The early-life environment is one of the most important factors affecting life-long health (Anand, 2000; van den Bergh et al., 2005; Lupien et al., 2009; Boksa, 2010; Strüber et al., 2014). In humans, early-life stress is associated with a preterm birth and a low birth weight, and can prime the neonate for further complications later in life that include psychiatric disorders, aged-related cognitive dysfunction, obesity, and hypertension (Barker et al., 1989; Fowden et al., 2005; Lemaire et al., 2006; Lahiri et al., 2009; Strüber et al., 2014). Animal studies also suggest that exposure to stressors or steroids during early-life alter the programming of the hypothalamic-pituitary-adrenal (HPA) axis, neurobehavior, and neuroimmune systems (Matthews, 2000; Mueller and Bale, 2008; Lupien et al., 2009; Brunton and Russell, 2010; Chen and Zhang, 2011; Lai and Huang, 2011; Strüber et al., 2014). Epigenetic modification has gained increasing attention in recent years because of its connection with early-life adversities (Weaver et al., 2004; Meaney et al., 2007; Mueller and Bale, 2008; Chen and Zhang, 2011; McClelland et al., 2011a,b; Murgatroyd and Spengler, 2011; Lucassen et al., 2013; Rabbe and Spengler, 2013). On the other hand, stress during development can have a significant epigenetic impact on the brain, and this relationship is bidirectional (Hunter, 2012).

Early-life stressors include prenatal, postnatal, and adolescence stress (Lupien et al., 2009; Schmidt, 2010). For example, in humans, early-life stress can include prenatal stressors such as exposure to exogenous glucocorticoids, maternal infection (King et al., 2005; Sørensen et al., 2009; Jenkins, 2013), and birth complications, as well as postnatal stressors such as exposure to exogenous glucocorticoids, maternal postpartum depression, loss of a parent, exposure to family conflict and violence, neglect, or physical maltreatment (De Bellis, 2002; King et al., 2005; Frodl et al., 2010). Both prenatal and postnatal stress can increase the likelihood of seizures in early life (Joels, 2009; Koe et al., 2009) and epileptogenesis in later life. This article focuses only on the influences of prenatal stress and postnatal stress.

The hippocampus develops primarily during the fetal period in both rodents and primates (Seress et al., 2001; Khalaf-Nazzal and Francis, 2013). The limbic system, which includes the hippocampus, amygdala, and anterior cingulate cortex are already formed during the third and fourth month. Dentate gyrus forms at late stages of embryogenesis, however small numbers of dentate gyrus cells are formed from mid-embyrogenesis making temporal matching and connectivity of cells from other hippocampal subfields (Deguchi et al., 2011). Rodents and primates differ in the timing at which the majority of the dentate granule cells are produced; however, both rodents and primates produce ~85% postnatally (Bayer, 1980a; Rakic and Nowakowski, 1981). A similar percentage of cornus ammonis (CA) 1–3 subfield neurons are produced during the last days of gestation in rodents, and during the first half of pregnancy in primates (Bayer, 1980b; Rakic and Nowakowski, 1981). The hippocampal subfields can be recognized with distinct molecular markers from embryonic stages (Khalaf-Nazzal and Francis, 2013).

In the rodent, maturation and full differentiation of the hippocampal formation takes place during early postnatal periods (Avishai-Eliner et al., 2002). During the first postnatal weeks, neuronal birth, differentiation, and migration are ongoing (Altman and Bayer, 1990; Gould and Cameron, 1996). Neurogenesis of granule cells peaks during the second week of life in rodents (Bayer, 1980a), and during the third month in humans (Seress et al., 2001). In addition, synaptogenesis and the establishment of enduring connectivity patterns continue for weeks in the rodent, and for years in humans (Avishai-Eliner et al., 2002).

Glucocorticoids are released from the adrenal glands in response to stress, readily cross the blood-brain barrier, and activate hippocampal glucocorticoids receptors (McEwen, 1998). Glucocorticoids interact with their receptors in multiple target tissues, especially the HPA axis. Glucocorticoids act via two intracellular receptors, the glucocorticoid receptor (GR) and the mineralocorticoid receptor (MR) to regulate gene transcription. In addition, glucocorticoids can change neural function via rapid nongenomic actions. GR and MR differ in ligand affinity and distribution (de Kloet et al., 2005): GR has a lower affinity than MR has, and therefore are more frequently occupied when corticosterone levels increase (de Kloet et al., 2005). The actions of glucocorticoids depend on the functionality of the balance between GR and MR in the brain (de Kloet et al., 2005).

There is a distinct ontogenic profile for GR and MR in the fetal rat brain (Diaz et al., 1998). GR mRNA is present in the anterior hypothalamus, hippocampus, and pituitary by gestational day 13 (Diaz et al., 1998), whereas MR mRNA is present in the hippocampus by gestational day 16 and the hypothalamus by day 17 (Diaz et al., 1998). GR and MR in the rat fetal brain are low throughout gestation, but increase rapidly after birth, consistent with the postnatal development of the brain in the rat (Diaz et al., 1998).

During pregnancy, the mother's HPA axis undergoes major changes (Lindsay and Nieman, 2005). Cortisol secretion increases steadily through gestation (Jung et al., 2011); thus, the normal physiological responses to stressors and the cortisol awakening response (i.e., basal HPA activity) are attenuated (Lindsay and Nieman, 2005). For most of the pregnancy, the baby and mother share a common corticotrophin-releasing hormone (CRH)-adrenocorticotropic hormone (ACTH)-cortisol axis (McLean et al., 1995).

By the end of the first week of life (Bohn et al., 1994; Vazquez et al., 1998), the number of MRs reaches adult levels. The number of GRs present during the first few week of life, however, is ~30% of adult levels, but approach adult levels after ~30 days of life. Both GR and MR are highly expressed in the developing brain, and have different and complex ontogenies that allow intricate brain development.

Between postnatal day 4 and 14, neonatal rat pups have low basal corticosterone levels and the corticosterone response to stressors is blunted, which constitutes the so-called stress hyporesponsive period (SHRP) (Levine, 2005). However, disruption of normal maternal behavior in rat during the SHRP can influence HPA axis development. In humans, the HPA axis is highly reactive and labile during early infancy, but organizes between 2 and 6 months of age through interactions between the infant and caregiver. The quality of caregiving that the infant receives predicts the infant's ability to self-regulate later in life. Sensitive caregiving is associated with better self-regulatory abilities and optimal functioning of the child's HPA system (Gunnar and Cheatham, 2002; Gunnar and Donzella, 2002).

Epileptogenesis is a process through which the normal brain develops epilepsy, and the hippocampus is implicated in the pathogenesis of both the initiation and propagation phases (Pitkänen and Lukasiuk, 2011). Mesial temporal lobe epilepsy (MTLE), the most common focal intractable epilepsy, is thought to be a multi-stage process of increasing epileptogenesis commencing in early life. The ongoing process of epileptogenesis and the course of epilepsy might be negatively influenced by the stress associated with the disease itself (Joels, 2009; Sawyer and Escayg, 2010). As a result, a negative loop might occur in which stress promotes epileptogenesis in predisposed individuals or lowers seizure threshold in epilepsy patients, thereby increasing the likelihood of exposure to stress, which in turn exacerbates the disease. Epidemiological data implicate stress in the cause of epilepsy and seizures in both children and adults (Temkin and Davis, 1984; Swinkels et al., 1998; Bosnjak et al., 2002).

Stress is a natural factor that may exacerbate or trigger seizures (Novakova et al., 2013; van Campen et al., 2013). HPA-related stress hormones, especially glucocorticoid and CRH, can affect excitatory and inhibitory processes in brain areas that are critically involved in seizure generation. Glucocorticoid exposure can alter plasticity in the hippocampus through increasing extracellular glutamate levels and calcium conductance (either voltage- or ligand-gated), alter expression of N-methyl-D-aspartate (NMDA) receptor subunits, and reduce glial uptake of glutamate, and thus, facilitate epileptiform discharges and seizures in animals. Glucocorticoids facilitate epileptiform discharges and seizures in animals. CRH is expressed in interneurons in both the developing and adult hippocampus and is released during stress (Sakanaka et al., 1987; Chen et al., 2001). Both glucocorticoids and CRH are important hormones that regulate the stress response and may contribute to seizure-induced loss of neurons, dendritic spines, and branching if it persists for a prolonged period (Ribak and Baram, 1996; Chen et al., 2012).

Negative life events and stress sensitivity are linked with childhood epilepsy (van Campen et al., 2012, 2013). In addition, epidemiological data implicate stress in the causation of epilepsy and seizures in children (Bosnjak et al., 2002). Specifically, early-life stress might create an enduring vulnerability to limbic epilepsy through altering glucocorticoids (Kumar et al., 2007), HPA axis (Joels, 2009), CRH (Baram and Hatalski, 1998), inflammation (Vezzani et al., 2013), membrane receptors such as gamma-aminobutyric acid (GABA) (Reddy, 2013), NMDA (Olney et al., 1991), and 2-amino-3-(3-hydroxy-5-methylisoxazol-4-yl) propionic acid (AMPA) receptors and neurotransmission (Rogawski, 2013), cellular electrophysiology, such as long-term potentiation (LTP) and long-term depression (Blaise et al., 2008), limbic area structures (Wong and Guo, 2013), and neuronal cell proliferation and neurogenesis (McCabe et al., 2001).

Prenatal stress is an important programming factor in brain development and function. A recent cross-sectional study indicated that 6% of pregnant women reported high levels of psychological stress during their pregnancies that resulted from conditions including depression, panic disorder, or domestic violence (Woods et al., 2009). Talge et al. reviewed several prospective studies related to prenatal maternal stress, and found a substantial number of emotional/behavioral problems in children, including attention deficit hyperactivity disorder, anxiety, and language delay, that were attributed to prenatal stress or anxiety in ~15% of the subjects (Talge et al., 2007).

Beck and Gavin treated pregnant mice with beta-2-thienylalanine solvent or a sham injection on gestational days 10–12. Audiogenic seizures were tested on postnatal day 23. An increase in audiogenic seizure frequencies were observed in injected mice, irrespective of the nature of the injected substance. This finding suggested that the act of manipulation, rather than the test substance, caused stress and increased seizure propensity (Beck and Gavin, 1976). Frye and Bayon exposed rats to 20 min of restraint stress toward the end of their pregnancy (Frye and Bayon, 1999). They found that the prenatally stressed offspring had more partial seizures and tonic-clonic seizures with long durations than did control rats. Edwards et al. examined how stress exposure at different times during gestation might affect later limbic system excitability and the propensity to develop epilepsy (Edwards et al., 2002). Pregnant dams were restrained under bright light for 45 min, three times a day during either early gestation (gestational days 5–12) or mid-late gestation (gestational days 12–20). Offspring of the stressed dams were then tested as an infant at postnatal day 10 or as adults, and were compared with offspring from non-stressed dams. Outcome measures assessed were the stimulation-induced seizure threshold, after-discharge threshold, and the rate of seizure development using electrical hippocampal kindling. Both prenatal stressors significantly lowered after-discharge threshold in pups, but this effect appeared to diminish by adulthood in the early gestational stress group. In addition, mid to late gestational stress accelerated kindling rates in all infant offspring and in adult males, but had no effect in adult female rats. Notably, Young et al. administered dexamethasone or betamethasone on gestational days 15–18, and tested the seizure threshold and kindling parameters (Young et al., 2006). They found prenatal betamethasone treatment increased seizure threshold for both models. Prenatal dexamethasone treatment increased kindling threshold, but not seizure threshold. Kindling rate was unaffected by either glucocorticoid treatment (Young et al., 2006). Velisek showed prenatal exposure to betamethasone decreased postnatal susceptibility to flurothyl-induced clonic seizures but not to kainic acid-induced seizures. Prenatal hydrocortisone decreased postnatal weight but did not affect seizure susceptibility (Velíšek, 2011). In their subsequent work, Yum et al. demonstrated that prenatal restraint stress (2 × 45 min) in rats on gestational day 15 would increase susceptibility to spasms on postnatal day 15 (Yum et al., 2012).

Shang et al. showed an association between the onset risk of infantile spasms and the degree of maternal stress (Shang et al., 2010). However, in a population-based cohort study in Denmark, Li et al. studied children who were hospitalized because of epilepsy and born to women who had lost a close relative during pregnancy 1 year before pregnancy (Li et al., 2008). In this study, no association was found between this particular form of prenatal stress and the risk of epilepsy.

Indirect evidence links prenatal stress and an increased likelihood of childhood seizures in children with autistic disorder. Minshew et al. pointed out that epilepsy is found in about one-third of patients with autistic disorder, a disorder related to prenatal stress (Kinney et al., 2008), compared with a prevalence of only 2–3% in the general population (Minshew et al., 2005). Table 1 summarizes the current rodent studies regarding the impact of prenatal stress on seizure occurrence and epileptogenesis.

Early-life adversity (childhood abuse and neglect, loss of parents, or extreme poverty) occurs worldwide and are all too common in the lives of children (Jones, 2008; Sandberg and Rutter, 2008). In the Dunedin Study birth cohort of 1037 children, followed prospectively for 32 years, maltreatment includes maternal rejection, harsh discipline, sexual abuse, physical abuse, and disruptive caregiver changes (Danese et al., 2009). For each child, the cumulative index counts the number of maltreatment indicators experienced during the first decade of life; 63.7% of children experienced no maltreatment, 26.7% experienced one form of maltreatment, and 9.6% experienced two or more forms of maltreatment (Danese et al., 2009). Clinical evidence from life-course epidemiology study points to the importance of early life experiences in shaping adult disease (Poulton et al., 2010).

Edwards et al. investigated the effects of maternal separation on kindling epileptogenesis utilizing a relatively benign separation protocol that included 60 min on postnatal days 4 and 5 (Edwards et al., 2002). The comparison group included the other littermates, which were briefly handled but not removed from the mother. This postnatal manipulation had no effect on after-discharge threshold or rapid hippocampal kindling rates when assessed at 2 weeks of age.

To investigate the effects of maternal separation on the long-term consequences of early-life status epilepticus, Lai et al. tested whether maternal separation for 1h affected the long-term sequelae of emotional disorders following seizure early in life (Lai et al., 2006). Lai et al. used maternal separation that involved 1h of isolation daily during postnatal days 2 and 9, and used lithium-pilocarpine-induced status epilepticus on postnatal day 10 rats. As adults, anxiety-related behavior was assessed using the elevated plus maze test and seizure susceptibility was assessed by pentylenetetrazol-induced seizures. Rats exposed to maternal separation and seizures demonstrated a reduced pentylenetetrazol threshold for seizure induction compared to non-handled rats or rats exposed to isolation or seizure alone. Metyrapone (a corticosterone synthesis inhibitor) treatment prior to seizure did not reverse this enhanced excitability, indicating a partial role of glucocorticoids in this context. Salzberg et al. examined the effects of maternal separation on limbic excitability and the development of amygdala kindling (Salzberg et al., 2007). Postnatal stress was induced by separating pups from their mothers for 180 min daily from postnatal days 2–14. The comparison condition was mother and pup separation for 15 min per day over the same period, an exposure referred to as early handling. At 8 weeks of age, equivalent to young adult life, rats were tested for the after-discharge threshold and subjected to rapid amygdala kindling. Rats exposed to early-life stress exhibited significantly lower seizure thresholds and an accelerated rate of kindling, compared to early handled rats. These effects on limbic excitability and epileptogenesis were specifically observed in female rats, whereas males did not demonstrate changes in epilepsy outcomes, despite demonstrating increases in anxiety-like behavior. Using the rat amygdala-kindling model, Kumar et al. demonstrated that early-life stress induced by maternal separation accelerates the progression of focal limbic seizures to secondary generalized convulsive seizures in adult rats (Kumar et al., 2011). Desgent et al. used a two-hit model of TLE characterized by two early-life insults: a freeze lesion-induced cortical malformation on postnatal day 1, and a prolonged hyperthermic seizure on postnatal day 10 (Desgent et al., 2012). They demonstrated that after both insults, females did not develop MTLE while all males did. This correlated with a rise in corticosterone levels on postnatal day 1 following the lesion, but only in males. Their data demonstrated sexual dimorphism in the long-term vulnerability for developing epilepsy in the lesion plus hyperthermia animal model of MTLE, and suggested that the response to early-life stress at postnatal day 1 contributed significantly to epileptogenesis in a sex-specific manner (Desgent et al., 2012). Ali et al. demonstrated changes in firing patterns in thalamocortical and hippocampal regions resulting from both maternal separation and amygdala kindling, which might reflect cellular changes underlying the enhanced vulnerability to kindling in rats that had been exposed to early-life stress (Ali et al., 2013).

Similarly, Leussis and Heinrichs cross-fostered El pups to CD-1 dams because CD-1 dams exhibit a higher quality of maternal care than El dams. El pups raised by CD-1 dams experienced delayed seizure onset and reduced seizure frequency, suggesting that early-life environment can play an important role in shaping the adult seizure phenotype (Leussis and Heinrichs, 2009). It should be noted the El mouse model has not been verified for its effect on early-life stress. In addition, El pups raised in a biparental environment with both the El dam and sire attending the pups received more parental attention than El pups raised by only the El dam, yet they showed an earlier development of seizures (Orefice and Heinrichs, 2008). Together, early-life environment can interact with a genetic predisposition to shape the future seizure phenotype.

Van Campen et al. studied the effect of stress on seizure frequency in childhood epilepsy. They found stress sensitivity was reported in half of the children with epilepsy. They suggested that experiencing negative life events might cause a larger response to daily stressors, thereby increasing the likelihood to induce epileptic activity in childhood (van Campen et al., 2012). Table 2 summarizes the current rodent studies regarding the impacts of postnatal stress upon seizure occurrence and epileptogenesis.

Epigenetic modifications regulate gene expression without altering the DNA sequence. Epigenetic changes involve DNA methylation at cytosine-guanine sequences-CpG sites, histone posttranslational modifications (histone methylation, acetylation, phosphorylation, ubiquitylation, sumoylation, and propionylation), and microRNAs (Gräff et al., 2011). Epigenetic mechanisms control nucleosome spacing and how they are condensed, which subsequently determines gene activity. Briefly, chromatin exists in an inactivated and condensed state (heterochromatin) that prevents gene transcription, but when activated to an open state (euchromatin), genes can be transcribed.

It is now clear in both humans and animals that glucocorticois and stress have a significant epigenetic impact, and the relationship between the stress response and epigenetics in the brain is bidirectional (Hunter, 2012). Epigenetic alterations have become especially attractive to researchers in recent years, as increasing evidence indicates that they can be induced by physical and social exposure early in life (Meaney et al., 2007). For some neurobiological disorders, exposure to environmental agents during early developmental stages can epigenetically disturb gene regulation in a long-term manner and cause significant pathological manifestations later in life. This process is the latent early-life associated regulation model by Lahiri et al. (2009).

Epigenetic dysregulation has been associated with prenatal IUGR and disease in both humans and rodents (Baserga et al., 2007, 2010; Friso et al., 2008). Prenatal stress can cause increased DNA methylation in the frontal cortex and hippocampus (Mychasiuk et al., 2011; Matrisciano et al., 2013) and a lower DNA methyltransferase 3a immunoreactivity in the dentate gyrus in offspring (Sierksma et al., 2013).

Variations in maternal care in the rat result in differences in hippocampal development and synaptic plasticity in the offspring (Macrì and Würbel, 2006). Observational studies provide evidence for two forms of maternal behaviors during the first week of lactation: licking/grooming (LG) and the arched-back nursing (ABN) posture (Liu et al., 1997; Francis et al., 1999). In the rat, the adult offspring of high LG-ABN mothers show increased hippocampal GR expression and enhanced glucocorticoid feedback sensitivity compared to animals reared by low LG-ABN mothers (Liu et al., 1997; Francis et al., 1999). In addition, adult offspring of high LG-ABN mothers exhibited modest HPA stress responses compared to animals reared by low LG-ABN mothers (Menard and Hakvoort, 2007). In hippocampus, offspring from high LG-ABN mothers had hypomethylation of CpG dinucleotides in the exon 1₇ GR promoter sequence, and increased histone acetylation that might account for higher transcription of the GR gene (Weaver et al., 2004). The maternal effect is mediated by enhanced serotonergic activity and an increased expression of NGFI-A, which binds the exon 1₇ GR promoter sequence (Weaver et al., 2007). Cross-fostering experiments showed a causal relationship between maternal care and changes in the exon 1₇ GR promoter methylation (Weaver et al., 2005, 2006).

As stated above, epigenetic modifications underpin the programming effects of early-life stress. Interestingly, a wealth of evidence indicates that dysregulation of epigenetic mechanisms occurs in several human epilepsy syndromes. Epigenetic mechanisms can influence the acute deployment of genes resulting from seizures themselves or can have gradual effects on the steady-state expression profile of candidate genes that persist into epilepsy. Epigenetic modifications can affect seizure and epilepsy in several ways (Lubin, 2012; Roopra et al., 2012).

Firstly, histone acetylation is involved in epileptogenesis in human epilepsy patients. Seizure activity results in gene expression changes, including alterations in mRNA levels for glutamate receptor 2 and BDNF, the two well-characterized epileptogenesis-related genes. Of interest, histone acetyltransferase-mediated increases in histone acetylation levels at the promoter regions of the glutamate receptor 2 and BDNF genes have been shown to correlate with their gene expression changes following seizures in an experimental animal model (Huang et al., 2002b).

Secondly, DNA methylation has been highlighted as a component of the methylation hypothesis of epileptogenesis (Kobow and Blumcke, 2011). DNA methyltransferase enzymes 1 and 3a specifically, were increased in neurons from the temporal neocortices of 25 MTLE patients (Zhu et al., 2012). Using a rat model of MTLE, Williams-Karnesky et al. identified an increase in hippocampal DNA methylation that correlates with an increased DNA methyltransferase activity, disruption of adenosine homeostasis, and spontaneous recurrent seizures. To test the effects of adenosine, they used bioengineered silk implants to deliver a defined dose of adenosine over 10 days to the brains of epileptic rats (Williams-Karnesky et al., 2013). Adenosine implants reversed DNA hypermethylation seen in the epileptic brain, inhibited sprouting of mossy fibers in the hippocampus, and prevented the progression of epilepsy for at least 3 months (Williams-Karnesky et al., 2013).

Thirdly, transcription factors are involved in epileptogenesis in human epilepsy patients. Repressor element-1 silencing transcription factor and neuronal restrictive silencer factor serve to repress gene expression through dynamic recruitment of epigenetic complexes (Qureshi and Mehler, 2009). Of interest, repressor element-1 silencing transcription has been implicated in the regulation of several epileptogenesis specific factors, including growth factors, neurotransmitter receptors, ion channels, circuit excitability, and neurogenesis (Huang et al., 1999; McClelland et al., 2011a,b; Roopra et al., 2012).

Fourthly, methyl-CpG-binding protein 2 can regulate neuronal activity and is itself controlled by activity (Roopra et al., 2012).

Taken together, early-life stress can prime seizure occurrence and increases epileptogenesis. In addition, epigenetic modification is a shared pathogenic substrate of early-life stress and epilepsy.

Seizure is one of the most common pediatric emergencies, with the highest incidence in the first year of life. Animal studies have demonstrated early-life seizures differ from adult seizures by the seizure behaviors, the electroencephalogram features, and their consequences. Notwithstanding the higher susceptibility to seizures, the immature brain is less vulnerable to seizure-induced injuries than the mature brain (Dube et al., 2001; Holmes and Ben-Ari, 2001; Huang et al., 2012). However, under some circumstances seizure in the immature brain can cause permanent brain damage (Dube et al., 2006).

For humans, most early-life seizures occur in premature and sick neonates (Scher et al., 1993; Miller et al., 2002; Scher, 2003) who are hospitalized and separated from their mothers, and thus, under stress (Field, 1994; Anand, 2000). Reciprocally, early-life stress may prime the occurrence of seizures and act via glucocorticoids, thereby potentiating the excitotoxic effects of concurrent neurological insults (Sapolsky, 1996), such as seizure (Huang et al., 2002a; Lai et al., 2006).

As stated above, early-life stress can prime the seizure occurrence and subsequent epileptogenesis. Currently, more attention is being paid to the effect of early-life stress on adult-onset seizure; however, little work has focused on the effect of early-life stress on the early-life seizure (Beck and Gavin, 1976; Edwards et al., 2002; Lai et al., 2006; Young et al., 2006; Velíšek, 2011; Yum et al., 2012). Indeed, to study the coexistence of early-life stress and early-life seizures is of both experimental and clinical importance.

Barker et al. noted that low birth weight was associated with an increased risk of adverse outcomes in adulthood, such as coronary heart disease, stroke, high blood pressure, and type 2 diabetes (Barker et al., 1989). Gluckman et al. proposed the concept of DOHaD by observing the enduring effects of the fetal environment on physical health and disease in adulthood. The process of fetal programming or developmental plasticity is one of the core assumptions of DOHaD (Gluckman et al., 2007). Gluckman et al. use the concept of predictive adaptive responses to describe the developing organism by making phenotypic responses during development to obtain an adaptive advantage (Gluckman et al., 2005). The fetus will predict and make adaptive responses to a broad range of environmental cues to aid fitness and survival in later life. If the prediction is correct, then there will be a good match between the phenotype adopted and the environment in which the organism will later live. If the prediction is poor, there will be a mismatch between the environment experienced and the phenotype induced. The authors propose that developmental mismatch triggers or exacerbates certain diseases and provide a useful explanation for the DOHaD phenomenon (Gluckman et al., 2005, 2007). Furthermore, the notion of epigenetic modifications is applied to the DOHaD approach (Waterland and Michels, 2007). The DOHaD approach has become so popular that an international society has been formed, and this society is actively promoting research and collaboration in this area.

Figure 1 depicts the path form early-life adversity to long-term neuropsychiatric disorders, along with the underlying molecular and cellular mechanisms and epigenetic modifications, with a match or mismatch adaptation that leads to the final outcome.

Early-life stress can elicit detrimental effects on hippocampal development by altering the HPA axis, neuroplasticity, and behavior. Developmental plasticity allows an organism to adapt to environmental changes in the critical stages of early life. As highlighted in this review, early-life stress programs the development of the HPA axis, exerts profound effects on neural plasticity, primes seizure occurrence, and increases epileptogenesis. Epigenetic modifications play an important role in both early-life stress and epilepsy.

A number of important points made throughout the manuscript are reinforced here. Reducing damage done by prenatal and postnatal stress may help reduce the cost of treating adult diseases. Protecting pregnant mothers from harmful stress exposure and supporting programs to reduce stress or anxiety during pregnancy might lead to improvements in the health and well-being of their children later in life. Ideally, intervention and prevention should be achieved before pregnancy begins. In terms of postnatal stress, psychosocial interventions in early life can affect brain development and thereby benefit children at risk. Other perinatal adversities such as perinatal infection, nutritional disorders, and toxin exposures must be cautiously avoided and treated. The potential therapeutic value of pharmacological agents, such as CRHR₁ antagonists, MR and GR antagonists should be explored.

Recently, an increasing number of studies have shown that early-life stress primes seizure occurrence and increases epileptogenesis. An increased understanding of the link between early-life stress and epilepsy could improve the care and treatment of patients with epilepsy, while also allowing better management of other stress-related neurological disorders.

In the future, we will need to better determine the developmental windows during which preventative or therapeutic interventions can reverse the adverse effects of developmental programming. It will also be important to better understand stress biomarkers, especially epigenetic biomarkers. An increasing number of studies have provided clues as to how early-life stress induces changes at the cellular, molecular, and epigenetic levels. Continued progress on these fronts will provide great insight into disease mechanisms, in turn leading to the potential identification of novel targets for therapy and prevention.