Inflammatory and neurodegenerative pathophysiology implicated in postpartum depression

Abstract

Postpartum depression (PPD) is the most common psychiatric complication associated with pregnancy and childbirth with debilitating symptoms that negatively impact the quality of life of the mother as well as inflict potentially long-lasting developmental impairments to the child. Much of the theoretical pathophysiology put forth to explain the emergence of PPD overlaps with that of major depressive disorder (MDD) and, although not conventionally described in such terms, can be seen as neurodegenerative in nature. Framing the disorder from the perspective of the well-established inflammatory theory of depression, symptoms are thought to be driven by dysregulation, and subsequent hyperactivation of the body’s immune response to stress. Compounded by physiological stressors such as drastic fluctuations in hormone signaling, physical and psychosocial stressors placed upon new mothers lay bare a number of significant vulnerabilities, or points of potential failure, in systems critical for maintaining healthy brain function. The inability to compensate or properly adapt to meet the changing demands placed upon these systems has the potential to damage neurons, hinder neuronal growth and repair, and disrupt neuronal circuit integrity such that essential functional outputs like mood and cognition are altered. The impact of this deterioration in brain function, which includes depressive symptoms, extends to the child who relies on the mother for critical life-sustaining care as well as important cognitive stimulation, accentuating the need for further research.

There are three types of postpartum dysphoric mood states: a common and transient maternity blues, a rare postpartum affective psychosis and a frequent postpartum depression (PPD) type (REF). PPD is defined as an episode of non-psychotic depression, according to the DSM5 criteria, occurring within one year of childbirth and affects ~20% of women within the first three months of the postpartum period (REF). This makes postpartum depression the most common psychiatric complication associated with pregnancy and childbirth as well as the most common complication of childbearing when compared to gestational diabetes or preterm birth. PPD remains difficult to study and its etiology is not known. Usually, a history of major depressive disorder (MDD) increases the risk of postpartum depression (REF). PPD is often undertreated, which can be detrimental for the child as untreated maternal depression can negatively impact both the development of the fetus and mother-infant bonding. Additionally, unmitigated maternal depression increases the risk of child anxiety and/or depression later in life, as well as promotes deleterious effects on cognitive and emotional development during infancy and childhood (REF). Indeed, mothers create an infant’s environment and are therefore determinant in shaping neuronal circuitries associated with a child’s discovery of the external world.

Although tremendous effort has led to the development of theories describing the neurobiological underpinnings of PPD, it is difficult to disentangle PPD neurobiology from MDD neurobiology (REF). Thus, it remains unknown how depressive symptomatology arises in humans (Wohleb et al., 2016), and if postpartum depression represents a distinct disease category differentiated from MDD (Cooper & Murray, 1998; Di Florio & Meltzer-Brody, 2015; Whiffen, 1992). There are numerous psychological and social risk factors associated with PPD, including stressors during pregnancy, lower social economic class, history of sexual abuse, and depression, which remains the strongest predictor of PPD.

Pregnancy presents a unique set of challenges to a woman’s health as her anatomy and physiology accommodate the health and development of the fetus. Women with predisposed sensitivities to certain environmental or physiological ques are especially vulnerable to the stress imposed during pregnancy. Depressive pathogenesis during the perinatal period is potentially influenced by risk factors such as increased psychosocial stress as well as the dramatic fluctuations in hormone levels that distinguish pregnancy and childbirth from a woman’s healthy baseline. Evidence from animal models suggests that prolonged stress or abnormal fluctuations in hormone levels can lead to chronic activation and subsequent dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis. Neuroendocrine hyperactivation has been shown to be associated with immune system dysregulation, damage to neurons, decreased neurotransmitter metabolism, disruptions to synaptic connectivity, and impairment of protective cellular repair processes (Das & Basu, 2008). These neurodegenerative sequelae disrupt neuronal communication, network integrity, and consequently, information processing, possibly resulting in alterations in healthy brain functions, including impairments in cognition (Salehinejad et al., 2017; Zuckerman et al., 2018) as well as depressed mood (Borsini et al., 2015). In this review, we provide evidence for the involvement of hormones, HPA axis activation, and inflammation in neurodegenerative processes theorized to underlie PPD pathophysiology.

Hormones, HPA axis activation, and cortisol-induced atrophy in postpartum depression

There is evidence that hormonal changes are frequent in women with a history of depression (Bloch et al., 2003). Furthermore, an increased risk of postpartum depression is also associated with a history of severe premenstrual syndrome (Buttner et al., 2013) and has been linked to changes in the serotonin transporter (Sanjuan et al., 2008). Decline in circulating estrogen after birth has long been suspected of playing a pathogenic role in PPD. Similarly, fluctuation of estrogen has been associated with depression (Douma et al., 2005). Estrogen has a neuroprotective role, enhancing serotonin function, and reducing inflammatory responses. Therefore, deprivation of estrogen enhances neurodegenerative changes caused by oxidative stress leading to loss of synaptic connectivity and cognitive decline. While studies have thus far been unable to provide evidence that serum sex hormones alone account for mood disturbances in PPD patient populations, evidence has accumulated that the dynamic interplay between estrogen and corticotropin-releasing hormone (CRH) may play a role in the emergence of depressive symptoms. Specifically, following birth, the major source of CRH and estrogen during pregnancy, the placenta, is no longer present, whereas the hypothalamic production of CRH is thought to be suppressed as a result of previous exposure to high levels of cortisol and concurrent estrogen deficiency (Swaab et al., 2005). The observed antidepressant effect of high-dose estrogen postpartum is believed to reestablish normal stress system secretion of CRH and norepinephrine (Gregoire et al., 1996). Additionally, progressively increasing levels of serum CRH-binding protein in the third trimester are thought to contribute to HPA axis hyperactivation (Magiakou et al., 1996).

Mutations in the MAOA and COMT genes, both of which have been linked to sex-specific differences in cortisol responses to a social stressor, have also been implicated in PPD (Bouma et al., 2012). Additionally, gene variants of Protein kinase C beta type (PRKCB), which is understood to be a regulator of the HPA axis indirectly through glucocorticoid receptors and CRH signaling, have also been associated with PPD (Costas et al., 2010).

Inflammation and cytokine involvement in postpartum depression

Pregnancy is associated with unique immunological responses to prevent the attack of the fetus by the maternal immune system. This translates into increased anti-inflammatory cytokines and decrease of pro-inflammatory cytokines during pregnancy in order to favor immunosuppression. After delivery, in contrast, this anti-inflammatory milieu shifts to a pro-inflammatory state, induced in response to the physical trauma and exertion associated with childbirth, and this increase of pro-inflammatory cytokines lasts at least for the first 72 h after delivery (Corwin et al., 2003; Hebisch et al.; Vassiliadis et al., 1998). It has been hypothesized that in conjunction with a hyporesponsiveness of the HPA axis, PPD is associated with an increase of pro-inflammatory cytokines, which have been shown to trigger depressive symptoms. Thus, women with a clinical history that included MDD exhibited elevated serum levels of IL-6 in the early puerperium compared to women without any prior MDD diagnosis (Maes et al., 2001) demonstrating that immune activation associated with childbirth may be exacerbated by depressive pathophysiological mechanisms. Additionally, serum IL-6 was found to positively correlate with measures of postpartum anxiety and depressive symptom severity in the early puerperium (Maes et al., 2000). Although data are missing on the role of inflammation in PPD, we speculate that similar dysregulations of the immune response as in MDD, including cytokine hyperactivation and cellular immunosuppression, exist in PPD. This has been reviewed extensively in Beurel et al. (2020). Changes in the regulators of T cells, leading to T cell apoptosis, have also been reported in depressed women before delivery (Krause et al., 2014). This is consistent with the decreased T cell function reported in depression (Beurel et al., 2020).

Pro-inflammatory cytokines induce activation of indoleamine-2,3 dioxygenase (IDO), which metabolizes tryptophan, the precursor of serotonin, into kynurenine which, in the context of neuroinflammation is metabolized into neurotoxic quinolinic acid by microglia cells (Jo et al., 2015; Steiner et al., 2011). Depression and anxiety symptoms in the early puerperium were shown to be directly linked to increased inflammation-induced degradation of tryptophan along the kynurenine pathway (Maes et al., 2002). Although inflammation likely contributes to the development of PPD, further studies are required to clarify the role of inflammation in PPD.

Cognitive impairment associated with postpartum depression

Neurogenesis is increased by stress early in gestation (Pawluski et al., 2015). Similarly, late pregnancy is associated with a decrease in cell proliferation, and this is also reversed by stress (Pawluski et al., 2011). However, this negative regulatory pressure imposed upon neuronal proliferation by stress is absent in postpartum females (Hillerer et al., 2014). The hormone essential for many maternal tasks, oxytocin, has been proposed to reduce the response to stress and activates serotonergic neurons (Yoshida et al., 2009). Oxytocin also regulates emotions and social interactions. And higher levels of oxytocin in midpregnancy predict PPD (Skrundz et al., 2011). Since, depression has been associated with reduced neurogenesis, and loss of neurons (Jo et al., 2015; Kreisel et al., 2014; Miller & Hen, 2015; Sapolsky, 2001; Whitney et al., 2009), it is possible that similar effects on neurogenesis and neurons occur in PPD. Consistent with this, women diagnosed with PPD have significantly reduced serum levels of brain derived neurotrophic factor (BDNF) compared to healthy women around 60 days after delivery (Gazal et al., 2012). As BDNF promotes neurogenesis, reduced BDNF would portend a decrease in neurogenesis in PPD.

Additional mechanisms that drive neuronal cell death have been proposed for PPD. For instance, elevated levels of monoamine oxidase (MAO) have been measured in patients diagnosed with PPD (Sacher et al., 2015). MAO carries out oxidative deamination of monoamine neurotransmitters and metabolism of serotonin. Byproducts of this reaction, including hydrogen peroxide, have been shown to cause overproduction of reactive oxygen species (ROS) which disrupt mitochondrial function, cause damage to cell physiology, and can lead to neuronal cell death (Vaváková et al., 2015). Overall, PPD appears to have a negative impact on neurons.

PPD experienced by the mother also often has negative consequences for the neurodevelopment of the child, likely leading to impaired cognitive development, including poorer language acquisition and use as well as underdeveloped social skills (Brand & Brennan, 2009; Stein et al., 2014). This suggests that maternal PPD negatively impacts multiple neural pathways in the child. However, more studies are needed to understand, at the neurobiological level, the consequences of PPD for infant neurodevelopment.

The loss of neurons leading to possible improper neuronal connections both in the mother and infant raises the question of whether PPD is associated with a neurodegenerative process. It is clearly established that there is a correlation between the length of time depressive symptoms remain untreated and hippocampal atrophy (Sheline et al., 2003), suggesting that progressive neurodegeneration may be involved in depressive symptomatology, and therefore in PPD. Additionally, the link between major depressive disorder and cognitive deficits is well-established (Bora et al., 2013; Bremner et al., 2000; Gazzaniga, 2014; Gonda et al., 2015; Miller & Cohen, 2001; Salehinejad et al., 2017; Sapolsky, 2001; Sheline et al., 2003; Zuckerman et al., 2018). However, the exact relationship between postpartum depression and cognitive impairment remains less clear. Although deficits in working memory have been measured in women exhibiting symptoms of both antepartum and postpartum depression (Hampson et al., 2015), other studies indicate that cognitive impairments may be linked to the peripartum period more broadly, even in women without depressive symptoms (Shin et al., 2018; Zheng et al., 2018), suggesting factors underlying cognitive impairment during and after pregnancy, while not being sufficient alone, may contribute to the subsequent development of depressive symptoms postpartum. Nevertheless, any degree of cognitive impairment, whether in the context of concurrent depressive symptoms or not, raises the alarm for the possibility of neurodegenerative processes at play. This is particularly true in the context of inflammation and oxidative stress, which have been shown to have detrimental effects on neurons and promote neurodegeneration (Borsini et al., 2015; Das & Basu, 2008; Jo et al., 2015; Kohman & Rhodes, 2013; McAfoose & Baune, 2009; Whitney et al., 2009; Yirmiya & Goshen, 2011).

Although PPD is sometimes considered as “non-classical” depression occurring around childbirth (Pitt 1968), PPD remains in the DSM-IV and DSM-5 under the category of major depressive disorder with a specifier about the onset of the disease occurring during pregnancy or after childbirth (Wisner et al 2010; DSM-5, 2013). It is indeed difficult to distinguish between the two diagnoses, due to various confounding factors. For instance, the definition of the postnatal period varies between individuals, and some of the depressive symptoms overlap with common infant care challenges (fatigue, lack of appetite, disrupted sleep, etc.). In addition, the Edinburgh Postnatal Depression Scale used to diagnose PPD seems to lack specificity and sensitivity and seems better at detecting MDD than PPD. This is not a surprise when 78% of PPD cases are relapses from MDD (Holden 1996). However, PPD is unique from MDD in the sense that PPD has been attributed to the gonadal hormonal changes accompanying childbirth (drastic elevation of hormones during pregnancy which drops dramatically after child birth) (Bloch et al 2003; Bloch et al 2005)*. This suggests that women sensitive to mood-destabilizing effects of reproductive hormone fluctuations might be more prone to PPD. Similarly, the nature of the psychological stress associated with delivery and child care might underlie the difference between PPD and MDD. Nonetheless, treatment options for PPD and MDD are similar and include SSRIs, psychotherapy, hormonal therapies, TMS and ECT (for review Batt et al 2020). SSRIs, however, seem to take longer to induce antidepressant effects in PPD than in MDD (Wisner et al 1999).

To conclude, PPD has similarities with depression in its manifestation yet the mechanisms contributing to PPD remain largely undetermined. This relative lack of knowledge may stem from difficulty in conducting clinical trials in this vulnerable population. Nevertheless, currently available evidence suggests an undercurrent of neurodegenerative processes may fuel PPD pathophysiology. Additionally, the impact of maternal PPD on the infant is not negligible. This outsized human impact highlights the importance of uncovering therapies for PPD.