Stress Effects on Memory

In essence, neural and behavioral development have an interdependent relationship that relies on reciprocal processing activities. Thus, one postulate is that if the process is somehow dramatically perturbed (either endogenously or exogenously), then the associated memory systems would take an altered developmental trajectory. Early damage to the hippocampus, including perinatal, results in a substantial loss of context-rich memory abilities (Bachevalier & Vargha-Khadem, 2005). High-dose prednisone treatment for children with asthma is associated with lower verbal memory skills than low-dose treatment (Bender, Lerner, & Poland, 1991). Given such findings, Nelson and Carver (1998) hypothesized that the developing nervous system is vulnerable to the deleterious chronic effects of stress on memory. Specifically, these investigators predicted that hippocampal-dependent memory should be negatively affected by substantial stress because of the effects of stress hormones on cell birth and death in the hippocampus, especially when this region is exposed to such stress during immature and vulnerable periods. Andersen and Teicher (2004) found in rats that early severe stress is associated with reduced synaptic numbers in the hippocampal region. Other developmental studies in rats indicate that fear conditioning with auditory versus visual stimuli has different ontogenetic profiles (Hunt & Campbell, 1997; Moye & Rudy, 1987), which suggests that an interaction exists between the developing hippocampus and the memory systems mediating auditory and visual learning (Stanton, 2000). However, the process in which early experience in humans influences the neurodevelopment that underlies the development of memory remains unclear (Howe, Cicchetti, & Toth, 2006; Howe, Toth, & Cicchetti, 2006).

Reduced memory performance has been shown to be more closely associated with chronic rather than acute stress ( Jelicic & Bonke, 2001) as well as severe stress (Anda et al., 2006; Bremner, Vermetten, Afzal, & Vythilingam, 2004). In the study by Anda and colleagues (2006) with an epidemiological sample of 8,708 adult HMO members, a 4.4-fold increase in risk for impaired autobiographical memory of childhood was observed for persons who endorsed four or more adverse childhood experiences (e.g., abuse, witnessing domestic violence, serious household dysfunction). In addition, a dose-response relationship was seen between the extent of adverse childhood experiences and the number of age periods affected for memory disturbance. Such deficits in memory do not appear to be a function of decreased intellectual capacity (Bremner et al., 2004; but see Perez & Widom, 1994). Adult studies have also demonstrated that stress-related, glucocorticoid exposure effects possess an inverted U-shape function with memory (Andreano & Cahill, 2006). That is, stress appears to have opposite effects on memory depending on the duration and chronicity. Consolidation—a process that effects a more stable and consistent form of memory—is typically found to be enhanced by low doses of glucocorticoids (Roozendaal, 2002, 2003). However, a recent study in healthy females suggests that the glucocorticoid cortisol, induced by an acute stress challenge, negatively influences long-term consolidation of declarative memories—at least with a 24-hour retention interval ( Elzinga, Bakker, & Bremner, 2005).

Because the hippocampus has a high density of glucocorticoid receptors, is particularly vulnerable to the effects of stress hormones and plays an integral role in the encoding and retrieval of declarative memory, this brain region has received significant attention from investigators of chronic and traumatic stress, especially in adults. Elzinga and Bremner (2002), for example, recently conceptualized PTSD as a memory disorder in which the intrusive memories and declarative memory function deficits (as well as the stress response) are intimately tied to the hippocampus. To date, few studies of the brain and memory have been conducted in traumatized individuals (Howe, Toth, & Cicchetti, 2006). Although findings thus far are suggestive, hippocampal abnormalities are not necessarily associated with memory deficits in individuals with PTSD (Jelicic & Merckelbach, 2004).

Bremner and colleagues have provided much of the evidence to support the association between hippocampal anomalies and memory impairment.

In one of the earliest studies in this field, Bremner et al. (1997) examined 17 adults with histories of CA and current PTSD and 17 control subjects matched for age, demographic characteristics, and alcohol abuse. The authors found that subjects with PTSD had a 12% smaller left hip-pocampal volume than control subjects and performed poorer on several verbal memory tasks. However, no memory tests were correlated with hip-pocampal volume. A follow-up study was conducted using positron emission tomography (PET) in 22 women with histories of child sexual abuse ( Bremner et al., 1999). Their findings demonstrated that abuse memories (reactivated via auditory, personalized event scripts) are associated with decreased blood flow in right hippocampus (as well as fusiform/inferior temporal and supramarginal gyri and visual association cortex) in women with PTSD relative to those without PTSD. Further support for the association between hippocampal-related memory dysfunction and PTSD comes from a more recent study by Bremner et al. (2003). In their study, women with CA-related PTSD showed greater decreases in blood flow in several brain areas, including the left hippocampus, during retrieval of emotionally valenced word pairs, but not with neutral word pairs. Moreover, the manifest memory deficits appear to be specific to the verbal (versus visual) domain, independent of intelligence, but negatively correlated with PTSD symptom levels and severity of CA (Bremner et al., 2004).

CA-related borderline personality disorder (BPD) is another condition documented to be related to memory impairments. In an earlier study by O'Leary, Brouwers, Gardner, and Cowdry (1991), 16 research outpatients with BPD were compared to a group of 16 healthy volunteers using a battery of neuropsychological tests. The individuals with BPD were significantly impaired on memory tests requiring uncued recall of complex, recently learned material, which was not attributable to attentional problems, psychomotor impairment, current major depression, or history of alcohol abuse. In 21 females with BPD, Driessen et al. (2000) found 16% and 8% bilateral volume reductions of the hippocampus and amygdala, respectively. However, memory performance for this group was related to depressive symptoms rather than hippocampal volume. Two studies by Schmahl and colleagues (2003, 2004) used PET on women with and without BPD and found blood flow differences in the anterior cingulate and dorsolateral prefrontal cortex that were dependent on the types of memory reactivated. That is, hypometabolism was observed in these brain regions while subjects listened to personal CA scripts whereas memories of abandonment were associated with increased blood flow in bilateral dorsolateral prefrontal cortex. In addition, abandonment memories were associated with greater decreases in right anterior cingulate. Lastly, Lange et al. (2005) found regional hypometabolism in temporal and medial parietal cortical regions—areas intimately tied to episodic memory consolidation and retrieval—as well as strong correlations between impaired memory performance and metabolic activity in ventromedial and lateral temporal cortices in individuals with BPD.

Evidence from our research program suggests that the effects of CA-related stress on memory do not necessarily hinge upon the presence of a current psychiatric diagnosis. First, support for this hypothesis comes from an electrophysiological study utilizing probe auditory evoked potentials (Schiffer et al., 1995). Adults with a history of CA who were all currently well-functioning and had no active DSM-IV Axis I diagnosis were asked to actively recall a neutral or work-related memory, and then a disturbing memory from childhood with related affect. For psychologically healthy individuals with no trauma histories, both hemispheres appeared to be equally involved in the recall of these memories. In contrast, adults with a history of CA demonstrated a marked suppression of the evoked potential response over the left hemisphere during recall of the neutral memory, which is indicative of increased left hemispheric processing. In addition, a robust shift in laterality was observed, with the evoked potential response being suppressed over the right hemisphere during recall of the disturbing memory, indicating enhanced right hemispheric activation. These findings are consistent with the contention that CA is associated with an increase in hemispheric laterality together with a decrease in hemispheric integration. A recent follow-up study by Schiffer et al. (2007) suggests that the emotional valence of individuals with histories of CA can be lateralized in either hemisphere, but that the laterality of negative emotions in the left hemisphere, in particular, is associated with poorer memory (especially for visual information).

Second, Navalta, Polcari, Webster, Boghossian, and Teicher (2006) examined 26 young women (mean age = 20.0 years) with histories of childhood sexual abuse—most of whom (73%) had no current DSM-IV Axis I diagnosis. We found a strong graded association between duration of CA by the most closely related perpetrator (parent > sibling > uncle > other family > friend > authority figure > stranger/acquaintance) and memory function. Specifically, 2-point reductions in short-term, verbal, visual, and global memory scores were observed for every year that abuse occurred (standard scores; mean = 100, standard deviation [SD] = 15). Neither current symptoms of depression, anxiety, or PTSD nor a history of these disorders had a significant influence on this strong relationship. Although no difference between CA subjects and the contrast group was found on overall global memory, the CA group did exhibit lower visual span, but they also displayed superior verbal span. If the findings that childhood sexual abuse produces a duration-dependent decrease in memory function were true, then some of the CA subjects might actually have had quite high memory scores if they had never been abused.

In contrast to studies with adults, findings from child studies on the relationship between stress exposure and memory are scant and negative overall. Recent comprehensive reviews by Howe, Cicchetti, et al. (2006) and Howe, Toth, et al. (2006) clearly illustrate that evidence to date does not support the premise that stress causes memory deficits in children. Specific to children with abuse histories, basic memory performance does not distinguish them from demographically similar comparison children who have not been abused. For example, children with abuse histories apparently perform as well as (and possibly better than) nonabused children on daily memory tasks (Howe, Cicchetti, Toth, & Cerrito, 2004). Eisen, Qin, Goodman, and Davis (2002) evaluated children's memory and suggestibility about a medical examination and clinical assessment and found that children who were abused were just as likely to confuse details about the examination as children without abuse histories. In a study by Howe et al. (2004), children who were abused and from low socioeconomic backgrounds were assessed for the development of false memories. Similar to the findings of Eisen and colleagues (2002), both children with and without abuse histories demonstrated false memories, and the number of such memories increased with age at an equivalent rate. Overall, children from lower socioeconomic backgrounds had the poorest memory performance, regardless of abuse status. In addition, abuse-related PTSD ( Beers & De Bellis, 2002; Moradi, Doost, Taghavi, Yule, & Dalgleish, 1999), dissociation (Eisen et al., 2002), and depression (Orbach, Lamb, Sternberg, Williams, & Dawud-Noursi, 2001) do not appear to be associated with memory deficits in children.

Recently, Porter, Lawson, and Bigler (2005) conducted a well-designed study examining memory and intellectual functioning in 24 children who were sexually abused compared to 24 physically and behaviorally healthy children with no abuse histories, matched for demographic characteristics. Five subtests of the Wechsler Intelligence Scale for Children (3rd ed.) and the Test of Memory and Learning (TOMAL) were administered. Mean scores illustrate that the CA group was functioning within the average range of memory and intellectual functioning. First-pass analyses indicate that the children with abuse histories scored lower than comparison children on several of the TOMAL indices, including verbal memory, attention/concentration, sequential recall, and free recall—even when statistically controlling for socioeconomic status. However, these differences were no longer apparent between the two groups after controlling for both socioeconomic status and intelligence. In addition, the presence of PTSD did not have a differential effect on cognitive function between CA subjects who currently had the disorder from those who did not. Porter and colleagues (2005) concluded that this sample of children who were sexually abused did not possess a distinct pattern of memory impairment.

Single Parenting Becoming the Best Parent For Your Child

Single Parenting Becoming the Best Parent For Your Child

Parenting is a challenging task. As a single parent, how can you juggle work, parenting, and possibly college studies single handedly and still manage to be an ideal parent for your child? Read the 65-page eBook Single Parenting Becoming The Best Parent For Your Child to find out how. Loaded with tips, it can inspire, empower, and instruct you to successfully face the challenges of parenthood.

Get My Free Ebook

Post a comment