Early-life stress and cognitive outcome
- First Online:
- Cite this article as:
- Hedges, D.W. & Woon, F.L. Psychopharmacology (2011) 214: 121. doi:10.1007/s00213-010-2090-6
- 749 Views
Early-life stress is associated with later neuropsychiatric illness. While the association between early-life stress and brain development is well recognized, relatively few studies have examined the association between exposure to early-life stress and cognitive outcome.
The objective of this paper is to examine the association between early-life stress and cognitive outcome in animal models and humans.
In this article, we review alterations in cognitive function associated with early-life stress in animals and then discuss the association of early-life stress and cognitive function in humans.
Findings suggest that early-life stress is associated with abnormal cognitive function in animals and humans. Furthermore, cognitive deficits associated with exposure to early-life stress in humans may persist into at least early adulthood, although animal models of enriched environments and studies of children adopted from institutionalized care into foster families suggest that certain social factors may at least partially reverse cognitive deficits following exposure to early-life stress.
Exposure to stress in early life may be associated with later deficits in cognitive function.
KeywordsEarly-life stressCognitionCognitive functionMemoryChildhood abuseNeglectPosttraumatic-stress disorderHippocampusNeuropsychology
Considerable evidence implicates abnormal development and events such as malnutrition in fetal and early life as risk factors for a variety of adult diseases (Harding 2001; Shonkoff et al. 2009), such as breast cancer (Ahlgren et al. 2004; McCormack et al. 2003; Titus-Ernstoff et al. 2002), non-insulin dependent diabetes (Harding 2001), coronary-artery disease (Barker 1995), and obesity (Lissau and Sorensen 1994). Furthermore, the socioeconomic environment during childhood is associated with adult stroke, stomach cancer, coronary-artery disease, lung disease (Smith et al. 1998), liver disease, and possibly chronic pain (Gilbert et al. 2009), findings that indicate events and conditions in early life can affect risk for disease in adulthood.
Early-life stress in humans, defined here as a period of severe and/or chronic trauma, environmental or social deprivation, or neglect occurring either in prenatal or postnatal life or both, is also associated with a variety of neuropsychiatric outcomes, including later depression (Coffino 2009; Heim and Nemeroff 2001; Kaufman et al. 2000), posttraumatic-stress disorder (PTSD), panic disorder, schizophrenia, substance abuse, attention-deficit disorder (Heim and Nemeroff 2001), eating disorders (Gilbert et al. 2009), and an abnormal stress response (Kajantie and Raikkonen 2010; Kaufman et al. 2000). Moreover, maltreatment during childhood including emotional abuse, sexual abuse, physical abuse, and neglect, is an important risk factor for the development of adult psychiatric illness (Edwards et al. 2003; Green et al. 2010). In fact, Green et al. (2010) estimate that childhood adversity is associated with approximately one quarter to one third of psychiatric disorders presenting after childhood in the USA.
In addition to the association between stress exposure in early-life and later neuropsychiatric illness, stress exposure in early life is associated with deficits in cognitive function in animal models, including impaired spatial learning and memory (Aisa et al. 2007, 2009a). Furthermore, early-life exposure to stress in humans may lead to deficits in cognitive function during adult life, including deficits in language, attention, memory, and executive functioning (Chugani et al. 2001; De Bellis et al. 2009a). Whereas cognitive outcome studies have generally included the domains of intellectual skills, attention and working memory, processing speed, visuospatial/constructional, language, and executive skills (Lezak et al. 2004), early-life stress studies in human cognitive function (e.g., Ammerman et al. 1986; Moradi et al. 1999) and animal models (e.g., Aisa et al. 2007; Vallee et al. 1999) have generally focused on memory and learning.
In that, approximately 772,000 children were reported victims of child abuse or neglect between October 1, 2007 through September 30, 2008 in the USA (US Department of Health and Human Services 2010) and because the actual number of abused and neglected children is likely much higher due to lack of complete reporting and substantiation, the long-term effects of exposure to stress in early life may impair cognitive function later in life for a large number of people. Although review papers have examined associations between stress exposure in early life and physical-health and mental-health outcomes (Gilbert et al. 2009; Heim et al. 2008) and between stress exposure in early-life and brain development (McCrory et al. 2010; Tomalski and Johnson 2010), few reviews have attempted to integrate the findings concerning an association between early-life stress and later cognitive deficits. As such, the main focus in this article is to examine the association between exposure to early-life stress and later cognitive deficits. To do so, we first review cognitive deficits associated with early-life stress in animal models and then consider possible cognitive deficits associated with stress exposure in early life in humans and argue that exposure to stress in early life may be a risk factor for cognitive deficits later in life. Despite the differences in neuromaturation and sensitivity to environmental insult associated with different developmental periods, we include both prenatal and postnatal studies when available because of the comparatively few studies specifically addressing exposure to early-life stress and cognitive deficits and because both prenatal and postnatal stress exposure may be relevant in understating early-life stress exposures on later cognitive deficits. The association between early-life stress exposure and cognitive deficits is further complicated because complex cognitive functions (e.g., executive skills) mature later than do more simple, basic ones (e.g., sensorimotor skills) (Pechtel and Pizzagalli 2010), findings that show the relevancy of considering the association between stress exposure and developmental stage. Finally, to examine the cognitive effects of social intervention on children at risk for cognitive deficits, we review findings of cognitive function in children placed from institutions into foster families.
Animal models of the effects of early-life stress on cognitive function
Changes induced by early-life stress exposure in hippocampal (Aisa et al. 2009b; Andersen and Teicher 2004; Blaise et al. 2008) and frontal regions (Braun et al. 2000; Spinelli et al. 2009) in animal models suggest the potential for early-life exposure to stress to have long-term effects on cognitive function in animals. Because both the hippocampus and frontal brain regions are involved in cognitive function and affected by early-life stress exposure in animals, early-life adversity may be a factor in determining cognitive function in adult animal models.
In line with this hypothesis, animal models of early-life stress have consistently shown long-term cognitive deficits. Using a maternal-separation paradigm, Aisa et al. (2009a, b; 2007) found that rats exposed to stress in early life performed poorly on a learning and memory task compared with control rats. Moreover, early-life stress was associated with progressive learning and memory decline in later life in a rodent model, with the decrease in learning and memory occurring most prominently during middle age (Brunson et al. 2005). To appreciate the extent of early-life stress on cognitive deficits, emerging evidence from prospective studies has shown the role of the environment in shaping stress vulnerability and impaired learning and memory in animals as far back as to the prenatal period (Hosseini-Sharifabad and Hadinedoushan 2007; Vallee et al. 1999; Yang et al. 2006). In particular, a longitudinal study found that rats exposed to prenatal stress had accelerated, age-related decline in spatial and working memory (Vallee et al. 1999).
In addition to the effects of early-life stress on later cognitive deficits, convergent findings have shown that environmental conditions such as environmental enrichment improves cognitive functioning in rats undergoing stress due to isolation rearing. Environmental enrichment appears to improve memory and learning in cognitive tasks, including spatial memory in the Morris water maze and recognition memory in the novel-object recognition test (Falkenberg et al. 1992; Kempermann et al. 1997; Nilsson et al. 1999; Rampon et al. 2000). Further showing the sensitivity of cognitive function in animal models to stress exposure in early life, environmental enrichment may reverse cognitive impairment resulting from early-life stress in animal models. For example, enriched social housing provided from postnatal days 50 to 77 reversed the deleterious effects on spatial learning from social isolation that had occurred from postnatal days 22 to 49 (Lu et al. 2003). Furthermore, environmental enrichment reversed the effects of prenatal stress or reduced maternal care on spatial and non-spatial learning and object recognition (Bredy et al. 2003a, b; Bredy et al. 2004; Koo et al. 2003). Similarly, the adult offspring of the mothers that had had comparatively higher pup licking and grooming and arched-back nursing had enhanced spatial learning and memory, a finding suggesting that that maternal behavior can influence cognitive function in the offspring into adulthood (Liu et al. 2000). Other groups have consistently replicated these results (e.g., Aisa et al. 2009a, b; Bazak et al. 2009; Bredy et al. 2003a) that together suggest the capacity for cognitive restoration or preservation through environmental enrichment at a later developmental stage.
Non-human primate models can be particularly helpful in studying the role of early-life stress on later cognitive deficits. Compared to rodents, non-human primate models show more similarities to humans in their complex social development and structures (Stevens et al. 2009), and it is well recognized that early social deprivation in non-human primates results in behavioral sequelae (Barr et al. 2003; Champoux et al. 2002; Coe et al. 2003; Higley et al. 1991, 1996a, b; Higley et al. 1996a, 1996b; Suomi 1997). To our knowledge, however, few studies of the association between early-life stress and later cognitive deficits in non-human primates are available, despite the compelling reasons to explicitly examine any association between early-life stress and later cognitive deficits in non-human primates. In one available study, maternal stress during gestation in rhesus monkeys resulted in delayed object permanence in rhesus monkeys exposed to prenatal stress than in non-exposed monkeys (Schneider 1992). Not all types of early-life stress in non-human primates may have deleterious effects, however. Squirrel monkey that had been exposed to mild early stress showed better prefrontal-dependent cognitive ability than nonstressed animals (Parker et al. 2005), a finding that emphasizes the importance of understanding both type and severity of the stress exposure and the developmental timing of the exposure. Collectively, animal models based primarily on rodents show deleterious, long-term effects associated with early-life stress on memory and learning later in life. Of additional relevance to understanding associations between early-life environments and later cognitive deficits are findings that environmental enrichment may preserve or reverse memory and learning impairments resulting from early-life stress in animal models.
Human studies of the effect of early-life stress on cognitive function
In addition to increasing the risk for psychiatric illness in adulthood (Edwards et al. 2003; Green et al. 2010; Kaufman et al. 2000), findings that exposure to stress in early life are associated with later abnormalities in brain structure and function (Bremner 2002; Bremner et al. 1997; Carrion et al. 2001; De Bellis et al. 2000; Teicher et al. 2004) suggest the possibility that early-life exposure to stress in humans may result in later cognitive deficits, particularly because at least some of the abnormalities in brain structure and function associated with stress exposure in childhood persist into adulthood (Andersen et al. 2008; Bremner et al. 1997; Bremner et al. 2004). However, the association between stress exposure in early life and brain structure and function in adulthood is likely complex as shown, for example, by the finding of an association between birth weight and adult hippocampal volume but only in women reporting comparatively low levels of maternal care (Buss et al. 2007).
Strengthening the possible association between early-life stress and later cognitive function is the observation that brain regions showing abnormalities after exposure to early-life stress are areas with known involvement in cognitive function (Bremner et al. 1999). For example, stress exposure during childhood appears to be associated with adult volume and functional deficits in the hippocampus, a brain region associated with memory and learning (Eichenbaum et al. 1992; Squire et al. 2004). In contrast to the number of studies associating stress and trauma exposure in childhood to later neuropsychiatric illness, however, comparatively few studies have examined the association between childhood exposure to stress and later cognitive function. Despite a relative paucity of studies examining the association between stress exposure in early-life and later cognitive deficits, several lines of evidence suggest that stress exposure in early life may subsequently influence cognitive function.
Maternal stress during gestation may affect cognitive function in the children. In one study (Buitelaar et al. 2003), prenatal stress was associated with cognitive deficits, including attention dysregulation, at the age of 8 months. In another study designed to assess the effects of prenatal stress independent of postnatal stress exposure, prenatal stress predicted mental development at ages 14 to 19 months (Bergman et al. 2007). In fact, the authors concluded that “fetal programming may be as relevant for emotional and cognitive outcomes as it has been shown to be in many other areas of human health and development” (p. 1460). The implications for prenatal stress on cognitive function in later childhood, adolescence, and adulthood, however, remain unknown as this appears to be an unstudied area.
Related to the hypothesis that stress exposure in early life is related to later cognitive deficits is the substantial evidence showing that cognitive deficits are also found in children with PTSD. Moradi et al. (1999) investigated memory function using the Rivermead Behavioural Memory Test in 18 children and adolescents (mean age is 14.3 for eight boys and ten girls) with PTSD from traffic accidents or personal violence within the 2 years prior to testing compared with 22 children and adolescents (mean age, 14.3 for 12 boys and ten girls) who did not have a history of emotional disorder or trauma and found that children and adolescents with PTSD exhibited poorer overall memory performance compared with controls. In another study, 14 medication-naive children (six girls and eight boys; mean age, 11.4) with PTSD compared with 15 healthy, demographically matched children (seven girls and eight boys; mean age, 12.1) performed more poorly on measures of attention and executive functioning (Beers and De Bellis 2002). The authors, however, did not replicate the findings of Moradi et al. (1999) of general memory deficits associated with childhood PTSD, possibly due a small sample size and differences in the selection of cognitive measures. Despite failure to replicate earlier findings of deficits in general memory, Beers and De Bellis (2002) demonstrated that children with PTSD are at increased risk of cognitive deficits.
De Bellis et al. (2009a) investigated cognitive functioning in neglected children with and without PTSD and a demographically matched group of non-neglected children. For this study, neglected children were those who had a history of neglect without sexual abuse according to a Social Services department. While there were no significant differences in cognitive abilities between the neglected children with PTSD and those without PTSD, the authors found that both groups of neglected children had significantly worse language, visuospatial, learning, memory, attention, and executive functioning than healthy controls. Additionally, neglected children, both with and without PTSD, performed significantly lower on academic achievement and intelligence tests than controls. In another comprehensive cognitive outcome study of children with and without PTSD-related maltreatment (neglect, witnessing interpersonal violence, and physical, sexual, and emotional abuse), De Bellis et al. (2009b) found that lower visual memory performance was associated with more posttraumatic-stress symptoms. Other studies of PTSD in children and adolescents with PTSD have found lower verbal and visual memory, attention, and concept formation (Schoeman et al. 2009; Yasik et al. 2007). Together, the available findings indicate that childhood PTSD and adverse experience in childhood appear associated with deficits in cognitive function during childhood and possibly adolescence, although no available studies appeared to have specifically examined adolescents despite the importance of understanding the extent to which deficits in cognitive function associated with PTSD and maltreatment persist beyond childhood. The extent to which cognitive deficits associated with childhood PTSD persist into later life is unknown and requires additional research.
Studies of children not selected based on a PTSD diagnosis have also found an association between childhood abuse and deficits in intellectual skills (Ammerman et al. 1986; Carrey et al. 1995; Huth-Bocks et al. 2001) and memory (Ammerman et al. 1986; Eisen et al. 2007), as well as attention, verbal memory and motor development during childhood (Ammerman et al. 1986), abnormalities due at least in part to the direct consequences of brain trauma from the abuse. Of additional interest, one study explored differential relationships between personal (physical and emotional abuse and neglect, school and parental stressors) and community (neighborhood problems or violence) stressors and cognitive functioning in 553 children between 10 to 12 years (Fishbein et al. 2009). The study found that overall exposure to psychosocial stress was associated with lower levels of executive functioning, such as problem-solving, cognitive flexibility, and decision-making skills. Additionally, exposure to personal stressors was more strongly associated with lower levels of problem-solving skills than community stressors. Furthermore, stressor types were differentially associated with performance on specific cognitive tasks; specifically, emotional and physical neglect and stress with parents and in school were associated with poorer intellectual and executive skills (Fishbein et al. 2009). Similarly, Pears et al. (2008) found differential effects of neglect versus physical abuse in preschoolers in foster care, while neglect was negatively associated with executive functioning, physical abuse was positively associated with executive domain outcomes (Pears et al. 2008). Thus, stressor types appear to be differentially associated with a specific cognitive profile, although having findings from only two studies temper such conclusions. Certainly, more work is needed to better understand whether different types of abuse result in different patterns of cognitive deficits.
Long-term cognitive function in adulthood following childhood trauma is markedly understudied. In one of the few studies examining the long-term cognitive sequelae of exposure to childhood trauma (physical and sexual abuse), Bremner et al. (1995) found that adults with PTSD who had been severely abused during childhood showed deficits in verbal memory compared with healthy adults matched for age, sex, race, handedness, height, weight, education, parental education, and years of alcohol abuse. Abuse severity was related to the degree of verbal memory impairment, but general intellectual ability did not differ between groups.
Additional evidence supporting an association between early-life stress and cognitive deficits in adulthood comes from several studies. In one study, 26 college women with a history of repeated childhood sexual abuse (five of whom had past or current PTSD) demonstrated diminished memory and response inhibition (a dimension of executive functioning) when compared with ten healthy female college women (Navalta et al. 2006). A retrospective study found an association between prenatal stress exposure and working-memory function in 32 young women (mean age = 25 years) compared with 27 healthy young (mean age = 24) women (Entringer et al. 2009). In a cross-sectional study, Bremner et al. (2004) examined cognitive function in three groups of adult women: those with a history of childhood sexual abuse and PTSD (mean age = 34), those with a history of childhood sexual abuse but without PTSD (mean age = 32), and non-abused women without PTSD (mean age = 32). Although there were no differences in overall intellectual ability, the PTSD group had significant deficits in verbal declarative memory compared to the other two groups. A prospective, longitudinal study by Perez and Widom (1994) found lower intelligence scores in young adults (n = 413) 20 years after the occurrence of childhood trauma compared to demographically matched controls (n = 286). In contrast, Bremner et al. (2003) found that women with PTSD associated with sexual abuse in early childhood (n = 10) compared to healthy non-abused control women (n = 11) did not differ in the recall of emotionally valenced words in a word-pair retrieval task, although the authors noted study power issues might have precluded detection of differences between groups.
Although comparatively few studies have examined the association between childhood stress exposure and cognitive deficits in adulthood, the available findings suggest that cognitive sequelae of stress exposure may persist at least into young adulthood. More longitudinal studies are needed to examine whether association between early-life stress exposure and cognitive deficits persists or increases with age.
Cognitive function in children removed from institutionalized settings
Emerging findings suggest that favorable social factors may ameliorate or reverse some of the effects of early stress exposure in children. In this regard, studies of institutionally raised children provide additional insight into understanding the effects of social environment after stress exposure in early childhood.
Children assigned to foster care had significant gains in cognitive function compared with those placed in institutions (Nelson et al. 2007). A longitudinal study showed that children with general cognitive deficits within the mental-retardation range at age 2 years removed from Romanian orphanages and adoption services and raised by families in the UK showed substantial cognitive improvement by age 4 years (Rutter 1998). Cognitive deficits at age 4 years, however, were present at age 6 years (O’Connor and Rutter 2000), suggesting persistent cognitive deficits from early childhood neglect despite a later more nurturing environment. In addition, Chugani et al. (2001) found that post-institutionalized children from Romania displayed a wide range of mild deficits in language, attention, memory, and executive functioning but reported that there were significantly reduced intellectual abilities in children who remained in the institution compared with never-institutionalized children and children taken out of the institution and placed into foster care. Likewise, Cohen et al. (2008) found that developmental delays associated with deprivation in the first year of life had largely but not completely resolved in all children by 2 years after adoption (which occurred at age 8 to 21 months), findings similar to those of van den Dries et al. (2010) showing improvement in cognitive deficits by 6 months after adoption from an institution but with persistence of some deficits. In children ages 8 to 11 years tested 5 to 11 years after adoption from an institution, continuing deficits in cognitive function were associated with the amount of time originally spent in the institution (Loman et al. 2009), findings highlighting the cognitive deficits associated with early institutionalization and the advantages of family placements for young abandoned children (Nelson et al. 2007).
Taken together, studies of children with a history of institutionalization suggest that intellectual and cognitive deficits most likely resulting from early environmental deprivation may improve following adoption, although many adopted children continue to experience chronic cognitive sequelae after adoption. Therefore, subsequent parenting practices may have the potential to modify some of the cognitive deficits associated with early-life exposure to stress.
Given the large numbers of abused children (US Department of Health and Human Services 2010), it is important to understand the outcome of abused children across a variety of domains. Particularly concerning in addition to the increased risk for psychiatric illness in adults with early-life stress (Edwards et al. 2003; Green et al. 2010; Kaufman et al. 2000) are the findings associating early-life stress with later cognitive deficits occurring in childhood that may persist into at least young adulthood. Based on the current review, cognitive deficits associated with early-life stress may occur in multiple domains, including intellectual function, memory, executive function, visuospatial skills, and language. In contrast to the now considerable body of data showing associations between exposure to stress during childhood and mental and physical illness (Gilbert et al. 2009; Green et al. 2010), there are relatively few studies examining the association between stress exposure in childhood and adult cognitive outcome. In those few available studies, only one study used a longitudinal design with a control group (Perez and Widom 1994), which enabled a clearer characterization of cognitive deficit following exposure to early-life stress. One direction for future research, therefore, is to include an assessment of cognitive function in prospective studies investigating the association between childhood stress exposure and adult neuropsychiatric, medical, and functional outcomes.
The developmental timing of stress exposure may have differential effects on cognitive outcome. In this regard, Andersen et al. (2008) found that sexual abuse occurring between ages 3 and 5 years and between 11 and 13 years was associated with volume deficits in the hippocampus during adulthood whereas sexual abuse occurring between nine and 10 years was associated with volume deficits in the corpus callosum in adulthood and sexual abuse between ages 14 and 16 years was associated with deficits in gray matter volume in the frontal cortex. Similarly, it is possible that the type, degree, and persistence of cognitive deficit associated with early-life stress may depend upon whether the stress exposure occurs prenatally or postnatally. Furthermore, it may be that cognitive outcome is dependent upon stress exposure during sensitive and critical periods of brain maturation both prenatally and postnatally. In this regard, Pechtel and Pizzagalli (2010) note that stress may affect a brain region differently depending upon the developmental level of the child. Only additional studies including the use of animal models and longitudinal studies in humans can adequately determine how developmental timing of stress exposure affects later cognitive outcome.
An important question is whether it is the stress exposure itself or the resulting PTSD, or both, that results in the later putative cognitive deficits. Trauma exposure in the absence of PTSD appears to be associated with deficits in hippocampal volume compared with healthy controls, although the presence of PTSD may be associated with additional hippocampal volume deficit (Woon et al. 2010), a finding consistent with the hypothesis that stress exposure even without clinically recognizable PTSD could impair later cognitive function. However, one study (Bremner et al. 2004) found deficits in verbal declarative memory only in adult PTSD subjects with a history of childhood maltreatment compared with adults with a history of childhood maltreatment but no PTSD and healthy controls. Although some of the studies examining the association between exposure to childhood stress and cognitive function in adulthood used subjects who also had PTSD (Bremner et al. 1995, 2004) and thus do not enable a determination of whether stress exposure alone was associated with cognitive deficits, one study found deficits in memory and response inhibition in adults with a history of childhood sexual abuse even though only five of 26 subjects had current or past PTSD (Navalta et al. 2006). Moreover, cognitive deficits persisted for up to two decades in the absence of PTSD in the longitudinal study of children with early-life stress (Perez and Widom 1994). Further support of the hypothesis that exposure to stress in early life is associated with cognitive deficits in the absence of PTSD are the studies of Bergman et al. (2007) and Buitelaar et al. (2003) showing that prenatal stress predicted nearly 20% of the variance in mental abilities at ages 14 to 19 months and 8 months, respectively, and prenatal stress exposure was associated with impaired working memory in another study (Entringer et al. 2009). Together, the available studies suggest that cognitive deficits are associated with stress exposure in early childhood including prenatal exposure even in the absence of overt PTSD. Nevertheless, there is little information about the association between stress exposure in early-life and later function in specific cognitive domains. These areas of inquiry may represent a useful extension to the study of early-life stress and cognitive deficits.
The findings of Brunson et al. (2005) and summarized by Bremner (2006) suggesting that changes in hippocampal volume from stress exposure and associated alterations in memory in animal models can be reversed by serotonin reuptake inhibitors, tianeptine, or phenytoin contain intriguing possibilities about the potential for preventing or ameliorating the long-term cognitive deficits possibly associated with early-life stress. It may be that pharmacological intervention at critical times could inhibit subsequent cognitive deficits improve existing deficits.
Equally intriguing as the evidence suggesting pharmacological reversal of hippocampal volume deficits associated with stress exposure in animal models are findings showing the beneficial effects on cognitive function of placement in foster care in formerly institutionalized children (Chugani et al. 2001; Cohen et al. 2008; Loman et al. 2009; Nelson et al. 2007; O’Connor and Rutter 2000; Rutter 1998; van den Dries et al. 2010). The evidence to date, however, indicates that at least some of the deficits in cognitive function associated with early-life institutionalization may persist even after placement into foster care.
Even if pharmacological, parenting, and social intervention could prevent or attenuate the cognitive deficits associated with exposure to stress and trauma in early childhood, however, the small numbers of abused children identified as such (Gilbert et al. 2009) presents a considerable barrier to using pharmacological intervention under current social constraints. Far better would be approaches designed to minimize childhood exposure to stress and trauma in the first place.
A crucial but as yet inadequately addressed question is whether the putative cognitive deficits associated with early-life stress persist into and progress with old age in humans, or even whether early-life stress is a risk factor for late-life dementia, a mental disorder characterized by the development of multiple cognitive deficits (American Psychiatric Association 2000). In an animal study, early-life stress was associated with progressively worse cognitive deficits (Vallee et al. 1999), a finding that raises the possibility that early-life stress could be a risk factor for late-life dementia. In a retrospective, multivariably adjusted cohort study, PTSD appeared to be an independent risk factor for dementia in elderly veterans. In this study, elderly veterans with PTSD had nearly twice the risk (adjusted hazard ratio, 1.77) of developing dementia as veterans without PTSD (Yaffe et al. 2010). Whether exposure to early-life stress with its possible cognitive sequelae results in a similar risk for dementia in late life is unclear. Furthermore, adult PTSD may accelerate aging (Yaffe et al. 2010; Yehuda et al. 2005), but it is unknown whether exposure to early-life stress with or without adult PTSD results in similar changes. While there is limited evidence on the relationship between exposure to early-life stress and late-life cognitive deficits, several meta-analyses have found an association between PTSD and a smaller total brain volume (Hedges and Woon 2010), a smaller hippocampus in adults with PTSD but not in children with PTSD (Woon and Hedges 2008), and a smaller right hippocampus in individuals exposed to trauma with or without PTSD, as well as a further hippocampal reduction found in the trauma-exposed individuals without PTSD compared with normal controls (Woon et al. 2010). Hippocampal and global brain atrophy are frequently observed on magnetic resonance images in individuals with dementia of the Alzheimer type (Barnes et al. 2009; Beckett et al. 2010; den Heijer et al. 2010; Mungas et al. 2005).
Several limitations inherent in this review of the association between exposure to stress in early-life and cognitive deficits require consideration. Relatively few studies have explicitly examined potentially modifying factors such as the effects of sex and timing, duration, nature, and intensity of the stress exposure. Secondly, stress exposure may be accompanied by other types of exposures such as psychiatric comorbidity, physical brain trauma, and undernutrition, factors that by themselves can influence cognitive outcome and confound association studies between early-life stress exposure and cognitive deficits. Third, there are few prospective studies about the association between early-life stress exposure and cognitive deficits; as such, the available retrospective cross-sectional studies are necessarily subject to selection, recall, and reporting biases and confounding. While prospective animal models are an alternative in this case to prospective human studies, which are difficult to do, difficulties extrapolating cognitive findings in animal models to humans are formidable. Carefully designed and executed, fully adjusted case-control studies may provide an acceptable compromise between prospective animal models and cross-sectional human studies, at least until additional prospective data are available.
In summary, the preponderance of the findings to date suggests that stress exposure in early life adversely affects brain development in regions such as the hippocampus, abnormalities that may persist into adult life. Because the hippocampus is crucial for memory function, it is reasonable to speculate that hippocampal abnormalities associated with early-life stress may predispose to cognitive abnormalities in childhood and adulthood. Indeed, children with PTSD (Beers and De Bellis 2002; De Bellis et al. 2009a; Schoeman et al. 2009; Yasik et al. 2007) and children not necessarily selected on the basis of PTSD but with a history of exposure to early-life stress (Eisen et al. 2007; Huth-Bocks et al. 2001; Kerouac et al. 1986; Perez and Widom 1994) do appear to have cognitive deficits compared with their non-abused peers, and the few studies investigating cognitive deficits in young adults with a history of early-life stress also have found cognitive deficits, including short-term verbal memory (Bremner et al. 1995), intelligence (Perez and Widom 1994), memory and response inhibition (Navalta et al. 2006), working memory (Entringer et al. 2009), and verbal declarative memory (Bremner et al. 2004). As such, early-life exposure to stress may be related to cognitive deficits in adulthood, although the relationship with exposure to early-life stress and old age and dementia remains almost entirely unstudied.
Conflicts of interest