Maternal psychological stress during pregnancy and newborn telomere length: a systematic review and meta-analysis

Telomeres are DNA–protein structures at the end of chromosomes that protect the terminal regions from degradation and intrachromosomal fusion [1]. The end replication problem results from DNA polymerase's inability to completely replicate the ends, which causes telomeres to shorten with each replication cycle in somatic cells [2]. Telomere length (TL) shortening is not only a biomarker but also appears to be a causal factor in various physical and mental illnesses and mortality [3]. Shorter TL has been linked to major group diseases such as chronic obstructive pulmonary disease (COPD) and idiopathic pulmonary fibrosis [4], infections [5], autoimmune diseases [6], some types of cancer [7], diabetes and metabolic syndrome [8], cardiovascular diseases [9] and psychological disorders [10].

It is hypothesised that TL at birth can significantly predict TL later in life [11]. Like many other situations, genetics and environmental variables affect infants’ TL; hereditability accounts for 70% of the variances in TL [12]. Environmental factors can also affect TL. Telomeres are highly vulnerable to impairment through alkylation, ultraviolet irradiation, and oxidative stress [13]. Numerous aspects of a person's lifestyle, including physical activity, smoking, and sleep, are associated with TL. One of the most critical health issues is psychological stress [14].

Stress is a process resulting in both psychological demands and biological changes that could put a burden on someone. Chronic psychological stress at high levels has been linked to a variety of diseases and unhealthy situations, including tumor growth [15], metabolic syndrome [16], cardiovascular diseases [17], and many others. Stress and illness have a complicated relationship. Individuals differ in their susceptibility to stress. A stressful incident that causes disease in one person might not in another. To manifest as a disease, events must interact with various underlying causes [18].

Stress has a known effect on the biology of telomeres. Research has demonstrated that cumulative exposure to oxidative stress speeds up telomere erosion [13]. Furthermore, an inverse correlation between TL and an inflammatory stress response has been observed [19]. Adversity-induced stress may be responsible for accelerated early-onset telomere shortening. Psychological stress is one predictor of lifelong TL [20]. The first observations showed that women with the highest levels of perceived stress have shorter telomeres than women with less stress. This shortness is equivalent to a decade of old age [21]. First-ever research by Entringer et al. suggests that maternal psychological stress during pregnancy may be linked to reduced TL in neonates at birth [22]. Maternal stress induces the release of hormones, such as cortisol and placental corticotropin-releasing hormone (CRH), as well as mediators of inflammatory and oxidative stress. These hormones and mediators enter the fetal circulation and cause changes in the placental metabolism [23]. Other research was conducted in various populations using comparatively similar methods. Some results calculated different effects, such as a study that did not repeat this inverse effect [24]. The studies' inconsistent findings led us to identify the need for a comprehensive synthesis of the literature. In this study, we focused on maternal psychological stress during pregnancy and its relation to newborn TL.

This article aims to systematically review and synthesize the evidence from the most recent literature (indexed up to February 10, 2023) and conduct a meta-analysis. To advance clinical practice and the design of subsequent investigations, we aimed to determine the impact of self-reported maternal stress on newborns’ TL. There are reviews that measure the effect of adverse childhood experiences on TL, finding a negative association between them [25]. In a systematic review and meta-analysis, childhood trauma was significantly associated with shorter TL [26]. The relationship between shorter TL and several types of stress, such as chronic social stress, post-traumatic stress disorder, and preterm infant stress have been investigated in other reviews [26‐28]. To our knowledge, no published review compiles the research results in which self-reported maternal stress and newborn TL were examined.

This systematic review and meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines 2020 [29]. The registration number in PROSPERO is CRD4202128517.

We reviewed the relationship between maternal-perceived stress during pregnancy and the TL of newborns and identified eight studies that fulfilled the inclusion criteria.

We conducted a thorough review of eight included studies. It should be mentioned that there were some discrepancies in the findings of the included studies regarding this potential association. In five studies an inverse association between maternal-perceived stress and newborn TLwas reported [22, 23, 30, 31, 33]. Entringer et al. published the first preliminary data in humans demonstrating a link between prenatal stress exposure and shorter TL in 27 young adults. An inverse significantrelationship was reported (β = -0.099; P = 0.04) [35]. In a cross-sectional study, Salihu et al. examined 71 women of low-socioeconomic status, indicating that this population has a higher risk of stress. They discovered an inverse significantrelationship (P < 0.05 in 5000 of the 5000 repetitions using repeated ANOVA) [36]. The design of the study prohibited claims about causality, so more longitudinal studies are needed. Verner’s comparably large cohort with 656 participants replicate previous reports linking maternal stress during pregnancy with offspring TL with a borderline significant relationship (β = -0.079, p = 0.044). In Verner’s study, women had at least one risk factor for preeclampsia [31]. Early hypoperfusion or placental ischemia–reperfusion injury in preeclampsia leads to increased oxidative stress. According to Send et al., maternal perceived stress during pregnancy was significantly associated with shorter telomeres in newborns (β = -0.14; P = 0.015) [13]. Marchetto et al. conducted a prospective cohort study of 24 mothers and found a significant negative association between maternal stress and newborn TL (β = -0.463, P = 0.04) [13].

In the study conducted by Izano et al., maternal-perceived stress had a marginal association with shorter newborn TL; however, after adjusting for multiple comparisons, the relationship remained insignificant (β = -0.07; P = 0.29). In addition to using a linear regression model, this study applied Targeted Minimum Loss-Based Estimation (TMLE) to evaluate the relationship between TL and maternal exposure to stress [32]. Ämmälä et al., with a sample of 1405 infants, could not confirm a correlation between maternal stress and leukocyte TL in newborns. Some possible causes were differences in the TL analysis methods between studies, variations in the statistical models, different cell populations in umbilical blood samples, and differences in measures of maternal well-being and the timing of measurements [24]. On the other hand, Enlow et al. reported an association in the opposite direction. They showed maternal stress correlated with longer TL among male newborns (β = 0.011) [34]. It was the only study exploring potential sex effects on these associations.

In six studies, the PSS-4 was used to assess stress levels. PSS-4 is one of the most common tools for evaluating psychological stress. It is a valuable tool for assessing stress during pregnancy and can help clinicians obtain an accurate evaluation of the degree of maternal stress for optimising health care among pregnant women [37]. Seven studies used qPCR to determine TL. Only one study used the Telomere Restriction Fragment tool. All these measures are validated in measuring self-reported maternal stress during pregnancy. Some studies investigated the effects of gender on TL, and some further evaluated the role of gender in the association between maternal stress and TL. Izano et al. observed that a high level of maternal perceived stress was correlated with shorter TL among girls but longer among boys [32]. They stated that the difference in the association between maternal stress and TL among males and females might suggest the different influences of financial strain, unplanned pregnancy, and perceived stress; high levels of perceived stress and unplanned pregnancy were associated with shorter TL among females but longer among males. Food insecurity was strongly associated with male newborns and not female ones [32]. Verner et al. reported that the female sex was a significant predictor of TL at birth, as TL was shown to be longer in girls [33]. Ämmälä et al. found no significant difference in the association between maternal stress and TL between the sexes [24]. Entringer et al. observed no significant difference between males and females regarding TL [22]. Enlow et al. found no sex differences in TL; however, they reported that increased maternal stress predicted longer newborn TL among male newborns [34]. Sex differences in TL are controversial, but it is widely accepted that women have longer telomeres than men [38]. However, as the difference is barely seen in newborns, the effects of sex may be primarily related to different rates of telomere attrition after birth which mainly depends on estrogen. Estrogen can increase telomerase activity and, on the other hand, decrease oxidative stress [39]. Collectively, the present studies provide evidence that this relationship should be investigated.

The effect of preeclampsia on newborn TL was investigated in some included studies. Verner et al. reported no significant effect [33], while Enlow et al. reported that preeclampsia in a prior pregnancy was negatively associated with newborn telomere [34]. It has been proposed that in pregnancies complicated with preeclampsia, the length of the telomeres is significantly shorter [40]. A previous study showed shorter telomeres in trophoblast in preeclampsia with decreased expression of the human telomerase reverse transcriptase (hTERT) component of the human telomerase enzyme [41]. Another study reported a significant decrease in levels of protective proteins such as Sirtunin1 (SIRT1) in placentas from preeclampsia resulting in shorter telomeres [42].

Participants in three studies were completely white, making them non-representative samples.. In the US studies, however, samples consisted of diverse races/ethnicities. Enlow et al. and Izano et al. showed that newborns' race/ethnicity was not associated with TL [34]. However, the association between maternal stressors and TL in newborns varies by race/ethnicity; perceived stress is accompanied by shorter newborn TL among White women [32]. The impact of the maternal psychosocial condition on newborn TL is associated with sociodemographic characteristics. Different mechanisms have been suggested for this variety. Pregnant women of a racial minority are at higher risk for nutritional deficiencies, which may be associated with reduced newborn TL [34]. Race differences in newborn TL may also vary by receipt of public assistance. Infants born to black mothers have longer telomeres than those born to white mothers.This is contrary to the assumption that a mother’s social disadvantage is associated with shorter TL in offspring [43].

While the current meta-analysis is a first, the findings should be considered along with its limitations. We included only four studies in our meta-analysis; the number of studies conducted on this topic is limited. Given the limitations we had such as restricted number of studies and the presence of moderate heterogeneity, we determined that the maximum likelihood model was the most suitable one to utilize. Differences in LTL analysis methods, stress measurement tools, statistical models, and study populations across studies posed challenges in combining data and led to some limitations in the generalizability of the findings.As these kinds of research require advanced tools and precise follow-up, the existing studies have been performed in a few high-income countries. In all included studies, the measure for stress was a self-reported questionnaire. No biological marker, such as maternal or cord blood oxidative stress marker, was examined. Furthermore, there was a wide range in the number of subjects included in the studies. We could not access the full text of one study despite contacting the corresponding author three times and other authors one time. Our study exclusively focused on English-written articles. Consequently, we recommend that future studies consider incorporating articles published in other languages to enhance the representation of research within the field.

In summary, we identified and systematically reviewed 8 studies that investigated the relationship between maternal-perceived stress and newborn TL; 4 studies were included in the meta-analysis. The findings of the studies overally point to aninverse statistically significant relationship among them, which was supported by our meta-analysis.. Individuals with shortened telomeres are more susceptible to developing age-related diseases, and this association between TL and chronic diseases is mainly determined in the first decades of life. This raises concern as maternal-perceived stress during pregnancy can increase the risk for health problems in their offspring from an early age. We recommend further studies to investigate this correlation in large sample sizes and determine the exact mechanisms and risk factors for shortened telomeres in newborns of mothers who experienced high stress levels during pregnancy. Subsequent studies may measure stress using biological markers. Further research is needed to understand the complex relationship between maternal psychological stress during pregnancy and TL, which may yield deeper insights and allow us to identify and protect pregnant women from stressful situations.

Telomere length varies significantly among newborns, which indicates the role of physiological and environmental factors during the fetal period. In this study, the analysis of the existing studies revealed an inverse statistically significant relationship between maternal-perceived stress and newborn TL. This research provides a better understanding of the underlying mechanisms and pathways involved. We suggest that early identification and assessment of pregnant women experiencing high levels of stress, along with health follow-ups of infants whose mothers experienced psychological stress during pregnancy, could be beneficial to support the well-being of both mothers and their infants. This systematic review and meta-analysis is a step towards implementing more studies considering some important limitations that were mentioned.