Researchers have found that early-life adversity can have long-term effects on people’s memory and thinking in later life.
Dr. Ruby Tsang from the University of Oxford showed that experiencing family financial hardship in early life, and poorer childhood health predicts greater memory and thinking decline in later life.
She presents the research today (Tuesday 24 March) at the Alzheimer’s Research UK Virtual Conference.
Although the charity’s annual Research Conference in Wales was canceled due to the coronavirus outbreak, the event is being recreated online to allow scientists to share the latest dementia research findings.
Research suggests that early-life experiences play an important role in mental health conditions like depression and anxiety, with exposure to early life adversity linked to a range of aspects of our lifestyle.
There’s limited research, however, on the influences, early-life experiences have on how memory and thinking skills decline later in life.
Now, using information from the Dementias Platform UK (DPUK) Data Portal, Dr. Tsang studied 15,309 volunteers, including over 5,000 ex-civil servants. Using questionnaires, the study volunteers answered questions about their childhood concerning their family socioeconomic status, their own health and whether they were the victim of abuse.
The researchers then looked at volunteer’s scores on a wide range of memory and thinking tests, including verbal fluency in mid-to later life.
Dr. Tsang examined the association between early adversity factors and a decline in memory and thinking. The analysis was carried out using the Dementias Platform UK (DPUK) Data Portal.
Dr. Tsang found three different patterns of memory and thinking, which reflect resilience to cognitive decline, gradual age-related decline and rapid cognitive decline. The results showed fewer years of education, having experienced family financial hardship in early life, and poorer childhood health predicted a greater decline in memory and thinking skills.
Dr. Ruby Tsang said: “Our socioeconomic status is closely intertwined with many aspects of our lifestyle and is particularly associated with our risk of various health conditions as we age. This research suggests that even in childhood, these experiences have a far-reaching and important influence on our cognitive performance.
We found key differences between men and women, with women more likely to be in the resilient group for commonly used screening tests that measure a range of memory and thinking skills.
“It’s fantastic to still be able to share this research funded by Dementias Platform UK in these times. It is important that we continue to communicate our research findings, and a virtual conference like this allows the work of early-career researchers like me to reach more people out there.”
Dr. Tsang found three different patterns of memory and thinking, which reflect resilience to cognitive decline, gradual age-related decline and rapid cognitive decline.
Dr. Carol Routledge, director of research at Alzheimer’s Research UK, said: “While we can’t change our past, keeping mentally and physically active, maintaining a healthy weight, not smoking, only drinking within recommended guidelines, eating a healthy diet, and keeping blood pressure and cholesterol in check can help to support brain health as we age.
“While this study didn’t investigate whether people went on to develop dementia, understanding the risk factors for declining memory and thinking could help us to design better strategies for keeping people’s brains healthy.
This research adds to growing evidence that suggests we need to protect brain health throughout life, just as we do with heart health.
“By considering life-long risk factors that influence cognitive health, we can support measures to help keep people’s brains healthy at any age.
With greater understanding of the factors influencing brain health and dementia risk, we will be better able to make breakthroughs that could make a real difference to people’s lives.”
Experiencing a traumatic event (e.g. death of a parent, friend, or family member, physical, emotional, or sexual abuse) during childhood is common1 and a significant public health problem2.
Additionally, the population of the United States is growing older, with the number of adults age 65 or older estimated to more than double to just under 100 million by the year 20603.
As many as 47% of older adults report experiencing a childhood trauma, with some evidence to suggest this prevalence is higher in recent cohorts4.
Despite this, the long-term effects of childhood trauma on health outcomes in older adulthood remains understudied. With older adults comprising the fastest growing segment of the population, understanding the associations between childhood trauma and later-life health outcomes will only grow in importance as adult survivors of childhood trauma continue to age.
Childhood trauma has been associated with poor brain health. Younger adults who have experienced childhood trauma are at increased risk of poor physical health and psychiatric distress, such as Generalized Anxiety Disorder (GAD), Posttraumatic Stress Disorder (PTSD), and mood disorders1, 5, 6.
Early life traumatic events may also negatively affect cognitive functioning7, 8, particularly slower information processing speed and worse attention, memory, and executive functioning in children and younger adults9–11.
Many of these studies find that general (e.g., unexpected death of a family member, witnessing physical violence), physical, and sexual traumas are particularly harmful to cognitive functioning12–14.
Additionally, experiencing one childhood traumatic event is a risk factor for experiencing subsequent traumas, and multiple traumatic events in childhood are associated with worse outcomes than a single traumatic event15.
Despite the strong implications of increased incidence of cognitive decline and disorders in later life, the extent to which childhood trauma is associated with cognitive performance in older adults is largely understudied.
Childhood adversity comorbid with depression was associated with greater declines in processing speed in 1,312 older adults in the Longitudinal Aging Study Amsterdam16. Longitudinal research with 846 community-dwelling older adults found that repeated early life sexual assault was associated with greater declines in verbal fluency and Trails-B performance in older adulthood 17.
Studies have been mixed, however, with some studies reporting no association between childhood trauma and worse cognitive performance in later life18 and other studies reporting childhood adversity was protective against decline on a cognitive performance composite score over time19.
No research has examined the association between cognitive performance and childhood trauma in clinical samples of older adults with anxiety disorders, despite the fact that anxious older adults typically report higher rates of childhood trauma compared to psychiatrically healthy older adults20.
Neurobiological mechanisms involving chronic inflammation21, decreased neuroplasticity22, and epigenetic modification of stress-related pathways23 may explain the possible association between early life trauma and poorer cognitive performance in later life.
The hypothalamus-pituitary-adrenal (HPA) axis is thought to play a particularly important role in this association.
The HPA axis activates under stress, resulting in elevated levels of the cortisol hormone and certain severe and/or chronic stressors, particularly if experienced early in life, may permanently alter HPA axis function24.
Studying this association is important, as chronically elevated cortisol levels have been associated with worse neuropsychological performance in later life25, 26.
The purpose of this study was to investigate the association between childhood trauma, cortisol, and cognitive performance in two samples of older adults.
The discovery cohort consisted of a case-comparison sample of older adults with GAD and age-equated psychiatrically healthy comparison subjects, and the replication sample consisted of older adults with a major depressive disorder or an anxiety disorder.
We hypothesized that childhood trauma would be associated with worse processing speed, attention, memory, and executive functioning.
We also hypothesized that childhood trauma would be associated with higher cortisol levels and cortisol would mediate the relationship between childhood trauma and cognitive performance.
Results
Sample characteristics and frequency of childhood trauma
Table 1 provides descriptive statistics of the discovery and replication samples. Compared to the discovery sample, participants in the replication sample completed significantly more years of education. When compared to the two clinical groups, the healthy comparison participants endorsed significantly less worry and less childhood trauma. Table 2 presents the frequencies of each self-reported childhood trauma.
Table 1.
Descriptive statistics and comparisons between the health comparisons (N=19) and GAD (N=57) participants in the discovery sample and replication sample (N=48)
| Discovery Sample M (SD) OR % (N) | Replication sample M (SD) or % (N) (N=48) | Omnibus F Or chi- square | Compare F GAD and Replication | ||
|---|---|---|---|---|---|
| Comparison (N=19) | GAD (N=57) | ||||
| Age | 75.32 (6.96) | 71.75 (8.32) | 72.33 (5.31) | 1.27 | 0.17 |
| Percent Women | 78.90 (15) | 68.40 (39) | 64.60 (31) | 1.30 | 0.17 |
| Education (Years) | 14.95 (2.55) | 14.12 (2.64) | 16.08 (2.70) | 6.25* | 14.10** |
| Childhood trauma full | 2.68 (3.00) | 4.61 (3.92) | 5.51 (3.79) | 7.33** | 1.44 |
| Childhood trauma general, physical, and sexual abuse | 1.89 (1.91) | 3.24 (2.60) | 3.55 (2.59) | 4.91* | 0.39 |
| Psychiatric and medical comorbidity | |||||
| PSWQ-A | 12.63 (4.52) | 26.18 (7.59) | 25.52 (7.55) | 23.26** | 0.16 |
| Hamilton Depression | 1.37 (1.67) | 11.12 (3.41) | … | 143.20** | … |
| PROMIS Depression | … | … | 18.08 (7.66) | … | … |
| CIRS-G | 8.42 (2.80) | 8.91 (3.95) | 7.77 (3.17) | 1.25 | 2.60 |
| PTSD Diagnosis | 0.0 (0) | 3.50 (2) | 4.30 (2) | 0.80 | 0.04 |
| Cognitive performance | |||||
| RBANS | |||||
| Total SS | 97.37 (10.13) | 94.25 (14.70) | … | 0.74 | … |
| Immediate Memory SS | 103.11 (15.37) | 97.60 (15.22) | … | 1.86 | … |
| Visuospatial Ability SS | 91.21 (18.12) | 83.16 (16.86) | … | 3.13 | … |
| Delayed Memory SS | 97.95 (11.28) | 94.72 (14.31) | … | 2.13 | … |
| Attention SS | 102.42 (10.52) | 101.33 (17.58) | … | 0.07 | … |
| Language SS | 97.26 (8.98) | 102.18 (13.68) | … | 0.80 | … |
| Coding SS | 103.04 (14.56) | 96.74 (19.09) | … | 1.73 | … |
| WAIS-III LN Sequencing SS | 104.47 (11.77) | 105.18 (11.30) | … | 0.05 | … |
| D-KEFS Sorting SS | 114.21 (14.84) | 111.05 (15.29) | … | 0.62 | … |
| Digit Span SS | 101.11 (16.78) | 101.73 (17.75) | 92.88 (15.41) | 5.67* | 7.29** |
| Trenerry Stroop Interference SS | 95.50 (14.90) | 94.57 (14.95) | … | 0.05 | … |
| Digit Vigilance Time SS | … | … | 95.31(13.26) | … | … |
| Digit Vigilance Errors SS | … | … | 92.84 (18.90) | … | … |
| List Learning Immediate Recall (# words) | … | … | 26.29 (7.77) | … | … |
| List Learning Delayed Recall (# words) | … | … | 5.12 (2.94) | … | … |
| Paragraph 1 Immediate Recall (# words) | … | … | 17.46 (6.15) | … | … |
| Paragraph 2 Immediate Recall (# words) | … | … | 16.23 (5.66) | … | … |
| Paragraph 1 Delayed Recall (# words) | … | … | 13.54 (6.63) | … | … |
| Paragraph 2 Delayed Recall (# words) | … | … | 12.41 (5.56) | … | … |
| D-KEFS Stroop Color-Word SS | … | … | 101.35 (18.27) | … | … |
| D-KEFS Verbal Fluency SS | … | … | 101.25 (16.71) | … | … |
| Max Cortisol | 3.16 (1.23) | 5.23 (1.78) | 4.99 (2.45) | 6.21* | 0.27 |
Open in a separate windowaChildhood trauma full = composite score of all subscales from the Early Trauma Inventory Self Report-Short Form (ETISR-SF)bChildhood trauma general, physical, and sexual abuse = composite score of these subscalescThe symbol … represents not applicabled *indicated p < 0.05e **indicates p < 0.01
Abbreviations: CIRS = Cumulative Illness Rating Scale-Geriatrics; D-KEFS = Delis-Kaplan Executive Function System; PSWQ-A = Penn State Worry Questionnaire- Abbreviated; SS = standard score; WAIS-III = Wechsler Adult Intelligence Scale version III
Table 2.
Items and frequency of endorsing each item from the Early Trauma Inventory Self Report- Short Form (ETISR-SF) between the healthy comparison (N=19) and GAD (N=57) participants in the discovery sample and replication sample (N=48)
| Traumatic Event | Discovery | Replication | |
|---|---|---|---|
| Comparison % (n) | GAD % (n) | % (n) | |
| General Trauma | |||
| 1. Exposed to life-threatening natural disaster? | 5.3 (1) | 1.8 (1) | 20.8 (10) |
| 2. Were you ever exposed to a life-threatening event due to war, terrorist attack, or political conflict? | 0.0 (0) | 1.8 (1) | 10.4 (5) |
| 3. Were you involved in a serious accident? | 0.0 (0) | 14.0 (8) | 25.0 (12) |
| 4. Did you ever suffer a serious personal injury or illness? | 31.6 (6) | 26.8 (15) | 29.2 (14) |
| 5. Did you ever experience the death or serious illness of a parent or a primary caretaker? | 26.3 (5) | 43.9 (25) | 40.0 (19) |
| 6. Did you experience the divorce or separation of your parents? | 5.3 (1) | 10.5 (6) | 16.7 (8) |
| 7. Did you experience the death or serious injury of a sibling? | 10.5 (2) | 12.3 (7) | 16.7 (8) |
| 8. Did you experience the death or serious injury of a friend? | 15.8 (3) | 38.6 (22) | 31.3 (15) |
| 9. Did you ever witness violence towards others, including family members? | 15.8 (3) | 26.3 (15) | 27.1 (13) |
| 10. Did anyone in your family ever suffer from mental or psychiatric illness or have a “breakdown”? | 5.3 (1) | 29.8 (17) | 25.0 (12) |
| 11. Did your parents or primary caretaker have a problem with alcoholism or drug abuse? | 15.8 (3) | 28.1 (16) | 25.0 (12) |
| 12. Did you ever see someone murdered? | 0.0 (0) | 1.8 (1) | 0.0 (0) |
| Physical Punishment | |||
| 1. Were you ever slapped in the face with an open hand? | 15.8 (3) | 15.8 (9) | 25.0 (12) |
| 2. Were you ever burned with hot water, a cigarette or something else? | 0.0 (0) | 0.0 (0) | 0.0 (0) |
| 3. Were you ever punched or kicked? | 0.0 (0) | 7.0 (4) | 16.7 (8) |
| 4. Were you ever hit with an object that was thrown at you? | 0.0 (0) | 7.0 (4) | 8.3 (4) |
| 5. Were you ever pushed or shoved? | 5.3 (1) | 12.3 (7) | 10.4 (5) |
| Emotional Abuse | |||
| 1. Were you often put down or ridiculed? | 26.3 (5) | 33.3 (19) | 50.0 (24) |
| 2. Were you often ignored or made to feel that you didn’t count? | 10.5 (2) | 26.3 (15) | 43.8 (21) |
| 3. Were you often told you were no good? | 10.5 (2) | 15.8 (9) | 27.1 (13) |
| 4. Most of the time were you treated in a cold, uncaring way or made to feel like you were not loved? | 21.1 (4) | 26.3 (15) | 43.8 (21) |
| 5. Did your parents or caretakers often fail to understand you or your needs? | 10.5 (2) | 35.1 (20) | 43.8 (21) |
| Sexual Events | |||
| 1. Were you ever touched in an intimate or private part of your body (e.g. breasts, thighs, genitals) in a way that surprised you or made you feel uncomfortable? | 15.8 (3) | 19.3 (11) | 22.9 (11) |
| 2. Did you ever experience someone rubbing their genitals against you? | 5.3 (1) | 10.5 (6) | 10.4 (5) |
| 3. Were you ever forced or coerced to touch another person in an intimate or private part of their body? | 5.3 (1) | 5.3 (3) | 2.1 (1) |
| 4. Did anyone ever have genital sex with you against your will? | 5.3 (1) | 1.8 (1) | 4.2 (2) |
| 5. Were you ever forced or coerced to perform oral sex on someone against your will? | 0.0 (0) | 1.8 (1) | 0.0 (0) |
| 6. Were you ever forced or coerced to kiss someone in a sexual rather than an affectionate way? | 5.3 (1) | 0.0 (0) | 4.2 (2) |
Association of childhood trauma with cognitive performance
Table 3 displays results from the linear regression models investigating self-reported trauma exposure in relation to cognitive performance in the discovery sample. After controlling for age, sex, education, physical health, and GAD status, childhood trauma was associated with a worse total index score on the RBANS.
This score was driven by worse performance on the Attention Index. Childhood trauma was also associated with worse Coding, Digit Span, Stroop Color-Word Interference performance, and sorting abilities.
After adjusting for PTSD diagnosis, depressive and worry symptoms, childhood trauma remained associated with worse attention performance, Stroop Color-Word Interference, and sorting performance. When examining these associations in the subsample for which APOE data was available the magnitude of the results did not significantly change when we adjust for APOE (see supplemental table 1).
See figure 1 for graphs of the estimated test performance for individuals with no trauma, those reporting the mean number of traumatic events, and those reporting high number of traumatic events.
Graph depicting the estimated performance, with 95% confidence interval, on cognitive tests and cortisol AUC by self-reported childhood trauma controlling for age, sex, education, health, GAD status, PTSD diagnosis, worry, and depressive symptoms.
a Participants were divided into three categories: no traumatic events (ETISR-SF = 0), mean number of traumatic events (ETISR-SF = 4) or high levels of traumatic events (+1 SD, ETISR-SF=8).
Table 3.
Discovery sample results from the hierarchical linear regression examining number of traumatic events reported in relation to cognitive test performance and hierarchical linear regression examining number of general, physical, and sexual traumatic events reported in relation to cognitive test performance.
| Predictor: Childhood trauma full composite score Outcome: | Step 1: Adjust for demographics and health (N=76) | Step 2: Adjust for depression and anxiety (N=76) | Step 3: Adjust for cortisol (N=57) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Trauma β | 95% CI | p | Trauma β | 95% CI | p | Trauma β | 95% CI | p | |
| RBANS | |||||||||
| Total IS | −0.99 | −1.84 to −0.14 | 0.03 | −0.81 | −1.67 to 0.06 | 0.07 | −0.60 | −1.65 to 0.44 | 0.25 |
| Attention IS | −1.70 | −2.66 to −0.73 | <0.01 | −1.59 | −2.59 to −0.58 | <0.01 | −1.55 | −2.80 to −0.30 | 0.02 |
| Coding SS | −1.33 | −2.47 to −0.20 | 0.02 | −1.10 | −2.27 to 0.07 | 0.06 | −0.88 | −2.34 to 0.59 | 0.23 |
| Digit SS | −1.53 | −2.60 to −0.47 | <0.01 | −1.48 | −2.60 to −0.36 | 0.01 | −1.68 | −3.13to −0.23 | 0.02 |
| Immediate Memory IS | −0.54 | −1.55 to 0.47 | 0.29 | −0.37 | −1.40 to 0.67 | 0.48 | 0.03 | −1.35to 1.41 | 0.97 |
| Visuospatial Function IS | −0.74 | −1.83 to 0.34 | 0.18 | −0.61 | −1.74 to 0.52 | 0.28 | −0.69 | −2.11 to 0.72 | 0.33 |
| Delayed Memory IS | −0.61 | −1.49 to 0.27 | 0.17 | −0.58 | −1.50 to 0.34 | 0.21 | −0.32 | −1.47 to 0.83 | 0.58 |
| Language IS | −0.12 | −0.90 to 0.66 | 0.76 | 0.06 | −0.72 to 0.85 | 0.87 | 0.37 | −0.64 to 1.38 | 0.46 |
| Letter Number Sequencing SS | −0.62 | −1.36 to 0.12 | 0.10 | −0.47 | −1.22 to 0.29 | 0.22 | −0.22 | −1.20 to 0.76 | 0.66 |
| Confirmed Correct Sorts SS | −1.14 | −2.04 to −0.24 | 0.01 | −1.19 | −2.14 to −0.24 | 0.01 | −1.11 | −2.08 to −0.13 | 0.03 |
| Stroop Color-Word interference SS | −1.00 | −1.92 to −0.08 | 0.03 | −1.00 | −1.95 to −0.04 | 0.04 | −1.61 | −2.83 to −0.39 | 0.01 |
| Predictor: Childhood general, physical, and sexual trauma composite score Outcome: | Step 1: Adjust for demographics and health (N=76) | Step 2: Adjust for depression and anxiety (N=76) | Step 3: Adjust for cortisol (N=57) | ||||||
| Trauma β | 95% CI | p | Trauma β | 95% CI | p | Trauma β | 95% CI | p | |
| RBANS | |||||||||
| Total IS | −1.50 | −2.78 to −0.22 | 0.02 | −1.25 | −2.54 to 0.03 | 0.06 | −0.81 | −2.39 to 0.76 | 0.30 |
| Attention IS | −2.31 | −3.78 to −0.84 | <0.01 | −2.14 | −3.66 to −0.62 | <0.01 | −1.89 | −3.80 to 0.02 | 0.053 |
| Coding SS | −1.82 | −3.54 to −0.11 | 0.04 | −1.55 | −3.30 to 0.18 | 0.08 | −1.11 | −3.32 to 1.09 | 0.32 |
| Digit SS | −2.39 | −3.99 to −0.80 | <0.01 | −2.30 | −3.96 to −0.64 | <0.01 | −2.45 | −4.63 to −0.28 | 0.03 |
| Immediate Memory IS | −0.87 | −2.40 to 0.64 | 0.25 | −0.59 | −2.14 to 0.95 | 0.45 | −0.01 | −2.08 to 2.06 | 0.99 |
| Visuospatial Function IS | −0.96 | 2.60 to 0.67 | 0.24 | −0.80 | −2.49 to 0.88 | 0.35 | −0.63 | −2.77 to 1.50 | 0.53 |
| Delayed Memory IS | −1.15 | −2.46 to 0.16 | 0.08 | −1.06 | −2.42 to 0.30 | 0.12 | −0.61 | −2.33 to 1.10 | 0.48 |
| Language IS | −0.22 | −1.39 to 0.95 | 0.71 | −0.05 | −1.22 to 1.12 | 0.93 | 0.40 | −1.12 to 1.92 | 0.60 |
| Letter Number Sequencing SS | −0.93 | −2.05 to 0.18 | 0.10 | −0.68 | −1.80 to 0.44 | 0.23 | −0.09 | 1.56 to 1.39 | 0.91 |
| Confirmed Correct Sorts SS | −1.99 | −3.33 to −0.66 | <0.01 | −2.00 | −3.39 to −0.60 | <0.01 | −1.79 | −3.24 to −0.34 | 0.02 |
| Stroop Color-Word interference SS | −1.68 | −3.07 to −0.29 | 0.02 | −1.72 | −3.14 to 0.30 | 0.02 | −2.73 | −4.56 to −0.89 | <0.01 |
Open in a separate windowaStep 1 includes age, sex, years of education, cirs-g, PTSD and GAD statusbStep 2 adds PSWQ-A and HRSD to the modelcStep 3 adds max cortisol to the modeldEstimates below p <0.05 are bolded
Abbreviations: CI = Confidence interval; GAD= Generalized Anxiety Disorder; HRSD = Hamilton Rating Scale for Depression; PSWQ-A= Abbreviated Penn State Worry Questionnaire; PTSD = Posttraumatic Stress Disorder; RBANS = Repeatable Battery for the Assessment of Neuropsychological Status; IS=Index Score; SS= Standard Score; Trauma β = estimated effect of Early Trauma Inventory Self Report- Short Form score on cognitive performance.
When examining composite of general, physical, and sexual traumas in relation to cognitive performance, a similar pattern of results emerged. The number of self-reported events was associated with a lower Total Index Score on the RBANS total and attention indices, Coding subtest, Digit Span subtest, Stroop Color-Word Interference, and Sorting. The effect size of the association between childhood trauma and worse cognitive performance was generally larger when only examining the composite of general, physical, and sexual traumatic events compared to the ETISR-SF total composite score.
In the replication sample, a similar pattern of results of childhood trauma on cognitive performance emerged (see Table 4). The number of traumatic events from the full childhood trauma inventory was associated with worse Digit Span performance. Number of traumatic events was also associated with worse verbal fluency and D-KEFS Color-Word Interference. Similar to the discovery sample, when examining only the total number of general, physical, and sexual traumas reported, the effect size of trauma approximately doubled for most outcomes.
Table 4.
Replication sample results from the hierarchical linear regression examining number of traumatic events reported in relation to cognitive test performance and hierarchical linear regression examining number of general, physical, and sexual traumatic events reported in relation to cognitive test performance in the replication sample (N =48).
| Childhood trauma full composite score | Childhood general, physical, and sexual trauma composite score | |||||
|---|---|---|---|---|---|---|
| Outcome | Trauma β | 95% CI | p | Trauma β | 95% CI | p |
| Attention | ||||||
| Digit Span (SS) | −1.40 | −2.74 to −0.06 | 0.04 | −2.68 | −4.61 to −0.74 | <0.01 |
| Digit Vigilance Time (SS) | −0.50 | −1.78 to 0.80 | 0.44 | −1.72 | −3.61 to 0.17 | 0.07 |
| Digit Vigilance Errors (SS) | −1.15 | −2.96 to 0.66 | 0.21 | −0.89 | −3.69 to 1.91 | 0.53 |
| Memory | ||||||
| List Immediate. Recall (# words) | −0.46 | −1.13 to 0.20 | 0.17 | −0.92 | −1.91 to 0.07 | 0.06 |
| List Delay Recall (# words) | −0.18 | −0.42 to 0.07 | 0.15 | −0.38 | −0.74 to −0.02 | 0.04 |
| Paragraph 1 Imm. Recall (# pieces) | −0.16 | −0.74 to 0.41 | 0.57 | −0.38 | −1.24 to 0.47 | 0.37 |
| Paragraph 2 Imm. Recall (# pieces) | −0.25 | −0.77 to 0.27 | 0.34 | −0.36 | −1.14 to 0.42 | 0.35 |
| Paragraph 1 Delay Recall (# pieces) | −0.11 | −0.70 to 0.48 | 0.72 | −0.44 | −1.32 to 0.43 | 0.32 |
| Paragraph 2 Delay Recall (# pieces) | −0.29 | −0.80 to 0.22 | 0.26 | −0.47 | −1.24 to 0.30 | 0.22 |
| Executive Functioning | ||||||
| Verbal Fluency Raw | −1.68 | −3.19 to −0.17 | 0.03 | −2.69 | −4.96 to −0.41 | 0.02 |
| Stroop Interference | −1.87 | −3.55 to −0.19 | 0.03 | −3.30 | −5.79 to −0.82 | 0.01 |
aBoth models include age, sex, years of education, cirs-g, PTSD diagnosis, GAD, PSWQ-A, and PROMIS Depression as covariates.
Abbreviations: CI = Confidence interval; GAD= Generalized Anxiety Disorder; PROMIS = Patient-Reported Outcomes Measurement Information System; PSWQ-A= Abbreviated Penn State Worry Questionnaire; PTSD = Posttraumatic Stress Disorder; RBANS = Repeatable Battery for the Assessment of Neuropsychological Status; SS= Standard Score; Trauma β = estimated effect of Early Trauma Inventory Self Report- Short Form score on cognitive performance.
Cortisol and childhood trauma
To examine whether cortisol levels were explaining the association between childhood trauma and worse cognitive performance, we first examined the association between childhood trauma and cortisol level.
The total number of events reported from the full childhood trauma inventory was not associated with higher peak (β = 0.06, 95% CI = −0.09; 0.20, p = 0.44) or area under the curve (β = 0.13, 95% CI = −0.73; 0.10, p = 0.76) cortisol. Trauma was also not associated with either peak or AUC cortisol when examining a composite of general, physical, or sexual traumas (p > 0.31).
Discussion
This cross-sectional study investigated the number of self-reported traumatic events experienced in childhood in relation to cognitive functioning in two samples of older adults.
Childhood traumatic events were associated with worse attention and poorer executive functioning.
These associations were independent of physical health, demographic variables, APOE ε4 genotype, and depressive and worry symptom severity.
Furthermore, the effect of trauma on cognitive performance was larger for those reporting general, physical, or sexual trauma, compared to emotional abuse. This association was present in three different groups of older adults: older adults with GAD, psychiatrically healthy older adults, and older adults with anxiety or depressive disorders.
These findings supported our initial hypothesis and were present in both the discovery and replication samples. Our second hypothesis was not supported, as the number of traumatic events in childhood was not associated with cortisol levels, and cortisol levels did not explain the association between childhood trauma and cognitive performance.
Early childhood trauma was not associated with current cortisol levels and cortisol did not explain the association between childhood trauma and worse cognitive performance, highlighting the potential importance of other physiological mechanisms.
Prior research has also documented associations between childhood trauma and increased inflammation in older adulthood and documented functional and structural changes to the hippocampus and frontal cortex in younger adults47.
Childhood trauma may affect neurotrophic factors such as Brain Derived Neurotrophic Factor48.
As a result, upon entering older adulthood and experiencing age-related cognitive decline, individuals with childhood trauma may be at risk for greater declines in cognitive performance due to lower brain and/or cognitive reserve stemming from the potential adverse effects of childhood trauma on brain development.
Future studies need to examine the association between childhood trauma and physiological biomarkers further. The extent to which childhood trauma is associated with structural and functional neuroanatomy, inflammation, and neurotrophic factors in older adulthood are important unanswered questions.
It is well established that depression is associated with worse cognitive performance in later life49, and a growing body of literature is demonstrating the association between anxiety and worse cognitive performance50, 51.
Future research should also examine the extent to which childhood trauma mediates the association between anxiety and cognitive performance in later life. The extent to which childhood trauma is associated with structural and functional neuroanatomy in older adulthood is an important unanswered question that needs to be studied further. Lastly, future research needs to examine the role of resilience factors, such as coping style and cognitive hardiness52, in moderating this association.
This study has several limitations. Childhood trauma was measured with retrospective recall. It is possible that individuals who are particularly distressed, or more cognitively impaired, may exhibit a recall bias and be more likely to endorse traumatic events.
Measures of biomarkers other than cortisol, such as inflammatory cytokines, neurotrophic factors, and neuroimaging, were not available, limiting the ability to examine other potential biological mechanisms. Third, although we controlled for diagnosis of PTSD, we did not have a measure of PTSD symptoms.
Fourth, the cognitive and clinical measures were not uniform across both studies, although they assessed similar domains with similar tests. Fifth, we were unable to examine the modifying role of ethnicity, and past research has found that the role of early life adversity and cognitive performance in later life may be different for different racial groups12.
Lastly, the cross-sectional design does not allow us to make conclusions of causality.
In conclusion, findings demonstrate the public health importance of preventing childhood trauma and developing interventions to help younger and older individuals with history of childhood trauma. Additionally, cortisol levels were not associated with childhood trauma, highlighting the importance of other possible physiological mechanisms.
Source:
Alzheimer’s Research UK
