Maternal Psychosocial Functioning, Obstetric Health History, and Newborn Telomere Length

Abstract

There is growing interest in elucidating the determinants of newborn telomere length, given its potential as a biomarker of lifetime disease risk affected by prenatal exposures. There is limited evidence that increased maternal stress during pregnancy predicts shorter newborn telomere length. However, the few studies published to date have been conducted primarily with small samples utilizing inconsistent definitions of maternal stress. Moreover, the potential influence of fetal sex as a moderator of maternal stress effects on newborn telomere length has been largely ignored despite compelling evidence of likely impact. In a prospective cohort study of pregnant women seeking routine prenatal care, we tested whether a range of maternal measures of stressor exposures, subjective feelings of stress, and mental health (depression, anxiety) were associated with newborn telomere length assessed from cord blood among 146 pregnant women and their newborn infants. We further examined whether the pattern of associations differed by infant sex. Sociodemographic and maternal and newborn health indicators were considered as potential covariates. When examined within the whole sample, none of the maternal psychosocial measures were associated with newborn telomere length. Among potential covariates, maternal history of smoking and preeclampsia in a previous pregnancy were negatively associated with newborn telomere length. In adjusted linear regression analyses that considered potential sex-specific effects, maternal depression, general anxiety, and pregnancy-specific anxiety symptoms were positively associated with newborn telomere length among males. Overall, the findings provide some evidence for an association between maternal psychosocial wellbeing in pregnancy and newborn telomere length in males, although in the opposite direction than previously reported. Maternal smoking and obstetric history prior to conception may be associated with shorter offspring telomere length.

1. Introduction

The measurement of telomeres—the repeating nucleotide sequences of variable number that protect against chromosome deterioration and regulate cellular and tissue function—has gained interest as a potential biomarker of health. Shorter telomere length has been associated with multiple indicators of poor health, including decreased immunocompetence, the development of chronic disease (e.g., cardiovascular disease, diabetes, obesity, inflammatory diseases, depression), abnormalities in brain structure and functioning, and earlier mortality (Baragetti et al., 2015; Factor-Litvak et al., 2016; Geronimus et al., 2015; Hochstrasser, Marksteiner, & Humpel, 2012; Mundstock et al., 2015; Qi Nan, Ling, & Bing, 2015; Rode, Nordestgaard, & Bojesen, 2015). Consequently, researchers have sought to elucidate factors that hasten telomere attrition rate as potential critical determinants of health. Telomere length at birth is hypothesized to establish an individual’s initial telomere length and subsequent attrition rate, thus influencing lifetime telomere biology (Drury et al., 2015; Entringer, de Punder, Buss, & Wadhwa, 2018; Factor-Litvak et al., 2016; Martens, Plusquin, Gyselaers, De Vivo, & Nawrot, 2016). Moreover, the setting of the telomere biology system at birth appears to account for the largest proportion of its attributable effects on health outcomes (Entringer et al., 2018). Therefore, processes that affect newborn telomere biology may be particularly important determinants of lifetime health.

A small but growing literature has begun to explore factors that may influence newborn telomere length. Although research suggests high heritability (Broer et al., 2013), environmental exposures appear to have major effects on telomere length at birth (Entringer et al., 2018; Hjelmborg et al., 2015). Maternal psychological well-being in pregnancy has been identified as one potentially consequential exposure. Pregnancy is a period of increased stress and vulnerability to the development or exacerbation of mental health difficulties (e.g., depression, anxiety) (Van den Bergh et al., 2017). Maternal stress and psychopathology during pregnancy increase offspring risk for numerous mental and physical health problems across the lifespan, as well as aberrations in functional and structural brain development, neurocognitive functioning, and stress reactivity (Van den Bergh et al., 2017). A few studies have specifically linked higher maternal stress and depressive and anxiety symptoms during pregnancy to shorter newborn telomere length (Bosquet Enlow et al., 2018; Entringer et al., 2013; Marchetto et al., 2016; Salihu et al., 2016; Send et al., 2017; Suh et al., 2019). Moreover, a number of biological correlates of stress, including disruptions to hypothalamic-pituitary-adrenal (HPA) axis functioning/cortisol output and increased oxidative stress and inflammation, have been implicated as causal factors in accelerated telomere erosion (Bosquet Enlow et al., 2019; Entringer et al., 2018). Importantly, maternal stress and mental health are modifiable via behavioral interventions in pregnancy. Furthermore, interventions that attenuate stress increase the activity of telomerase, a cellular holoenzyme responsible for the maintenance of telomeric integrity (Entringer et al., 2018).

Despite significant indirect evidence supporting the potential influence of maternal psychosocial functioning on newborn telomere length, direct evidence is relatively limited and inconsistent. The first study to document an association between maternal stress and newborn telomere length showed a negative association between maternal scores on a 4-item self-report questionnaire of pregnancy-specific stress assessed in early pregnancy (M=9.2 weeks gestation) and cord blood peripheral blood mononuclear cells in a small sample (N=27 dyads) (Entringer et al., 2013). Another small study (N=24 dyads) showed a negative association between maternal self-reported exposure to potentially stressful life events over the course of pregnancy (high vs. low on the Holmes and Rahe Stress Scale) and cord blood telomere length (Marchetto et al., 2016). Salihu and colleagues (Salihu et al., 2016) found that cord blood telomere length was negatively associated in a dose-response pattern with maternal self-reported subjective evaluation of stress (low/normal/high) over the last month of pregnancy, assessed via the Perceived Stress Scale administered during hospital admission for delivery (N=71 dyads). In a relatively large study (N=410 dyads), Send and colleagues (Send et al., 2017) also found that cord blood telomere length was negatively associated with maternal scores on the Perceived Stress Scale collected during the 3^rd trimester of pregnancy but not with maternal lifetime history of psychiatric disorder. In another large study (N=656 dyads), Verner and colleagues (Verner et al. 2020) linked shorter cord blood telomere length to maternal stress, operationalized as the mother’s average score from multiple stress assessments across pregnancy, including the Perceived Stress Scale, perceived stress from a visual analogue scale for stress, and items related to negative mood reactivity to pregnancy-related events from the Pregnancy Experience Scale. The study further found that maternal psychological resilience during pregnancy, computed by regressing maternal positive emotional state on maternal stress (i.e., the ability to maintain positivity and satisfactory social relationships in the face of stress), was positively associated with newborn telomere length. Suh and colleagues (Suh et al., 2019) reported an association between high maternal depressive (Center for Epidemiological Studies-Depression) and/or anxiety (State-Trait Anxiety Inventory [STAI]-Trait subscale) symptoms (described as “high prenatal maternal stress,” with categorization based on sample distribution) at 36 weeks gestation and shorter cord blood telomere length (N=68 dyads). Bosquet Enlow and colleagues (Bosquet Enlow et al., 2018) failed to find an association of maternal exposure to negative life events (Crisis in Family Systems-Revised, exposures over prior 6 months, assessed at M=19.4 weeks gestation, SD=8.6 weeks gestation) or posttraumatic stress symptoms (Posttraumatic Stress Disorder Checklist-Civilian Version, symptoms over prior month) in pregnancy with cord blood telomere length but found that elevated maternal depressive symptoms (≥ 13 on Edinburgh Postnatal Depression Scale, over prior week) was associated with shorter telomere length in newborn males but not females (N=151).

As evident above, operationalization of maternal stress has varied widely in newborn telomere research in terms of (a) the underlying construct being assessed (e.g., exposure to negative life events, subjective feelings of general stress, specific concerns regarding pregnancy, psychiatric symptoms), (b) the nature of scoring (e.g., categorical based on sample distribution, categorical based on measure cutoffs/diagnostic criteria, continuous symptom scoring; scores from individual measures versus combined scores across measures), (c) timing of assessment (e.g., early, middle, or late pregnancy, around labor/delivery, averaged over pregnancy); and (d) period of coverage of exposure (e.g., prior week, month, 6 months, lifetime). Moreover, with rare exception (Bosquet Enlow et al., 2018), studies have largely focused on one assessment of maternal stress or mental health in relation to newborn telomere length. Due to the limited number of studies conducted and the methodological differences across studies, definitive statements cannot be made about the effects of maternal psychosocial wellbeing on newborn telomere length. More research is needed to determine the specific facets of maternal psychological functioning that influence newborn telomere biology.

Although data suggest that maternal psychosocial effects on newborn telomere length may be moderated by fetal sex, assessment has been limited. One study examined whether maternal risk and protective factors were differentially associated with newborn telomere length by newborn sex and reported that male, but not female, fetuses appeared susceptible to a range of exposures, including maternal depressive symptoms, socioeconomic status, and health behaviors in pregnancy and trauma in childhood (Bosquet Enlow et al., 2018). These findings are consistent with studies in children and adults that have found sex differences in the effects of various risk factors (e.g., internalizing disorders, parental caregiving quality, family violence exposure) on telomere attrition, with males more susceptible in many, although not all studies (Drury et al., 2015; Enokido et al., 2014; Moller et al., 2009; Shalev et al., 2014; Zalli et al., 2014). These findings also are consistent with studies documenting sex differences in fetal responses to prenatal stress exposures, with males and females showing varying vulnerability and different adaptation strategies, depending on the nature and timing of the exposure and the developmental outcome of interest (Davis, Sandman, Buss, Wing, & Head, 2013; Doyle et al., 2015; Gabory, Attig, & Junien, 2009; Ostlund et al., 2016; Van den Bergh et al., 2017). Some posit that male fetuses are more vulnerable to maternal distress in pregnancy, whereas others that females are more responsive to a range of in utero exposures (Doyle et al., 2015). Notably, over the course of gestation, male and female fetuses experience different patterns of cortisol exposure, which may be influenced by maternal experiences of stress and have implications for telomere biology (Bosquet Enlow et al., 2019; DiPietro, Costigan, Kivlighan, Chen, & Laudenslager, 2011). In fact, one recent study showed sex-specific effects of maternal cortisol levels during the 3^rd trimester of pregnancy on newborn telomere length, with higher levels associated with longer telomere length among females (Bosquet Enlow et al., 2019). Together, these data suggest that maternal psychosocial wellbeing may have sex-specific effects on newborn telomere biology, but more research is needed to confirm this hypothesis and to specify the nature of any sex effects.

The overall objectives of the current study were to test associations of a range of maternal stress and psychopathology measures with newborn telomere length and to explore potential sex-specific associations. A secondary objective was to examine whether any stress effects were maintained after consideration of sociodemographic and health factors that have been associated with newborn telomere length (Drury et al., 2015; Factor-Litvak et al., 2016; Martens et al., 2016). Particular attention was given to obstetric health, given evidence that obstetric conditions may influence newborn telomere biology via the same biological mediators hypothesized to account for maternal stress effects (e.g., oxidative, endocrine, metabolic, immune) (Entringer et al., 2018; Fragkiadaki et al., 2016; Shalev et al., 2014; Xu et al., 2014). We hypothesized that (a) elevated scores on maternal stress and mental health measures predict shorter newborn telomere length; (b) newborn males, compared to females, are more susceptible to maternal stress and mental health effects on telomere length; and (c) the effects of maternal psychosocial functioning on newborn telomere length is independent of sociodemographic and health factors.

2. Methods

2.1. Participants and Procedures

Participants were 146 pregnant women enrolled in the Spontaneous Prematurity and Epigenetics of the Cervix (SPEC) Study, a prospective pregnancy cohort (N=1165) designed to assess the association between epigenetic changes of the cervix and preterm birth. Between 2013 and 2019, pregnant women (M=18.0, SD=4.0, weeks of gestation) receiving routine prenatal care were recruited from a prenatal clinic in an urban hospital in Boston, MA. Inclusion criteria included: 1) ≥18 years of age and 2) <28 weeks of gestation. Beyond these criteria, there were no exclusions for the SPEC Study. SPEC participants completed demographic and psychosocial questionnaires at enrollment or via email, typically within two weeks of enrollment. Partway through recruitment for the SPEC Study, a protocol for cord blood biospecimen collection was incorporated. Inclusion criteria for the current analyses were 1) availability of newborn telomere length data derived from cord blood (see Section 2.2.1.), 2) availability of relevant psychosocial data during pregnancy (see Section 2.2.2.), and 3) singleton gestation pregnancy. Of the 1165 pregnant women enrolled in SPEC, 797 did not provide cord blood specimens, 203 did not complete any psychosocial questionnaires, and 11 completed the psychosocial questionnaires in the postpartum period, leaving 154 participants; of these, 4 cord blood samples were from a multiple gestation pregnancy and 4 cord blood biospecimens failed to produce usable telomere length data, leaving a final analytic sample of 146. There were no differences between SPEC participants who did and did not provide data necessary for the current analyses on sociodemographic or health variables, including maternal age, race/ethnicity, educational attainment, household income, pre-pregnancy BMI, smoking history, or parity, or infant sex, birthweight, or gestational age. Study procedures were approved by the Institutional Review Board at Beth Israel Deaconess Medical Center, and participants provided written informed consent.

2.2. Measures

2.2.1. Newborn telomere length.

Newborn telomere length was assessed from umbilical cord blood leukocyte DNA, a valid index of newborn telomere length (Entringer et al., 2013; Okuda et al., 2002). Cord blood was drawn from the umbilical cord vein at delivery after the clinician collected any blood samples required for clinical reasons, after clamping the cord. Cord blood samples were collected in EDTA-tubes, stored at +4C for <4 hours, and then immediately centrifuged to separate buffy coat from plasma. The buffy coat fraction was then stored at −80C for up to 4 years prior to DNA extraction. There were no other freeze thaw cycles. Genomic DNA was isolated from the leukocytes using the Qiagen DNeasy kit (Venlo, Netherlands). DNA yield and purity were quantified using an Implen NanoPhotometer Pearl (Westlake Village, CA) to ensure that the optical density ratios were within expected ranges. When plating, the amount of DNA was quantified with Thermo Fisher Scientific’s Quant-iT PicoGreen dsDNA Assay Kit (Waltham, MA) to ensure that the amount of DNA in each reaction was equal.

Telomere length was determined using real-time polymerase chain reaction (PCR), which a recent meta-analysis determined is a valid technique for quantifying telomere length (Ridout et al., 2017). Telomere length was measured using the quantitative real-time method described by Cawthon (Cawthon, 2009) and modified by Pavanello and colleagues (Pavanello et al., 2011). This method measures the relative telomere length by determining the ratio of telomeric repeat copy number (T) to a single copy gene (S) copy number (T/S ratio) in a given sample relative to a reference pooled DNA used to generate a standard curve, which is inserted in each PCR run. Details regarding the telomere assaying procedure are provided in the Supplemental Material, Detailed Cord Blood Telomere Length Assaying Procedures. Because PCR provides a relative measure of telomere length, the term “relative telomere length” is used when describing the current analyses.

2.2.2. Maternal stress and psychopathology predictors.

2.2.2.1. Stressor exposures during pregnancy.

Maternal stressor exposures during pregnancy were assessed using the Crisis in Family Systems-Revised (CRISYS-R) survey, which inquires about exposures to negative life events during the prior six months (Berry, Shalowitz, Quinn, & Wolf, 2001). The CRISYS-R is suitable for sociodemographically diverse populations, has good test/retest reliability, and has been utilized in several studies as a measure of prenatal stress (Enlow et al., 2017). The 80-item survey encompasses 11 domains (financial, legal, career, stability in relationships, medical issues pertaining to self, medical issues pertaining to others, safety in the community, safety in the home, housing problems, difficulty with authority, discrimination), with multiple items assessing each domain. Participants rated endorsed items as positive, negative, or neutral events. Research suggests increased vulnerability when exposed to negative events across multiple domains, including for telomere biology (Enlow et al., 2017; Ridout et al., 2017). Thus, the number of domains with one or more negative events endorsed was summed to create a negative life events domain score (possible range 0–11), as done in prior research (Enlow et al., 2017). Higher scores indicate greater exposure to stressors.

2.2.2.2. Psychological stress during pregnancy.

Perception of psychological stress was assessed using the 4-item version of the Perceived Stress Scale (PSS; Cohen, 1995), a validated measure of stress appraisal that assesses the degree to which the respondent felt her life was unpredictable, uncontrollable, and overwhelming relative to her coping resources in the prior month. Each item was scored on a 5-point scale ranging from “never” (0) to “very often” (4), and the scores were summed to obtain a total score (possible range 0–16). The PSS is the most widely used stress appraisal measure, with documented reliability and validity (Pizzagalli, 2007).

Concerns specifically related to the pregnancy were assessed using the 7-item Pregnancy-Related Anxiety Scale (PRAS) (Rini, Dunkel-Schetter, Wadhwa, & Sandman, 1999; Wadhwa, Sandman, Porto, Dunkel-Schetter, & Garite, 1993). The scale captures worries about fetal growth and development, health, and labor and delivery. Each item was scored on a 4-point scale ranging from “not at all” (1) to “very much” (4), and the scores were summed to obtain a total score (possible range 7–28). Pregnancy-related anxiety appears distinct from general anxiety and depression symptoms, with robust associations with maternal and offspring outcomes, including cord blood DNA methylation (Cardenas et al., 2019; Rini et al., 1999; Wadhwa et al., 1993).

2.2.2.3. Psychopathology during and prior to current pregnancy.

Maternal self-report of depressive symptoms during pregnancy were assessed using the Edinburgh Postnatal Depression Scale (EPDS; (Gibson, McKenzie-McHarg, Shakespeare, Price, & Gray, 2009). The EPDS is a 10-item self-report questionnaire designed to measure the presence of depressive symptoms in the past 7 days in women during the perinatal period. Each EPDS item was scored for severity from 0 to 3 and then summed to provide a total score (possible range 0–30). The EPDS has demonstrated high internal consistency and validity for detecting major depression in the perinatal period.

Maternal self-report of general anxiety symptoms during pregnancy was measured using the total score of the 10-item trait scale of the State-Trait Anxiety Inventory (STAI; (Spielberger, Gorsuch, Lushene, Vagg, & Jacobs, 1983). Items inquire about the respondent’s general feelings of worry and anxiety. Each STAI item was scored for severity from 0 to 3 and then summed to provide a total score (possible range 0–30). The STAI has demonstrated high internal consistency, stability, and validity, including in the peripartum period (e.g., (Grant, McMahon, & Austin, 2008 2008).

Maternal diagnosis of depression or anxiety during the current pregnancy was ascertained via review of medical records by a trained medical professional (“depression diagnosis, medical record” and “anxiety diagnosis, medical record”). Maternal diagnosis of depression by a professional prior to the current pregnancy (“diagnosed depression, prior to pregnancy”) was ascertained via maternal report in response to the following question: “Before this pregnancy, did you ever see a health care professional who said that you were depressed (yes/no)?” Women were also asked if they met the major symptom criterion for depression prior to the current pregnancy (“self-reported depression, prior to pregnancy”): “Before this pregnancy, was there ever a period of time when you were feeling depressed or down or when you lost interest in pleasurable activities most of the day, nearly every day, for at least 2 weeks (yes/no)?”

2.2.2.4. Psychological resilience during pregnancy.

Maternal psychological resilience was assessed by the 25-item Connor-Davidson Resilience Scale (CD-RISC; (Connor & Davidson, 2003), which has been used in other pregnancy studies (Grobman et al., 2018; Johnson et al., 2018). The scale inquires about the respondent’s psychological coping resources, available social supports, and feelings of control and ability to manage difficult situations over the past month. Items were scored on a 5-point scale ranging from “not true at all” (0) to “true nearly all the time” (4) and summed to create a total score (possible range 0–100). Higher scores reflect greater levels of psychological resilience.

2.2.3. Covariate ascertainment.

Given the lack of an established set of covariates to be used in telomere research, variables that prior literature suggests may be associated with newborn telomere length were considered as potential confounding variables. Participants completed sociodemographic and health history questionnaires that provided self-reported race/ethnicity, education, annual household income, smoking history, pre-pregnancy body mass index, and obstetric history. We obtained medical data pertaining to the current pregnancy through a combination of querying of the electronic health record and manual chart review. Smoking history was based on self-report as to whether the participant was currently smoking (yes/no) or had smoked 50 or more cigarettes in her lifetime prior to the current pregnancy (yes/no). Pre-pregnancy BMI was calculated by dividing self-reported pre-pregnancy weight (kg) by height squared (meter). Because maternal stress and psychopathology questionnaires were not completed at the same time point in pregnancy across participants, timing of questionnaire completion was considered as a potential covariate; this variable was coded as 1^st trimester, 2^nd trimester, or 3^rd trimester.

2.3. Data Analysis Plan

A post hoc effect size calculation for our sample size of 146 determined that we would be able to detect correlations as small as 0.23 with 80% power (null hypothesis r=0.00 and p=.05). Linear regression analyses with full information maximum likelihood estimation to account for missing data tested whether maternal stressor exposures, psychological stress, psychopathology, or psychological resilience predicted newborn relative telomere length. Separate models were fit for each of the maternal predictors (CRISYS; PSS; PRAS; EPDS; STAI; depression diagnosis, medical record; anxiety diagnosis, medical record; diagnosed depression, prior to pregnancy; self-reported depression, prior to pregnancy; CD-RISC). The models included terms for newborn sex and the interaction of newborn sex and the maternal predictor in order to report separate effects for males and females. Given that the relevant extant literature has relied on analyses unadjusted for potential confounders, prior to conducting the linear regression analyses, supplemental correlational analyses were run to test the associations between the maternal psychosocial measures and newborn telomere length in the sample as a whole, as well as split by infant sex, to allow comparisons with prior studies. To determine the covariates to be included in the linear regression analyses, associations of each of the potential covariates with newborn relative telomere length were examined in bivariate correlational analyses; potential covariates that were categorical were only considered if there were at least five participants in each category. Covariates that were associated with newborn relative telomere length at p<.05 were included in the linear regression analyses. Data analyses were performed using STATA, version 16.1 (StataCorp, College Station, TX, USA) and IBM SPSS Statistics for Windows, version 24 (IBM Corp., Armonk, NY, USA).

3. Results

3.1. Sample Characteristics and Descriptive Data

Table 1 displays the sample characteristics, including potential covariates. The majority of the sample was non-Hispanic White (64.4%) and of higher socioeconomic status, as indicated by maternal educational attainment and household income. The infants were primarily of normal birthweight (97.3% ≥2500 grams) and gestational age (92.5% ≥37 weeks). Pregnancies were largely uncomplicated by obstetric health issues. The majority of maternal questionnaires were administered during the 2^nd trimester (77%), and the remaining during the 3^rd trimester (16%) or 1^st trimester (6%). Table 2 displays the descriptive data for the maternal predictors and newborn relative telomere length. Newborn relative telomere length was normally distributed with no outliers and was not associated with number of days elapsed between cord blood collection and DNA extraction. There were no sex differences in newborn relative telomere length, males M=1.04, SD=0.24, females M=0.97, SD=0.25, t(144)=1.72, p=.09, or in any of the maternal predictors or potential covariates, ps>.05. As shown in Table 2, 13.0% and 17.8% of participants, respectively, had a depression or anxiety diagnosis during pregnancy; these rates are similar to that documented in the general population, including during pregnancy (Biaggi, Conroy, Pawlby, & Pariante, 2016). The distribution of stress and psychopathology symptom scores suggest overall low to moderate stress/psychopathology in the sample, similar to many of the prior studies in this area. Supplemental Table 1 displays the correlation coefficients among the stress and psychopathology measures, which were low to moderate, justifying their separate consideration in analyses.

Table 1.

Sample characteristics (N = 146)

	na	%	M	SD
Maternal age (years)			32.7	5.1
Maternal race/ethnicity
White	94	64.4
Black	19	13.0
Asian	13	8.9
Hispanic	6	4.1
Other	12	8.2
Maternal education
Less than high school	3	2.1
High school/GED	11	7.5
Some college	23	15.8
College degree	39	26.7
Graduate degree	63	43.2
Annual household income
< $10,000	10	6.8
$10,000-$24,999	14	9.6
$25,000-$49,999	14	9.6
$50,000-$74,999	12	8.2
$75,000-$99,999	14	9.6
$100,000+	74	50.7
Maternal prenatal body mass index (BMI), kg/m2			26.5	6.8
Maternal smoking, current pregnancy (yes)	2	1.4
Maternal lifetime smoking ≥ 50 cigarettes (yes)	23	15.8
Parity (nulliparous)	75	51.4
Prior obstetric / health history
Preterm birth	11	7.5
Preeclampsia	6	4.1
Pre-existing diabetes	1	0.7
Health indicators of current pregnancy
Chronic hypertensionb	6	4.1
Gestational hypertensionc	6	4.1
Preeclampsia	5	3.4
Gestational diabetes	4	2.7
Newborn sex (male)	74	50.7
Newborn birthweight (grams)			3407	474
Newborn gestational age (weeks)			39.3	1.5

^aData were missing for 0 to 16 participants across study variables.

^bDefined as hypertension identified prior to 20^th week of pregnancy.

^cDefined as hypertension diagnosed after 20^th week of pregnancy.

Table 2.

Descriptive data for the maternal predictors and cord blood relative telomere length

	na	%	M	SD	Median	Interquartile Range
Stressor exposures during pregnancy (CRISYS-R)			1.22	1.57	1	0 – 2
Psychological stress (PSS)			8.33	2.12	8	8 – 9
Pregnancy-related anxiety (PRAS)			14.03	5.11	13	10 – 16
Depressive symptoms (EPDS)			5.91	4.98	5	2 – 8.25
Anxiety symptoms (STAI)			7.16	4.70	6	4 – 10
Depression diagnosis, medical record	19	13.0
Anxiety diagnosis, medical record	26	17.8
Diagnosed depression, prior to pregnancy	36	24.7
Self-reported depression, prior to pregnancy	49	33.6
Psychological resiliency (CD-RISC)			75.49	12.80	76	66 – 86
Cord blood relative telomere length (T/S ratio)b			1.00	0.25	0.98	0.79 – 1.19

Note. CRISYS-R = Crisis in Family Systems-Revised; PSS = Perceived Stress Scale; PRAS = Pregnancy-Related Anxiety Scale; EPDS = Edinburgh Postnatal Depression Scale; STAI = Spielberger Trait Anxiety Inventory; CD-RISC = Connor-Davidson Resilience Scale.

^aData were missing for 0 to 15 participants across study variables.

^bTelomere length is represented by the ratio between the average of three values obtained from telomere amplification and from β-globin amplification (T/S ratio).

3.2. Bivariate Associations of Maternal Psychosocial Variables and Potential Covariates with Newborn Relative Telomere Length

Unadjusted correlational analyses did not reveal any significant associations between any of the maternal psychosocial measures and newborn infant telomere length in the sample as a whole or among females (Supplemental Table 2). Among males, newborn relative telomere length was positively associated with pregnancy-related anxiety symptoms (PRAS), depressive symptoms during pregnancy (EPDS), general anxiety symptoms (STAI), and self-reported depression prior to the current pregnancy. Among the potential covariates, newborn relative telomere length was negatively associated with maternal history of smoking, maternal preeclampsia in a prior pregnancy, and timing of maternal questionnaire completion (Supplemental Table 3); post-hoc correlation coefficient difference tests via the Fisher-to-z transformation showed no sex differences in the association between these covariates and newborn telomere length, ps≥.70. None of the other tested potential covariates were associated with newborn telomere length (Supplemental Table 3) and therefore were not considered further in analyses.

3.3. Linear Regression Analyses Predicting Newborn Relative Telomere Length from Maternal Stress and Psychopathology

Based on bivariate analyses, all linear regression analyses were adjusted for maternal history of smoking, preeclampsia in a prior pregnancy, and timing of maternal questionnaire completion. Maternal depressive symptoms in pregnancy (EPDS), maternal self-reported depression prior to pregnancy, pregnancy-related anxiety (PRAS), and general anxiety (STAI) each were positively associated with newborn relative telomere length among males. These effects were significantly different than the effects observed in females (interaction effects ps<.05), which were all non-significant. For maternal psychological resilience (CD-RISC), the sex interaction term was significant; however, the association between maternal psychological resilience and newborn telomere length was not significant among males or females. In all models, maternal history of smoking and preeclampsia in a prior pregnancy remained significantly associated with shorter newborn telomere length. Table 3 displays the beta effect sizes and 95% confidence intervals for each maternal psychosocial predictor and the p-value for each maternal predictor x newborn sex interaction term for each of the linear regression analyses.

Table 3.

Betas and 95% confidence intervals for each maternal predictor and the p-value for each maternal predictor x newborn sex interaction term for the linear regression analyses predicting cord blood relative telomere length

	Males	Females	Interaction p-value
Maternal predictor	β [95% CI]	β [95% CI]
Stressor exposures during pregnancy (CRISYS-R)	0.017 [−0.020,0.054]	−0.005 [−0.038,0.028]	0.38
Psychological stress (PSS)	0.011 [−0.018,0.040]	−0.011 [−0.034,0.013]	0.25
Pregnancy-related anxiety (PRAS)	0.012 [0.0005,0.023]*	−0.009 [−0.018,0.001]	0.006
Depressive symptoms (EPDS)	0.013 [0.003,0.023]*	−0.005 [−0.016,0.006]	0.02
Anxiety symptoms (STAI)	0.012 [0.0005,0.023]*	−0.007 [−0.018,0.004]	0.02
Depression diagnosis, medical record	0.144 [−0.006,0.293]	−0.004 [−0.174,0.166]	0.19
Anxiety diagnosis, medical record	0.062 [−0.070,0.195]	0.026 [−0.122,0.175]	0.72
Diagnosed depression, prior to pregnancy	0.117 [−0.005,0.239]	0.009 [−0.112,0.129]	0.21
Self-reported depression, prior to pregnancy	0.112 [0.006,0.219]*	−0.089 [−0.199,0.022]	0.01
Psychological resiliency (CD-RISC)	−0.003 [−0.007,0.002]	0.005 [−0.00003,0.009]	0.03

Note. CRISYS-R = Crisis in Family Systems-Revised; PSS = Perceived Stress Scale; PRAS = Pregnancy-Related Anxiety Scale; EPDS = Edinburgh Postnatal Depression Scale; STAI = Spielberger Trait Anxiety Inventory; CD-RISC = Connor-Davidson Resilience Scale. All linear regression analyses adjusted for maternal lifetime history of smoking, preeclampsia in a prior pregnancy, and timing of completion of the maternal questionnaires.

^*p < 0.05.

4. Discussion

The objectives of this study were to test associations of a range of maternal stress and psychopathology measures with newborn telomere length and to explore potential sex effects on these associations in a prospective cohort study of pregnant women seeking routine prenatal care. Contrary to our hypotheses, when examined in the sample as a whole, we did not find evidence for maternal psychosocial effects on newborn telomere length, despite considering measures of maternal stressor exposures, subjective stress, and mental health. However, there was evidence for sex effects, with maternal depression symptoms prior to pregnancy and depressive, pregnancy-related anxiety, and general anxiety symptoms in pregnancy positively associated with newborn telomere length among males. Analyses further indicated main effects for maternal lifetime history of smoking and preeclampsia in a prior pregnancy, with both negatively associated with newborn telomere length in the sample as a whole.

Inconsistencies in findings across studies of maternal stress and newborn telomere length may be attributable to a number of factors. Studies that have documented associations have almost exclusively tested one measure of “stress,” primarily in small sample sizes, with the operationalization of stress varying widely. Specifically, among the studies purporting to show maternal stress effects on newborn telomere length, one operationalized stress as pregnancy-specific stress/anxiety (Entringer et al., 2013), one as exposure to potentially stressful life events (Holmes and Rahe Stress Scale), two as subjective evaluation of stress (Salihu et al., 2016; Send et al., 2017), one as a combination of perceived stress and negative mood reactivity to pregnancy-related events (Verner et al., 2020), and one as elevated depressive and/or anxiety symptoms (Suh et al., 2019). The few studies that have explored whether maternal psychological symptoms/diagnoses are associated with newborn telomere length have reported mixed findings (Bosquet Enlow et al., 2018; Send et al., 2017; Suh et al., 2019). Also, in the current and previous studies, with the exception of Verner and colleagues (Verner et al., 2020), maternal stress/mental health was assessed once, with variation as to when during pregnancy. Moreover, the various assessments across studies asked participants to report exposures/symptoms according to different timelines (e.g., in prior week, in prior 6 months, over lifetime). An extensive literature suggests that the timing of maternal psychosocial exposures influences the magnitude and nature of effects on a wide range of infant outcomes, and that timing effects vary by sex (Van den Bergh et al., 2017). Notably, telomerase activity varies over the course of pregnancy (Fragkiadaki et al., 2016), which may influence stress exposure effects on fetal telomere biology. Interestingly, in a study examining associations between maternal perinatal stress, assessed at multiple timepoints via the Perceived Stress Scale, and preschoolers’ buccal telomere length in 111 mother-child dyads, Carroll and colleagues (Carroll, Mahrer, Shalowitz, Ramey, & Dunkel Schetter, 2020) found that maternal stress during the 3^rd trimester was most robustly associated with shortened child telomere length, with modest effects for the 2^nd trimester and no effects for the 12 months prior to conception or the 1^st month postpartum. Moreover, the association between maternal 3^rd trimester stress and child telomere length was maintained after adjusting for concurrent maternal stress. Although this study did not assess newborn telomere length, the findings suggest not only that prenatal stress exposures might have long-term effects on offspring telomere length, but also that the timing of exposure during pregnancy may be critical. Notably, other environmental exposures (e.g., air pollution) have been associated with both shorter and longer telomere length at birth, depending on the timing of the exposure and fetal sex (Rosa et al., 2019). The extant literature does not provide guidance as to how the timing and severity of maternal stress and psychopathology during pregnancy may influence newborn telomere biology. Thus, the evidence to date for effects of maternal psychosocial functioning in pregnancy on newborn telomere length is far from conclusive. Considerably more research is needed that explicates (a) the specific parameters of maternal stress and mental health (e.g., type; timing; severity, including symptoms vs. clinical diagnoses) that affect newborn telomere length; (b) the biological mechanisms by which such exposures exert influence; (c) factors that modify the magnitude and direction of exposure effects; and (d) the role of fetal sex in these processes.

Prior studies linking maternal psychosocial measures to newborn telomere length have reported that higher levels of stress/psychopathology predict shorter newborn telomeres. In contrast, in the current study, elevated maternal symptoms on some of the psychosocial measures were associated with longer newborn telomere length, and only among males. Because telomere length at birth establishes an individual’s initial telomere length setting, shortened telomeres may be expected to predict poorer health (Factor-Litvak et al., 2016; Martens et al., 2016). However, some suggest that longer telomeres at birth predict more rapid telomere attrition across the life course and thus may be an indicator of vulnerability (Drury et al., 2015). There is indirect evidence for this hypothesis from research of individuals born preterm, who have longer telomere length at birth and at term equivalent age compared to individuals born full-term but accelerated telomere attrition and increased risk for the development of diseases of aging in later life (Smeets, Codd, Samani, & Hokken-Koelega, 2015; Vasu et al., 2017). There is a major gap in the literature as to the value of newborn telomere length for predicting later morbidity and mortality. Studies are needed that test whether newborn telomere length and/or telomere attrition rate in early life predicts later physical and mental health outcomes. Findings from such research would clarify the health consequences of relatively shorter and longer telomere length at birth and subsequent attrition rate and aid in the interpretation of in utero exposure effects on newborn telomere biology.

Interestingly, among the sociodemographic and health covariates tested, those that emerged as significant were related to maternal health prior to the infant’s conception, operated similarly among male and female infants, and showed associations in the expected direction (i.e., poorer health predicting shorter newborn telomere length). The association between maternal lifetime smoking history and newborn telomere length is not attributable to the effects of smoking during pregnancy, as only two women endorsed current smoking, and they were not represented in the group coded for a positive prior smoking history. Maternal preeclampsia in a prior pregnancy also was negatively associated with newborn telomere length. These findings suggest that these health conditions may have persistent effects on maternal biological systems involved in fetal telomere programming. Such a hypothesis is consistent with evidence that preeclampsia increases a woman’s risk for a range of adverse health outcomes months to years after delivery (Bokslag, van Weissenbruch, Mol, & de Groot, 2016). However, these findings should be interpreted with substantial caution and require replication due to the low number of women with a preeclampsia history in the current sample. Notably, newborn telomere length was not associated with maternal obstetrical health during the current pregnancy. Some prior research has linked gestational diabetes and hypertension/preeclampsia with disruptions in newborn telomere biology, but other studies have failed to find effects (Entringer et al., 2018; Fragkiadaki et al., 2016; Verner et al., 2020; Xu et al., 2014). Given the small number of women in the current study with current or prior obstetrical complications, these findings should be pursued in future research to ascertain the impact of current and past obstetrical health on newborn telomere biology and the mechanisms by which such factors operate.

We did not find evidence for a sex difference in mean newborn telomere length, consistent with some prior studies, although others have found longer telomere length among females (Bosquet Enlow et al., 2018; Factor-Litvak et al., 2016; Martens et al., 2016; Okuda et al., 2002; Verner et al., 2020; Wojcicki et al., 2016). In later life, females consistently demonstrate longer telomere length (Chen et al., 2011). These findings, in combination with emerging evidence for sex-specific effects of environmental exposures on telomere length beginning in pregnancy (Bosquet Enlow et al., 2018; Bosquet Enlow et al., 2019; Rosa et al., 2019), suggest that males and females may be differentially susceptible to environmental influences on telomere biology. Potential contributory mechanisms include fetal sex differences in the production of maternal cortisol over pregnancy and strategies for adapting to maternal stress reactivity and stress hormones (Bosquet Enlow et al., 2019; Davis et al., 2013; DiPietro et al., 2011; Doyle et al., 2015; Gabory et al., 2009; Ostlund et al., 2016). The effects of sex-specific responses to environmental insults on telomere biology may accumulate across time, leading to the eventual emergence of observable differences in telomere length between males and females. Although our findings were in the opposite direction than hypothesized, they were consistent with a prior study suggesting that, compared to females, males may be more susceptible to the influence of maternal psychosocial functioning on newborn telomere biology (Bosquet Enlow et al., 2018). Limited studies in adults also suggest that males may be more vulnerable than females to the effects of psychosocial factors (e.g., psychological distress, reduced social support, early life adversity) on telomere attrition (Shalev et al., 2014; Zalli et al., 2014). Research is needed to establish the mechanisms of sex effects on telomere biology from fetal development through adulthood and to determine the relevance of such effects for sex disparities in various disease states across the lifespan.

In the current study, maternal education, income, and race/ethnicity were not associated with newborn telomere length. Whereas some studies have found higher maternal socioeconomic status to be associated with longer newborn telomere length (Bosquet Enlow et al., 2018; Martens et al., 2016; Wojcicki et al., 2016), others have failed to find such effects (Drury et al., 2015; Factor-Litvak et al., 2016). Race/ethnicity differences in telomere length have varied widely in nature and degree across studies of infants as well as children and adults (Drury et al., 2015; Factor-Litvak et al., 2016; Martens et al., 2016; Needham, Fernandez, Lin, Epel, & Blackburn, 2012; Okuda et al., 2002). Race/ethnicity and socioeconomic status are often confounded in research (Geronimus et al., 2015), complicating efforts to understand the influence of specific sociodemographic characteristics on telomere parameters. Sociodemographic characteristics of the current sample have implications for interpreting the findings. The majority were of high socioeconomic status by measures of maternal education and household income. Although there was some racial/ethnic diversity, the majority were non-Hispanic White. The effects of maternal psychosocial exposures on newborn telomere length may vary by sociodemographic characteristics due to a variety of mechanisms. For example, maternal dietary deficiencies have been associated with shortened newborn telomere length (Entringer, Buss, & Wadhwa, 2015; Entringer et al., 2018; Kim et al., 2018), and low-income and racial/ethnic minority women are at heightened risk for nutritional deficiencies during pregnancy (Brunst, Wright, et al., 2014). Moreover, more optimal maternal nutrition during pregnancy may mitigate the effects of maternal stress on child outcomes, potentially through diet-driven reductions in oxidative stress and inflammation (Brunst, Enlow, et al., 2014; Lipton et al., 2017). In addition to more optimal nutrition, the women in our sample may have experienced lower levels of discrimination, less exposure to prior trauma, higher levels of social support, and better access to obstetric and mental health care, all of which may have moderated the effects of the examined psychosocial exposures on newborn telomere length. Research in older women suggests that engagement in greater health behaviors (nutrition, sleep, leisure time) attenuates or even eliminates the association between stress and telomere attrition (Puterman, Lin, Krauss, Blackburn, & Epel, 2015). Notably, with the possible exception of Verner and colleagues (Verner et al., 2020), the prior studies that both documented links between maternal psychosocial functioning and newborn telomere length and reported participant sociodemographic data had samples with a higher percentage of participants of racial minority background and of low socioeconomic status than represented in the current sample. Future studies should explore the role of sociodemographic factors in moderating associations between maternal psychosocial functioning and newborn telomere length and incorporate measures of potential compensatory factors to determine whether any deleterious maternal stress effects on newborn telomere biology may be mitigated by behavioral health interventions.

Notably, data suggest that, under certain conditions, stress exposures may actually enhance health outcomes across a range of indicators, including immune functioning and HPA axis regulation (Dhabhar, 2019; Jessop, 2019). This may be relevant for fetal telomere biology given that the mechanisms via which maternal exposures are hypothesized to influence newborn telomere length include maternal inflammation and HPA axis reactivity, as well as oxidative stress, epigenetic changes resulting in altered gene expression, mitochondrial dysfunction, and changes in telomerase production (Entringer et al., 2011; Entringer et al., 2013; Lupien, McEwen, Gunnar, & Heim, 2009; Shiels et al., 2011; Steptoe et al., 2011; Van den Bergh et al., 2017). The reported levels of stress and mental health difficulties in the current sample were generally low to moderate. Although highly speculative, the current findings may suggest that low to moderate levels of maternal stress in pregnancy may have positive, adaptive effects on fetal telomere biology, particularly among males. This hypothesis assumes that longer telomere length at birth is protective, which, as mentioned above, is controversial.

This study contributes to the small literature exploring determinants of newborn telomere length. It is unique in its consideration of a wide range of maternal psychosocial exposures and indicators of maternal health. Other strengths include the relatively large sample for a study of newborn telomere length determinants and consideration of potential sex effects. Limitations may include the use of cord blood as an indicator of newborn telomere length. Although other studies have validated the use of cord blood for this purpose (e.g., (De Carli et al., 2017; Okuda et al., 2002), and all of the above cited studies relied on cord blood for assessing newborn telomere length (Bosquet Enlow et al., 2018; Bosquet Enlow et al., 2019; Entringer et al., 2013; Marchetto et al., 2016; Salihu et al., 2016; Send et al., 2017; Suh et al., 2019; Verner et al., 2020), concerns have been raised about its validity. For example, cord blood may be contaminated by maternal blood during labor or sample collection, with maternal DNA present in 1–17% of umbilical cord samples (Scaradavou, Carrier, Mollen, Stevens, & Rubinstein, 1996). Future studies should consider the use of other newborn DNA sources (e.g., blood spots, saliva, buccal cells) and sampling procedures that do not risk or can control for contamination by maternal factors (Bosquet Enlow et al., 2019) and should confirm the validity of these various methods. The current study utilized buffy coat as the source of leukocyte DNA for assessing newborn telomere length. Most telomere research has relied upon leukocyte DNA isolated from peripheral blood mononuclear cells, although some studies have utilized buffy coat, including studies of newborn telomere length (e.g., Suh et al., 2019). Noted strengths of buffy coat include low biovariability and excellent DNA yield and quality, including higher yield compared to whole blood (Lin, Smith, Esteves, & Drury, 2019). However, buffy coat samples may contain mixed immune cell types (Lin et al., 2019), and yields of DNA from frozen buffy coat may be diminished and more difficult to process than the all-cell-pellet of whole blood, may be vulnerable to processing techniques, and may have elevated inter-individual variability (Goldman et al., 2018). Buffy coat fractions were stored frozen for up to 4 years prior to DNA extraction. The impact of long-term storage on telomere length quality has not been examined; however, extracting all samples at the same time under the same conditions, using DNA extraction and assay reagents from the same lot, may be more critical than storage time (Lin et al., 2019). The method utilized for quantifying telomere length, real-time PCR, presents both strengths and challenges. It requires small amounts of DNA and is relatively inexpensive and less labor intensive than other methods (Montpetit et al., 2014). However, PCR may be more vulnerable to measurement error/variation than other methods (e.g., telomere restriction fragmentation) and, because it produces relative telomere length rather than absolute kilobase length estimates, may limit comparability with findings from other studies (Montpetit et al., 2014). In this study, PCR efficiency was observed to be elevated (see Supplemental Material: Detailed Cord Blood Telomere Length Assaying Procedures for more detail), possibly due to residual contamination from the extraction, non-specificity of the primers, and/or primer-dimers. However, the elevated efficiency should not impact nor bias the results given the manner in which telomere length was calculated. We did not make adjustments to our alpha level despite testing several different maternal risk factors; therefore, our findings should be replicated. While we did ascertain maternal smoking history, we did not measure cotinine to confirm maternal self-report or ask about exposure to environmental tobacco smoke during pregnancy, which could affect newborn telomere length. We did not consider paternal factors, which may contribute to newborn telomere length via various processes, including heritability, paternal characteristics (e.g., age), and assortative mating (Broer et al., 2013). Future studies would be strengthened by considering paternal factors, which have been given minimal attention in the newborn telomere literature. The relatively limited representation of individuals of minority racial/ethnic backgrounds and varied socioeconomic status may restrict generalizability of the findings; the effects of maternal psychosocial exposures on newborn telomere length should be explored in samples representing a range of sociodemographic characteristics.

4.1. Conclusions

In a prospective cohort study of pregnant women seeking routine prenatal care, we found some evidence for sex-specific effects of maternal psychosocial functioning on newborn telomere length. Previous studies have linked elevated maternal stress with shortened newborn telomere length; these studies have been inconsistent in their operationalization of stress and primarily conducted in small samples without consideration of sex effects. In the current study of a relatively healthy, high socioeconomic status sample, the evidence suggested elevated maternal stress/mental health symptoms predicted longer newborn telomere length among males. Findings also indicated that maternal health and obstetric history prior to conception may influence newborn telomere length in male and female newborns. More research is needed to determine the clinical significance of both shorter and longer telomere length at birth for later health outcomes and the potential impact of compensatory factors and behavioral interventions in disrupting any deleterious influences of maternal physical and mental health on fetal telomere biology.