Abstract: Studies have suggested that maternal antenatal depression and anxiety has been associated with increased risk of preterm birth. The exact mechanisms underlying this process are not fully understood. This paper discusses the existing literature on the subject, including the relationship between preterm birth and maternal depression, and suggests a neuroendocrine mechanism by which HPA axis dysregulation and alterations in maternal cortisol may result in fetal epigenetic and hormonal changes, including alterations in placental 11β-HSD2 expression, regulation of glucocorticoid receptor NR3C1 gene expression, and placental CRH abnormalities. The manner in which these factors dynamically interact to increase the risk of preterm birth is examined, as well as the possibility that abnormal CRH specifically is disrupting the normal partuition response due to its role in the mechanism of birth.
Introduction
In the United States, one of the most prevalent problems in obstetrics today is preterm birth, primarily due to its association with adverse maternal and fetal health outcomes (Wadhwa et al., 2011). The incidence of preterm birth is about 12-13% in the United States compared to 5-9% in other developed countries. In the face of increasing knowledge and literature on its etiology, preterm birth rates continue to ascend, rising from 9.5% in 1981 to 12.7% in 2005 (Goldenberg et al., 2008). The cause of preterm birth is complex and currently poorly characterized. From 1971-2011, preterm birth did not significantly decline (Wadhwa et al., 2011), and currently preterm birth contributes to 75% of infant mortality in the United States (Goldenberg et al., 2008). Determining its etiology requires a cross-disciplinary understanding of endocrinology, embryology, epigenetics, neurobiology, clinical psychology, and obstetrics (Wadhwa et al., 2011).
There exists a significant link between depression, stress, and preterm birth, even after adjusting for behaviors associated with depression and stress (including drug, cigarette smoking, and alcohol use) (Dole et al., 2003). Clinical depression is diagnosed in 16% of all pregnant women in the United States, and in general, 35% of pregnant women show at least some symptoms of depression (Goldenberg et al., 2008). Pregnant women experiencing high levels of psychological or social stress generally double their chances of preterm birth (Dole et al., 2003). In this review paper, we will discuss the mechanisms by which maternal depression and stress can influence preterm birth, especially in regards to the physiological response of the fetus, the maternal and fetal stress response, as well as the associated epigenetics. Here we attempt to delineate the neuroendocrine mechanisms by which maternal depression and stress may act in epigenetically-controlled hormonal pathways to increase the risk of preterm birth.
Preterm Birth in Prenatally Depressed Mothers: A Clinical Look
There have been many studies using various diagnostic assessment tools to associate maternal depression and stress during pregnancy with a greater risk of preterm birth. Li et al. performed a prospective cohort study that showed a significant association between maternal depression during pregnancy and preterm birth, as observed in a population of 791 pregnant women. Prenatal depression and severity was measured according to the Center for Epidemiological Study Depression scale (CESD). Scores greater than 16 indicated “significant depressive symptoms,” while those greater than 22 indicated “severe depressive symptoms.” Preterm birth was found in 8.3% of women with depressive symptoms (CESD >16), as opposed to only 4.1% of women without depressive symptoms (CESD <16). These findings indicate more than a two-fold greater risk of delivering prematurely if presenting with symptoms of depression during pregnancy.
Also, increasing severity of depression as measured by CESD score correlated with higher rates of preterm birth—the incidence of preterm birth seen in pregnant women with significant and severe depressive symptoms were 5.8% and 9.3%, respectively (Li et al., 2009). One limitation of this study was that it measured depressive symptoms at only one time point during pregnancy rather than continuously, which meant that the potential for symptom progression was not taken into account. Strengths include its prospective design and its ability to enroll subjects who were not taking antidepressants at the time of pregnancy, avoiding any independent effects that may have resulted from the medication.
Furthermore, Dayan et al. revealed that prenatal depressive symptoms, as measured using the Edinburgh Postnatal Depression Scale (EPDS), have also been positively associated with higher rates of spontaneous preterm birth. Among a cohort of 641 pregnant women who were prospectively recruited, those with high depression scores (EPDS greater than 14) had a premature delivery incidence of 9.7%, compared to only 4.0% among the non-depressed women (EPDS less than 14). These results similarly show a significant association between depression and preterm birth, with the rate of premature delivery more than doubling among the depressed group (Dayan et al., 2006). These investigators controlled for potential confounding variables, including enrollment-related stress and coexisting antidepressant use, which add strength to these findings. Nevertheless, there are some points that should be kept in mind when evaluating these studies’ findings. First, these two studies purely displayed correlation, not causation. Second, differing methods of measuring depression were used that may cloud the ability to estimate the degree of correlation with complete accuracy.
In addition to depressive symptoms, Mancuso et al. presented the association between prenatal stress in the mother and greater risk of preterm birth, while also showing a connection with maternal corticotropin-releasing hormone (CRH) levels. At 28 to 30 weeks into gestation, greater pregnancy-specific anxiety was reported among the pregnant women who had premature deliveries in comparison to those who gave birth at 37 weeks or more. There was also a positive correlation between pregnancy-specific anxiety and prenatal CRH levels at 28 to 30 weeks gestation. Moreover, there was found to be a concurrent relationship between CRH levels and earlier gestational delivery. CRH levels were measured to be significantly higher at two independent time points throughout pregnancy among the women who delivered prematurely (Mancuso et al., 2004). Despite some limitations that might relate to the tools used in measuring pregnancy-specific stress, these results further replicate the findings by Lilliecreutz et al. The observed association of maternal depression, stress, and elevated CRH in pregnancy with higher risk of preterm birth necessitates investigation into altered neuroendocrine mechanisms, such as the Hypothalamic-Pituitary-Adrenal (HPA) axis, as a potential mediator in the onset of adverse birth outcomes.
HPA Axis and its Role in Parturition
The HPA axis is implicated in various physiological neuroendocrine responses to stress, mood, energy storage and expenditure, circadian rhythms, and immunity. In response to stress, CRH is released from the paraventricular nucleus of the hypothalamus into the hypothalamo-hypophyseal portal system. When acted upon by CRH, the anterior pituitary gland secretes adrenocorticotropic hormone (ACTH), which in turn stimulates the zona fasciculata of the adrenal cortex to release glucocorticoids, specifically cortisol in humans. The effects of cortisol are many; most importantly it increases blood glucose levels, heart rate and blood pressure, and suppresses the immune system (Wadhwa et al., 2011).
During pregnancy, the placenta is a “temporary endocrine structure” that plays an acute role as the interface between the mother and the fetus, especially in regards to its implication in the parturition response through its contribution to elevated CRH levels (Monk et al., 2012). The human fetal pituitary releases ACTH when it is activated by placental CRH. ACTH then acts upon the fetal adrenal cortex to drive the synthesis and release adrenal glucocorticoids and steroids. Upon fetal maturation, this activation of the fetal HPA axis is involved in a positive feedback loop that increases the release of placental CRH. Placental production of ACTH can also be stimulated via the paracrine route through the action of placental CRH (Robinson et al., 1988). Placental CRH amplifies the effect of estrogens on the uterus and the cervix, and interacts with both prostaglandins and oxytocin, which are known to induce myometrial contractions during parturition (Wadhwa, 2001).
Due to the vast amount of effects that placental CRH exerts at the end of pregnancy, it has been referred to as a “placental clock” in regards to its huge role in the onset of parturition (McLean and Smith, 1999). Therefore, we believe that dysregulation of normal functioning of the HPA axis, specifically relating to placental CRH levels, might be strongly implicated in the mechanisms behind preterm birth.
Dysregulation of the HPA Axis: Depression’s Effect on Altered Fetal Physiology in Utero
Because parental psychiatric illness can be a risk factor for developmental abnormalities in offspring, dysregulation of maternal HPA axis during pregnancy has the potential to harm the fetus in utero. Several studies measure fetal heart rate (FHR) and its response to stressors as an indicator of maternal HPA axis dysregulation. Monk et al. measured baseline FHR and FHR variability during a psychological Stroop color-word matching challenge in 3rd trimester pregnant women with depression, anxiety, comorbid for depression and anxiety, and no history of psychiatric disease. Results showed that fetuses of comorbid mothers had greater increases in FHR relative to the mothers with just depression or anxiety and healthy mothers. In addition, mothers were asked to do a paced-breathing task, which showed a similar increase in FHR in all subjects. Thus, the study was able to demonstrate that increased FHR was not due to increased cardiorespiratory activity secondary to psychological stress, but rather due to the psychological stress itself (Monk et al., 2011). However the results of this study did not correlate to Monk et al.’s previous study in 2004, which showed that depression alone in mothers also increased FHR greater than in mothers with anxiety alone (Monk et al, 2004). Both analyses were conducted on a small patient population, and thus necessitate verification with replication.
In a similar study, Allister et al. sought to measure FHR variability in response to a vibroacoustic stimulus (VAS) in depressed mothers. The group conducted a study with 20 participants who were between 32 and 36 weeks of gestation. Ten participants were mothers with depression and the remaining had no prior history of depression. Baseline FHR, FHR variability, and FHR while fetuses were exposed to a VAS were measured. Fetuses of mothers with depression had elevated FHR at baseline, took longer times to react to VAS, and took a longer time to return to baseline after the stimulus compared to fetuses of mothers without depression (6.5 minutes compared to 2 minutes). Thus, maternal depression is implicated in possible desensitization of FHR. Therefore, this finding further strengthens a possible association between prenatal stress and HPA axis dysregulation, through the observed effects on fetal physiology (Allister et al., 2001). One area of interest involved in the HPA axis that this paper explores is the effect of maternal cortisol levels on preterm birth.
High Maternal Cortisol as a Possible Mediator
Field et al. found that high maternal cortisol is positively associated with increased risk of preterm birth. Maternal urinary cortisol was measured at 20 weeks of gestation and subjects were grouped into equally-numbered high and low cortisol groups. 32% of the high cortisol mothers, as opposed to none of the low cortisol mothers, gave birth prematurely, and all of the low birth weight births occurred in the high cortisol group. (Field et al., 2006). Another study found that both maternal CRH and cortisol at 24 weeks of gestation were higher in women who gave birth prematurely (Mercer et al., 2006).
Studies have also established that there is a corresponding rise in placental CRH peptide and mRNA when glucocorticoids are present in the maternal plasma (Robinson et al., 1988). Robinson et al. conducted a study investigating the controlling mechanisms for placental CRH expression in the primary cultures of purified human trophoblast cells. The study found that trophoblast cells cultured for 24-72 hours increased CRH mRNA content and peptide secretion into media by 2- to 5-fold (from 1 uM to 10 uM) in the presence of glucocorticoid dexamethasone (1uM for 24 hr). CRH is released by the placenta during pregnancy, interfaces with prostaglandin and enhances the effects of oxytocin, which is responsible for stimulating myometrial contractions during parturition (Wadhwa et al., 2001). Moreover, there is a plethora of clinical studies that have examined the correlation between plasma concentrations of CRH and spontaneous preterm labor, and the literature suggests that women in preterm labor have notably elevated levels of CRH (Wadhwa et al., 2011). Therefore, depending on the periodicity of stress and the resultant increase in CRH synthesis associated with high maternal glucocorticoid levels during pregnancy, stress may be a potential contributor to early onset of parturition and preterm birth (Wadhwa et al., 2011).
In addition to maternal cortisol levels being correlated with both higher placental CRH levels and greater risk of preterm birth, studies have also found a positive association between maternal plasma cortisol and amniotic fluid cortisol (Baibazarova et al., 2012). Glover et al. obtained similar results at higher levels of maternal anxiety, and found that the correlation increased with the level of maternal anxiety. While this study referenced down-regulation of 11β-HSD2 as a potential cause, 11β-HSD2 levels were not measured (Glover et al., 2009). It is unclear whether anxiety levels, which were measured 15 minutes prior to amniocentesis, persisted during the period when the amniotic fluid was actually drawn.
Given the observed stimulatory positive feedback effect of glucocorticoids on placental CRH release during pregnancy that was mentioned earlier, we hypothesize that fetal exposure to high cortisol levels might contribute to the high placental CRH levels found in prenatally stressed mothers. As previously mentioned, HPA axis dysregulation has been associated with both maternal prenatal depression and elevation in maternal CRH levels, which have been shown to correlate with measurable physiologic effects on the fetus (Baibazarova et al., 2012; ). Therefore, a plausible mechanism for “transmission” from mother to fetus calls into question placental regulation and its role in mediating the relationship between maternal psychological stress and fetal hormonal response. Thus, we will further explore fetal cortisol levels and, in particular, the regulatory role of the placenta in mediating fetal glucocorticoid exposure.
Placental Dysfunction and High Fetal Cortisol Levels as Possible Mediators
The placental enzyme 11β-hydroxysteroid-dehydrogenase-2 (11β-HSD2) inactivates cortisol by converting it to cortisone (Seth et al., 2015). 11β-HSD2 is expressed in high levels in various fetal tissues as well as in the placenta (La Marca-Ghaemmaghami et al., 2015). Its presence maintains fetal cortisol levels that are 5 to 10 times lower than maternal levels (La Marca-Ghaemmaghami et al., 2015). Epinephrine and norepinephrine, which are released during stress, inhibit 11β-HSD2, resulting in increased fetal exposure to cortisol (La Marca-Ghaemmaghami et al., 2015). Therefore, altered activity and expression of 11β-HSD2 may contribute to fetal overexposure to glucocorticoids, and ultimately may be implicated in preterm birth.
Seth et al. found that placental 11β-HSD2 gene expression was negatively associated with maternal depression and maternal anxiety as assessed during the first and third trimesters. The effect was more pronounced for both anxiety and depression in the third trimester. 11β-HSD2 gene expression was measured by sampling placental tissue immediately after delivery. However, this was a pilot study with a small sample size, so the results were not statistically significant and further studies with larger sample sizes are warranted (Seth et al., 2015).
La Marca-Ghaemmaghami et al. showed a correlation between sympathetic overactivation associated with chronic stress and decreased levels of 11β-HSD2. The researchers investigated the autonomic stress response to amniocentesis in 34 women and its association with amniotic fluid cortisol and E/(E+F), or cortisone/(cortisone+cortisol). E/(E+F) was used to assess fetoplacental 11β-HSD2 activity because it measures the proportion of cortisol converted to inactive cortisone. The ratio of low (LF) to high frequency (HF) heart rate variability was used to measure sympathovagal balance (La Marca-Ghaemmaghami et al., 2015). A low ratio indicates a parasympathetic-dominant state and thus lower stress levels (Taelman et al., 2008). LF/HF numbers obtained 50 minutes after amniocentesis were negatively associated with amniotic E/(E+F) and positively associated with amniotic cortisol levels (La Marca-Ghaemmaghami et al., 2015). A higher acute LF/HF response during amniocentesis was positively associated with amniotic E/(E+F) and negatively with amniotic cortisol levels. These results aligned with previous animal studies suggesting acute stress exposure upregulated 11β-HSD2 and lowered cortisol levels, while long-term stress yielded lower 11β-HSD2 levels and elevated amniotic cortisol (La Marca-Ghaemmaghami et al., 2015).
The sample size was relatively small in the La Marca-Ghaemmaghami study, and mood was evaluated at times temporally attenuated from the amniocentesis process itself. Additionally, the sample was skewed toward older mothers, so was not necessarily representative of pregnant women as a whole. Overall, this study suggests that while 11β-HSB2 may be upregulated in response to short-term maternal cortisol increases, over the long term, it is insufficient to handle maternal glucocorticoid overproduction. (La Marca-Ghaemmaghami et al., 2015).
O’Donnell et al. found prenatal state and trait anxiety were negatively associated with placental 11β-HSD2 mRNA expression. However, the negative association with prenatal depression was statistically insignificant. Depression and anxiety inventories were given the day prior to scheduled Caesarian section, and 11β-HSD2 mRNA expression was determined via placental samples obtained immediately after delivery. Self-reported anxiety and depression, however, probably is not best measured the day prior to a major surgery. (O’Donnell et al., 2012)
The relationship between decreased 11β-HSD2 activity and preterm birth has also been demonstrated in a study done in Finland which examined pregnant women who were high consumers of liquorice (more than 500 mg per week) (Strandberg et al., 2000). Liquorice synthesized from the rhizome of the plant Glycyrrhiza galabra contains glycyrrhizin which inhibits the action of 11β-HSD2 (Strandberg et al., 2000). The study found that women who were high consumers had significantly higher risk of preterm birth. In a multivariate analysis, high consumption of liquorice during gestation more than doubled the risk of the fetus being born prior to 38 weeks. It is established here that blocking the ability of 11β-HSD2 to convert cortisol to an inactive form exposes the fetus to higher levels of glucocorticoids, and increases the probability of preterm birth (Strandberg et al., 2000).
The issue of regulation of 11β-HSD2 levels in response to circulating glucocorticoids is of clinical significance not only because 18% of women suffer from depression antenatally (Andersson et al., 2004) but also because glucocorticoids are often administered in pregnancy (Marciniak et al., 2011), and the effects of maternal stress and/or circulating cortisol levels warrants additional research. Unfortunately, due to ethical considerations, fetal 11β-HSD2 can only be measured via amniocentesis or immediately after birth via cord blood or biopsy (Seth et al., 2015). Fetal cortisol levels are subject to similar constraints. Nevertheless, more tightly controlled studies assessing maternal stress longitudinally and more rigorous control of confounding factors are warranted.
The Role of Epigenetics
Epigenetic mechanisms involve changes in gene expression and regulation without alterations in the primary DNA sequence. In methylation specifically, DNA is epigenetically modified by the attachment of a methyl group to a cytosine residue, converting it to 5-methylcytosine (Monk et al., 2012). Increased rates of gene expression are generally associated with low DNA methylation of the promoter region of a gene (Non et al., 2014), whereas downregulation of expression is associated with high amounts of DNA methylation. As indicated by recent molecular biology research, environmental events can induce epigenetic regulation of gene activity, linking exposure to prenatal stress to changes to fetal behavior and physiological outcomes (Monk et al., 2012; see Figure 2). In particular, the effect of epigenetic modifications during pregnancy and their relationship to preterm birth is an area of intense investigation (Wadhwa et al., 2011). Thus, it is vital to explore gene-environmental interactions as a potential contributor to preterm births.
Non et al.’s (2014) study demonstrated that there is genome-wide DNA methylation in neonates exposed to untreated maternal depression during pregnancy. Utilizing the IIlumina Infinium HumanMethylation450 BeadChip for methylation analysis, they compared the methylation levels of umbilical cord blood of neonates of untreated depressed mothers, mothers with depression treated with selective serotonin reuptake inhibitors (SSRIs), and with non-depressed mothers. In their study, they determined that neonates exposed to untreated maternal depression or anxiety had notably high DNA methylation levels across 42 CpG sites compared to neonates of non-depressed women (Non et al, 2014), thus demonstrating differential degrees of methylation with increased stress.
Regulation of the glucocorticoid receptor gene NR3C1 is associated with alterations in the HPA axis stress response (Oberlander et al., 2008). A study conducted by Oberlander et al. demonstrated that prenatal exposure to heightened levels of NR3C1 methylation at a predicted NGFI-A binding site were correlated with elevated maternal depression alone or accompanied with anxiety during the third trimester (Oberlander et al., 2008). Oberlander et al. measured degrees of methylation of a CpG-rich region of the NR3C1 gene from cord blood mononuclear cells extracted from mothers in the study with untreated depression, mothers who were depressed but treated with SSRIs, as well as non-depressed mothers (Oberlander et al., 2008).
The degree of methylation was quantified utilizing bisulfite pyrosequencing (Oberlander et al., 2008). To allow for visualization of methylated cytosine cytosine residues at each of the CpG sites, genomic DNA was withdrawn from maternal and newborn leukocytes and exposed to bisulfite, which converted all unmethylated cytosine residues to thymidine (Oberlander et al., 2008). In order to confirm the functional effects of increased methylation of NR3C1 genes, salivary basal cortisol levels were measured in three-month-old infants whose cord blood was previously measured for methylation levels of the NR3C1 gene. Basal cortisol levels were measured in these infants preceding exposure to a non-noxious stressor, and subsequent levels were also gauged (Oberlander et al., 2008). The results found that elevated methylation of NR3C1 is correlated with heightened salivary cortisol stress response in 3-month-old infants, when accounting for differences in prenatal SSRI exposure, postnatal age, and pre- and postnatal maternal depression status (Oberlander et al., 2008). Since these results were found in 3-month-old infants, methylation of glucocorticoid receptor gene expression may be associated with elevated activity of the HPA axis in the fetus as well.
As discussed earlier, during fetal development, the placenta is known to play a crucial role in the maintenance of the intrauterine environment. It is extremely vulnerable to maternal distress, and changes in placental regulation may result in a dysregulation of gene expression (Pena et al., 2012). Fetal gene expression is found to deviate from control groups across the genome in untreated maternal depression (Non et al, 2014), and in some instances the increased expression of certain gene products such as 11β-HSD2 can be linked to preterm birth (Strandberg, 2000), elevation of the HPA axis in fetuses (Pena et al., 2012), and increased stress response in neonates (Oberlander et al., 2008).
Moreover, in a an animal study conducted with rats who were subjected to restraint stress, Pena et al. (2012) found that placental CpG methylation patterns were directly influenced by maternal stress. Utilizing real-time quantitative PCR to evaluate gene expression, prenatal stress was found to be correlated with an appreciable reduction in 11β-HSD2 mRNA levels, and within at specific CpG sites at the 11β-HSD2 gene promoter in the placenta there were marked increases in mRNA levels of DNA methyltransferase as well as increased rates of DNA methylation (Pena et al., 2012). In pregnancy, maternal stress and depression induces upregulation of transcription of DNA methyltransferases DNMT1 and DNMT3a, which leads leads to methylation and reduced expression of 11β-HSD2 (Pena et al., 2012), which is consistent with increased placental CpG methylation. This suggests that the genetic regulation of placental 11β-HSD2 levels presents a significant link between maternal depression and increased risk of preterm birth (Pena et al., 2012). These studies affirm the mounting evidence which suggest that maternal psychosocial state, via epigenetic pathways, can induce HPA axis dysregulation, alter placental function, and increase the risk of preterm birth.
Conclusion
Studies discussed in this review provide considerable evidence that altered neuroendocrine processes, specifically maternal-fetal HPA axis dysregulation, are implicated in the positive association between maternal stress and depression during pregnancy and the risk of preterm birth. After first describing the role that stress hormones play in the parturition response, observed correlations found in the literature between depression-associated levels of glucocorticoids, placental CRH activity, and preterm birth were highlighted. This may be a potential explanation for the premature triggering of the positive feedback response of placental CRH, which as mentioned earlier is thought to be largely involved in the initiation of labor (Wadhwa et al., 2011).
Furthermore, this paper explores the mechanism by which the placenta regulates maternal and fetal hormonal levels, as measured by amniocentesis. While study results are mixed, the results seem to indicate that 11β-HSD2 is crucial in maintaining a stable hormonal environment for the fetus, and that maternal stress symptoms are heavily intertwined with fetal health, as shown by the evidence linking maternal stress states to preterm birth. However, as amniocentesis declines in popularity (Obstet Gynecol 2004;103:1255-1260.), new methods will have to be improvised to measure fetal hormone levels. Amniocentesis only measures hormonal levels at one snapshot in time; however, because depression and stress are chronic mediators of neuroendocrine processes, studies where hormonal levels are sampled at many points are ideal for studying these complex relationships. Despite logistical difficulties, the interplay between maternal and fetal cortisol and 11β-HSD2 levels and their impact on preterm birth is an interesting area of research worthy of further study.
Although there are many studies that attempt to elucidate the link between epigenetic modifications in the fetus and maternal depression, the biological significance of these differential genetic expression patterns remains unclear. Studies that attempt to characterize the variance in expression of candidate genes, such as NR3C1, yield differing results across studies and different tissues (Non et al., 2014). While the literature has shown that epigenetics plays a role in fetal development, the evidence of both increased and decreased DNA methylation across different CpG sites, when comparing between neonates who were exposed to untreated maternal depression and neonates of nondepressed mothers (Non et al., 2014), suggests that epigenetic processes are multifaceted and dynamic. Furthermore, evaluating how these epigenetic changes translate into phenotypic expression and elucidating the biological significance of these differential gene expressions is an area that requires further investigation.
The studies reviewed in this paper demonstrate that maternal psychological status has an effect on fetal physiology and adverse birth outcomes. Thus it is important to monitor maternal mental health not only postpartum, but prenatally as well. This calls into question the effectiveness of current prenatal screening techniques that normally take into account family history, genetic screening, diet, and medical conditions such as diabetes, seizures, and hypertension to name a few. The prevalence of gestational diabetes, for example, can be as high as 9.2% (DeSisto et al., 2010); however, as previously mentioned, 16% of pregnant women in the United States have been diagnosed with clinical depression. Thus, prenatal screening could prove to be an opportune time to assess maternal psychological health and its implications on fetal health.
Although HPA axis dysregulation is strongly implicated in the mechanism behind preterm birth in prenatally depressed mothers, a direct causation is unconvincing. There are various other biochemical pathways involved in parturition, including immune, inflammatory and vascular processes, whose potential alterations might also contribute. More specifically, pro-inflammatory cytokines released during infection have been shown to stimulate prostaglandin and metalloprotease activity, two organic compounds other than placental CRH that participate in the parturition response. Moreover, maternal vascular diseases, such as underlying hypertension and preeclampsia, have also been linked to premature delivery (Wadhwa et al., 2001). This lends itself to the conclusion that preterm birth is likely a result of a multifaceted response, involving a dynamic integration of several biological pathways that can each be independently altered.