Which medication would the nurse identify as being both for cervical ripening during labor and as a stomach protectant?

Biol Res Nurs. 2020 Apr; 22(2): 157–168.

Nicole S. Carlson, PhD, CNM,1 Jennifer K. Frediani, PhD, RD, ACSM-CES,1 Elizabeth J. Corwin, PhD, RN,1,2 Anne Dunlop, MD, MPH,1,3,4 and Dean Jones, PhD5

Abstract

Objectives:

The purpose of this study was to evaluate differences in the metabolic pathways activated in late-pregnancy serum samples among African American women who went on to have term (≥37 weeks) labor induction requiring high total oxytocin doses to complete first-stage labor compared to those in similar women with low-oxytocin labor inductions.

Study Design:

Case–control study (N = 27 women with labor induction with successful cervical ripening: 13 requiring the highest total doses of synthetic oxytocin to progress from 4- to 10-cm cervical dilation and 14 requiring the lowest total doses) with groups balanced on parity and gestational age. Serum samples obtained between 24 and 30 weeks’ gestation were analyzed using ultra-high-resolution metabolomics. Differentially expressed metabolites between high-oxytocin induction cases and low-oxytocin induction comparison subjects were evaluated using linear regression with xmsPANDA. Metabolic pathways analysis was conducted using Mummichog Version 2.0, with discriminating metabolites annotated using xMSannotator Version 1.3.

Results:

Labor processes were similar by group with the exception that cases received over 6 times more oxytocin between 4- and 10-cm cervical dilation than comparison women. Induction requiring high total doses of synthetic oxytocin was associated with late-pregnancy serum levels of metabolites from the linoleate and fatty acid activation pathways in term, African American women.

Conclusion:

Serum levels of several lipid metabolites predicted more complicated labor induction involving higher doses of synthetic oxytocin to complete first-stage labor. Further investigation in larger, more diverse cohorts of women is needed to identify potential targets to prevent failed labor induction.

Keywords: induction of labor, mechanisms, metabolomics, parturition, oxytocin

During the past few decades, the use of labor induction has increased substantially (Declercq et al., 2013). Currently, 21.6% of women in the United States end their pregnancy with a labor induction (Martin et al., 2018). Labor induction for women without advanced cervical dilation typically involves a two-part approach. First, the cervix is ripened with a combination of medical and/or transcervical dilators (Penfield & Wing, 2017). Then, an intravenous infusion of synthetic oxytocin is initiated. Oxytocin infusions are typically continued until birth as long as both the woman and fetus tolerate contractions. Labor personnel titrate oxytocin infusions to stimulate regular uterine contractions resulting in progressive cervical dilation, increasing them gradually and typically capping them at 20 mU/min in most facilities (“ACOG Practice Bulletin No. 107: Induction of Labor,” 2009).

Although these processes of a typical labor induction are well known and standardized in many facilities, huge variability exists in the success of labor induction. For example, rates of cesarean delivery following labor induction among low-risk, nulliparous women averaged 32% in a sample of 240 California hospitals, with some facilities reporting cesarean rates as high as 60% (Main, 2018). In another analysis from a large, nationally representative sample of women in the United States, failed labor induction prior to 6-cm cervical dilation was the indication for 53% of all nulliparous cesarean births (Zhang et al., 2010). In an attempt to help prevent failed labor induction and its downstream consequences, much work has focused on improving the standardization of labor-induction protocols with an emphasis on allowing sufficient time for women to complete cervical ripening and dilation (Grobman et al., 2018; Spong et al., 2012). However, additional key factors, including maternal characteristics such as obesity and advanced age, are independently associated with prolonged or difficult labor inductions (Beckwith et al., 2017; Ellis et al., 2019; Hill et al., 2015; Roloff et al., 2015). Moreover, the burden of unsuccessful labor induction is not distributed equally among women of different racial or ethnic backgrounds. In the present study, we focused on non-Hispanic Black women who have the highest rates of induction compared to women of other racial/ethnic groups in the United States and are thus more likely to experience poor labor outcomes resulting from prolonged or difficult labor following induction (Creanga et al., 2014, 2017; Singh et al., 2018).

Unfortunately, the physiologic mechanisms underlying differences in women’s response to labor-induction interventions remain largely unknown. However, there are hints from several investigations that uterine contractile frequency and strength during labor may be linked to the maternal metabolic milieu during pregnancy, with implications for labor induction complications (Parkington et al., 2014; Shmygol et al., 2007; Zhang et al., 2007). For example, researchers observed that cholesterol had an inhibitory effect in vitro on both spontaneous and oxytocin-induced contractions (Moynihan et al., 2006; Zhang et al., 2007), which are essential for labor initiation (cervical ripening, uterine contractile synchronization) and labor progression (contractile frequency and force). These inhibitory effects appear to be related to uterine-signaling interference caused by the incorporation of cholesterol in the lipid bilayers of myometrial cell membranes (Wray, 2015). In a rat model, high-fat, high-cholesterol diet was associated with unstable, asynchronous contractions during labor that exhibited a blunted response to oxytocin (Muir et al., 2016). Animals fed the high-fat diet experienced decreased levels of circulating prostaglandins and blunted expression of myometrial gap junctions near the time of labor. Myometrial gap junctions act as intracellular conduits for cell-to-cell communication in the uterus and are essential for coordinated contractions during labor (Cunningham et al., 2018). Thus, increased levels of cholesterol and other dietary fats during pregnancy appear to detrimentally alter prostaglandin levels and myometrial gap junctions near the time of labor, causing asynchronous uterine contractions.

There is also evidence that elevated levels of free fatty acids in pregnancy may contribute to decreases in uterine contractile strength (Gam et al., 2017). When excess dietary free fatty acid is consumed, it is sequestered as triglyceride in new adipocytes or within fat droplets in nonadipose tissues (ectopic fat; Frayn, 2010). Interestingly, Gam and colleagues (2017) found that myometrial cells taken from obese women at term had ectopic fat deposits and reduced muscle fiber compared to cells from normal weight women. It is not known how these changes might influence clinical characteristics (synthetic-oxytocin dosing requirements) or outcomes (labor duration, rates of failed labor induction). However, this evidence of a link between labor difficulty and maternal physiologic factors suggests the possibility that observed differences in labor-induction success may be partially predicted by the maternal metabolic environment.

Thus, we conducted a case–control study to identify metabolic profiles activated in serum collected during late pregnancy from African American women who experienced difficult labor induction (i.e., high dose of synthetic-oxytocin infusion to complete first-stage labor from 4- to 10-cm cervical dilation following successful cervical ripening). We chose to focus on total dose of oxytocin rather than labor length with the rationale that total oxytocin dose captures both the duration of first-stage labor and the oxytocin titration, thus better reflecting myometrial response to oxytocin infused during induction. We hypothesized that African American women with difficult labor inductions would have distinct late-pregnancy metabolic profiles compared to similar women who had uncomplicated labor induction (i.e., low dose of synthetic oxytocin required to complete first-stage labor following successful cervical ripening).

Methods and Materials

Study Design and Ethical Approvals

The present investigation was a matched case–control study utilizing serum samples and medical-record data on labor/birth outcomes of African American women participating in the Emory University African American Vaginal, Oral, and Gut Microbiome in Pregnancy Cohort Study (Corwin et al., 2017). In that longitudinal investigation (the parent study), women were enrolled at 8–14 weeks of pregnancy then followed until birth to uncover intrarace risk and protective factors of preterm birth. Investigators collected blood and microbiome samples in early (8–14 weeks) and late (24–30 weeks) pregnancy. For the current investigation, we focused on a subset of women from the parent study who achieved term gestations (gestational age ≥37 0/7 weeks). Utilizing blood samples from late pregnancy for high-throughput metabolomics, we conducted expanded medical-record reviews to collect information on the course of the mothers’ term labor. We obtained institutional review board approval from Emory University for this project.

Study Participants

Women from the parent study (African American, age 18–40 years) were eligible for the present investigation if they had a term labor induction (gestational age ≥ 37 0/7 weeks at hospital admission) with a nonanomalous singleton fetus in vertex presentation, were not diabetic (no gestational diabetes mellitus nor preexisting diabetes) or hypertensive, and completed the first stage of labor. We excluded labor inductions performed for pregnancy complications that may have altered the labor course (e.g., placenta abruption) and inductions that did not require cervical ripening. The rationale for focusing on women who required cervical ripening was that it enabled us to evaluate the labor course in a group of women who started with similar levels of physiologic preinduction cervical preparation. Women included had inductions for the following indications: elective induction, gestational age > 41.0 weeks, and decreased fetal movement.

To obtain information on each woman’s demographic factors, prenatal complications, parity, height and weight, gestational age at hospital admission, and labor induction, we conducted abstractions of the medical record. Using a standardized chart-abstraction tool, two experienced nurse-practitioners collected information from each participant’s medical record. To confirm the accuracy of data abstracted from medical records, we randomly selected 20 charts for repeat medical-record review, which represented 6.9% of the abstractions performed, as recommended in protocols for chart abstraction reliability testing (To et al., 2008). We entered clinical data into REDCap and downloaded it into R Statistical Software Version 3.5.2 (R Foundation Vienna, Austria) for data cleaning and statistical analyses.

Parity was recorded as a discrete variable and categorized as nulliparous (parity < 1 previous vaginal birth) or multiparous (parity > 1 previous vaginal birth) for analysis. Gestational age at hospital admission was calculated based on the estimated date of delivery, which had been assigned for all participants at the 8–14-week antepartum visit based on the last menstrual period and/or ultrasound before 14 weeks’ gestation according to standard clinical criteria (“Committee Opinion No. 700: Methods for Estimating the Due Date,” 2017). Maternal body mass index (BMI) was calculated from participants’ height at first prenatal visit and their prepregnant weight (weight before pregnancy or in the first trimester) and again using their weight at hospital admission for labor induction (or at last prenatal visit occurring at 36 weeks’ gestation or later, if hospital admission weight was not available). Cervical-ripening methods, doses, and timing were collected from medication administration records.

At the two hospitals where participants labored, providers, most of whom practiced at both locations, used a shared induction protocol. Cervical-ripening methods were limited to misoprostol 25 μg vaginal, dinoprostone 10 mg vaginal insert, and transcervical catheters. Induction duration was calculated for each woman from the time when the first cervical-ripening method was placed until birth. Additionally, we calculated duration of time for cervical ripening (placement of first cervical-ripening method until 4-cm cervical dilation) and for first-stage labor (cervical dilation increase from 4 to 10 cm per clinician examination). Synthetic oxytocin titrations (mU/min) and doses (mU) were calculated from time-stamped recordings made every 15 min during oxytocin infusion.

Cases of labor induction requiring the highest total doses of synthetic oxytocin from 4- to 10-cm dilation and comparison subjects with labor inductions requiring the lowest total doses of synthetic oxytocin were selected based upon highest versus lowest tertile of ranked total dose of oxytocin, matching on parity and gestational age at hospital admission. Both parity and gestational age were also included as covariates in multivariate analyses to control for possible confounding caused by these matching factors in a separate analysis (Pearce, 2016). Although matching in case–control studies improves statistical precision, the matching process can make comparison subjects similar to cases for both the matching factor and the exposure (in this case, metabolites in second-trimester serum). Controlling for matching factors corrects this issue.

Independent risk factors for higher oxytocin requirements during labor induction including nulliparity, macrosomia, low Bishop score, maternal obesity, and/or maternal age > 35 years (Batinelli et al., 2018; Ellis et al., 2019) were either used as exclusion criteria, balanced in case/comparison groups, used as covariates in multivariate analyses, or not controlled in analyses if they occurred during labor induction as part of the causal pathway (Supplemental Table 1; Rogers et al., 2018; Snowden et al., 2018).

We summarized all categorical variables describing maternal demographic, pregnancy, and labor characteristics using frequencies and percentages across the total sample and in case/comparison groups. For describing continuous data, we used median and interquartile range for nonnormal distributions and mean and standard deviation for normally distributed variables. For categorical data, we used likelihood ratio tests, while for continuous data we used either Mann–Whitney U test or t test to assess statistical significance of group differences. We set a two-tailed α of .05 to determine statistical significance.

High-Resolution Untargeted Metabolomics

We ran high-resolution, untargeted metabolomics assays on stored serum samples collected from participants at a gestational age of 24–30 weeks. Samples were stored at −80°C until analysis. We analyzed samples, after thawing, using a high-resolution metabolomics platform described in previous publications (Hoffman et al., 2014; Li et al., 2013; Neujahr et al., 2014; Roede et al., 2013; Soltow et al., 2013). We analyzed all samples in triplicate and evaluated reproducibility. For each batch of 40 samples, we ran reference standards at the beginning, middle, and end. To minimize batch effects, we randomized samples. Prior to data analysis, we used ComBat (Johnson et al., 2007) to correct for batch effects. We extracted and aligned raw data files using apLCMS (Yu et al., 2013) and xMSanalyzer (Uppal et al., 2013).

Metabolomics Data Extraction

Metabolites are detected by mass spectrometry in a charged state in the gaseous phase and expressed as mass-to-charge ratio (m/z). We used computational methods designed for high-resolution mass spectrometry to extract and quantify metabolites from the liquid chromatography mass-spectrometry data (LC-MS/MS; Uppal et al., 2013; Yu et al., 2013). The resulting data contained ions defined by mass-to-charge ratio (m/z) and retention time (RT) which we then filtered by detection-quality metrics. We then performed differential analyses using the xmsPANDA package Version 1.0.7.5 (Smyth, 2004; Uppal, 2019).

We ranked metabolites using a static score according to how likely they were to differentiate between cases of high-dose oxytocin inductions versus low-dose oxytocin inductions. We filtered metabolites that were missing in more than 80% of samples overall or 80% of samples in one group then log2 transformed and quantile-normalized all metabolites. To select metabolites differentiating cases and comparison samples, we used linear regression with a p value < .05. We used the Benjamini–Hochberg false discovery rate (FDR) method for multiple-hypothesis correction with a threshold q value of .20 to evaluate individual metabolites (Hochberg & Benjamini, 1990). In addition, we evaluated metabolic pathway differences and visualization of metabolome-wide associations with inductions requiring high-dose oxytocin using raw p values.

We repeated analyses after adjusting for parity and gestational age at induction onset. We visualized metabolome-wide association using Type 1 (−log10p vs. m/z) and Type 2 (−log10p vs. RT) Manhattan plots with respect to molecular mass and chemical properties. To visualize the relationship of metabolites differentiating women with and without high-oxytocin dosing during induction, we used two-way hierarchical clustering analysis (HCA). To visualize cases and comparison subjects according to differentiating metabolites, we also used principle component analysis (PCA; Meng et al., 2016).

Metabolic Pathway Analysis

We used Mummichog Version 2.0 to conduct metabolic pathway analysis with the top-ranked (p < .05) metabolites (Li et al., 2013). Metabolic pathway analysis provides a two-step approach in metabolomics investigations to protect against both type I and type II errors (Uppal et al., 2016). Significant metabolic patterns, even those that were not identified in conventional pathways or single-metabolite analyses, can be identified using Mummichog as clustered metabolic reactions. Therefore, even if no particular metabolite was dramatically different by group (i.e., met the FDR threshold), we are still able to observe differences in the action of metabolic networks if multiple metabolites from the same functional pathway were different by subject group. We identified distinct metabolic profiles or networks that were associated with cases of high-oxytocin inductions at a raw p < .05 (Desert et al., 2015; Serkova et al., 2006; Xie et al., 2012; Zhou et al., 2013). We conducted metabolic pathway enrichment analyses using significant metabolites from univariate analysis and then again with metabolites that were significant after adjustment for parity and gestational age. Only metabolic pathways with a minimum overlap size of 4 (i.e., number of metabolites with significant discrimination between case and comparison samples within a particular metabolic pathway) were evaluated for further analyses.

Metabolite Annotation

We used xMSannotator Version 1.3 to annotate metabolites identified as part of metabolic pathways that significantly differentiated case and comparison samples in Mummichog (Uppal et al., 2017). xMSannotator uses a multistage clustering algorithm to assign confidence scores to identify the chemical name of metabolites. Confidence scores for metabolite identification are based on the Schymanski criteria (Schymanski et al., 2014). Research has shown that xMSannotator correctly identifies 80% of features with high or medium confidence scores in human metabolomics studies (Uppal et al., 2017). Only metabolites annotated by xMSannotator as high-confidence matches are reported in this manuscript.

Results

We completed medical-record abstractions for 287 participants of the parent study meeting inclusion criteria (term labor, singleton nonanomalous fetus in vertex position). Interrater reliability testing to confirm chart-abstraction accuracy showed greater than 98.0% accuracy on a random sample of 6.9% of the abstracted charts, thus meeting medical-abstraction accuracy standards of >95% (Mi et al., 2013; Yawn & Wollan, 2005). After we excluded women who did not have a labor induction or who were diabetic or hypertensive, 67 labors remained. Of these, 48 labors involved use of at least one cervical-ripening agent. We ranked labors according to the total dose of synthetic oxytocin women received during labor from 4- to 10-cm cervical dilation. Focusing on the highest versus the lowest tertile total oxytocin dose during first-stage labor, we selected 13 cases of labor inductions with cervical ripening involving high doses of synthetic oxytocin and 14 comparison women with the lowest total doses of synthetic oxytocin, matching on parity and gestational age at labor-induction onset.

Labor Induction

Cases and comparison women were similar on maternal demographic and pregnancy/labor characteristics and on labor-outcome variables, with the exception of synthetic oxytocin infusion dose/titration and duration of first-stage labor. Total dose of synthetic oxytocin during first-stage labor was higher among cases (mean 8,862.16 mU among cases, SD = 5,545.80 vs. mean 1,305.86 mU among comparison women, SD = 1,186.35, p < .001, Table 1). However, cases and comparison women started their inductions at similar cervical dilation/effacement (0 –1 cm dilation, 30% effacement at the mean in both groups) and had equal likelihood of needing an additional method of cervical ripening (Table 1). In addition, the total duration of labor (induction initiation until birth) and cervical ripening (induction initiation until 4 cm dilation) was not different by group, although cases took longer to move from 4- to 10-cm cervical dilation than comparison women (11.57 hr mean for cases, SD = 4.55 vs. 5.74 hr mean among comparison women, SD = 3.13). Thus, cases and comparison women started their inductions with similar cervical progression and received similar treatment for cervical ripening, but, nevertheless, cases required significantly more time and significantly more synthetic oxytocin to move from 4- to 10-cm cervical dilation. Labor interventions like amniotomy and epidural analgesia can change labor duration, especially between 4- and 10-cm cervical dilation (Gerli et al., 2011; Nozomi & Shigeko, 2012); however, cases and comparison women in this study were equally likely to have these interventions during labor (Table 1).

Table 1.

Maternal, Pregnancy, and Labor Characteristics/Outcomes.

Metabolomics Results

We identified a total of 16,482 metabolites in mass spectral data from LC-MS/MS analysis of serum from participants. After metabolite data extraction and quality filtering, we included 8,896 for further analysis. Of these, 140 differentiated between cases and comparisons at a raw p < .05 in univariate analysis. Separation of cases and comparisons was visible in HCA and PCA visualizations (Figure 1), but no individual metabolite was significant at an FDR q value of < .20. Although FDR adjustment protects against type I error, it may exclude true positives (type II error). Therefore, we also analyzed metabolites with a raw p < .05 using pathway enrichment analysis (Mummichog; Chandler et al., 2016). Based on metabolites identified in both univariate and multivariate analysis, we found that several lipid metabolism pathways differentiated cases and comparison subjects after adjustment for parity and gestational age: linoleate metabolism, fatty acid activation/biosynthesis, fatty acid metabolism, vitamin E metabolism, carnitine shuttle, and glycerophospholipid metabolism (p < .05 with an overlap size ≥ 4 for all pathways, Table 2).

Hierarchical cluster analysis (HCA, two-way), principle components analysis (PCA), and volcano plot showing intensities of the significant metabolites differentially expressed between cases of labor inductions requiring the highest total doses of synthetic oxytocin and controls with labor inductions requiring the lowest total doses of synthetic oxytocin. (A) HCA showing top 140/8,813 metabolites with a raw p < .05. Each row represents a participant, and each column a metabolite. Red hues indicate metabolites with higher intensities, and blue hues represent metabolites with lower intensities. High-dose oxytocin-infusion induction of labor is represented in green and low-dose oxytocin-infusion induction in red across the x-axis. (B) PCA for same samples. Red triangles represent labor-induction controls (low dose) and green triangles represent labor-induction cases (high dose). ANOVA = analysis of variance. (C) Volcano plot showing metabolites upregulated or downregulated in cases versus controls. Red dots indicate metabolites that were downregulated in controls, and blue dots represent metabolites that were upregulated in controls. (D) Mummichog Version 2.0-generated metabolic pathways overrepresented in serum from 51 compounds significantly differentiating cases from controls after adjustment for parity and gestational age. Dotted line indicates p < .05.

Table 2.

Metabolic Pathways Significantly Different in Serum From African American Women With High-Oxytocin Requirements Versus Low-Oxytocin Requirements During Induction of Labor.

Cross-referencing the individual metabolites identified by Mummichog, we identified several high-confidence annotations putatively matching pathway metabolites (Figure 2). As measured in the untargeted LC-MS/MS, serum levels of alpha-linolenic acid (m/z 279.2316, RT 194.6, HMBD 01388), a metabolite of the linoleate pathway, were lower among women who went on to experience high-oxytocin-dose inductions compared to comparison samples (2.92-fold change, p = .045). In the fatty acid activation pathway, there was a high-confidence annotation of (Z)-13-octadecenoic acid (m/z 283.263, RT 41.2, HMDB 41480). As measured in untargeted LC-MS/MS, serum levels of this metabolite were also lower among cases compared to comparison samples (0.36-fold change, p = .023).

Serum levels of metabolites differentiating high-oxytocin induction cases (n = 13) from low-oxytocin induction controls (n = 14). (A) Serum levels of alpha-linolenic acid (m/z 279.2316, RT 194.6, HMBD 01388), a three-carbon polyunsaturated fatty acid metabolite in the linoleate pathway, is decreased in cases of high-oxytocin induction versus low-oxytocin induction controls, log2 change = 2.92, p = .045. (B) Serum levels of (Z)-13-octadeconoic acid (m/z 283.263, RT 41.2, HMDB 41480), a metabolite in the fatty acid activation pathway, is decreased in cases (n = 13) of high-oxytocin induction versus low-oxytocin induction controls (n = 14), log2 change = 0.36, p = .023.

Because many of the significant pathways differentiating cases and comparisons were potentially related to maternal obesity and because obesity in pregnant women has been shown to cause metabolic changes that are distinguishable in a variety of body fluids and tissue samples (Fattuoni et al., 2018; Hellmuth et al., 2017), we repeated metabolite and pathway analyses after adjusting for maternal BMI at the time of hospital admission. However, there was no difference in the identity of significant pathways differentiating cases and comparison subjects in the BMI-adjusted analysis compared to the parity- and gestational-age-adjusted analysis (BMI-adjusted analysis not shown).

Discussion

In the present case–control study, late-pregnancy serum levels of metabolites from the linoleate and fatty acid activation pathways predicted high-oxytocin term labor induction following successful cervical ripening in African American women. In addition, we observed associations between high-oxytocin induction and metabolic pathways controlling fatty acid biosynthesis, vitamin E, carnitine shuttle, and glycerophospholipid levels. Although labor induction is an increasingly common procedure (Declercq et al., 2013), research has demonstrated wide variations in success associated with maternal factors such as elevated BMI (Batinelli et al., 2018; Ellis et al., 2019). However, little is known regarding the mechanisms underlying a difficult labor-induction phenotype.

Interestingly, all of the metabolic pathways meeting a priori overlap and significance criteria for predicting the difficult labor-induction phenotype in the present sample were related to the metabolism of lipids. Research has not traced differences in labor-induction processes to particular maternal biomarkers, but there is evidence suggesting that maternal obesity prolongs labor induction and decreases the effectiveness of both cervical-ripening medications and synthetic oxytocin during induction (Beckwith et al., 2017; Ellis et al., 2019; Hill et al., 2015; Roloff et al., 2015).

In the present study, a metabolite in the linoleate pathway, alpha-linolenic acid, was decreased in high-oxytocin induction cases compared to comparison samples. Alpha-linolenic acid is an omega-3 short-chain polyunsaturated fatty acid (PUFA; Holman, 1958). Along with the omega-6 short-chain PUFA linoleic acid, it is one of the two essential fatty acids in humans. Together, alpha-linolenic acid and linoleic acid are converted into several long-chain PUFAs, including eicosapentaenoic acid, docosahexaenoic acid, leukotriene, and prostaglandin E2 (Office of Dietary Supplements, 2018). Long-chain PUFAs are important for the timing of labor (Baguma-Nibasheka et al., 1999; Olsen et al., 1986). Investigators using a rat model showed that a high-fat, high-cholesterol diet compared to a normal diet during pregnancy decreased both plasma and uterine levels of omega-3 PUFAs, thereby decreasing circulating levels of prostaglandins and expression of myometrial gap junctions and causing decreased uterine contractility coordination and strength (Muir et al., 2016, 2018). Thus, our observation that women with high-oxytocin labor inductions had low levels of metabolites in the linoleate pathway is corroborated by evidence from animal studies and suggests a link between obesity-inducing diets and labor difficulty. We theorize that cases in the present study were more likely to require high doses of synthetic oxytocin from 4- to 10-cm cervical dilation because they had difficulty coordinating uterine contractions (decreased myometrial gap junctions) and increasing the force and frequency of contractions (decreased prostaglandin levels) during labor. In response, their care providers titrated oxytocin infusions higher and higher during labor, attempting to stimulate a regular and productive pattern of uterine contractions. Despite those efforts, cases in our study required nearly twice as much time to achieve full cervical dilation as comparison women (11.57 hr [SD = 4.55] vs. 5.74 hr [SD = 3.13], respectively, p < .001).

We also found evidence of decreased fatty acid activation among cases compared to comparisons in our study. Serum levels of a metabolite from the fatty acid activation pathway, (Z)-13-octadecenoic acid, were relatively lower in more complicated induction cases versus comparison women. Octadecenoic acids, commonly found in olive oil, are long-chain PUFAs (Hastings et al., 2016) that accelerate rates of fatty acid oxidation (Lim et al., 2013), thereby reducing systemic metabolic dysfunction caused by free fatty acids being inappropriately stored in nonadipose tissues (Badoud et al., 2015; Gonçalves-de-Albuquerque et al., 2016; Zhang et al., 2015). In labor, accumulation of lipids in uterine tissue is associated with reductions in myometrial muscle mass, thus theoretically decreasing the ability of the uterus to contract forcefully (Gam et al., 2017). It is therefore possible that the decreased fatty acid activation we observed among women with more complicated inductions caused them to instead store free fatty acids in their uterus as ectopic fat, resulting in decreased myometrial muscle mass. If this scenario occurred prior to labor induction, their uterus may have been less able to contract forcefully in response to natural or synthetic oxytocin, resulting in longer labor duration and higher oxytocin requirements during the active phase of labor. Further investigation is needed to expand upon these preliminary findings of a link between certain forms of PUFAs and labor-induction difficulty.

A strength of this study was the inclusion of only African American women, the group with the highest rates of labor-induction and pregnancy-related complications in the United States (Creanga et al., 2017; Metcalfe et al., 2017; Min et al., 2015; Singh et al., 2018). By focusing our study on women of one race, we avoided the confounding effects of between-race differences that have been observed using high-throughput methodologies like metabolomics (Patel et al., 2013). Results from the present investigation suggest several new research directions in the search for ways to decrease complications of labor induction in this population of women.

A second strength of the present investigation was the careful consideration of clinical and maternal factors that may have confounded the relationship between metabolic predictors and labor induction processes and outcomes. Research has identified multiple maternal factors that are associated with difficult labor induction (Batinelli et al., 2018; Ellis et al., 2019). We conducted a subanalysis in the present study to clarify the influence of maternal BMI on metabolic-pathway results and found no difference in important pathways identified in the analysis adjusted for BMI compared to that adjusted for parity and gestational age. In our choice of exclusion criteria and when selecting cases and comparisons, we focused on balanced groups of low-risk women who started induction with a similar cervical status and were cared for by the same group of providers using common induction protocols. Further, we created comparison groups using total oxytocin doses infused during the same period of time during women’s labor induction (from 4- to 10-cm cervical dilation). We also compared use of labor interventions including cervical ripening, amniotomy, and epidural across groups to be sure that there were no significant differences in the processes of induction.

The present study had several limitations. Untargeted high-resolution LC-MS/MS is highly sensitive for detecting a broad coverage of low- and high-abundance metabolites from endogenous, environmental, and dietary sources that might be relevant to difficult labor-induction physiology. However, this technique provides little information on the precise structural identity of differentiating metabolites. Although we utilized the additional steps of identifying significant metabolic pathways and conducting annotations with stringent criteria, additional studies using targeted assays would be necessary to verify metabolite identities and allow their levels to be compared across samples from different groups of women. This study is also limited by a small sample size. In order to focus on a precisely defined phenotype of labor induction while balancing important maternal factors, we limited this preliminary study to a sample of 27 women. Future investigations utilizing a prospective design with larger samples of women are needed to verify and further these findings.

Conclusion

In summary, in the present metabolomics study, we identified several lipid metabolism pathways activated in the serum of African American women in late pregnancy that predicted more complicated labor induction involving longer durations and higher doses of synthetic oxytocin to complete first-stage labor following successful cervical ripening. Further investigation is necessary to determine whether lower levels of omega-3 short-chain polyunsaturated fatty acid and fatty acid activation metabolites are specifically linked to difficult labor induction in African American women or are important for predicting labor-induction processes in all women. In addition, future studies building upon these preliminary results should involve targeted analyses to identify potential agents for therapeutic interventions to prevent or effectively treat difficult labor induction. For example, in the study mentioned earlier utilizing a rat model, investigators found that changing to a low-fat, low-cholesterol diet at conception improved labor outcomes (Muir et al., 2018). As labor induction and augmentation become increasingly common in the modern care of pregnant women, increased knowledge and tools are needed to bring more women safely to an uncomplicated vaginal birth.

Supplemental Material

SUPPLEMENTARY - Metabolic Pathways Associated With Term Labor Induction Course in African American Women:

SUPPLEMENTARY for Metabolic Pathways Associated With Term Labor Induction Course in African American Women by Nicole S. Carlson, Jennifer K. Frediani, Elizabeth J. Corwin, Anne Dunlop and Dean Jones in Biological Research For Nursing

Acknowledgment

The authors would like to thank Rebecca M. Mitchell, PhD, for her assistance with R programming for data cleaning and analysis.

Footnotes

Author Contributions: Nicole S. Carlson contributed to conception, design, acquisition, and interpretation; drafted the manuscript; critically revised the manuscript; gave final approval; and agreed to be accountable for all aspects of work ensuring integrity and accuracy. Jennifer K. Frediani contributed to conception, design, acquisition, analysis, and interpretation; critically revised the manuscript; gave final approval; and agreed to be accountable for all aspects of work ensuring integrity and accuracy. Elizabeth J. Corwin contributed to conception, design, acquisition, and interpretation; critically revised the manuscript; gave final approval; and agreed to be accountable for all aspects of work ensuring integrity and accuracy. Anne Dunlop contributed to conception, design, acquisition, and interpretation; critically revised the manuscript; gave final approval; and agreed to be accountable for all aspects of work ensuring integrity and accuracy. Dean Jones contributed to conception, design, analysis, and interpretation; critically revised the manuscript; gave final approval; and agreed to be accountable for all aspects of work ensuring integrity and accuracy.

Declaration of Conflicting Interests: The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Drs. Nicole Carlson and Jennifer Frediani were supported by grant number K01NR016984 from the National Institute of Nursing Research during manuscript production. This article was supported by funding from the National Institutes of Health National Institute of Nursing Research (R01NR014800), the Office of the Director (3R01NR014800, UG3OD023318), the National Institute of Environmental Health Sciences (P50ES926071), and the U.S. Environmental Protection Agency (83615301 Mod. 2).

ORCID iDs: Nicole S. Carlson

Elizabeth J. Corwin

Supplemental Material: Supplemental material for this article is available online.

References

• ACOG Practice Bulletin No. 107: Induction of labor. (2009). Obstetrics & Gynecology, 114, 386–397. 10.1097/AOG.0b013e3181b48ef5 [PubMed] [CrossRef] [Google Scholar]
• Badoud F., Perreault M., Zulyniak M. A., Mutch D. M. (2015). Molecular insights into the role of white adipose tissue in metabolically unhealthy normal weight and metabolically healthy obese individuals. FASEB Journal, 29, 748–758. 10.1096/fj.14-263913 [PubMed] [CrossRef] [Google Scholar]
• Baguma-Nibasheka M., Brenna J. T., Nathanielsz P. W. (1999). Delay of preterm delivery in sheep by Omega-3 long-chain polyunsaturates. Biology of Reproduction, 60, 698–701. 10.1095/biolreprod60.3.698 [PubMed] [CrossRef] [Google Scholar]
• Batinelli L., Serafini A., Nante N., Petraglia F., Severi F. M., Messina G. (2018). Induction of labour: Clinical predictive factors for success and failure. Journal of Obstetrics and Gynaecology Research, 38, 352–358. 10.1080/01443615.2017.1361388 [PubMed] [CrossRef] [Google Scholar]
• Beckwith L., Magner K., Kritzer S., Warshak C. R. (2017). Prostaglandin versus mechanical dilation and the effect of maternal obesity on failure to achieve active labor: A cohort study. Journal of Maternal-Fetal and Neonatal Medicine, 30, 1621–1626. 10.1080/14767058.2016.1220523 [PubMed] [CrossRef] [Google Scholar]
• Chandler J. D., Hu X., Ko E.-J., Park S., Lee Y.-T., Orr M., Fernandes J., Uppal K., Kang S. M., Jones D. P., Go Y.-M. (2016). Metabolic pathways of lung inflammation revealed by high-resolution metabolomics (HRM) of H1N1 influenza virus infection in mice. American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, 311, R906–R916. 10.1152/ajpregu.00298.2016 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Committee Opinion No 700: Methods for estimating the due date. (2017). Obstetrics & Gynecology, 129, e150–e154. 10.1097/aog.0000000000002046 [PubMed] [CrossRef]
• Corwin E. J., Hogue C. J., Pearce B., Hill C. C., Read T. D., Mulle J., Dunlop A. L. (2017). Protocol for the Emory University African American vaginal, oral, and gut microbiome in pregnancy cohort study. BMC Pregnancy Childbirth, 17, 161 10.1186/s12884-017-1357-x [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Creanga A. A., Bateman B. T., Kuklina E. V., Callaghan W. M. (2014). Racial and ethnic disparities in severe maternal morbidity: A multistate analysis, 2008–2010. American Journal of Obstetrics & Gynecology, 210, 435.e431–438 10.1016/j.ajog.2013.11.039 [PubMed] [CrossRef] [Google Scholar]
• Creanga A. A., Syverson C., Seed K., Callaghan W. M. (2017). Pregnancy-related mortality in the United States, 2011–2013. Obstetrics & Gynecology, 130, 366–373. 10.1097/aog.0000000000002114 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Cunningham F. G., Dashe J. S., Bloom S. L., Spong C. Y., Hoffman B. L. (2018). Williams obstetrics (25th ed). McGraw Hill Medical. [Google Scholar]
• Declercq E. R., Sakala C., Corry M. P., Applebaum S., Herrlich A. (2013). Listening to Mothers III: Report of national surveys of women’s childbearing experiences conducted October–December 2012 and January–April 2013. http://transform.childbirthconnection.org/wp-content/uploads/2013/06/LTM-III_NMSO.pdf [PMC free article] [PubMed]
• Desert R., Canlet C., Costet N., Cordier S., Bonvallot N. (2015). Impact of maternal obesity on the metabolic profiles of pregnant women and their offspring at birth. Metabolomics, 11, 1896–1907. 10.1007/s11306-015-0836-1 [CrossRef] [Google Scholar]
• Ellis J., Brown C. M., Barger B., Carlson N. S. (2019). Influence of maternal obesity on labor induction: A systematic review and meta-analysis. Journal of Midwifery & Women’s Health, 64, 55–67. 10.1111/jmwh.12935 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Fattuoni C., Mando C., Palmas F., Anelli G. M., Novielli C., Parejo Laudicina E., Savasi V. M., Barberini L., Dessì A., Pintus R., Fanos V., Noto A., Cetin I. (2018). Preliminary metabolomics analysis of placenta in maternal obesity. Placenta, 61, 89–95. 10.1016/j.placenta.2017.11.014 [PubMed] [CrossRef] [Google Scholar]
• Frayn K. N. (2010). Metabolic regulation: A human perspective (3rd ed). Wiley-Blackwell. [Google Scholar]
• Gam C. M. B. F., Larsen L. H., Mortensen O. H., Engelbrechtsen L., Poulsen S. S., Qvortrup K., Mathiesen E. R., Damm P., Quistorff B. (2017). Unchanged mitochondrial phenotype, but accumulation of lipids in the myometrium in obese pregnant women. Journal of Physiology, 595, 7109–7122. 10.1113/JP274838 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Gerli S., Favilli A., Acanfora M. M., Bini V., Giorgini C., Di Renzo G. C. (2011). Effect of epidural analgesia on labor and delivery: A retrospective study. Journal of Maternal-Fetal & Neonatal Medicine, 24, 458–460. 10.3109/14767058.2010.498904 [PubMed] [CrossRef] [Google Scholar]
• Gonçalves-de-Albuquerque C. F., Medeiros-de-Moraes I. M., Oliveira F. M. d. J., Burth P., Bozza P. T., Castro Faria M. V., Silva A. R., Castro-Faria-Neto H. C. d. (2016). Omega-9 oleic acid induces fatty acid oxidation and decreases organ dysfunction and mortality in experimental sepsis. PLoS One, 11, e0153607 10.1371/journal.pone.0153607 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Grobman W. A., Bailit J., Lai Y., Reddy U. M., Wapner R. J., Varner M. W., Thorp J. M., Jr, Leveno K. J., Caritis S. N., Prasad M., Tita A. T. N., Saade G., Sorokin Y., Rouse D. J., Blackwell S. C., Tolosa J. E., Eunice Kennedy Shriver National Institute of Child Health and Human Development Maternal-Fetal Medicine Units Network. (2018). Defining failed induction of labor. American Journal of Obstetrics and Gynecology, 218, 122 e121–122 e128 10.1016/j.ajog.2017.11.556 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Hastings J., Owen G., Dekker A., Ennis M., Kale N., Muthukrishnan V., Turner S., Swainston N., Mendes P., Steinbeck C. (2016). ChEBI in 2016: Improved services and an expanding collection of metabolites. CHEBI:25634—octadecenoic acid. Nucleic Acids Research, 44, D1214–S1219. [PMC free article] [PubMed] [Google Scholar]
• Hellmuth C., Lindsay K. L., Uhl O., Buss C., Wadhwa P. D., Koletzko B., Entringer S. (2017). Association of maternal prepregnancy BMI with metabolomic profile across gestation. International Journal of Obesity, 41, 159–169. 10.1038/ijo.2016.153 [PubMed] [CrossRef] [Google Scholar]
• Hill M., Reed K. L., Cohen W. R. (2015). Oxytocin utilization for labor induction in obese and lean women. Journal of Perinatal Medicine, 43, 703–706. 10.1515/jpm-2014-0134 [PubMed] [CrossRef] [Google Scholar]
• Hochberg Y., Benjamini Y. (1990). More powerful procedures for multiple significance testing. Statistics in Medicine, 9, 811–818. [PubMed] [Google Scholar]
• Hoffman J. M., Soltow Q. A., Li S., Sidik A., Jones D. P., Promislow D. E. (2014). Effects of age, sex, and genotype on high-sensitivity metabolomic profiles in the fruit fly, Drosophila melanogaster. Aging Cell, 13, 596–604. 10.1111/acel.12215 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Holman R. T. (1958). Essential fatty acids. Nutrition Reviews, 16, 33–35. 10.1111/j.1753-4887.1958.tb00660.x [PubMed] [CrossRef] [Google Scholar]
• Johnson W., Li C., Rabinovic A. (2007). Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics (Oxford, England), 8, 118–127. [PubMed] [Google Scholar]
• Li S., Park Y., Duraisingham S., Strobel F. H., Khan N., Soltow Q. A., Jones D. P., Pulendran B. (2013). Predicting network activity from high throughput metabolomics. PLoS Computational Biology, 9, e1003123 10.1371/journal.pcbi.1003123 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Lim J. H., Gerhart-Hines Z., Dominy J. E., Lee Y., Kim S., Tabata M., Xiang Y. K., Puigserver P. (2013). Oleic acid stimulates complete oxidation of fatty acids through protein kinase A-dependent activation of SIRT1-PGC1alpha complex. Journal of Biological Chemistry, 288, 7117–7126. 10.1074/jbc.M112.415729 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Main E. (2018). California maternal quality care collaborative comments on the ARRIVE trial [Press release]. https://www.cmqcc.org/sites/default/files/Arrive%20Trial%20Statement%20Final.pdf
• Martin J. A., Hamilton B. E., Osterman M. J., Driscoll A. K., Drake M. S. (2018). Births: Final data for 2017. https://www.cdc.gov/nchs/data/nvsr/nvsr67/nvsr67_08-508.pdf
• Meng C., Zeleznik O. A., Thallinger G. G., Kuster B., Gholami A. M., Culhane A. C. (2016). Dimension reduction techniques for the integrative analysis of multi-omics data. Briefings in Bioinformatics, 17, 628–641. 10.1093/bib/bbv108 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Metcalfe A., Wick J., Ronksley P. (2017). Racial disparities in comorbidity and severe maternal morbidity/mortality in the United States: An analysis of temporal trends. Acta Obstetrics & Gynecology Scandanavia, 97, 89–96. 10.1111/aogs.13245 [PubMed] [CrossRef] [Google Scholar]
• Mi M. Y., Collins J. E., Lerner V., Losina E., Katz J. N. (2013). Reliability of medical record abstraction by non-physicians for orthopedic research. BMC Musculoskeletal Disorders, 14, 181 10.1186/1471-2474-14-181 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Min C. J., Ehrenthal D. B., Strobino D. M. (2015). Investigating racial differences in risk factors for primary cesarean delivery. American Journal of Obstetrics and Gynecology, 212, 814.e1–814.e14. 10.1016/j.ajog.2015.01.029 [PubMed] [CrossRef] [Google Scholar]
• Moynihan A. T., Hehir M. P., Glavey S. V., Smith T. J., Morrison J. J. (2006). Inhibitory effect of leptin on human uterine contractility in vitro. American Journal of Obstetrics and Gynecology, 195, 504–509. [PubMed] [Google Scholar]
• Muir R., Ballan J., Clifford B., McMullen S., Khan R., Shmygol A., Quenby S., Elmes M. (2016). Modelling maternal obesity: The effects of a chronic high-fat, high-cholesterol diet on uterine expression of contractile-associated proteins and ex vivo contractile activity during labour in the rat. Clinical Science, 130, 183–192. 10.1042/CS20150539 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Muir R., Liu G., Khan R., Shmygol A., Quenby S., Gibson R. A., Muhlhausler B., Elmes M. (2018). Maternal obesity-induced decreases in plasma, hepatic and uterine polyunsaturated fatty acids during labour is reversed through improved nutrition at conception. Scientific Reports, 8, 3389 10.1038/s41598-018-21809-9 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Neujahr D. C., Uppal K., Force S. D., Fernandez F., Lawrence C., Pickens A., Bag R., Lockard C., Kirk A. D., Tran V., Lee K., Jones D. P., Park Y. (2014). Bile acid aspiration associated with lung chemical profile linked to other biomarkers of injury after lung transplantation. American Journal of Transplantation, 14, 841–848. 10.1111/ajt.12631 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Nozomi S., Shigeko H. (2012). Current status of amniotomy in relationship to labor progress and delivery outcomes; A retrospective study. Journal of the Japanese Academy of Midwifery, 26, 72–82. [Google Scholar]
• Office of Dietary Supplements. (2018). Omega-3 fatty acids. Health Information. https://ods.od.nih.gov/factsheets/Omega3FattyAcids-Consumer/
• Olsen S. F., Hansen H. S., Sorensen T. I., Jensen B., Secher N. J., Sommer S., Knudsen L. B. (1986). Intake of marine fat, rich in (n-3)-polyunsaturated fatty acids, may increase birthweight by prolonging gestation. Lancet, 2, 367–369. 10.1016/s0140-6736(86)90055-3 [PubMed] [CrossRef] [Google Scholar]
• Parkington H. C., Stevenson J., Tonta M. A., Paul J., Butler T., Maiti K., Chan E. C., Sheehan P. M., Brennecke S. P., Coleman H. A., Smith R. (2014). Diminished hERG K+ channel activity facilitates strong human labour contractions but is dysregulated in obese women. Nature Communications, 5, 4108 10.1038/ncomms5108 [PubMed] [CrossRef] [Google Scholar]
• Patel M. J., Batch B. C., Svetkey L. P., Bain J. R., Turer C. B., Haynes C., Muehlbauer M. J., Stevens R. D., Newgard C. B., Shah S. H. (2013). Race and sex differences in small-molecule metabolites and metabolic hormones in overweight and obese adults. Omics, 17, 627–635. 10.1089/omi.2013.0031 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Pearce N. (2016). Analysis of matched case-control studies. British Medical Journal, 352, i969 10.1136/bmj.i969 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Penfield C. A., Wing D. A. (2017). Labor induction techniques: Which is the best? Obstetrics and Gynecology Clinics of North America, 44, 567–582. 10.1016/j.ogc.2017.08.011 [PubMed] [CrossRef] [Google Scholar]
• Roede J. R., Uppal K., Park Y., Lee K., Tran V., Walker D., Strobel F. H., Rhodes S. L., Ritz B., Jones D. P. (2013). Serum metabolomics of slow vs. rapid motor progression Parkinson’s disease: A pilot study. PLoS One, 8, e77629 10.1371/journal.pone.0077629 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Rogers A. J. G., Harper L. M., Mari G. (2018). A conceptual framework for the impact of obesity on risk of cesarean delivery. American Journal of Obstetrics and Gynecology, 219, 356–363. 10.1016/j.ajog.2018.06.006 [PubMed] [CrossRef] [Google Scholar]
• Roloff K., Peng S., Sanchez-Ramos L., Valenzuela G. J. (2015). Cumulative oxytocin dose during induction of labor according to maternal body mass index. International Journal of Gynaecology & Obstetrics, 131, 54–58. 10.1016/j.ijgo.2015.04.038 [PubMed] [CrossRef] [Google Scholar]
• Schymanski E. L., Jeon J., Gulde R., Fenner K., Ruff M., Singer H. P., Hollender J. (2014). Identifying small molecules via high resolution mass spectrometry: Communicating confidence. Environmental Science & Technology, 48, 2097–2098. 10.1021/es5002105 [PubMed] [CrossRef] [Google Scholar]
• Serkova N. J., Jackman M., Brown J. L., Liu T., Hirose R., Roberts J. P., Maher J. J., Niemann C. U. (2006). Metabolic profiling of livers and blood from obese Zucker rats. Journal of Hepatology, 44, 956–962. 10.1016/j.jhep.2005.07.009 [PubMed] [CrossRef] [Google Scholar]
• Shmygol A., Noble K., Wray S. (2007). Depletion of membrane cholesterol eliminates the Ca2+-activated component of outward potassium current and decreases membrane capacitance in rat uterine myocytes. Journal of Physiology-London, 581, 445–456. 10.1113/jphysiol.2007.129452 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Singh J., Reddy U. M., Huang C. C., Driggers R. W., Landy H. J., Grantz K. L. (2018). Racial/ethnic differences in labor induction in a contemporary US cohort: A retrospective cohort study. American Journal of Perinatology, 35, 361–368. 10.1055/s-0037-1607285 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Smyth G. K. (2004). Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Statistical Applications in Genetics and Molecular Biology, 3, Article 3 10.2202/1544-6115.1027 [PubMed] [CrossRef] [Google Scholar]
• Snowden J. M., Tilden E. L., Odden M. C. (2018). Formulating and answering high-impact causal questions in physiologic childbirth science: Concepts and assumptions. Journal of Midwifery & Women’s Health, 63, 721–730. 10.1111/jmwh.12868 [PubMed] [CrossRef] [Google Scholar]
• Soltow Q. A., Zeanah E. H., Lira V. A., Criswell D. S. (2013). Cessation of cyclic stretch induces atrophy of C2C12 myotubes. Biochemical and Biophysical Research Communications, 434, 316–321. 10.1016/j.bbrc.2013.03.048 [PubMed] [CrossRef] [Google Scholar]
• Spong C. Y., Berghella V., Wenstrom K. D., Mercer B. M., Saade G. R. (2012). Preventing the first cesarean delivery: Summary of a joint Eunice Kennedy Shriver National Institute of Child Health and Human Development, Society for Maternal-Fetal Medicine, and American College of Obstetricians and Gynecologists Workshop. Obstetrics & Gynecology, 120, 1181–1193. 10.1097/AOG.0b013e3182704880 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• To T., Estrabillo E., Wang C., Cicutto L. (2008). Examining intra-rater and inter-rater response agreement: A medical chart abstraction study of a community-based asthma care program. BMC Medical Research Methodology, 8, 29 10.1186/1471-2288-8-29 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Uppal K. (2019). Predictive and network discovery analysis (xmsPANDA): R package for biomarker discovery and biomarker-driven network analysis for studying disease mechanisms using metabolomics. https://rdrr.io/github/kuppal2/xmsPANDA/
• Uppal K., Soltow Q. A., Strobel F. H., Pittard W. S., Gernert K. M., Yu T., Jones D. P. (2013). xMSanalyzer: Automated pipeline for improved feature detection and downstream analysis of large-scale, non-targeted metabolomics data. BMC Bioinformatics, 14, 15 10.1186/1471-2105-14-15 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Uppal K., Walker D. I., Jones D. P. (2017). xMSannotator: An R package for network-based annotation of high-resolution metabolomics data. Analytical Chemistry, 89, 1063–1067. 10.1021/acs.analchem.6b01214 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Uppal K., Walker D. I., Liu K., Li S., Go Y. M., Jones D. P. (2016). Computational metabolomics: A framework for the million metabolome. Chemical Research in Toxicology, 29, 1956–1975. 10.1021/acs.chemrestox.6b00179 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Wray S. (2015). Insights from physiology into myometrial function and dysfunction. Experimental Physiology, 100, 1468–1476. 10.1113/EP085131 [PubMed] [CrossRef] [Google Scholar]
• Xie B., Waters M. J., Schirra H. J. (2012). Investigating potential mechanisms of obesity by metabolomics. Journal of Biomedicine and Biotechnology, 2012, 805683 10.1155/2012/805683 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Yawn B. P., Wollan P. (2005). Inter-rater reliability: Completing the methods description in medical records review studies. American Journal of Epidemiology, 161, 974–977. 10.1093/aje/kwi122 [PubMed] [CrossRef] [Google Scholar]
• Yu T., Park Y., Li S., Jones D. P. (2013). Hybrid feature detection and information accumulation using high-resolution LC-MS metabolomics data. Journal of Proteome Research, 12, 1419–1427. 10.1021/pr301053d [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Zhang J., Kendrick A., Quenby S., Wray S. (2007). Contractility and calcium signaling of human myometrium are profoundly affected by cholesterol manipulation: Implications for labor? Reproductive Sciences, 14, 456–466. 10.1177/1933719107306229 [PubMed] [CrossRef] [Google Scholar]
• Zhang J., Troendle J., Reddy U. M., Laughon S. K., Branch D. W., Burkman R., Landy H. J., Hibbard J. U., Haberman S., Ramirez M. M., Bailit J. L., Hoffman M. K., Gregory K. D., Gonzalez-Quintero V. H., Kominiarek M., Learman L. A., Hatjis C. G., van Veldhuisen P., & Consortium on Safe Labor. (2010). Contemporary cesarean delivery practice in the United States. American Journal of Obstetrics and Gynecology, 203, 326.e321–326.e310. [PMC free article] [PubMed] [Google Scholar]
• Zhang S., Chen G., Li N., Dai M., Chen C., Wang P., Tang H., Hoopes S. L., Zeldin D. C., Wang D. W., Xu X. (2015). CYP2J2 overexpression ameliorates hyperlipidemia via increased fatty acid oxidation mediated by the AMPK pathway. Obesity (Silver Spring), 23, 1401–1413. 10.1002/oby.21115 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Zhou Y., Qiu L., Xiao Q., Wang Y., Meng X., Xu R., Wang S., Na R. (2013). Obesity and diabetes related plasma amino acid alterations. Clinical Biochemistry, 46, 1447–1452. 10.1016/j.clinbiochem.2013.05.045 [PubMed] [CrossRef] [Google Scholar]

Articles from Biological Research for Nursing are provided here courtesy of SAGE Publications