Child abuse and physical health in adulthood
by Tracie O. Afifi, Harriet L. MacMillan, Michael Boyle, Kristene Cheung, Tamara Taillieu, Sarah Turner, and Jitender Sareen
Child abuse is recognized as having adverse life-long consequences.Note 1Note 2 In Canada, 32% of a nationally representative adult sample indicated that they had experienced physical abuse, sexual abuse, and/or exposure to intimate partner violence during childhood.Note 3 While considerable research has examined associations between a history of child maltreatment and mental disorders,Note 4Note 5Note 6Note 7Note 8Note 9Note 10 less is known about associations with physical health.
Studies of the relationship between child abuse and physical health have tended to involve clinical or at-risk samples not representative of the general population, one or two types of abuse, and a limited number of physical conditions.Note 11Note 12Note 13 Few studies have examined sex differences in the relationship between child abuse and physical health. In addition, no studies have investigated associations between several types of abuse in childhood and multiple physical health outcomes in a representative sample of Canadian adults.
The current analysis addresses many of these gaps. It is based on a representative national sample of Canadian adults and examines three types of child abuse (physical abuse, sexual abuse, and exposure to intimate partner violence), 13 physical conditions, and self-perceived health. To investigate relationships between specific types of child abuse and specific physical conditions and sex differences, the models are adjusted for potential confounders, including sociodemographic characteristics, smoking, obesity, mental disorders, and other physical conditions and types of child abuse.
The objectives are to estimate: 1) the prevalence of abuse in childhood among Canadians aged 18 or older, by the presence of chronic physical conditions; 2) associations between different types of child abuse and self-perceived health; 3) associations of child abuse types and number of child abuse types with physical conditions; and 4) sex differences in associations between abuse in childhood and physical health in adulthood. It is hypothesized that: child abuse is associated with increased odds of physical conditions and poorer self-perceived health; a dose-response relationship exists, with exposure to more types of child abuse associated with higher odds of physical conditions; and findings vary by sex.
Data and methods
Data are from the 2012 Canadian Community Health Survey—Mental Health (CCHS—2012), a national cross-sectional survey conducted by Statistics Canada. A multi-stage stratified cluster sampling design was used to obtain a representative sample of respondents aged 15 or older in the 10 provinces. Fewer than 3% of the population were excluded from the sampling frame: people living on reserves and other Aboriginal settlements, full-time members of the Canadian Forces, and the institutionalized population.Note 14
The national household- and individual-level response rates were 79.8% and 68.9%, respectively.Note 14 Most interviews (87%) were conducted in person using computer-assisted interviewing. Because of the sensitive nature of the child abuse questions, only respondents aged 18 or older were asked these items (n = 23,395). Respondents were informed about the privacy and confidentiality provisions, and the voluntary nature of the CCHS—2012, and were required to give their consent before participating.Note 15 The data were de-identified; that is, identification of respondents is not possible. In addition, all files were vetted by Statistics Canada to ensure that privacy and confidentiality guidelines were maintained.
The CCHS—2012 assessed physical abuse, sexual abuse, and exposure to intimate partner violence using The Childhood Experiences of Violence Questionnaire (CEVQ).Note 16 All items referred to occurrences before age 16. Responses were provided on an ordinal scale. In accordance with the CEVQ guidelines, responses were dichotomized as the presence or absence of history of child abuse.
Physical abuse was present if at least one of the following criteria was met: 1) being slapped on the face, head or ears or hit or spanked with something hard three times or more; 2) being pushed, grabbed, shoved, or having something thrown at the respondent to hurt them three times or more; or 3) being kicked, bit, punched, choked, burned, or physically attacked at least once.
Sexual abuse was present if respondents experienced either of the following at least once: 1) attempts or being forced into unwanted sexual activity by being threatened, held down, or hurt; or 2) sexually touched, meaning unwanted touching or grabbing, kissing, or fondling against their will.
Exposure to intimate partner violence was present if respondents saw or heard parents, step-parents or guardians hitting each other or another adult in the home three times or more.Note 16
Variables were computed to determine the experience any child abuse (one or more types) and the number of types experienced (zero to three types).
Age group, sex, visible minority (self-reported yes or no), birthplace (self-reported Canada or other), level of education, past-year household income (in categories), marital status, current smoking status, and obesity were included as covariates in the models. Smoking status was coded as: never-smoker, former occasional smoker, former daily smoker, current occasional smoker, or current daily smoker. Body mass index (BMI) was calculated from self-reported height and weight (excluding pregnant women). BMI classifications were based on Health Canada and World Health Organization (WHO) recommendations, and modified using the Canadian standard.Note 17 Obesity was defined as a BMI of 30 or more.
Lifetime axis I disorders diagnoses were made using the WHO version of the Composite International Diagnostic Interview, which was based on criteria of the Diagnostic and Statistical Manual of Mental Disorders Fourth Edition (DSM-IV).Note 17Note 18 The disorders were: depression, bipolar disorder, generalized anxiety disorder, alcohol abuse/dependence, and drug abuse/dependence. Additional mental conditions were determined from self-reports by asking respondents if they had a long-term health condition diagnosed by a health professional that had lasted or was expected to last six months or more: obsessive-compulsive disorder, post-traumatic stress disorder, panic disorder, a phobia, a learning disability, attention deficit disorder, and an eating disorder. Mental disorders and conditions were included as covariates in the adjusted models to determine if a significant relationship existed between child abuse and physical health, independent of these mental disorders.
The analysis assessed 13 self-reported, physician-diagnosed chronic physical conditions. Respondents were asked if they had a long-term health condition diagnosed by a health professional that was expected to last or had lasted six months or more. The specific conditions were: asthma; arthritis (excluding fibromyalgia); back problems (excluding fibromyalgia and arthritis); high blood pressure; migraine headaches; chronic bronchitis/emphysema/chronic obstructive pulmonary disease (respondents aged 35 or older); diabetes; epilepsy; heart disease; cancer; stroke; bowel disease (Crohn’s disease, ulcerative colitis, irritable bowel syndrome, bowel incontinence); and chronic fatigue syndrome.
Respondents also rated their physical health as excellent, very good, good, fair, or poor.
The analyses were weighted to ensure that the sample was representative of the Canadian population aged 18 or older. To address the complexity of the survey design, bootstrapping was applied to all analyses as a variance estimation technique. The distribution of sociodemographic characteristics, smoking, and obesity by any physical condition was computed. The prevalence of physical conditions across different types of child abuse was then determined. Multinomial logistic regression models were used to examine relationships between child abuse and self-perceived general health, adjusting for sociodemographic characteristics (adjusted odds ratio (AOR-1)). Logistic regression models were used to examine associations between abuse in childhood abuse and physical conditions in adulthood. Nested models were first adjusted for sociodemographic characteristics, smoking, and obesity (AOR-1), and then for mental disorders, other physical conditions, and other types of child abuse (AOR-2). Finally, sex differences in the relationships between abuse in childhood and physical health in adulthood were examined using interaction terms and adjusting for sociodemographic characteristics. Lack of multicollinearity among covariates was determined for all models.
A number of sociodemographic and lifestyle factors were associated with having a physical condition. The prevalence was higher among women than men and rose with age (Table 1). Higher education and higher household income were associated with a reduced prevalence. Relative to never-smokers, occasional and daily smokers and current daily smokers had elevated odds of having a physical condition (ORs from 1.1 to 2.2). Obesity was also associated with having a physical condition.
All types of child abuse were associated with an increased likelihood of having a physical condition, with odds ranging from 1.4 for being slapped on the face, head or ears or hit or spanked with something hard to 2.0 for sexual abuse (Table 2).
When sociodemographic characteristics were taken into account, physical abuse (AOR = 1.2; 95% CI = 1.1 to 1.4), sexual abuse (AOR = 1.4; 95% CI = 1.1 to 1.6), and exposure to intimate partner violence (AOR = 1.3; 95% CI = 1.1 to 1.6) in childhood were associated with increased odds of obesity in adulthood (data not shown).
As well, people who experienced child abuse were more likely than those who had not to report their health as less than excellent (very good, good, fair, poor) (Table 3).
Experiencing any type of child abuse was associated with increased odds of arthritis, back problems, high blood pressure, migraine headaches, chronic bronchitis/emphysema/COPD, cancer, stroke, bowel disease, and chronic fatigue syndrome, when adjusting for sociodemographic characteristics, smoking, and obesity (AOR-1) (Table 4). Each type of child abuse was associated with six conditions—arthritis, back problems, migraine headaches, cancer, bowel disease, and chronic fatigue syndrome. Additionally, physical abuse was associated with high blood pressure and stroke; sexual abuse was associated with chronic bronchitis/emphysema/COPD and diabetes; and exposure to intimate partner violence was associated with chronic bronchitis/emphysema/COPD. Regardless of further adjustments for mental disorders, other physical health conditions, and other types of child abuse, significant relationships remained between physical abuse and back problems, migraine headaches, and stroke, and between exposure to intimate partner violence and migraine headaches (AOR-2).
A general trend emerged—as the number of types of child abuse experienced increased, so did the odds of most conditions (Table 5).
When interaction effects between sex and any child abuse were examined, significant effects were noted for back problems (AOR for the interaction term = 1.2; 95% CI = 1.0 to 1.5), chronic bronchitis/emphysema/COPD (AOR for the interaction term = 2.0; 95% CI = 1.3 to 3.0), cancer (AOR for the interaction term = 1.8; 95% CI = 1.3 to 2.4), and chronic fatigue syndrome (AOR for the interaction term = 2.3; 95% CI = 1.1 to 2.5). In all cases, the effects were slightly stronger for women than for men.
The findings of this study support the research hypotheses—child abuse was associated with increased odds of physical conditions and poor self-perceived health; a dose-response pattern was observed, with the number of child abuse types experienced corresponding to increased odds of some conditions; and sex differences emerged for associations between child abuse and back problems, chronic bronchitis/emphysema/COPD, cancer, and chronic fatigue syndrome, with slightly stronger effects for women. Overall, associations between child abuse and physical conditions were non-specific; no clear differentiation was apparent between particular types of child abuse and specific physical conditions.
Even when sociodemographic characteristics, obesity, and smoking were taken into account, all child abuse types were associated with several physical conditions. Further adjustments for mental conditions, other physical conditions, and other child abuse types yielded unique relationships between physical abuse and back problems and stroke; no other unique relationships between specific types of child abuse and specific physical conditions were noted. Physical abuse and exposure to intimate partner violence were both associated with migraine headaches when adjusting for sociodemographic characteristics, obesity, smoking, mental conditions, other physical conditions, and other types of child abuse.
The findings of this study are generally consistent with the literature. Results of surveys, most of which were conducted in the United States, show associations between child maltreatment and adult physical health, including reduced health-related quality of life, pain, poor or fair self-rated health, and frequent emergency department visits.Note 11Note 12Note 13, Note 19-32 For instance, a study based on data collected in 2004 and 2005 by the U.S. National Epidemiological Survey on Alcohol and Related Conditions reported that people who had been grabbed, shoved, hit, and slapped when they were children had increased odds of developing cardiovascular disease, arthritis, and obesity in adulthood; more serious child maltreatment (severe physical abuse, sexual abuse, exposure to intimate partner violence, and/or neglect) was associated with each of the seven physical conditions that were assessed—arteriosclerosis or hypertension, hepatic disease, diabetes, other cardiovascular disease, gastrointestinal disease, arthritis, and obesity.Note 33 In addition, the prevalence of obesity and having any physical condition was greater among women who experienced child maltreatment, and among men who experienced harsh physical punishment.Note 33
Despite the general consistency of CCHS—MH findings with other research, a few exceptions are apparent. No relationship was found between child abuse and asthma, although it has been noted in previous work.Note 31Note 34Note 35 This may, in part, be attributable to differences in sampling (population versus clinical samples), measurement (self-reported asthma versus physician-diagnosed), and study design (cross-sectional versus prospective). A recent systematic review indicated that evidence supporting the relationship between childhood adversity and asthma onset remains inconsistent.Note 36
Several explanations for the relationship between child abuse and poor physical health are possible. Infants who experience abuse have high hormonal reactivity to stress.Note 37Note 38 The link between child abuse and physical health may also reflect physiologic responses to violence. Exposure to abuse may affect the hypothalamus-pituitary adrenal axis, leading to excess secretion of cortisol and consequent physiological responses such as increased heart rate and blood pressure.Note 25Note 39Note 40Note 41Note 42 A review of the literature concluded that, based on neuroimaging and neurochemical and proton spectroscopy data, individuals exposed to abuse and violence have acute, subacute, and chronic changes in the brain, particularly related to the hypothalamic-pituitary-adrenal axis.Note 43 Child abuse can also alter patterns of sleep,Note 44 which may worsen physical symptoms.Note 25 As well, child abuse and neglect are associated with alterations in the immune system—markers of systemic inflammation appear to be more common among children and adults exposed to maltreatment.Note 25Note 40Note 45Note 46Note 47 Indirect pathways between child abuse and physical health may also exist through mental conditions or physical co-morbidities. Individuals exposed to maltreatment in childhood may have emotional, cognitive, and behavioural responses such as substance abuse and overeating, which may be another pathway to health problems.Note 25Note 40Note 45Note 46 In the present study, these indirect pathways may be indicated by the AOR-2 models that become non-significant when adjusting for mental health conditions, other physical conditions, and other types of child abuse.
Strengths and limitations
The strengths of this study include: the use of recent data from a nationally representative Canadian sample; examination of three types of child abuse, 13 physical conditions, and self-perceived health; and adjustment for sociodemographic characteristics, obesity, smoking, mental disorders, other physical conditions, and other types of child abuse.
However, the results should be interpreted in the context of several limitations. The cross-sectional design precludes causal inferences about relationships between child abuse and physical conditions. Also, retrospective accounts of exposure to abuse in childhood may be subject to recall or reporting bias. Evidence supports the valid recall of adverse childhood experiences,Note 48Note 49Note 50 but ethical and practical constraints make it difficult to collect such information prospectively from a general population sample. Moreover, the CHMS—MH did not cover emotional abuse and neglect in childhood. As well, child abuse is associated with an increased likelihood of violence in adult relationshipsNote 51 and could contribute to health problems in adulthood, but intimate partner violence was not covered in the survey. In addition, physical conditions were based on self-reports and were not independently confirmed. Nonetheless, agreement between self-reports and physician-diagnosed conditions has been shown to be high.Note 52 BMI was computed using self-reported height and weight, which tend to introduce measurement error. Finally, the study adjusted for current sociodemographic characteristics rather than characteristics before age 16.
The results of this study underscore the association between three major types of child abuse and physical health conditions and perceived general health in adulthood. Given the relative paucity of information about the relationship between child abuse and physical health compared to mental health, clinicians may have a limited awareness of the role of child abuse in relation to physical conditions. It is important that such awareness is increased, since exposure to abuse in childhood may have implications for treatment of physical illnesses. From a public health perspective, it is increasingly recognized that prevention of child abuse has major implications for reduction in mental health problems, but it is also possible that reducing child abuse may lead to better physical health outcomes.
This analysis was supported by a Canadian Institutes of Health Research operating grant (Afifi, MacMillan, Boyle, and Sareen), a CIHR New Investigator Award (Afifi), a Manitoba Health Research Council establishment award (Afifi), a Manitoba Health Research Council Chair Award (Sareen). Harriet L. MacMillan is supported by the Chedoke Health Chair in Child Psychiatry. Michael Boyle has a Canada Research Chair in the Social Determinants of Child Health.