Association of Childhood Trauma Exposure with Inflammatory Biomarkers Among Midlife Women

Abstract

Background: Childhood abuse has been associated with poor health outcomes in adulthood. However, the physiologic pathways by which abuse is linked to health are not fully elucidated. Inflammation plays a significant role in the pathophysiology of multiple chronic diseases. We tested whether childhood trauma exposure was related to increased systemic inflammation in midlife women.

Materials and Methods: Participants were 304 nonsmoking perimenopausal and postmenopausal women aged 40 to 60 years and free of cardiovascular disease. They completed questionnaires assessing psychosocial and behavioral factors, including childhood trauma, anthropometric measures, wrist actigraphy sleep measurements, and a fasting blood draw for inflammatory markers high-sensitivity C-reactive protein (hsCRP) and interleukin-6 (IL-6). Associations between childhood trauma and inflammatory markers were tested in linear regression models controlling for age, race/ethnicity, education, body mass index, anti-inflammatory medication use, and alcohol consumption. Other covariates considered included sleep continuity and depressive symptoms.

Results: A total of 44.8% of the sample experienced at least one type of childhood abuse/neglect. Women with a history of emotional abuse had higher IL-6 levels than women without this history in multivariate models (β = 0.077, standard error = 0.032, p = 0.017). Results were not accounted for by covariates and persisted additionally controlling for depressive symptoms and sleep. Childhood abuse/neglect was not related to hsCRP.

Conclusions: Childhood emotional abuse was associated with higher levels of IL-6 in midlife women. Assessing childhood trauma exposure along with inflammatory markers may be important for the development of prevention strategies at midlife to prevent chronic diseases later in life.

Introduction

Exposure to childhood trauma is common among women. National estimates indicate that 9.1 per 1,000 U.S. children were found to be victims of childhood abuse or neglect in 2016, with girls comprising slightly over half of this statistic (51.0%).¹ In the landmark CDC-Kaiser Permanente study on adverse childhood experiences (ACEs), which surveyed over 17,000 participants, 13.1% of women reported experiencing emotional abuse, 27.0% reported physical abuse, and 24.7% reported sexual abuse.^2,3 A growing body of research indicates that childhood adversity is associated with an increased risk of poor health outcomes, including mental illness, cardiovascular disease, diabetes, metabolic syndrome, and other chronic illnesses.^4–7 Elucidating the physiological correlates of such early life stressors may therefore have important implications for the prevention and treatment of chronic medical conditions in women later in life.

Inflammatory processes have been linked to a wide variety of medical disorders.^8,9 Furthermore, there is compelling evidence that chronic inflammation may represent a potential mechanism for how stress can influence the development of pathological conditions.¹⁰ For example, acute stress has been shown to induce secretion of proinflammatory cytokines and acute-phase proteins.¹¹ Trauma exposure early in life is linked to chronic dysregulation of the body's response to stress. Childhood trauma exposure has been associated with higher expression of proinflammatory genes in late adulthood.¹² Individuals with a history of childhood trauma have been found to have elevated circulating levels of inflammatory biomarkers.¹³ C-reactive protein (CRP), an acute-phase reactant synthesized in the liver, is a well-studied inflammatory biomarker that has previously been linked to childhood abuse.¹⁴ Interleukin-6 (IL-6) is a proinflammatory cytokine that has also been linked to childhood abuse.¹³ However, studies have inconsistently considered key covariates such as adiposity. Adiposity may be a critical factor to consider as childhood abuse increases risk for adult obesity,¹⁵ adiposity plays a key role in inflammatory responses to stress,¹⁶ and adiposity has accounted for relationships between childhood abuse and adult inflammation in prior work.^17,18

Although the association between childhood trauma exposure and inflammatory biomarkers has been extensively studied, few studies have focused on the relationship between childhood trauma and inflammation during the menopause transition, a period often heralding the onset of chronic health conditions such as cardiovascular disease, metabolic syndrome, diabetes, and cancer among women.¹⁹ Low-grade chronic inflammation may be an important mediator connecting early childhood trauma and poor health outcomes in this population. CRP and IL-6 levels measured in apparently healthy women predict future cardiovascular events.⁸ In a subset of participants from the Study of Women's Health Across the Nation (SWAN) longitudinal cohort study, midlife women who reported a history of child abuse and neglect were found to have elevated levels of CRP over a 7-year follow-up period; these changes were largely accounted for by changes in adiposity.²⁰ It should be noted that this study only studied CRP as a marker of inflammation, and further testing of the effects of abuse and neglect on other inflammatory biomarkers in this subpopulation is warranted. In addition, although different subtypes of childhood abuse may impact differentially on mental and physical health, few studies have examined the relationships between specific types of abuse and CRP and IL-6 levels.

In the present study, we investigated the association between early childhood trauma and the inflammatory biomarkers CRP and IL-6 in a cohort of 304 midlife women transitioning through menopause. We hypothesized that childhood abuse and neglect would be associated with increased levels of CRP and IL-6 in midlife women after adjusting for key covariates, including adiposity.

Materials and Methods

Study sample

Participants were recruited for a study investigating the relationship between menopausal symptoms and CVD risk.²¹ The sample consisted of 304 late perimenopausal (2–12 months amenorrhea) and postmenopausal (≥12 months amenorrhea) nonsmoking women between the ages of 40 and 60 years from the surrounding community (Pittsburgh, PA). Half of the study sample reported daily hot flashes, and half reported no hot flashes in the past 3 months. Exclusion criteria included hysterectomy and/or bilateral oophorectomy; endometrial ablation, endarterectomy, or lymph node removal; history of cardiovascular disease, stroke, arrhythmia, pheochromocytoma, gynecological cancer, kidney failure, seizure disorder, Parkinson's disease, or Raynaud's phenomenon; currently undergoing chemotherapy or dialysis; current pregnancy; or use of the following medications in the past 3 months: hormone therapy (i.e., exogenous estrogen and/or progesterone), selective serotonin reuptake inhibitors, serotonin norepinephrine reuptake inhibitors, gabapentin, insulin, beta-blockers, calcium channel blockers, alpha-2 adrenergic agonists, or other antiarrhythmic agents.

Design and procedures

Midlife women were recruited from the community via local advertisements, electronic and paper mailings, and online message boards. Following telephone and in-person screening procedures, participants underwent physical measurements and 3 days of ambulatory monitoring, including hot flash measurement by sternal skin conductance and sleep measurement by actigraphy. All protocols and instruments were approved by the University of Pittsburgh Institutional Review Board. Participants provided written informed consent.

Measures

Childhood trauma exposure

Childhood trauma history was assessed via the 28-item Childhood Trauma Questionnaire (CTQ), a previously validated self-report inventory of child abuse and neglect experienced at or before age 18.²² The CTQ included questions regarding five dimensions of childhood maltreatment: (1) emotional abuse (e.g., “People in my family said hateful or insulting things to me”), (2) physical abuse (e.g., “People in my family hit me so hard that it left me with bruises or marks”), (3) sexual abuse (e.g., “Someone molested me”), (4) physical neglect (e.g., “I didn't have enough to eat”), and (5) emotional neglect (e.g., “I did not feel loved”). Items were rated on a 5-point Likert scale, ranging from 1 = “never true” to 5 = “very often true,” producing scores of 5 to 25 for each subscale. The CTQ has strong test/retest reliability (0.79–0.86), internal consistency (α = 0.75–0.94), and convergent validity.²² Validated cut points²³ were used for each category of abuse/neglect (emotional abuse, 10; physical abuse, 8; sexual abuse, 8; physical neglect, 8; emotional neglect, 15; sensitivity and specificity ≥0.85 relative to structured clinical interviews for each of the five subscales). Abuse/neglect was considered both continuously and according to cut points for analyses.

Inflammatory biomarkers

Phlebotomy was performed after a 12-hour overnight fast. All women were free of acute illness when blood was drawn. Plasma CRP was measured using a high-sensitivity CRP reagent set (Beckman Coulter, Brea, CA) analyzed on an AU400 from Olympus America, Inc. (Melville, NY) (intra-assay CV: 5.5% and interassay CV: 3.0%). IL-6 was measured with a high-sensitivity ELISA kit from R&D Systems (Minneapolis, MN) (intra-assay CV: 9.1%, interassay CV: 10.2%, detection range 0.16–10 pg/mL, and sensitivity 0.11 pg/mL).

Covariates

Demographics (age, education, race/ethnicity) and health behaviors (e.g., alcohol use) were assessed by self-report questionnaires and in-person interview. Medical history was obtained from in-person interviews, and medical conditions and medication use documented. Height and weight were measured, and a body mass index (BMI) was calculated (kg/m²). Depressive symptoms were assessed with the Center for Epidemiologic Studies Depression Scale.²⁴ Sleep was assessed via wrist actigraphy. Participants wore an Actiwatch 2 (Respironics, Inc., Murrysville, PA) on their nondominant wrist and completed a daily sleep diary for three consecutive days. Information from the sleep diary was used to define sleep intervals for actigraphy as per guidelines.^25,26 Actigraphy data were collected in 1-minute epochs and analyzed with Philips Actiware v6.0 software. Actigraphy-assessed wake after sleep onset (WASO), a measure of sleep continuity, was the primary sleep variable of interest due to the previous documented associations of sleep continuity with IL-6 in this sample, as well as WASO being the sleep disturbance most commonly associated with menopause.^27,28 Other sleep variables examined were sleep time as assessed by actigraphy (average over 3 days) and subjective sleep quality as assessed by the Pittsburgh Sleep Quality Index.²⁹

Statistical analysis

CRP, IL-6, and continuous abuse/neglect exposures were log transformed to meet model assumptions of normality. Differences between participants by included/excluded status were tested using linear regression, Wilcoxon rank-sum, or chi-square tests. Associations between abuse/neglect and inflammatory markers were evaluated using linear regression, with each abuse/neglect exposure considered separately. For primary analyses, each subdomain of childhood trauma was dichotomized according to clinical cut points of CTQ scores and considered categorical variables. Covariates for models (i.e., age, race/ethnicity, education, BMI, anti-inflammatory medications, and alcohol consumption) were selected based on their prior documented associations with dependent variables and associations with the outcome at p < 0.10, with anti-inflammatory medications forced into models. Additional secondary models considered the number of abuse subtypes (0, 1, ≥2) in relation to outcomes; included additional covariate depressive symptoms and actigraphy-assessed sleep continuity (WASO), given their previously documented links to elevated IL-6;²⁷ and considered childhood trauma subdomains as continuous variables in relation to outcomes. Leisure physical activity and other medications were considered covariates but did not meet p < 0.10 criteria and so were not included in final models. Interactions between childhood abuse/neglect and sleep variables were tested by cross product terms in multivariate models. Residual analysis and diagnostic plots were conducted to verify model assumptions. Analyses were performed with IBM SPSS Statistics v25 (IBM Corporation). Models were two sided at α = 0.05.

Results

Participants were on average 54 years old, postmenopausal, white, overweight, and college-educated (Table 1). A total of 44.8% (n = 134) of the sample experienced at least one type of childhood abuse or neglect. The most common form of abuse was emotional abuse (23.9%, n = 72), followed by physical neglect (20.5%, n = 62), physical abuse (20.2%, n = 61), emotional neglect (17.9%, n = 54), and sexual abuse (13.6%, n = 40). Furthermore, emotional abuse, physical abuse, sexual abuse, physical neglect, and emotional neglect were significantly and positively correlated with each other (ρ's: 0.23–0.57, p's < 0.0001; Supplementary Table S1). In addition, significant correlations were observed between CRP and education and BMI, as well as between IL-6 and race/ethnicity, education, BMI, and alcohol consumption, and thus, these variables were considered covariates in all models (Supplementary Table S2).

Table 1.

Sample Characteristics

	Childhood abuse/neglect
	Yes (n = 134)	No (n = 165)
Age, M (SD)	53.77 (4.14)	54.27 (3.90)
Race/ethnicity, n (%)a
White	88 (65.7)	129 (78.2)
African American, Hispanic, other	46 (34.3)	36 (21.8)
BMI, M (SD)	29.19 (7.59)	28.87 (6.04)
Education, n (%)a
High school, vocational school	21 (15.6)	21 (12.7)
College education or higher	109 (81.3)	143 (86.7)
Physical activity, leisure time, IPAQ, M (SD)	53.14 (27.4)	56.82 (30.74)
Depressive symptoms (CESD), M (SD)a	10.00 (8.53)	6.01 (7.50)
hsCRP, mg/L, M (SD)	2.25 (2.28)	2.12 (2.12)
IL-6, pg/mL, M (SD)	2.21 (2.20)	1.96 (2.42)
Anti-inflammatory medications, n (%)	23 (17.2)	30 (18.2)
Average sleep time, minutes, M (SD)	370.07 (75.44)	373.89 (58.90)
WASO, minutes, M (SD)	49.88 (28.11)	43.01 (22.44)
Sleep quality (PSQI), M (SD)a	6.31 (3.30)	4.93 (2.59)
Current alcohol use, n (%)
Yes	90 (67.2)	122 (73.9)
No	44 (32.8)	43 (26.1)

Depressive symptoms, IL-6, log transformed; physical activity square root-transformed for comparison.

^aDiffers between trauma exposure group at p < 0.05.

BMI, body mass index; BTQ, Brief Trauma Questionnaire; CESD, Center for Epidemiologic Studies Depression Scale; hsCRP, high-sensitivity C-reactive protein; IL-6, interleukin-6; M, mean; PSQI, Pittsburgh Sleep Quality Index; SD, standard deviation; WASO, wake after sleep onset.

Women with a history of childhood emotional abuse had higher levels of IL-6 compared with women without this history (β = 0.077, standard error [SE] = 0.032, p = 0.017) (Table 2, Fig. 1, Supplementary Table S3). Other types of childhood abuse or neglect were not significantly related to IL-6. Neither childhood abuse nor neglect was related to CRP.

Table 2.

Relationship Between Childhood Trauma and Inflammatory Markers

	hsCRP B (SE)	IL-6 B (SE)
Emotional abuse (yes)	−0.007 (0.047)	0.077 (0.032)a
Physical abuse (yes)	−0.035 (0.050)	0.051 (0.035)
Sexual abuse (yes)	−0.056 (0.060)	0.001 (0.041)
Physical neglect (yes)	−0.026 (0.051)	0.059 (0.035)
Emotional neglect (yes)	−0.031 (0.053)	0.026 (0.036)
Any childhood abuse/neglect (yes)	0.000 (0.042)	0.039 (0.028)

Models adjusted for age, race, education, BMI, anti-inflammatory medications, and alcohol consumption.

^ap < 0.05.

B, beta; SE, standard error.

FIG. 1.

Adjusted mean IL-6 (pg/mL) by abuse exposure. Means adjusted for age, race, education, BMI, anti-inflammatory medications, and alcohol consumption. *p < 0.05 relative to no exposure. BMI, body mass index; IL-6, interleukin-6. Error bars represent standard errors.

We considered several secondary analyses. We considered the number of abuse subtypes; the number of abuse exposures was not significantly related to IL-6 (1: β = 0.031, SE = 0.035, p = 0.571; ≥2: β = 0.088, SE = 0.017, p = 0.100, relative to no abuse). We considered additional covariates, further adjusting for depressive symptoms, actigraphy-assessed WASO and sleep time, subjective sleep quality (Pittsburgh Sleep Quality Index [PSQI] score), and menopause status (perimenopausal vs. postmenopausal) or time since the final menstrual period in multivariate models; findings were unchanged (data not shown). We considered abuse/neglect exposures as continuous variables in relation to outcomes; findings were comparable with those with categorical predictors (data not shown). Models were also repeated excluding participants who reported rheumatoid arthritis (n = 6) or hepatitis C (n = 1) or with extreme IL-6 values (>15 pg/mL, n = 1), as they may indicate acute infection; findings were unchanged (data not shown).

Discussion

In this sample of midlife women, women with a history of childhood emotional abuse had higher levels of circulating IL-6. This association remained significant after adjustment for multiple covariates. These findings suggest that childhood emotional abuse may impact the inflammatory immune system, providing a possible mechanism by which childhood trauma confers increased vulnerability to the development of chronic diseases in midlife and older ages.

Previous work has demonstrated a relationship between childhood abuse and inflammation.¹³ In contrast to several other studies, childhood abuse and neglect were not significantly related to CRP in this sample, but we did observe a strong association with IL-6, consistent with prior findings. The discrepancy between these markers could possibly be explained by differences in their locations in inflammatory signaling pathways, as IL-6 is an upstream proinflammatory cytokine that induces hepatic production of CRP and other acute-phase reactants.³⁰ Similarly, another study found a history of childhood abuse associated with increased IL-6 but not CRP responses to daily stressors in adulthood,³¹ and another work found that emotional abuse, but not other forms of childhood maltreatment, was associated with blunted responsivity of cortisol, a potent anti-inflammatory hormone, in adults later in life.³² Thus, IL-6 might be particularly responsive to childhood emotional abuse.

We found a significant association between childhood emotional abuse and IL-6, but did not observe a significant relationship between other forms of childhood abuse and IL-6 or between childhood abuse and CRP. Our findings diverge somewhat from other works, particularly a large meta-analysis examining childhood abuse and CRP and IL-6, in which significant associations were found between childhood trauma and both CRP and IL-6.¹³ Notably, in this meta-analysis, effect sizes were small and slightly greater for IL-6 than for CRP. Thus, the somewhat less robust findings observed here could have been due, in part, to our smaller sample size as well as our relatively healthy study sample (e.g., nonsmoking, no clinical cardiovascular disease, and no current use of selective serotonin reuptake inhibitors or serotonin norepinephrine reuptake inhibitors). The meta-analysis also included subjects who were considerably younger, in contrast to our study that focused on midlife women transitioning through menopause.

Notably, few studies have investigated the association between childhood abuse and inflammatory markers in midlife menopausal women. A notable exception is SWAN, which included 326 midlife women between 42 and 52 years of age and who had at least one menstrual period in the past 3 months, were not using reproductive hormones, and had not undergone a hysterectomy or bilateral oophorectomy. In SWAN, midlife women with a history of childhood abuse or neglect had higher CRP over an 8-year period than women without this history.²⁰ BMI accounted for most of these associations with the exception of associations between emotional abuse or neglect and CRP, which were independent of BMI. The present study extends these findings by examining an additional inflammatory marker upstream of CRP and lends support to emotional abuse being specifically related to increased systemic inflammation in midlife women, even after adjusting for BMI. It is possible that emotional abuse may represent a chronic stressor over several years, due to inadequate emotional support in early life leading to later difficulties with trust, attachment, and fear of interpersonal rejection that persist into adulthood.³³ Importantly, our study found that childhood emotional abuse was the most commonly reported type of childhood abuse among this sample of midlife women. Other studies, including the CDC-Kaiser ACE study, have reported relatively lower rates compared with other forms of childhood abuse,² which could reflect differences in the assessment of abuse between the studies and demographic compositions of the two study populations. Further studies could assess associations of inflammatory markers with duration of emotional abuse exposure or persistence of emotional stressors in adulthood.

We considered several possible explanatory factors and confounders that could account for associations between childhood abuse and inflammatory markers. Adjusting for demographic factors of age, race/ethnicity, education level, anti-inflammatory medications, or BMI did not change associations between emotional abuse and IL-6, nor did further controlling for depression. In other works, lower sleep continuity was related to higher levels of circulating IL-6 in these women,²⁷ and therefore, we examined whether sleep continuity accounted for relationships between childhood trauma and inflammation. The association between emotional abuse and IL-6 persisted after adjustment for sleep indices, suggesting that emotional abuse may be an independent risk factor for elevated IL-6 in midlife women.

Although the underlying physiological mechanism for how childhood trauma may impact the inflammatory immune system is unclear, a number of possible explanations have been proposed. Neural and endocrine responses to stress appear to enhance transcription of proinflammatory genes, suggesting that exposure to persistent stress may lead to a chronically inflammatory phenotype.³⁴ In addition, the hypothalamic/pituitary/adrenal (HPA) axis is a powerful modulator of inflammatory activity as immune activation progressively stimulates secretion of glucocorticoids, which eventually terminates the inflammatory response.³⁵ Early life stress may induce changes in epigenetic regulation of genes relevant to the HPA axis. In particular, childhood trauma is associated with increased methylation of the glucocorticoid receptor, resulting in impaired negative feedback of the HPA axis and an exacerbated inflammatory response.³⁶ Moreover, women with a history of childhood abuse have increased adrenocorticotropic hormone and heart rate responses to stress, suggesting persistent sensitization of the HPA axis and autonomic nervous system to mild stress in adulthood.³⁷ Thus, multiple pathways may underlie the increased systemic inflammation observed with emotional abuse in our sample.

Although the pathways by which elevated inflammation may lead to poor health outcomes remain to be determined, these findings have relevance for the development of targeted therapeutic strategies. For instance, some research indicates that depressed patients with increased systemic inflammation are less likely to respond to conventional antidepressants and may benefit from adjunctive treatment with anti-inflammatory medications.³⁸ In addition, elevated IL-6 levels have been implicated in the pathogenesis of chronic disease processes such as coronary heart disease through a combination of autocrine, paracrine, and endocrine mechanisms.³⁹

This study has several notable strengths. We studied a large and well-characterized sample of midlife women transitioning through menopause, a period signaling the onset of women's susceptibility to developing chronic health conditions. We were able to adjust for multiple covariates and findings persisted. Finally, we studied inflammatory biomarkers that show robust associations with poor health outcomes. To our knowledge, this is the first study to demonstrate a relationship between childhood abuse and IL-6 in midlife women transitioning through menopause, a critical transition occurring before the onset of many major diseases such as cardiovascular disease.

Several limitations should also be acknowledged. First, we focused primarily on childhood abuse/neglect but did not assess other possible forms of childhood adversity, which may also impact inflammation. In addition, the CTQ, although a validated instrument for assessing childhood trauma history, is a retrospective and subjective measure as it relies on participants' recall and self-reported ratings of childhood abuse/neglect, and there is a potential for recall bias. Furthermore, our inclusion criteria (e.g., no major comorbidities such as cardiovascular disease or stroke, nonsmoking, intact uterus, and no use of selective serotonin reuptake inhibitors or serotonin norepinephrine reuptake inhibitors) may have yielded a lower risk group than the general population, which in turn may have biased observed associations toward the null. Multiple types of abuse and neglect were considered, which resulted in multiple tests; future work should replicate these findings. Finally, as this is a cross-sectional study, we cannot rule out the possibility of an unmeasured confounding variable that accounts for the relationship between childhood trauma exposure and IL-6.

Conclusions

Childhood emotional abuse was associated with increased levels of circulating IL-6 in midlife women independent of key covariates. The menopause transition period is a critical transition in women's lives. Our study underscores the importance of screening for childhood abuse in considering the development of chronic disease. Assessing a history of childhood emotional abuse, a relatively underrecognized form of abuse compared with other forms of maltreatment,⁴⁰ may be particularly important. Early detection of childhood trauma history may be important for risk assessment of chronic inflammation-related health conditions in this group and the development of targeted prevention and therapeutic strategies.

Author Disclosure Statement

J.K.N. has no conflict of interests to declare; R.C.T.: Astellas Pharma (consulting), Pfizer (consulting), Procter & Gamble (consulting).

Funding Information

This work was supported by the National Institutes of Health (NIH), National Heart Lung and Blood Institute (R01HL105647, K24123565 to R.C.T.), and the University of Pittsburgh Clinical and Translational Science Institute (NIH Grant UL1TR000005). The National Institutes of Health funded this work and approved the initial study design but was not involved in the conduct of the study; collection, management, analysis, or interpretation of the data; nor the preparation, review, or approval of the article.

Supplementary Material

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3