Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Apr 22.
Published in final edited form as: Matern Child Health J. 2025 Oct 24;29(12):1759–1770. doi: 10.1007/s10995-025-04186-4

Mid-trimester allostatic load and spontaneous preterm birth in a cohort of pregnant women living with HIV in Zambia

Katelyn J Rittenhouse 1,§, Bellington Vwalika 2, Yuri V Sebastião 1, Rachel S Resop 1, Humphrey Mwape 3, Kristina De Paris 4, Mwansa K Lubeya 2, Margaret P Kasaro 3, Jeffrey S A Stringer 1, Joan T Price 1
PMCID: PMC13099250  NIHMSID: NIHMS2161161  PMID: 41134424

Introduction

Maternal HIV infection (Wedi et al., 2016) and exposure to antiretroviral therapy (ART) (Brocklehurst & French, 1998) are associated with an elevated risk of preterm birth, the leading cause of neonatal morbidity and mortality worldwide (Chawanpaiboon et al., 2019; L. Liu et al., 2015) and a critical causal factor for disease in later life (Mikkola et al., 2005). In sub-Saharan-Africa, where HIV is endemic, preterm birth rates are among the highest in the world (Joint United Nations Programme on HIV/AIDS, 2014) and capacity to care for premature newborns is often lacking (Lawn et al., 2014; L. Liu et al., 2016). Despite focused research efforts, the causes of preterm birth remain unknown in more than 50% of cases (Kramer et al., 2013; Medley et al., 2018).

Preterm birth is a complex syndrome with several distinct phenotypes (Goldenberg et al., 2008) that can be broadly divided into those arising spontaneously and those initiated by a provider. A growing body of research suggests that adverse perinatal programming by chronic or repeated stress may increase the risk of spontaneous preterm birth and fetal growth restriction (Hobel, 2004; Rich-Edwards & Grizzard, 2005; Zhu et al., 2010). Allostatic load is part of a theoretical framework that attempts to capture multi-system physiological “wear-and-tear,” or describe the detrimental effects of accumulating stress resulting in a degradation or dysregulation of physiological function (McEwen, 1998). By using an allostatic load index representing neuroendocrine, immune, metabolic, and cardiovascular system functioning, numerous studies have demonstrated greater prediction of maternal morbidity and mortality compared to traditional stress metrics used in biomedical practice (Juster et al., 2010). As such, allostatic load has been proposed as a useful concept for quantifying the pathophysiology and prediction of adverse pregnancy outcomes, including spontaneous preterm birth (Olson et al., 2015). Further, allostatic load may prove a useful tool in identification of targets for prevention of spontaneous preterm birth.

Although the relationship between allostatic load and spontaneous preterm birth has previously been reported, this relationship has only been reported in high-income countries (Accortt et al., 2017; Barrett et al., 2018; Hux et al., 2014; McKee et al., 2017; Sayre, 2016; Wallace et al., 2013a, 2013b; Wallace & Harville, 2013). There remains a paucity of data among women with HIV or in African cohorts (Mulligan et al., 2012; Premji, 2014; Rodney & Mulligan, 2014). We performed a nested case-cohort study to examine the association between allostatic load and spontaneous delivery before term among pregnant women with HIV. We hypothesized that mid-trimester allostatic load would be associated with higher risk of spontaneous preterm birth.

Methods

Study Design and Population

We conducted a secondary analysis of data and biological specimens collected from participants in a clinical trial of progesterone for the prevention of preterm birth among 800 pregnant women with HIV in Zambia. The current analysis was restricted to a case-cohort sample of 152 trial participants. Detailed elements of the sampling design are provided below.

IPOP Study

The Improving Pregnancy Outcomes with Progesterone (IPOP) study was a randomized, double-blind, placebo-controlled study conducted in Lusaka, Zambia to evaluate whether intramuscular 17 alpha-hydroxyprogesterone caproate (17P) would reduce the risk of preterm birth among women with HIV. Detailed procedures and primary analysis results have been previously published. Briefly, from February 2018 to January 2020, 800 eligible participants were enrolled and randomized (1:1) to receive 17P or placebo weekly injections, starting between 16-24 weeks and continuing until delivery or 37 weeks. Women aged 18 years or older with a viable intrauterine singleton gestation at less than 24 weeks, confirmed HIV-1 infection, and receiving or intending to initiate ART in pregnancy were eligible for inclusion. Women with a history of spontaneous preterm birth, planned or in situ cervical cerclage, known uterine or major fetal anomaly, indication for planned preterm delivery, or evidence of preterm labor or rupture of membranes at the time of screening were excluded. Antenatal care was provided at a study-run clinic. Written informed consent for collection of clinical data and biological specimens and related analyses was obtained prior to study enrollment. Ultrasounds were performed at screening to assign gestational age and again at 16-24 weeks to measure transvaginal cervical length. At delivery, parturition was classified as either spontaneous (e.g. cervical dilation, contractions) or provider-initiated (no signs of spontaneous labor prior to initiation of induction of labor or cesarean) (Villar et al., 2012). Despite high adherence and retention, randomization to 17P had no effect on the primary composite outcome: 36 (9%) of 399 participants assigned to 17P had preterm birth or stillbirth, compared to 36 (9%) of 401 participants assigned to placebo (Price et al., 2019, 2021).

Case-Cohort Sample Size and Selection

The sample size for the case-cohort was determined by the number of IPOP trial participants who experienced the outcome of interest (spontaneous preterm birth before 37 gestational weeks) and practical reasons (e.g., projected cost of analysis of allostatic load markers). First, we observed an overall risk of spontaneous preterm birth among IPOP participants of 6.4% (n=51/800). Conversely, 93.6% of IPOP participants (n=749/800) did not have a spontaneous preterm birth (91.8%, n=734, with a term birth; and 1.9%, n=15, with a provider-initiated preterm birth). Second, we selected a simple random sample of n=107 IPOP participants to ensure a sample containing at least 100 participants without spontaneous preterm birth (100/0.936 = 107). The resulting simple random sample of n=107 participants contained 6 participants with spontaneous preterm birth (the cases), and 101 participants without spontaneous preterm birth (the non-cases; comprising 99 term births and 2 provider-initiated preterm births). Finally, we selected all remaining cases of spontaneous preterm birth that were not previously selected into the random sample (n=45). The final case-cohort sample comprised 152 participants (51 cases with spontaneous preterm birth, and 101 non-cases). To account for the sampling design, all subsequent analyses of spontaneous preterm birth risk associated with exposures of interest were weighted by the inverse probability of selection into the sample. Spontaneous preterm birth cases were given a weight of 1 because all 51 cases in the overall IPOP study population were selected into the sample. Non-cases were weighted by the inverse of the probability of selection among non-cases in the overall IPOP study population (749/101) (Figure 1).

Figure 1.

Figure 1.

Open in a new tab

IPOP case-cohort sample selection

Ethical Approval

The IPOP trial is registered with clinicaltrials.gov (NCT03297216) and has ongoing approval by the University of Zambia Biomedical Research Ethics Committee and the University of North Carolina Institutional Review Board. All participants provided written informed consent prior to study enrollment.

Exposure and Outcome Variables

The outcome of interest for this analysis was spontaneous preterm birth, defined as spontaneous initiation of labor (or spontaneous membrane rupture) with delivery before 37 weeks of gestation. This outcome was operationalized as a dichotomous variable for spontaneous preterm birth (cases) and non-spontaneous preterm birth, such that non-cases included both term births (n=99) and provider-initiated preterm births (n=2).

The primary exposure of interest was allostatic load index, a composite of multiple individual markers. We selected 15 markers from four domains – cardiovascular, immune, metabolic, and neuroendocrine – based on their inclusion in previous allostatic load studies and their availability in the IPOP trial data set. The following markers were included: systolic blood pressure (mmHg), diastolic blood pressure (mmHg), pulse (bpm), body mass index (kg/m2), mid-upper arm circumference (cm), total cholesterol (mg/dL), high-density lipoprotein (mmol/L), triglycerides (mmol/L), hemoglobin A1c (mmol/L), albumin (g/L), C-reactive protein (mg/L), cortisol (nmol/L), progesterone (nmol/L), estradiol (pg/mL), and 25-OH Vitamin D (ng/L). Body mass index and mid-upper arm circumference were measured at the screening visit (before 24 gestational weeks). All other clinical and laboratory data were collected from the 24-week visit, except for two participants who delivered before this mid-trimester visit for whom all allostatic load markers were obtained from their screening visit.

Laboratory Analyses

Blood specimens were collected at 24 weeks of pregnancy, processed immediately, and stored at −80°C according to standard operating procedures (Castillo et al., 2018; Price et al., 2019). Total cholesterol (mg/dL), high-density lipoprotein (HDL, mmol/L), triglycerides (mmol/L) and albumin (g/L) were measured in serum separated from whole blood by clinical chemistry analysis (AGD 2260 Clinical Chemistry Analyzer). C-reactive protein (CRP, mg/L), cortisol (nmol/L), progesterone (nmol/L), estradiol (pg/mL) and 25-OH Vitamin D (ng/L) were measured in serum by immunochemistry analysis (Maglumi Immunoassay Analyzer). Hemoglobin A1c (mmol/L) was measured in stored whole blood (Mispa-I2 analyzer). Quantification of biomarker concentrations was performed at a local contract laboratory (Sanket Diagnostics Ltd., Lusaka, Zambia).

Statistical Analyses

Summary measures and plots for the distribution of each individual marker were estimated in the study sample, by primary outcome (spontaneous preterm birth vs term birth or provider-initiated preterm birth), and baseline covariates. We examined the distribution of each marker in its original measurement scale and distribution of each marker’s Z-score (standardized measures). Z-scores were calculated by standardizing each marker to a mean of zero and a standard deviation of one. For any given marker, the Z-score measures the difference between a participant’s value from the reference mean value, divided by the reference standard deviation. The reference mean and standard deviation were specific to each gestational age (week) at sample collection. A single participant had BMI and MUAC measured at 6 weeks, and another had her blood collection at 26 weeks, for which we used reference values from 7 weeks and 25 weeks, respectively, to calculate the Z-scores. The Z-scores for HDL, albumin, progesterone, and 25-OH Vitamin D were reversed to ensure that high values for all markers reflect greater dysregulation. Composite allostatic load indexes (ALI) were then generated by combining individual marker Z-scores. A 15-variable allostatic load index (ALI-15) was calculated by summing the Z-scores of all 15 markers. Additionally, a subset of seven markers (systolic blood pressure, maternal heart rate, high-density lipoprotein, triglycerides, hemoglobin A1c, albumin, and 25-OH Vitamin D) that met a criterion of difference in Z-scores between cases and non-cases of >0.10 were summed to generate a 7-variable allostatic load index (ALI-7).

Log-binomial regression was used to estimate the risk and risk ratio of spontaneous preterm birth associated with individual allostatic load markers and composite indexes. We first ran a separate model containing each individual marker’s Z-score as a categorical predictor variable defined by Z-score quartiles. We then ran separate models containing each composite index (ALI-15 and ALI-7) as categorical predictor variables defined by Z-score quartiles. We then estimated time-to-event curves and hazard ratios associated with composite allostatic load index quartile groups and employed robust errors to account for the case-cohort design. The individual marker Z-scores were categorized into quartile groups based on the 0-25th, 26-50th, 51-75th, and >75th percentiles of the standard normal distribution, respectively. The composite indexes were categorized into quartiles groups based on the 0-25th, 26-50th, 51-75th, and >75th percentiles in the study sample, respectively. Finally, we generated locally weighted regression (LOESS) plots of the risk of spontaneous preterm birth as a function of composite allostatic load index. Separate models were run, each containing the ALI-15 and ALI-7 as a continuous predictor variable.

As a sensitivity analysis, we used exploratory factor analysis to estimate weighted allostatic load index scores. In this analysis, the composite allostatic load index was obtained by calculating a weighted sum of each individual markers’ Z-score. The Z-score for each individual was weighted by its factor analysis loading, which measured how well they were associated with the underlying construct of allostatic load (Li et al., 2019). Similar to the approach for ALI-15 and ALI-7, we separately estimated a weighted ALI-15 using all of the markers, and a weighted ALI-7 using the same subset of 7 individual markers (systolic blood pressure, maternal heart rate, high-density lipoprotein, triglycerides, hemoglobin A1c, albumin, and 25-OH Vitamin D) that had a difference in Z-scores between cases and non-cases of >0.1. We then repeated all analyses of the association between allostatic load and spontaneous preterm birth using the weighted allostatic load index scores (weighted ALI-15, weighted ALI-7). All analyses were weighted by the inverse probability of selection into the study sample to account for the case-cohort sampling design. All data management, descriptive statistics and analyses were performed using SAS 9.4 (SAS, Inc., Cary, NC).

Results

A total of 152 participants (51 cases of spontaneous preterm birth and 101 non-cases) were included in the case-cohort. Demographic characteristics of the unweighted and weighted samples are summarized in Table 1. Using weighted baseline participant data, median age was 29.0 years (interquartile range, IQR, 24.0-33.0) and 78% were parous. Median gestational age at enrollment was 19.0 weeks (IQR 15-21). Approximately three-quarters (77%) of study participants were on preconceptional ART and viral load was undetectable in 55% of participants (median among those with detectable viral load: 946 copies/mL, IQR 80-16000). Most participants completed primary education (53%), were married or cohabitating (80%), and had electricity in their households (91%).

Table 1.

Demographic characteristics of the unweighted and weighted participant populations

% (n) or
Median (IQR)		 Weighted† % (n)
or Median (IQR)
Participants 100% (152) 100% (800)
Maternal age 29.0 (24.0, 33.0) 29.0 (24.0, 33.0)
Parous 78.9% (120) 78.4% (627)
Prior stillbirth 4.2% (5) 2.8% (18)
Prior preterm birth (N=120) 2.5% (3) 1.5% (9)
Gestational age at enrollment, weeks 19.0 (15.0, 21.0) 19.0 (15.0, 21.0)
Gestational age at follow-up visit 2, weeks (N=150) 24.0 (24.0, 25.0) 24.0 (24.0, 25.0)
Short cervix (<2.5cm) 0.7% (1) 0.1% (1)
Preconceptional ART 75.0% (114) 76.8% (614)
Viral load undetectable 53.9% (82) 55.2% (441)
Viral load, copies/mL (N=70 detectable) 1108 (85.0, 19900) 946.0 (80.0, 16000)
CD4 count, cells/uL 557.5 (367.5, 772.0) 568.0 (370.0, 772.0)
ART regimen (any with protease inhibitor) 1.3% (2) 1.1% (8)
Randomized to 17-Hydroxyprogesterone group 49.3% (75) 49.5% (396)
Syphilis seropositive 15.8% (24) 15.0% (120)
Bacteriuria 14.5% (22) 14.0% (112)
Hemoglobin < 11g/dL 15.1% (23) 16.5% (132)
Did not complete primary 45.0% (68) 47.4% (376)
Neither married nor cohabitating 20.4% (31) 19.9% (159)
Electricity in household 89.5% (136) 90.8% (726)
Flush/pour toilet facility in household 43.4% (66) 42.7% (342)
Electricity used for cooking 32.2% (49) 32.6% (261)
Alcohol use during pregnancy 10.5% (16) 10.0% (80)
Open in a new tab
†

Weighted by the inverse probability of selection into the case-cohort from the overall trial population (N=800)

Allostatic load component descriptive statistics and distribution of Z-scores by spontaneous preterm birth outcome status are presented in Table 2 and Supplemental Figures S1A-E. In general, there were no substantial differences in individual allostatic load marker Z-scores by outcome status. A moderate difference was observed for maternal heart rate whereby participants who experienced spontaneous preterm birth had a mean Z-score that was 27% greater than those who did not (mean Z-score difference: 0.27; 95% CI for the difference: −0.09, 0.64). Six additional markers had absolute differences in Z-scores that ranged from 0.13 to 0.19 (systolic blood pressure, high-density lipoprotein, triglycerides, hemoglobin A1c, albumin, and 25-OH Vitamin D) and were therefore included in the 7-variable composite allostatic load index along with maternal heart rate.

Table 2.

Allostatic load descriptive statistics and distribution of Z-scores by case status (n=152 unweighted; N=800 weighted)

Na Mean (95% CI) Difference (95% CI)
Systolic blood pressure TB/PIPTB 101 0.082 (−0.113, 0.277) 0
SPTB 51 −0.056 (−0.298, 0.187) −0.138 (−0.449, 0.174)
Diastolic blood pressure TB/PIPTB 101 0.024 (−0.149, 0.197) 0
SPTB 51 0.019 (−0.246, 0.284) −0.005 (−0.321, 0.312)
Maternal heart rate TB/PIPTB 101 0.006 (−0.175, 0.186) 0
SPTB 51 0.278 (−0.037, 0.594) 0.272 (−0.091, 0.636)
Body mass index TB/PIPTB 101 −0.024 (−0.203, 0.154) 0
SPTB 51 0.021 (−0.312, 0.353) 0.045 (−0.332, 0.422)
Mid-upper arm circumference TB/PIPTB 101 −0.025 (−0.206, 0.155) 0
SPTB 51 −0.040 (−0.286, 0.205) −0.015 (−0.320, 0.289)
High-density lipoprotein TB/PIPTB 101 −0.042 (−0.231, 0.146) 0
SPTB 51 0.085 (−0.195, 0.366) 0.128 (−0.210, 0.466)
Triglycerides TB/PIPTB 101 −0.045 (−0.232, 0.142) 0
SPTB 51 0.102 (−0.182, 0.386) 0.147 (−0.193, 0.488)
Total cholesterol TB/PIPTB 101 0.041 (−0.144, 0.226) 0
SPTB 51 −0.065 (−0.356, 0.226) −0.106 (−0.451, 0.239)
Hemoglobin A1C TB/PIPTB 101 −0.042 (−0.221, 0.137) 0
SPTB 51 0.091 (−0.213, 0.394) 0.133 (−0.220, 0.485)
Albumin TB/PIPTB 101 −0.063 (−0.264, 0.137) 0
SPTB 51 0.127 (−0.115, 0.369) 0.190 (−0.124, 0.505)
C-reactive protein TB/PIPTB 101 0.023 (−0.159, 0.205) 0
SPTB 51 −0.023 (−0.324, 0.278) −0.047 (−0.398, 0.305)
Cortisol TB/PIPTB 101 0.032 (−0.168, 0.233) 0
SPTB 51 −0.053 (−0.300, 0.194) −0.085 (−0.404, 0.233)
Progesterone TB/PIPTB 101 −0.018 (−0.217, 0.180) 0
SPTB 51 0.033 (−0.221, 0.286) 0.051 (−0.271, 0.373)
Estradiol TB/PIPTB 101 0.022 (−0.158, 0.202) 0
SPTB 51 −0.021 (−0.328, 0.285) −0.043 (−0.399, 0.312)
25-hydroxyvitamin D TB/PIPTB 101 −0.043 (−0.247, 0.161) 0
SPTB 51 0.086 (−0.148, 0.320) 0.129 (−0.182, 0.439)
Open in a new tab

SPTB: spontaneous preterm birth; TB/PIPTB: term birth/provider-initiated preterm birth

Means, differences, and 95% CIs estimated from marginal linear models with robust standard errors to account for sampling weights.

The Z-scores for HDL, albumin, progesterone, and 25-OH Vitamin D were reversed to ensure that high values for all markers reflect greater dysregulation.

Unadjusted risks and risk ratios of spontaneous preterm birth by individual and composite allostatic load Z-scores are presented in Figures 2 (risk) and Supplemental Figure S2 (risk ratios), respectively. The highest quartile groups of ALI-15 and ALI-7 were found to have the highest risk of preterm birth at 8% (95% CI 4-15) and 11% (95% CI 6-19), respectively, compared to a range of 5-7% among the other quartile groups (Figure 2; Supplemental Table S2). Compared to participants in the second quartile group, participants in the highest quartile groups of ALI-15 and ALI-7 quartile groups had a risk ratio of spontaneous preterm birth of 1.2 (95% CI 0.5-2.9) and 2.4 (95% CI 1.0-5.8), respectively (Supplemental Figure S3 and Table S2). In locally weighted regression plots, the risk of spontaneous preterm birth increased with both ALI-15 and ALI-7 but the association was stronger (i.e., steeper curve) for ALI-7 (Supplemental Figure S3). In time-to-event analyses (Figure 3), the hazard ratio of spontaneous preterm birth for participants in the highest quartile groups of ALI-15 and ALI-7 was 1.2 (95% CI: 0.6-3.0) and 2.5 (95% CI 1.2-5.4), respectively. When analyses were repeated using the weighted ALI-15 and ALI-7 there were no clear trends for an association with spontaneous preterm birth (Figures 2, S3, S4, S5).

Figure 2.

Figure 2.

Open in a new tab

Unadjusted risk of spontaneous preterm birth associated with each allostatic load marker Z-score or composite index group

Cut-offs for individual marker Z-score groups correspond to quartiles of the standard normal distribution; cut-offs for composite allostatic load indexes (ALI) correspond to quartiles in the study sample; ALI-15: sum of all 15 Z-scores; ALI-7: sum of 7 Z-scores (SBP, MHR, HDL-i, Triglycerides, HbA1c, Albumin-i, and 25(OH)D-i). The Z-scores for HDL, albumin, progesterone, and 25-OH Vitamin D were reversed to ensure that high values for all markers reflect greater dysregulation.

Figure 3.

Figure 3.

Open in a new tab

Risk (proportion) of spontaneous preterm birth over time by allostatic load index scores.

Tick marks on each curve represent censored observations. Vertical line at 91 days indicates the time by which all participants in the sample would have reached term gestation. Participants who did not experience outcome of preterm birth were censored at time of delivery or stillbirth.

Allostatic load indexes: ALI-15 is the sum of all 15 individual marker Z-scores, ALI-7 is the sum of 7 Z-scores (SBP, MHR, HDL-i, Triglycerides, HbA1c, Albumin-i, and 25(OH)D-i); ALI group cut-offs correspond to quartiles in the study sample.

Supplemental Tables S1 and S2 provide summaries of the distribution of allostatic load indexes by participant characteristics, and risk estimates of spontaneous preterm birth by participant characteristics, respectively. Except for prior stillbirth (RR: 2.7; 95% CI 0.6-12.5) and prior preterm birth (RR: 3.4; 0.5-23.0), there were no substantial differences in risk of spontaneous preterm birth by any other study covariates (Supplemental Table S2). Given the small number of participants with previous stillbirth (n=5) or previous preterm birth (n=3) we did not adjust for potential confounding by these factors.

Discussion

In this nested case-cohort analysis, we found that a high allostatic load index was associated with spontaneous preterm birth among pregnant women with HIV in Zambia. Our analysis is novel in that we focused on allostatic load in pregnancy among HIV-infected women in sub-Saharan Africa. Early identification of women at the highest risk of preterm birth using mid-trimester allostatic load may provide the opportunity for preventative intervention, which could have substantial public health benefit.

Despite a well-established relationship between maternal stress and preterm birth (Hobel et al., 2008; Simmons et al., 2010; The Community Child Health Network et al., 2015; Wadhwa et al., 2011; Wallace & Harville, 2013), data on the relationship between antenatal allostatic load and preterm birth to date are limited and mixed (Barrett et al., 2018; McKee et al., 2017; Premji et al., 2022; Sayre, 2016; Wallace et al., 2013; Wallace & Harville, 2013). This is likely due to substantial heterogeneity in biomarkers evaluated, variable measurement timing during the antenatal period, and methods used to derive the allostatic load index. For instance, we found 2 studies that previously evaluated allostatic load in the second trimester (McKee et al., 2017; Wallace & Harville, 2013), whereas others have evaluated allostatic load preconceptionally (Barrett et al., 2018; Wallace et al., 2013) and in the third trimester (Sayre, 2016). Additionally, a wide range of algorithms have been used to generate allostatic load indexes, including but not limited to, approaches that are count-based (Lindfors et al., 2006; Maloney et al., 2009; Maselko et al., 2007), utilize standard deviation cut-offs (Goodman et al., 2005) and Z-statistic weights (Karlamangla et al., 2006), factor-analytic approaches (Wiley et al., 2016), and item response theory (S. H. Liu et al., 2021). There remains a lack of consensus on the most appropriate allostatic load formulation to use (Juster et al., 2011; Seplaki et al., 2005). These factors limit direct comparison to other studies. A further complicating factor in comparing our study to prior studies is that studies to date evaluating the relationship between allostatic load and preterm birth have been conducted in populations of women who are primarily white, have postsecondary education, and receive prenatal care in urban settings in the United States.

In our study, we find an association between allostatic load and preterm birth in our restricted index (ALI-7), but not in the comprehensive composite index including all allostatic load components evaluated (ALI-15). Unfortunately, comparison of our findings with other studies is limited by the scarcity of previous investigations of allostatic load in African and HIV-infected populations. Physiologic “wear and tear” may have unique expression in different populations. For instance, in a 2020 study by McGowan et al (McGowan & Norris, 2020) evaluating the association between allostatic load and birth weight, linear growth, and weight gain in South African children and adolescents, levels of individual allostatic load biomarkers were found to differ by sex. A recent study of American young adults found that the allostatic load score among Black women was significantly higher than in both white women and Black men, adjusting for multiple covariates including the level of education attained, a difference that the authors attribute primarily to social factors differentially affecting young Black women in the U.S. (Richardson et al., 2021).

Moreover, a caveat of using allostatic load for preterm birth risk stratification is that interventions to improve the measure are limited. Interventions that have demonstrated a significant reduction in allostatic load include osteopathic manipulation treatment (Nuño et al., 2019), second-generation anti-psychotics (Berger et al., 2018), sleep seminars, Tai Chi Chih, and cognitive behavioral therapy (Carroll et al., 2015), and self-expression programs with guided mentor support (Ye et al., 2017) (reviewed in Rosemberg et al., 2020)). Other interventions, such as dietary (Soltani et al., 2018) and government assistance (McClain et al., 2018) programs, have been attempted but were unsuccessful in decreasing allostatic load. Whether similar interventions to those shown to be beneficial would be feasible in resource-constrained settings among pregnant women living with HIV would need to be examined in future studies.

A strength of our study is its performance in a population of African, HIV-infected women, a group that is largely absent from the allostatic load literature to date. The only study we found evaluating allostatic load in an African population quantified its association with early-life growth in young, urban South African adults (McGowan & Norris, 2020). The only study to date in a HIV-infected population was performed in men and non-pregnant women evaluating the relationship between adverse childhood experiences and allostatic load in the United States. Additional research into physiologic stress in African and HIV-infected populations is needed.

A further strength of our study is our random sampling design, representative of the larger IPOP study population, and use of inverse probability weighting in all analyses. Additionally, we evaluated 15 allostatic load components from multiple physiologic domains in our analysis, more than the 5 to 10 included in prior allostatic load studies in antenatal cohorts (Accortt et al., 2017; Barrett et al., 2018; Hux et al., 2014; McKee et al., 2017; Sayre, 2016; Wallace et al., 2013a, 2013b; Wallace & Harville, 2013). Further, as there is evidence of dampening of biological and psychological adaptations to stress in the third trimester, which serves to protect the pregnant individual and fetus from adverse health implications (Glynn et al., 2008; Premji, 2014), we chose to assess allostatic load in the second trimester.

A limitation of our analysis is the exclusion of certain women at high risk of preterm birth from the parent IPOP trial – and consequently from our analysis – including those with a history of spontaneous preterm birth and planned or in situ cervical cerclage (Price et al., 2019, 2021). The IPOP study also had a lower than expected preterm birth rate as compared to both contemporaneous Zambian cohorts (Price, Phiri, et al., 2020; Price, Vwalika, et al., 2020) and regional estimates (Chawanpaiboon et al., 2019). The exclusion criteria and lower than expected preterm birth risk may limit the generalizability of our findings. In addition, although our primary outcome was spontaneous preterm birth, we included two provider-initiated preterm births in our non-case group. These cases represented less than 2% of the comparator group and were retained to preserve the integrity of the random sampling design and maximize use of available data. While we recognize that provider-initiated and spontaneous preterm births may arise through distinct pathways, we believe the inclusion of these cases had minimal impact on our estimates. Nonetheless, future studies with larger sample sizes may benefit from excluding or separately analyzing provider-initiated preterm births to enhance etiologic specificity. Further, due to budget restrictions, we only examined a single pregnancy timepoint and 15 markers. Examining rate changes in allostatic load during pregnancy may capture dysregulation more comprehensively than a single timepoint measurement (Doan, 2021). Future research should evaluate the change in allostatic load over the course of pregnancy (Li et al., 2020). Due to the retrospective analysis of stored specimens, we were unable to ensure early morning blood collection, an important limitation for inclusion of cortisol given its diurnal fluctuations. Lastly, while this analysis focuses on the association between allostatic load and preterm birth, we recognize a continuum of interrelated disease processes (e.g. hypertensive disorders, diabetes) associated with cumulative life stress that contributes to a common pathway of adverse outcomes (e.g. fetal growth restriction, stillbirth, neonatal demise).

While our study focuses on pregnant women living with HIV in Lusaka, Zambia, it is important to contextualize this sample within the broader Zambian population. According to the 2018 Zambia Demographic and Health Survey (Zambia Statistics Agency, Ministry of Health Zambia, and ICF, 2019), the national HIV prevalence among women aged 15-49 was 14%, with Lusaka Province exhibiting a higher prevalence of 19%. Additionally, urban areas like Lusaka have greater access to healthcare services, including antenatal care, compared to rural regions. They also have improved access to clean drinking water (92% in urban areas compared to 58% in rural areas). These factors may influence both the health status and healthcare utilization patterns of our study population. Therefore, while our findings provide valuable insights into an understudied group, caution should be exercised when generalizing results to the entire Zambian population.

In conclusion, this study finds that a mid-trimester allostatic load index is associated with spontaneous preterm birth in Zambian pregnant women living with HIV. This finding may at least partially explain the elevated risk of adverse pregnancy outcomes in this high-risk population. However, a more nuanced understanding of the maternal stress response in pregnancy is needed to better identify optimal operationalization and timing of allostatic load biomarker collection to more accurately reflect the relevant window of exposures and the biological pathways that result in birth disparities. Populations at highest risk of adverse pregnancy outcomes should be prioritized.

Supplementary Material

Supplemental Figure 1b

Supplemental Figure S1B. Distribution of allostatic load components (left) and their standardized values (Z-scores, right) by case status: body mass index (BMI), mid-upper arm circumference (MUAC), high-density lipoprotein (HDL)

SPTB: spontaneous preterm birth; TB/PIPTB: term birth/provider-initiated preterm birth

Supplemental Figure 1c

Supplemental Figure S1C. Distribution of allostatic load components (left) and their standardized values (Z-scores, right) by case status: triglycerides, total cholesterol, hemoglobin A1C (HbA1c)

SPTB: spontaneous preterm birth; TB/PIPTB: term birth/provider-initiated preterm birth

Supplemental Figure 1a

Supplemental Figure S1A. Distribution of allostatic load components (left) and their standardized values (Z-scores, right) by case status: systolic blood pressure (SBP), diastolic blood pressure (DBP), and maternal heart rate (MHR).

SPTB: spontaneous preterm birth; TB/PIPTB: term birth/provider-initiated preterm birth

Supplemental Figure 2

Supplemental Figure S2. Unadjusted risk ratio of spontaneous preterm birth associated with each allostatic load marker Z-score or composite index group

Cut-offs for individual marker Z-score groups correspond to quartiles of the standard normal distribution; cut-offs for composite allostatic load indexes (ALI) correspond to quartiles in the study sample; ALI-15: sum of all 15 Z-scores; ALI-7: sum of 7 Z-scores (SBP, MHR, HDL-i, Triglycerides, HbA1c, Albumin-i, and 25(OH)D-i). The Z-scores for HDL, albumin, progesterone, and 25-OH Vitamin D were reversed to ensure that high values for all markers reflect greater dysregulation.

Supplemental Figure 1d

Supplemental Figure S1D. Distribution of allostatic load components (left) and their standardized values (Z-scores, right) by case status: albumin, C-reactive protein (CRP), and cortisol

SPTB: spontaneous preterm birth; TB/PIPTB: term birth/provider-initiated preterm birth

Supplemental Figure 1e

Supplemental Figure S1E. Distribution of allostatic load components (left) and their standardized values (Z-scores, right) by case status: progesterone, estradiol, and 25-hydroxyvitamin D (25(OH)D)

SPTB: spontaneous preterm birth; TB/PIPTB: term birth/provider-initiated preterm birth

Supplemental Figure 3

Supplemental Figure S3. Locally weighted regression plot of the risk of spontaneous preterm birth as a function of allostatic load index scores.

Horizontal line indicates the overall risk of spontaneous preterm birth in the case-cohort (0.064; 6.4%). The individual circles indicated represent individual participants who experience either spontaneous preterm birth (top) or term/provider-initiated spontaneous preterm birth (bottom).

Allostatic load indexes: ALI-15 is the sum of all 15 individual marker Z-scores, ALI-7 is the sum of 7 Z-scores (SBP, MHR, HDL-i, Triglycerides, HbA1c, Albumin-i, and 25(OH)D-i).

Supplemental Figure 4

Supplemental Figure S4. Locally weighted regression plot of the risk of spontaneous preterm birth as a function of allostatic load weighted index scores.

Horizontal line indicates the overall risk of spontaneous preterm birth in the case-cohort (0.064; 6.4%). The individual circles indicated represent individual participants who experience either spontaneous preterm birth (top) or term/provider-initiated spontaneous preterm birth (bottom).

Weighted allostatic load indexes: ALI-15 is the weighted sum of all 15 individual marker Z-scores, ALI-7 is the weighted sum of 7 Z-scores (SBP, MHR, HDL-i, Triglycerides, HbA1c, Albumin-I, and 25(OH)D-i)

Supplemental Figure 5

Supplemental Figure S5. Risk (proportion) of spontaneous preterm birth over time by allostatic load weighted index scores.

Tick marks on each curve represent censored observations. Vertical line at 91 days indicates the time by which all participants in the sample would have reached term gestation. Participants who did not experience outcome of preterm birth were censored at time of delivery or stillbirth.

Weighted allostatic load indexes: ALI-15 is the weighted sum of all 15 individual marker Z-scores, ALI-7 is the weighted sum of 7 Z-scores (SBP, MHR, HDL-i, Triglycerides, HbA1c, Albumin-I, and 25(OH)D-i); Group cut-offs correspond to quartiles in the study sample.

10

Conflicts of Interest and Source of Funding:

The authors declare no competing interests. The research presented in this article was funded by the US National Institutes of Health (R01 HD087119 and K01 TW010857) and the Bill and Melinda Gates Foundation (OPP1172799). Additional support was provided by the US National Institutes of Health through the UNC Center for AIDS Research (P30 AI50410). Trainee support has been provided through the US National Institutes of Health for JTP and KJR (T32 HD075731, L30 HD119794, D43 TW009340). The conclusions and opinions expressed in this article are those of the authors and do not necessarily reflect those of the funders. The corresponding author had full access to all data in the study and had final responsibility for the decision to submit for publication.

References

  1. Accortt EE, Mirocha J, Dunkel Schetter C, & Hobel CJ (2017). Adverse Perinatal Outcomes and Postpartum Multi-Systemic Dysregulation: Adding Vitamin D Deficiency to the Allostatic Load Index. Maternal and Child Health Journal, 21(3), 398–406. 10.1007/s10995-016-2226-3 [DOI] [PubMed] [Google Scholar]
  2. Barrett ES, Vitek W, Mbowe O, Thurston SW, Legro RS, Alvero R, Baker V, Bates GW, Casson P, Coutifaris C, Eisenberg E, Hansen K, Krawetz S, Robinson R, Rosen M, Usadi R, Zhang H, Santoro N, & Diamond M (2018). Allostatic load, a measure of chronic physiological stress, is associated with pregnancy outcomes, but not fertility, among women with unexplained infertility. Human Reproduction, 33(9), 1757–1766. 10.1093/humrep/dey261 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Berger M, Juster R-P, Westphal S, Amminger GP, Bogerts B, Schiltz K, Bahn S, Steiner J, & Sarnyai Z (2018). Allostatic load is associated with psychotic symptoms and decreases with antipsychotic treatment in patients with schizophrenia and first-episode psychosis. Psychoneuroendocrinology, 90, 35–42. 10.1016/j.psyneuen.2018.02.001 [DOI] [PubMed] [Google Scholar]
  4. Brocklehurst P, & French R (1998). The association between maternal HIV infection and perinatal outcome: A systematic review of the literature and meta-analysis. British Journal of Obstetrics and Gynaecology, 105(8), 836–848. 10.1111/j.1471-0528.1998.tb10227.x [DOI] [PubMed] [Google Scholar]
  5. Carroll JE, Seeman TE, Olmstead R, Melendez G, Sadakane R, Bootzin R, Nicassio P, & Irwin MR (2015). Improved sleep quality in older adults with insomnia reduces biomarkers of disease risk: Pilot results from a randomized controlled comparative efficacy trial. Psychoneuroendocrinology, 55, 184–192. 10.1016/j.psyneuen.2015.02.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Castillo MC, Fuseini NM, Rittenhouse K, Price JT, Freeman BL, Mwape H, Winston J, Sindano N, Baruch-Gravett C, Chi BH, Kasaro MP, Litch JA, Stringer JSA, & Vwalika B (2018). The Zambian Preterm Birth Prevention Study (ZAPPS): Cohort characteristics at enrollment. Gates Open Research, 2, 25. 10.12688/gatesopenres.12820.2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chawanpaiboon S, Vogel JP, Moller A-B, Lumbiganon P, Petzold M, Hogan D, Landoulsi S, Jampathong N, Kongwattanakul K, Laopaiboon M, Lewis C, Rattanakanokchai S, Teng DN, Thinkhamrop J, Watananirun K, Zhang J, Zhou W, & Gülmezoglu AM (2019). Global, regional, and national estimates of levels of preterm birth in 2014: A systematic review and modelling analysis. The Lancet Global Health, 7(1), e37–e46. 10.1016/S2214-109X(18)30451-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Doan SN (2021). Allostatic load: Developmental and conceptual considerations in a multi-system physiological indicator of chronic stress exposure. Developmental Psychobiology, 63(5), 825–836. 10.1002/dev.22107 [DOI] [PubMed] [Google Scholar]
  9. Glynn LM, Schetter CD, Hobel CJ, & Sandman CA (2008). Pattern of perceived stress and anxiety in pregnancy predicts preterm birth. Health Psychology, 27(1), 43–51. 10.1037/0278-6133.27.1.43 [DOI] [PubMed] [Google Scholar]
  10. Goldenberg RL, Culhane JF, Iams JD, & Romero R (2008). Epidemiology and causes of preterm birth. The Lancet, 371(9606), 75–84. 10.1016/S0140-6736(08)60074-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Goodman E, McEwen BS, Huang B, Dolan LM, & Adler NE (2005). Social Inequalities in Biomarkers of Cardiovascular Risk in Adolescence: Psychosomatic Medicine, 67(1), 9–15. 10.1097/01.psy.0000149254.36133.1a [DOI] [PubMed] [Google Scholar]
  12. Hobel CJ (2004). Stress and preterm birth. Clinical Obstetrics and Gynecology, 47(4), 856–880; discussion 881-882. 10.1097/01.grf.0000142512.38733.8c [DOI] [PubMed] [Google Scholar]
  13. Hobel CJ, Goldstein A, & Barrett ES (2008). Psychosocial stress and pregnancy outcome. Clinical Obstetrics and Gynecology, 51(2), 333–348. 10.1097/GRF.0b013e31816f2709 [DOI] [PubMed] [Google Scholar]
  14. Hux VJ, Catov JM, & Roberts JM (2014). Allostatic load in women with a history of low birth weight infants: The national health and nutrition examination survey. Journal of Women’s Health (2002), 23(12), 1039–1045. 10.1089/jwh.2013.4572 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Joint United Nations Programme on HIV/AIDS. (2014). The gap report. UN Joint Programme on HIV/AIDS (UNAIDS). https://www.unaids.org/sites/default/files/media_asset/UNAIDS_Gap_report_en.pdf [PubMed] [Google Scholar]
  16. Juster R-P, McEwen BS, & Lupien SJ (2010). Allostatic load biomarkers of chronic stress and impact on health and cognition. Neuroscience and Biobehavioral Reviews, 35(1), 2–16. 10.1016/j.neubiorev.2009.10.002 [DOI] [PubMed] [Google Scholar]
  17. Juster R-P, Sindi S, Marin M-F, Perna A, Hashemi A, Pruessner JC, & Lupien SJ (2011). A clinical allostatic load index is associated with burnout symptoms and hypocortisolemic profiles in healthy workers. Psychoneuroendocrinology, 36(6), 797–805. 10.1016/j.psyneuen.2010.11.001 [DOI] [PubMed] [Google Scholar]
  18. Karlamangla AS, Singer BH, & Seeman TE (2006). Reduction in Allostatic Load in Older Adults Is Associated With Lower All-Cause Mortality Risk: MacArthur Studies of Successful Aging. Psychosomatic Medicine, 68(3), 500–507. 10.1097/01.psy.0000221270.93985.82 [DOI] [PubMed] [Google Scholar]
  19. Kramer MS, Lydon J, Goulet L, Kahn S, Dahhou M, Platt RW, Sharma S, Meaney MJ, & Séguin L (2013). Maternal stress/distress, hormonal pathways and spontaneous preterm birth. Paediatric and Perinatal Epidemiology, 27(3), 237–246. 10.1111/ppe.12042 [DOI] [PubMed] [Google Scholar]
  20. Lawn JE, Blencowe H, Oza S, You D, Lee ACC, Waiswa P, Lalli M, Bhutta Z, Barros AJD, Christian P, Mathers C, Cousens SN, & Lancet Every Newborn Study Group. (2014). Every Newborn: Progress, priorities, and potential beyond survival. The Lancet, 384(9938), 189–205. 10.1016/S0140-6736(14)60496-7 [DOI] [PubMed] [Google Scholar]
  21. Li Y, Dalton VK, Lee SJ, Rosemberg M-AS, & Seng JS (2020). Exploring the validity of allostatic load in pregnant women. Midwifery, 82, 102621. 10.1016/j.midw.2019.102621 [DOI] [PubMed] [Google Scholar]
  22. Li Y, Rosemberg M-AS, Dalton VK, Lee SJ, & Seng JS (2019). Exploring the optimal allostatic load scoring method in women of reproductive age. Journal of Advanced Nursing, 75(11), 2548–2558. 10.1111/jan.14014 [DOI] [PubMed] [Google Scholar]
  23. Lindfors P, Lundberg O, & Lundberg U (2006). Allostatic Load and Clinical Risk as Related to Sense of Coherence in Middle-Aged Women. Psychosomatic Medicine, 68(5), 801–807. 10.1097/01.psy.0000232267.56605.22 [DOI] [PubMed] [Google Scholar]
  24. Liu L, Oza S, Hogan D, Chu Y, Perin J, Zhu J, Lawn JE, Cousens S, Mathers C, & Black RE (2016). Global, regional, and national causes of under-5 mortality in 2000-15: An updated systematic analysis with implications for the Sustainable Development Goals. The Lancet, 388(10063), 3027–3035. 10.1016/S0140-6736(16)31593-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Liu L, Oza S, Hogan D, Perin J, Rudan I, Lawn JE, Cousens S, Mathers C, & Black RE (2015). Global, regional, and national causes of child mortality in 2000-13, with projections to inform post-2015 priorities: An updated systematic analysis. The Lancet, 385(9966), 430–440. 10.1016/S0140-6736(14)61698-6 [DOI] [PubMed] [Google Scholar]
  26. Liu SH, Juster R-P, Dams-O’Connor K, & Spicer J (2021). Allostatic load scoring using item response theory. Comprehensive Psychoneuroendocrinology, 5, 100025. 10.1016/j.cpnec.2020.100025 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Maloney EM, Boneva R, Nater UM, & Reeves WC (2009). Chronic Fatigue Syndrome and High Allostatic Load: Results From a Population-Based Case-Control Study in Georgia. Psychosomatic Medicine, 71(5), 549–556. 10.1097/PSY.0b013e3181a4fea8 [DOI] [PubMed] [Google Scholar]
  28. Maselko J, Kubzansky L, Kawachi I, Seeman T, & Berkman L (2007). Religious Service Attendance and Allostatic Load Among High-Functioning Elderly. Psychosomatic Medicine, 69(5), 464–472. 10.1097/PSY.0b013e31806c7c57 [DOI] [PubMed] [Google Scholar]
  29. McClain AC, Xiao RS, Gao X, Tucker KL, Falcon LM, & Mattei J (2018). Food Insecurity and Odds of High Allostatic Load in Puerto Rican Adults: The Role of Participation in the Supplemental Nutrition Assistance Program During 5 Years of Follow-Up. Psychosomatic Medicine, 80(8), 733–741. 10.1097/PSY.0000000000000628 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. McEwen BS (1998). Stress, Adaptation, and Disease: Allostasis and Allostatic Load. Annals of the New York Academy of Sciences, 840(1), 33–44. 10.1111/j.1749-6632.1998.tb09546.x [DOI] [PubMed] [Google Scholar]
  31. McGowan CJ, & Norris SA (2020). Associations of early-life growth with health using an allostatic load score in young, urban African adults: Birth to Twenty Plus Cohort. Journal of Developmental Origins of Health and Disease, 11(4), 360–368. 10.1017/S2040174419000667 [DOI] [PubMed] [Google Scholar]
  32. McKee KS, Seplaki C, Fisher S, Groth SW, & Fernandez ID (2017). Cumulative Physiologic Dysfunction and Pregnancy: Characterization and Association with Birth Outcomes. Maternal and Child Health Journal, 21(1), 147–155. 10.1007/s10995-016-2103-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Medley N, Vogel JP, Care A, & Alfirevic Z (2018). Interventions during pregnancy to prevent preterm birth: An overview of Cochrane systematic reviews. The Cochrane Database of Systematic Reviews, 11, CD012505. 10.1002/14651858.CD012505.pub2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Mikkola K, Ritari N, Tommiska V, Salokorpi T, Lehtonen L, Tammela O, Pääkkönen L, Olsen P, Korkman M, & Fellman V (2005). Neurodevelopmental outcome at 5 years of age of a national cohort of extremely low birth weight infants who were born in 1996-1997. Pediatrics, 116(6), 1391–1400. 10.1542/peds.2005-0171 [DOI] [PubMed] [Google Scholar]
  35. Mulligan C, D’Errico N, Stees J, & Hughes D (2012). Methylation changes at NR3C1 in newborns associate with maternal prenatal stress exposure and newborn birth weight. Epigenetics, 7(8), 853–857. 10.4161/epi.21180 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Nuño V, Siu A, Deol N, & Juster R-P (2019). Osteopathic Manipulative Treatment for Allostatic Load Lowering. Journal of Osteopathic Medicine, 119(10), 646–654. 10.7556/jaoa.2019.112 [DOI] [PubMed] [Google Scholar]
  37. Olson DM, Severson EM, Verstraeten BSE, Ng JWY, McCreary JK, & Metz GAS (2015). Allostatic Load and Preterm Birth. International Journal of Molecular Sciences, 16(12), 29856–29874. 10.3390/ijms161226209 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Premji S (2014). Perinatal distress in women in low- and middle-income countries: Allostatic load as a framework to examine the effect of perinatal distress on preterm birth and infant health. Maternal and Child Health Journal, 18(10), 2393–2407. 10.1007/s10995-014-1479-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Premji SS, Pana GS, Cuncannon A, Ronksley PE, Dosani A, Hayden KA, Lalani S, Musana JW, Shaikh K, Yim IS, & Maternal-infant Global Health Team (MiGHT) Collaborators in Research. (2022). Prenatal allostatic load and preterm birth: A systematic review. Frontiers in Psychology, 13, 1004073. 10.3389/fpsyg.2022.1004073 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Price JT, Phiri WM, Freeman BL, Vwalika B, Winston J, Mabula-Bwalya CM, Mulenga HB, & Stringer JSA (2020). Vaginal progesterone to prevent preterm delivery among HIV-infected pregnant women in Zambia: A feasibility study. PLOS ONE, 15(1), e0224874. 10.1371/journal.pone.0224874 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Price JT, Vwalika B, Freeman BL, Cole SR, Mulenga HB, Winston J, Mbewe FM, Chomba E, Mofenson LM, Rouse DJ, Goldenberg RL, & Stringer JSA (2019). Intramuscular 17-hydroxyprogesterone caproate to prevent preterm birth among HIV-infected women in Zambia: Study protocol of the IPOP randomized trial. BMC Pregnancy and Childbirth, 19(1), 81. 10.1186/s12884-019-2224-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Price JT, Vwalika B, Freeman BL, Cole SR, Saha PT, Mbewe FM, Phiri WM, Peterson M, Muyangwa D, Sindano N, Mwape H, Smithmyer ME, Kasaro MP, Rouse DJ, Goldenberg RL, Chomba E, & Stringer JSA (2021). Weekly 17 alpha-hydroxyprogesterone caproate to prevent preterm birth among women living with HIV: A randomised, double-blind, placebo-controlled trial. The Lancet HIV, 8(10), e605–e613. 10.1016/S2352-3018(21)00150-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Price JT, Vwalika B, Rittenhouse KJ, Mwape H, Winston J, Freeman BL, Sindano N, Stringer EM, Kasaro MP, Chi BH, & Stringer JS (2020). Adverse birth outcomes and their clinical phenotypes in an urban Zambian cohort. Gates Open Research, 3, 1533. 10.12688/gatesopenres.13046.2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Richardson LJ, Goodwin AN, & Hummer RA (2021). Social status differences in allostatic load among young adults in the United States. SSM - Population Health, 15, 100771. 10.1016/j.ssmph.2021.100771 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Rich-Edwards JW, & Grizzard TA (2005). Psychosocial stress and neuroendocrine mechanisms in preterm delivery. American Journal of Obstetrics and Gynecology, 192(5 Suppl), S30–35. 10.1016/j.ajog.2005.01.072 [DOI] [PubMed] [Google Scholar]
  46. Rodney NC, & Mulligan CJ (2014). A biocultural study of the effects of maternal stress on mother and newborn health in the Democratic Republic of Congo: Biocultural Study of Stress and Health in the DRC. American Journal of Physical Anthropology, 155(2), 200–209. 10.1002/ajpa.22568 [DOI] [PubMed] [Google Scholar]
  47. Rosemberg M-AS, Granner J, Li Y, & Seng JS (2020). A scoping review of interventions targeting allostatic load. Stress, 23(5), 519–528. 10.1080/10253890.2020.1784136 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Sayre MEM (2016). Maternal Allostatic Load During Pregnancy: Predicting Length of Gestation [Doctoral dissertation, University of Kentucky; ]. 10.13023/ETD.2016.158 [DOI] [Google Scholar]
  49. Seplaki C, Goldman N, Glei D, & Weinstein M (2005). A comparative analysis of measurement approaches for physiological dysregulation in an older population. Experimental Gerontology, 40(5), 438–449. 10.1016/j.exger.2005.03.002 [DOI] [PubMed] [Google Scholar]
  50. Simmons LE, Rubens CE, Darmstadt GL, & Gravett MG (2010). Preventing Preterm Birth and Neonatal Mortality: Exploring the Epidemiology, Causes, and Interventions. Seminars in Perinatology, 34(6), 408–415. 10.1053/j.semperi.2010.09.005 [DOI] [PubMed] [Google Scholar]
  51. Soltani H, Keim N, & Laugero K (2018). Diet Quality for Sodium and Vegetables Mediate Effects of Whole Food Diets on 8-Week Changes in Stress Load. Nutrients, 10(11), 1606. 10.3390/nu10111606 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. The Community Child Health Network, Ramey SL, Schafer P, DeClerque JL, Lanzi RG, Hobel C, Shalowitz M, Chinchilli V, & Raju TNK, (2015). The Preconception Stress and Resiliency Pathways Model: A Multi-Level Framework on Maternal, Paternal, and Child Health Disparities Derived by Community-Based Participatory Research. Maternal and Child Health Journal, 19(4), 707–719. 10.1007/s10995-014-1581-1 [DOI] [PubMed] [Google Scholar]
  53. Villar J, Papageorghiou AT, Knight HE, Gravett MG, Iams J, Waller SA, Kramer M, Culhane JF, Barros FC, Conde-Agudelo A, Bhutta ZA, & Goldenberg RL (2012). The preterm birth syndrome: A prototype phenotypic classification. American Journal of Obstetrics and Gynecology, 206(2), 119–123. 10.1016/j.ajog.2011.10.866 [DOI] [PubMed] [Google Scholar]
  54. Wadhwa PD, Entringer S, Buss C, & Lu MC (2011). The Contribution of Maternal Stress to Preterm Birth: Issues and Considerations. Clinics in Perinatology, 38(3), 351–384. 10.1016/j.clp.2011.06.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Wallace ME, & Harville EW (2013). Allostatic load and birth outcomes among white and black women in New Orleans. Maternal and Child Health Journal, 17(6), 1025–1029. 10.1007/s10995-012-1083-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Wallace M, Harville E, Theall K, Webber L, Chen W, & Berenson G (2013a). Neighborhood poverty, allostatic load, and birth outcomes in African American and white women: Findings from the Bogalusa Heart Study. Health & Place, 24, 260–266. 10.1016/j.healthplace.2013.10.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Wallace M, Harville E, Theall K, Webber L, Chen W, & Berenson G (2013b). Preconception biomarkers of allostatic load and racial disparities in adverse birth outcomes: The Bogalusa Heart Study. Paediatric and Perinatal Epidemiology, 27(6), 587–597. 10.1111/ppe.12091 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Wedi COO, Kirtley S, Hopewell S, Corrigan R, Kennedy SH, & Hemelaar J (2016). Perinatal outcomes associated with maternal HIV infection: A systematic review and meta-analysis. The Lancet HIV, 3(1), e33–48. 10.1016/S2352-3018(15)00207-6 [DOI] [PubMed] [Google Scholar]
  59. Wiley JF, Gruenewald TL, Karlamangla AS, & Seeman TE (2016). Modeling Multisystem Physiological Dysregulation. Psychosomatic Medicine, 78(3), 290–301. 10.1097/PSY.0000000000000288 [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Ye ZJ, Qiu HZ, Liang MZ, Liu ML, Li PF, Chen P, Sun Z, Yu YL, Wang SN, Zhang Z, Liao KL, Peng CF, Huang H, Hu GY, Zhu YF, Zeng Z, Hu Q, & Zhao JJ (2017). Effect of a mentor-based, supportive-expressive program, Be Resilient to Breast Cancer, on survival in metastatic breast cancer: A randomised, controlled intervention trial. British Journal of Cancer, 117(10), 1486–1494. 10.1038/bjc.2017.325 [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Zambia Statistics Agency, Ministry of Health Zambia, and ICF. (2019). Zambia Demographic and Health Survey 2018. Lusaka, Zambia, and Rockville, Maryland, USA: Zambia Statistics Agency, Ministry of Health, and ICF. [Google Scholar]
  62. Zhu P, Tao F, Hao J, Sun Y, & Jiang X (2010). Prenatal life events stress: Implications for preterm birth and infant birthweight. American Journal of Obstetrics and Gynecology, 203(1), 34.e1–8. 10.1016/j.ajog.2010.02.023 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Figure 1b

Supplemental Figure S1B. Distribution of allostatic load components (left) and their standardized values (Z-scores, right) by case status: body mass index (BMI), mid-upper arm circumference (MUAC), high-density lipoprotein (HDL)

SPTB: spontaneous preterm birth; TB/PIPTB: term birth/provider-initiated preterm birth

NIHMS2161161-supplement-Supplemental_Figure_1b.jpg (413.3KB, jpg)
Supplemental Figure 1c

Supplemental Figure S1C. Distribution of allostatic load components (left) and their standardized values (Z-scores, right) by case status: triglycerides, total cholesterol, hemoglobin A1C (HbA1c)

SPTB: spontaneous preterm birth; TB/PIPTB: term birth/provider-initiated preterm birth

NIHMS2161161-supplement-Supplemental_Figure_1c.jpg (423.6KB, jpg)
Supplemental Figure 1a

Supplemental Figure S1A. Distribution of allostatic load components (left) and their standardized values (Z-scores, right) by case status: systolic blood pressure (SBP), diastolic blood pressure (DBP), and maternal heart rate (MHR).

SPTB: spontaneous preterm birth; TB/PIPTB: term birth/provider-initiated preterm birth

NIHMS2161161-supplement-Supplemental_Figure_1a.jpg (407.1KB, jpg)
Supplemental Figure 2

Supplemental Figure S2. Unadjusted risk ratio of spontaneous preterm birth associated with each allostatic load marker Z-score or composite index group

Cut-offs for individual marker Z-score groups correspond to quartiles of the standard normal distribution; cut-offs for composite allostatic load indexes (ALI) correspond to quartiles in the study sample; ALI-15: sum of all 15 Z-scores; ALI-7: sum of 7 Z-scores (SBP, MHR, HDL-i, Triglycerides, HbA1c, Albumin-i, and 25(OH)D-i). The Z-scores for HDL, albumin, progesterone, and 25-OH Vitamin D were reversed to ensure that high values for all markers reflect greater dysregulation.

NIHMS2161161-supplement-Supplemental_Figure_2.jpg (66.7KB, jpg)
Supplemental Figure 1d

Supplemental Figure S1D. Distribution of allostatic load components (left) and their standardized values (Z-scores, right) by case status: albumin, C-reactive protein (CRP), and cortisol

SPTB: spontaneous preterm birth; TB/PIPTB: term birth/provider-initiated preterm birth

NIHMS2161161-supplement-Supplemental_Figure_1d.jpg (409.6KB, jpg)
Supplemental Figure 1e

Supplemental Figure S1E. Distribution of allostatic load components (left) and their standardized values (Z-scores, right) by case status: progesterone, estradiol, and 25-hydroxyvitamin D (25(OH)D)

SPTB: spontaneous preterm birth; TB/PIPTB: term birth/provider-initiated preterm birth

NIHMS2161161-supplement-Supplemental_Figure_1e.jpg (424.2KB, jpg)
Supplemental Figure 3

Supplemental Figure S3. Locally weighted regression plot of the risk of spontaneous preterm birth as a function of allostatic load index scores.

Horizontal line indicates the overall risk of spontaneous preterm birth in the case-cohort (0.064; 6.4%). The individual circles indicated represent individual participants who experience either spontaneous preterm birth (top) or term/provider-initiated spontaneous preterm birth (bottom).

Allostatic load indexes: ALI-15 is the sum of all 15 individual marker Z-scores, ALI-7 is the sum of 7 Z-scores (SBP, MHR, HDL-i, Triglycerides, HbA1c, Albumin-i, and 25(OH)D-i).

NIHMS2161161-supplement-Supplemental_Figure_3.jpg (187.1KB, jpg)
Supplemental Figure 4

Supplemental Figure S4. Locally weighted regression plot of the risk of spontaneous preterm birth as a function of allostatic load weighted index scores.

Horizontal line indicates the overall risk of spontaneous preterm birth in the case-cohort (0.064; 6.4%). The individual circles indicated represent individual participants who experience either spontaneous preterm birth (top) or term/provider-initiated spontaneous preterm birth (bottom).

Weighted allostatic load indexes: ALI-15 is the weighted sum of all 15 individual marker Z-scores, ALI-7 is the weighted sum of 7 Z-scores (SBP, MHR, HDL-i, Triglycerides, HbA1c, Albumin-I, and 25(OH)D-i)

NIHMS2161161-supplement-Supplemental_Figure_4.jpg (197.6KB, jpg)
Supplemental Figure 5

Supplemental Figure S5. Risk (proportion) of spontaneous preterm birth over time by allostatic load weighted index scores.

Tick marks on each curve represent censored observations. Vertical line at 91 days indicates the time by which all participants in the sample would have reached term gestation. Participants who did not experience outcome of preterm birth were censored at time of delivery or stillbirth.

Weighted allostatic load indexes: ALI-15 is the weighted sum of all 15 individual marker Z-scores, ALI-7 is the weighted sum of 7 Z-scores (SBP, MHR, HDL-i, Triglycerides, HbA1c, Albumin-I, and 25(OH)D-i); Group cut-offs correspond to quartiles in the study sample.

NIHMS2161161-supplement-Supplemental_Figure_5.jpg (267.5KB, jpg)
10
NIHMS2161161-supplement-10.pdf (225.2KB, pdf)

RESOURCES