The relationship between childhood trauma and poor sleep health in adulthood

Abstract

Objective

Childhood trauma has been related to adverse behavioral, mental, and health outcomes later in life. Sleep may be a potential mechanism through which childhood trauma is related to adverse health. The current retrospective study aimed to characterize the relationship between childhood trauma exposure and sleep health, a novel multidimensional measure of sleep.

Methods

Participants (N = 161; Mean(SD) age = 59.85(9.06); 67.7% female) retrospectively reported trauma exposure using the Trauma History Questionnaire. Childhood trauma was defined as the number of reported traumatic events prior to 18-years of age. Trauma exposure post-18 years of age and across the lifespan was also recorded. Sleep health was derived both from diary- and actigraphy-assessed measures of sleep regularity, timing, efficiency, and duration, subjective sleep satisfaction, and daytime sleepiness from the Epworth Sleepiness Scale. The relationships between childhood trauma exposure and sleep health was examined using hierarchical linear regression, controlling for relevant covariates.

Results

In unadjusted models, a greater number of childhood trauma exposures was associated with poorer diary- and actigraphy-measured sleep health in adulthood. After adjustment for current stress, depression history, and other sociodemographic covariates, greater childhood trauma remained significantly associated with poorer sleep health (diary: β = −0.20, ΔR² = .032; actigraphy: β = −0.19, ΔR² = .027). Trauma exposure post-18 years of age and across the lifespan did not relate to diary- or actigraphy-based sleep health.

Conclusion

Childhood trauma may affect sleep health in adulthood. These findings align with the growing body of evidence linking childhood trauma to adverse health outcomes later in life.

Introduction

The prevalence of childhood trauma is alarmingly high (1,2) and is increasingly being considered a medically-relevant contributor to morbidity and mortality given the accumulating evidence showing relationships between childhood trauma and adverse behavioral (3,4), mental (5,6), and health outcomes (7,8). Because of this, it is important that we characterize the pathways through which childhood trauma affects health.

One such pathway may be sleep. Several retrospective cohort studies report a significant “dose-response” relationship between childhood trauma and sleep problems later in life (9–11). A systematic review of these studies noted significant effects of trauma on sleep apnea, nightmare-related distress, and insomnia symptoms (12). Sleep and sleep problems have also been related to a range of adverse behavioral, mental, and health outcomes similar to those associated with childhood trauma (13–16). Finally, normative sleep has been shown to be critical for physiological (17,18) and emotional (19) restitution and may represent an endogenous nocturnal coping mechanism necessary to stave off the physiological and emotion consequences of traumatic stress not adequately mitigated by psychological and behavioral coping mechanisms employed during wakefulness. For example, the quantity and quality of consolidated rapid-eye movement sleep following a traumatic experience has been associated with the presence and severity of post-traumatic stress disorder symptoms at follow-up (20,21). Consequently, chronically compromised sleep may contribute to long-term deterioration of physiological and emotional health. For these reasons, it is possible that sleep may be a pathway through which childhood trauma promotes poor health later in life and, if so, may also represent a potential behavioral target for intervention to reduce the negative, long term health effects of childhood trauma. To explore this possibility, a better understanding of the impact of childhood trauma on sleep is necessary. However, the heterogeneity of sleep outcome variables evaluated in previous studies, including the frequent use of single-item measures, significantly limits comparisons across studies and interpretation of results with regard to how sleep outcomes relate to health (12).

Sleep health (22) is an important construct designed to conceptualize and quantify sleep in a manner consistent with other health-related variables, e.g., cardiovascular-, metabolic-, psychiatric-health. As such, sleep health has been conceptualized as a multivariable construct including measures of sleep regularity, satisfaction, timing, efficiency, and duration, and daytime sleepiness. In support, structural equation modeling has shown these six dimensions successfully load onto a single underlying latent sleep health construct (model fit indices: CFI = .995, TLI = .964, SRMR = .026; 23). As such, this construct offers several advantages over traditionally evaluated, single-item measures of sleep. First, sleep health is a continuous measure that frames sleep in a positive light and can be implemented and interpreted across a range of health-related contexts and populations (e.g., patients, healthcare providers, educators). Second, quantifying sleep health on a continuous scale makes available a wealth of research opportunities not previously available in experimental designs where sleep health was dichotomized on the basis of disease presence or absence (e.g., insomnia, sleep apnea). Sleep health is not simply the absence of a clinical sleep disorder, but can vary incrementally between individuals allowing for the exploration of correlates of individual differences in sleep health. Third, each of the six sleep health components has been shown to independently relate to health outcomes (please see Table 1 in reference (22)) including cardiovascular and metabolic diseases (24), psychiatric illness (13), cognitive performance (25), and mortality (26–29). Consequently, sleep health establishes a framework for future research to characterize sleep-related pathways to disease. To date, independent studies have significantly related sleep health to blood pressure and BMI (23), self-reported health (30), and clinically-significant depression symptoms (31). Finally, sleep health aligns with broader health agendas and public policy and represents a target for the improvement of sleep, regardless of whether a clinically significant sleep disorder is present. The current retrospective study aimed to examine the relationships between retrospectively-reported childhood trauma exposure and sleep health in adulthood. Based on evidence linking childhood trauma to suboptimal sleep and sleep disorders (12), it was hypothesized that individuals reporting increased trauma exposure during childhood would have poorer subjectively- (e.g., diary) and behaviorally-measured (e.g., actigraphy) sleep health.

Methods

Participants

The current study represents a secondary analysis of data collected to analyze the relationships between sleep, depression and cardiovascular disease risk. Participants were recruited from a larger cohort of four separate studies conducted at the University of Pittsburgh between 1982–1999. One study [MH024652 (32)] examined sleep in adults without personal or first-degree family history of psychiatric disorders; the remaining three studies assessed sleep in individuals with major depressive disorder (MDD) [MH029618, MH049115, MH041884 (33–36)]. Of the 339 participants re-contacted from 2010–2014 to participate in the current study, 177 consented. For the purpose of the present report, data were available for 161 participants. This study was approved by the University of Pittsburgh Institutional Review Board and all participants provided written informed consent. Financial compensation was provided to participants.

Procedure

Upon entry to the current study participants underwent mental and physical health assessments. During the mental health assessment, lifetime psychiatric history was assessed by trained clinicians using the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I; 37). During the physical health assessment, a detailed medical history was taken, along with objective anthropometric measurements [e.g., body mass index (BMI) in kg/m²]. Participants were instructed to complete a daily diary and wear a wrist actigraph over a 2-week period to track health and sleep-wake behaviors. Finally, participants returned to the laboratory and completed a battery of questionnaires, a cardiovascular assessment, and a 2-night polysomnography (PSG) laboratory sleep study (cardiovascular and PSG data presented elsewhere (38)).

Study Measures

Sleep Health

Our protocol afforded the opportunity to calculate both subjective sleep health, as assessed by daily diary self-reports, and behaviorally-assessed sleep health, as assessed by wrist actigraphy. For the subjective assessment of sleep health, the six components of sleep health (22) were derived from the Pittsburgh Sleep Diary (39), which participants completed for up to two weeks (mean = 9.75 days, range: 6–14 days), and the Epworth Sleepiness Scale (40), a validated measure of daytime sleepiness. Actigraphy-based sleep health was derived from actigraphy-assessed measures of sleep duration, regularity, timing, and efficiency; the sleep quality and daytime sleepiness components were measured by self-report, as appropriate to the protocol. Participants wore the Actiwatch-2^® actigraph (Philips-Respironics, Bend, OR) 24-hours per day on the non-dominant wrist for a minimum of 7 consecutive days (mean = 9.29 days, range: 7–14 days). Nocturnal sleep intervals were identified by automated algorithms and confirmed by event markers and daily diaries. Actigraphy data were analyzed with Actiware 6.1 software, using the medium sensitivity setting and scored in 30-second epochs. Table 1 outlines the operationalization of the six components of sleep health and the rationale for chosen cut-off values. In general, cut-off values were chosen from larger, community-based samples of older adults and from published meta-, and psychometric-analyses (41–44). Each component of sleep health was dichotomized as 0 or 1, with a higher score indicting better sleep health. Thus, total sleep health scores range from 0–6, with a value of 6 indicating optimal sleep health and a value of 0 indicating poorest sleep health.

Table 1.

Sleep Health Score Derivation

Sleep Health Component	Operationalization	Scoring	Score 0a (N)	Score 1a (N)	Cut-off Reason
Regularity	Standard deviation of calculated sleep midpoint from daily sleep diary	Male ≤ 29min = 1 >29min = 0 Female ≤ 26min = 1 >26min = 0	129 108	32 30	Median values from MrOS and SOF studies; larger community-based samples of elderly participants (41)
Satisfactionb	Average self-reported sleep quality from daily sleep diary, 0 (very bad) - 100 (very good) scale	Median Split > 66.66 = 1 < 66.66 = 0	81	80	Median split in current sample
Sleepinessb (daytime)	Total score on Epworth Sleepiness Scale	≤ 10 = 1 > 10 =0	30	131	A cut-off of 10 has been previously found to indicate excessive daytime sleepiness (40)
Timing	Average of calculated sleep midpoint from daily sleep diary, midnight = 0min	Male ≥147 OR ≤ 218min = 1 <147 OR > 218min = 0 Female ≥161 OR ≤ 234min = 1 <161 OR > 234min = 0	91 80	70 58	Median values from MrOS and SOF studies; larger community-based samples of elderly participants (41)
Efficiency	Average calculated sleep efficiency from daily sleep diary	> 85% = 1 < 85% = 0	31 33	130 105	Sleep restriction therapy aims to achieve sleep efficiency of 85% (42,43)
Duration	Average self-reported sleep duration from daily sleep diary	7hr – 8hr = 1 < 7hr = 0 > 8hr = 0	104 87	57 51	Meta-analysis of 27 elderly cohort studies (N > 70,000) reported increased all-cause mortality in sleep durations ≤ 7hrs and ≥ 8hrs (44)
Total Sleep Health Score Range: 0 (poor) – 6 (optimal)

Notes:

^aTop number represents diary-based values, bottom number represents actigraphy-based values,

^bNo actigraphy based-values are provided because these subjective measures were used in both diary- and actigraphy-based sleep health derivations

Trauma History

Trauma history was measured using the Trauma History Questionnaire (THQ; 45,46). The THQ is a retrospective self-reported inventory of “high-magnitude stressor events” applicable to the general population. The THQ is relatively stable over time. The stability coefficient for total number of trauma events is 0.76 over 2 weeks (47), while values for individual trauma categories range from 0.54–0.92 over a 2–3 month period (48). The inventory is comprised of 24 yes/no questions, grouped into three categories: crime-related trauma, general disasters, and unwanted physical/sexual trauma. When an event is endorsed, participants provide frequency of occurrence and age of exposure(s). To test the hypothesis of the current study, trauma exposure before 18 years of age (number of endorsed event types multiplied by reported frequency of each event type) was calculated from the THQ. Trauma exposures across the lifetime and after 18 years of age were also quantified.

Possible Covariates

Current stress at the time of study participation was measured using a self-reported inventory of chronic and ongoing stress. Participants rated the extent to which each of eleven stressor categories (e.g., financial, relationship, work-related stress) caused stress over the prior 12 months. The rating scale was 0 (no stress) to 3 (most upsetting stress); total scores ranged from 0–33. Diagnostic information from the original four studies and the mental health assessment of the current study were used to determine lifetime depression status. Individuals with no history of MDD at either time point were categorized as having “no MDD history.” Those who endorsed a single major depressive episode across the lifespan were grouped in a “single episode MDD” category. Finally, participants endorsing more than one lifetime major depressive episode were coded as “recurrent MDD.” In follow-up supplementary analyses, Hamilton Depression Rating Scale scores, less sleep items (items 4–6), were used in lieu of lifetime depression history (49,50). The total score from a modified 12-item version of the Interpersonal Support Evaluation (51) was used to assess perceived availability of interpersonal support. Participants responded to counterbalanced (positive versus negative) items using a 4-point Likert-style scale anchored by “definitely false” and “definitely true,” with higher scores indicating a greater availability of interpersonal support (range: 0–48). This scale has been shown to display high internal reliability (Cronbach’s α range = 0.88–0.90) and acceptable test-retest reliability (r-range: 0.70–0.87). Socio-economic status was measured using education level and subjective social standing. Education was self-reported using an 8-level scale that ranged from middle school to post-graduate school. Subjective social standing was assessed using a 9-rung social standing ladder (52). Participants were asked to rate their perceived social standing, relative to their community ranging from 1 (highest) to 9 (lowest). Daily alcohol use, in drinks per day, was assessed via daily diary. Sex, age, and BMI, and time 1 parent cohort were also identified as potential covariates.

Statistical Analysis

A small group of participants (n = 13) indicated that their traumatic experiences before the age of 18 were too numerous to count. To preserve rank order of reported trauma exposures in statistical analyses, these participants were given a value equal to the 95^th percentile (53). It should be noted that, in sensitivity analyses, removal of these participants did not change the reported results (results not shown). Hierarchical linear regression was used to examine the relationship between childhood trauma exposure and both diary- and actigraphy-based sleep health. An unadjusted model tested whether increased childhood trauma related to poorer sleep health. In a second model, adjustments were made for time 1 parent cohort and covariates found to be associated with sleep health in the present sample: lifetime depression status, current stress, subjective social standing, and interpersonal support, p values < .10. Finally, a fully adjusted model was tested that included all a priori potential covariates. This series of analyses was repeated using both lifetime trauma exposure and trauma exposure post-18 years of age as independent variables. Finally, a series of follow-up analyses using binary logistic regression were undertaken to determine which dimensions of sleep health were significantly associated with childhood trauma. An unadjusted model was tested followed by two adjusted models that were identical to those outlined above. Lifetime depression status was dummy-coded before inclusion in all regression analyses. All data are presented as mean (standard deviation), unless otherwise stated. Statistical analyses were carried out using SPSS version 23 (IBM; Chicago, IL). For all analyses, statistical significance was operationalized as α < .05. For analyses with actigraphy-based sleep health, 138 participants were available for analysis; discrepancies in participant number (as reported in the tables) represent occasional missing data points.

Results

Sample Characteristics

Participants (N = 161; 67.7% females; 98.1% non-Hispanic White) were, on average, 59.85 (9.06) years old and had a BMI of 30.11 (7.05). Fifty-nine participants had no lifetime history of MDD, 17 participants met criteria for a single lifetime MDD episode, and 66 individuals were classified as recurrent MDD patients. On average, current self-reported stress score was 5.09 (4.17) and scores of interpersonal support and subjective social standing were 41.12 (6.44) and 3.71 (1.71), respectively. Average sleep health score was 3.11 (1.18) and 3.09 (1.34) for diary- and actigraphy-based measures, respectively. Average Hamilton Depression Rating Scale score was 4.47 (4.77) and average childhood trauma exposure was 1.03 (1.77) events.

Childhood Trauma and Sleep Health

The amounts of trauma exposure reported pre- and post-18 years of age were significantly related (r = .191, p = .016), although the relationship was not particularly strong. Sample sleep characteristics and results of the linear regression analysis of childhood trauma and sleep health can be found in Tables 2 and 3, respectively. In both unadjusted and adjusted models, childhood trauma was related to both diary- and actigraphy-based sleep health such that greater childhood trauma was associated with a lower sleep health scores (β-range = −.19 to −.27; all p ≤.07). After controlling for time 1 parent cohort, gender, age, lifetime depression history, subjective social standing, BMI, current stress, daily alcohol use, and interpersonal support, diary-based sleep health decreased by −.15 units (95% CI −0.28 to −0.01, p = .031) per traumatic event (adjusted model 2, Table 3). For actigraphy-based sleep health, sleep health decreased −.17 (95% CI −0.35 to 0.01, p = .07) units per childhood traumatic event in the same fully adjusted model (adjusted model 2, Table 3). Increased childhood trauma was significantly associated with increased risk of reporting poor diary-assessed sleep efficiency after controlling for time 1 parent cohort, depression history, current stress, interpersonal support, and subjective social standing (OR = 1.34, p = .041) in sleep health component analyses. No single component of actigraphy-based sleep health was significantly related to childhood trauma (all p’s ≥ .07).

Table 2.

Sleep Characteristics of Sample

Sleep Measure	N	Diary Mean(SD)	N	Actigraphy Mean(SD)
Duration, min*	158	406.86 (61.44)	136	414.00 (52.73)
Efficiency, %*	158	89.57 (7.45)	136	87.42 (5.21)
Timing, min from midnight*	158	203.10 (64.78)	136	200.83 (62.47)
Regularity, SDmin*	158	46.59 (25.95)	136	50.17 (43.87)
WASO, min	158	28.06 (25.43)	136	42.39 (22.59)
Latency, min	158	19.41 (15.90)	136	8.70 (7.05)
Fragmentation Index	--	--	136	20.32 (9.47)
Satisfaction, 0–100*	158	65.04 (14.63)	--	--
ESS Scale Score, 0–24*	158	6.94 (3.92)	--	--

Notes: SD = standard deviation, WASO = wake time after sleep onset, ESS = Epworth Sleepiness Scale

^*included in sleep health variable,

Table 3.

Linear Regression Analysis of Trauma and Sleep Health

	N	β	t	p	ΔR2
Childhood Trauma
Diary-based Measure
Unadjusted	158	−.24	−3.11	.002	.058
Adjusted Model 1	138	−.20	−2.39	.018	.036
Adjusted Model 2	133	−.20	−2.19	.031	.032
Actigraphy-based Measure
Unadjusted	135	−.27	−3.18	.002	.070
Adjusted Model 1	118	−.21	−2.21	.029	.036
Adjusted Model 2	114	−.19	−1.84	.069	.027
Adult Trauma
Diary-based Measure
Unadjusted	159	−.08	−0.98	.33	.006
Adjusted Model 1	139	.08	0.95	.35	.006
Adjusted Model 2	134	.09	1.06	.29	.008
Actigraphy-based Measure
Unadjusted	136	.15	−1.71	.09	.021
Adjusted Model 1	119	−.04	−0.45	.66	.002
Adjusted Model 2	115	−.03	−0.36	.72	.001
Lifetime Trauma
Diary-based Measure
Unadjusted	159	−.13	−1.59	.11	.016
Adjusted Model 1	139	.04	0.41	.68	.001
Adjusted Model 2	134	.05	0.57	.57	.002
Actigraphy-based Measure
Unadjusted	136	−.20	−2.35	.020	.039
Adjusted Model 1	119	−.09	−0.94	.35	.007
Adjusted Model 2	115	−.07	−0.76	.45	.005

Model 1 adjusted for current stress, depression history, subjective social standing, time 1 parent cohort, and interpersonal support

Model 2 adjusted for model 1 + sex, age, daily alcohol use, and BMI

Lifetime Trauma and Sleep Health

Regression analyses revealed that neither lifetime trauma exposure nor trauma exposure post-18 years of age was related to diary- or actigraphy-based sleep health (all p’s > .05; Table 3).

Sensitivity Analyses

In sensitivity analyses, substituting Hamilton Depression Rating Scale scores for lifetime depression status did not change any of the reported results (Table S1, Supplemental Digital Content). Similarly, substituting subjective social standing with education did not alter the observed results (Table S2) Finally, analyses restricted to participants with complete data (e.g., all variables of interest and covariates), yielded identical results to those reported (Table S3).

Discussion

Our findings are consistent with the hypothesis that greater childhood trauma is associated with poorer sleep health in adulthood. Regression analysis revealed that increased retrospectively reported childhood trauma exposure was associated with poorer diary- and actigraphy-based sleep health, after multivariate adjustment for current stress, interpersonal support, depression history and other sociodemographic covariates. Notably, associations among trauma and sleep health were observed for childhood trauma only; neither lifetime trauma exposure nor trauma exposure after 18 years of age was significantly related to sleep health. Overall, these results align with evidence that childhood trauma may influence sleep and waking health later in life.

Several retrospective cohort studies report a significant association between childhood trauma and indices of suboptimal sleep later in life. In the Finnish Health and Social Support cohort, a significant graded association was found between family-related childhood trauma and self-reported sleep quality; endorsing multiple traumatic events significantly increased the odds of lower sleep quality later in life (9). A significant association between childhood abuse and subjective sleep quality was observed in the Midlife Development in the US cohort (10). In a California-based health maintenance organization cohort, endorsing any number of adverse childhood experiences (e.g., abuse, violence) significantly increased the odds of self-reported troubles falling asleep and daytime sleepiness (11). Reporting frequent insufficient sleep was significantly more likely in individuals reporting ≥1 adverse childhood experience in the Behavioral Risk Factor Surveillance Survey (2). Finally, a systematic review of retrospective cohort studies (12) and two prospective studies (54,55) support the hypothesis that childhood trauma is associated with indices of poor sleep later in life.

The present finding that retrospectively-reported childhood trauma is significantly related to poor sleep health, after multivariate adjustment including lifetime depression history, current stress, interpersonal support, and other covariates, replicates previous research and provides several novel additions to the literature. First, both diary- and actigraphy-based sleep health measures were associated with childhood trauma, but not adult or lifetime trauma. This finding not only speaks to the potentially heightened impact that childhood trauma has on sleep later in life, but also provides more robust evidence in support of previous findings linking childhood trauma and self-reported sleep disturbances using single-item measures. Previous research has often operationalized sleep using single-item or dichotomous sleep measures which limits the ability to interpret results vis-á-vis health, compare results across studies, and fails to capture the multidimensional nature of sleep (12,22). Sleep health, a composite measure of sleep duration, efficiency, satisfaction, timing, regularity, and daytime sleepiness, captures the multidimensional nature of sleep. The composite measure of sleep health related to childhood trauma after adjustment for a range of potential covariates including lifetime depression history and current stress. That childhood trauma was not systematically related to any one sleep health dimension after covariate adjustment, yet was associated with a multivariable composite measure of sleep speaks to the complex relationship between trauma and sleep and provides incentive for future research to probe the impact of trauma across many dimensions of sleep. It should be noted that the summation procedure used in the present study represents one method for computing sleep health. While others have used a similar approach (31), latent factor analysis (23) and survey style assessments (30) have also been used to quantify sleep health. Although conceptually similar, each approach is statistically unique and underscores the need for a psychometrically valid method operationalizing sleep health. Second, as each of these dimensions has been independently associated with clinical health outcomes (22), these results provide a framework for future research aimed at understanding the role of sleep as a potential mediator of the adverse health impact of childhood trauma. For example, in three separate studies sleep health has been associated with clinically-relevant depression symptoms (31), self-reported health (30), and blood pressure and BMI (23). Finally, not unlike a cardiovascular disease risk profile (56), sleep health is relevant to everyone, irrespective of whether a clinical sleep disorder is present or not, and is easily understandable. Thus, interventions to promote sleep health might be developed to reach a larger number of at-risk individuals. More specific to the aims of the present paper, sleep health interventions might also be developed as part of public health campaigns to improve sleep health in target populations (e.g., individuals with childhood trauma exposure).

Equally important is understanding the possible physiological, behavioral, and psychosocial pathways whereby childhood trauma adversely affects sleep health later in life. For example, childhood trauma may negatively affect cortisol regulation (57), to the detriment of sleep (58). Studies in children have shown trauma exposure to be associated with an attenuated cortisol awakening response (59,60) and a failure of cortisol to decline in the evening (61). Although still debated, trauma has been related to both augmented (57,62) and diminished (63,64) cortisol stress responses. Similarly, individuals who experience trauma in childhood have upregulated indicators of sympathetic nervous system activity, such as urinary catecholamines (65,66). Behaviorally, environments where trauma is present promote the uptake of behaviors that may impair sleep such as smoking or alcohol use (3,4,67). Similarly, a traumatic childhood environment may interfere with the learning of habits related to coping with stress and possibly lead to more stress in the future (68). Although these explanations are plausible, our results showed that the significant association between childhood trauma and sleep health withstood adjustment for daily alcohol use, current stressors, and subjective social status. Finally, the development of a mental health disorder after trauma exposure may lead to poor sleep. For example, both depression and post-traumatic stress disorder (PTSD) are associated with poor sleep (69–71). Although the current study excluded individuals with PTSD, we found that the relationship between childhood trauma and sleep health was independent of clinically-assessed lifetime MDD history.

The present results must be considered in light of several limitations. First, it is possible that recall bias may have influenced retrospective reports of trauma. However, previous studies designed to assess retrospective recall bias of trauma have failed to produce compelling evidence of systematic recall bias (72–74). Second, the sample was composed of mostly non-Hispanic White participants. Thus, generalizability to other racial/ethnic cohorts may be limited. Third, the present results do not provide information regarding the lifetime development of sleep health as a function of trauma. It would be informative to understand if trauma affects sleep health immediately or if sleep health evolves as a function of trauma across the lifespan. Fourth, sample size limitations precluded analyses regarding type of childhood trauma. Although some studies have reported a consistent negative impact of childhood trauma on future sleep, irrespective of type (e.g., physical, emotion, or sexual abuse; 9–11), other have not (75), leaving unanswered the question as to whether different types of trauma in childhood differentially affect sleep in adulthood. Given that types of childhood trauma have been shown to differentially affect biological stress responses (76–79), more fine-grained analyses of childhood trauma type are warranted in future studies.

In conclusion, in this retrospective study, greater trauma exposure prior to 18 years of age was associated with poorer sleep health in adulthood after controlling for a range of factors including lifetime MDD history and current stress. These results demonstrate that, although time may pass, childhood trauma may continue to be associated with sleep health in mid- and late-life. Given the alarming prevalence of childhood trauma experience, these data also underscore the need for research focused on understanding the negative impact childhood trauma has on sleep health and its downstream consequences to health and functioning later in life.

Figure 1.

Odds of reporting a problem (score = 0) with each of the sleep health dimensions as a function of childhood trauma exposure (N = 138 for diary-based sleep health components and N=118 for actigraphy-based sleep health components).

Note: Odds ratios and confidence intervals are adjusted for Time 1 parent cohort, depression history, current stress, interpersonal support, and subjective social standing. D=Diary-based, A=Actigraphy-based *denotes significantly increased risk, p<.05

Abbreviations

MDD

Major Depressive Disorder

BMI

Body Mass Index

PSG

Polysomnography

SD

Standard Deviation

THQ

Trauma History Questionnaire

OR

Odds Ratio

PTSD

Post-Traumatic Stress Disorder