TT | 128 (68.4) | |
| rs2254298 (202) | AA | 9 (4.5) |
| AG | 57 (28.2) | |
| GG | 136 (67.3) | |
| rs53576 (204) | AA | 43 (21.1) |
| AG | 101 (49.5) | |
| GG | 60 (29.4) | |
| Catechol- O -methyltransferase | ||
| rs4633 (201) | AA | 56 (27.9) |
| AG | 92 (45.8) | |
| GG | 53 (26.4) | |
| rs4646312 (201) | CC | 25 (12.4) |
| CT | 82 (40.8) | |
| TT | 94 (46.8) | |
| rs6269 (204) | AA | 90 (44.1) |
| AG | 82 (40.2) | |
| GG | 32 (15.7) |
Initial and final model
Significant univariate predictors of latent labor rate, active labor rate, and the cervical dilation at which transition occurs are shown in Table 3 . Homozygousity for G at β2AR gene was associated with later transition to active labor, but the affect was no longer significant when corrected for demographic variables ( Table 3 ). Parameters that describe the initial and final adjusted model are given in Table 4 . The initial model had a median predicted error of 0.42 cm, which suggested slight over prediction, and a median absolute predicted error of 0.98 cm showed accuracy within 1-cm cervical dilation. After construction of the final model, which was improved with a probability value of 1.62E-06; the median predicted error was reduced to 0.23 cm, and the median absolute predicted error was reduced to 0.87, with an improvement of 21% after correction for measurement error that had been calculated to be 0.45 cm.
| Initial population fit | P value | Active rate of labor, cm/hr | Latent rate of labor, cm/hr | Transition point, cm | Effect size for continuous variables |
|---|---|---|---|---|---|
| Active rate | |||||
| Birthweight | .002 | 0.43 | 0.069 | 5.0 | −0.000198 |
| Delivery type | .01 | ||||
| Vaginal | 0.43 | 0.066 | 5.0 | ||
| Cesarean | 0.25 | ||||
| Latent rate | |||||
| Birthweight | .002 | 0.43 | 0.069 | 5.0 | −0.000198 |
| Oxytocin receptor | |||||
| rs2228485 | .02 | 0.43 | 0.071 | 5.2 | |
| Genotype CC | 0.037 | ||||
| Catechol- Ο -methyltransferase | |||||
| rs4633 | .02 | 0.41 | 0.069 | 4.8 | |
| Genotype TT rs6269 | 0.047 | ||||
| Genotype AA rs4646312 | 0.053 | ||||
| Genotype TT | 0.054 | ||||
| Transition point | |||||
| Oxytocin receptor | |||||
| rs2228485 | .03 | 0.42 | 0.062 | 4.8 | |
| Genotype CC | 6.0 | ||||
| Oxytocin receptor | |||||
| rs2254298 | .05 | 0.42 | 0.067 | 5.4 | |
| Genotype GG | 4.8 | ||||
| β2 adrenergic receptor at amino acid 27 d | |||||
| rs1042713 | .05 | 0.42 | 0.071 | 5.1 | |
| Genotype GG | 6.3 | ||||
| Oxytocin receptor | |||||
| rs53576 | .02 | 0.42 | 0.067 | 4.8 | |
| Genotype GG | 5.5 | ||||
| β2 adrenergic receptor at amino acid 27 | |||||
| rs1042713 | .05 | 0.42 | 0.071 | 5.1 | |
| Genotype GG | 6.3 | ||||
| Birthweight | .02 | 0.43 | 0.073 | 5.2 | −0.00089 |
| Treatment variables | |||||
| Meperidine administration | .002 | 0.55 | 0.065 | 4.6 | 0.715 a |
| Induction of labor | .006 | 0.40 | 0.066 | 5.0 | 1.25 a |
a Time constant changes reflect the overall first stage of labor that progresses 72% as fast as the nominal rate in women treated with meperidine and 125% as fast as the nominal rate in women in whom labor is induced.
| Parameter | Parameter value | 95% CI | P value | Median predicted error | Median absolute predicted error |
|---|---|---|---|---|---|
| Initial model | N/A | 0.42 | 0.98 | ||
| Active labor rate, cm/hr | 0.41 | N/A | |||
| Latent labor rate, cm/hr | 0.07 | N/A | |||
| Transition point, cm | 4.98 | N/A | |||
| Final model | 1.62E–06 ix | 0.23 | 0.87 | ||
| Active labor rate, cm/hr | 0.53 | 0.41–0.66 | |||
| Latent labor rate, cm/hr | 0.07 | 0.05–0.09 | |||
| Transition point, cm | 4.52 | 3.96–5.13 | |||
| Effect of birthweight on active labor rate, cm/hr/kg | −0.22 | −0.37 to −0.08 | |||
| Time constant after meperidine administration, unitless | 0.76 | 0.60–0.93 | |||
| Time constant after induction of labor, unitless | 1.23 | 1.08–1.41 | |||
| Transition point (oxytocin receptor a rs53576 genotype GG) | 5.02 | 0.03–1.02 b | |||
| Latent labor rate (catechol- O -methyltransferase c rs4633 genotype TT) | 0.05 | (0.001–0.04) b |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
