|Sponsored by the HealthcareTrainingInstitute.org providing Quality Education since 1979|
Section 8 (Web #22)
The purpose of the current study was to test for gender differences in the associations of physical abuse in childhood with health problems in adulthood. Data are from a large, nationally representative survey conducted in the United States with 8000 men and 8000 women. This study extends the literature on child abuse by including a large sample of men and by specifically testing the differential effects of physical abuse on health problems for males and females.
Data for this study came from the National Violence Against Women Survey. Respondents were told that the survey was about personal safety. The study was conducted by the Center for Policy Research and was jointly funded by the National Institute of Justice and the Centers for Disease Control and Prevention. Data were collected between November 1995 and May 1996. The sample of 8000 men and 8000 women was derived using random-digit dialing among households with telephones in all 50 states and the District of Columbia. The participation rates for women and men were 72% and 69%, respectively. Before the interview, respondents were informed that their participation was voluntary and that their answers would be kept confidential. Computer-assisted telephone interviewing was used. A Spanish-language version of the survey was used for Spanish-speaking respondents. For more information on the sampling procedures, see Tjaden and Thoennes.
Predictor Variable: Physical Abuse in Childhood. Physical abuse in childhood was assessed using 12 questions from the Conflict Tactics Scales. Respondents were asked if they had experienced as a child (child was not defined) any of 12 violent behaviors perpetrated by a parent, stepparent, or guardian. The violent behaviors included having something thrown at [them]; being pushed, grabbed, or shoved; having hair pulled; being slapped or hit; being kicked or bitten; being choked or experiencing attempted drowning; being hit with some object; being beaten; being threatened with a gun; being threatened with a weapon other than a gun; having a gun used on [them]; and having another type of weapon used on [them]. Respondents were classified as victims (1 = yes; 47%) or nonvictims (0=no; 53%) based on whether they had experienced any of these violent behaviors as a child.
Dependent Variables: Health Problems in Adulthood. We included in the analyses only those health problems for which we were able to determine the onset and thus could create the correct temporal sequence (i.e., occurrence of childhood abuse preceded onset or occurrence of health problems). Physical injury assessed whether after the age of 17 years respondents had sustained a serious injury (e.g., head injury) that was disabling or interfered with their normal activities (8% said yes). Chronic physical health condition assessed whether after the age of 17 years respondents had acquired a chronic physical health problem (e.g., high blood pressure) that was disabling or interfered with their normal activities (11% said yes). Chronic mental health condition assessed whether after the age of 17 years respondents had acquired a chronic mental health problem (e.g., depression) that was disabling or interfered with their normal activities (2% said yes). Alcohol use assessed whether respondents drank alcohol every day or nearly every day during the past 12 months (6% said yes). Drug use assessed respondents' past-month use of tranquilizers (5% said yes), prescription pain killers (10% said yes), antidepressants (4% said yes), and illegal drugs such as marijuana, crack, heroin, or angel dust (3% said yes). Current perceptions of unfavorable physical health assessed whether respondents perceived their health to be fair or poor (12% said yes) rather than good, very good, or excellent.
Demographics and Other Violence Experienced. We controlled for 6 demographic variables: gender, marital status, employment status, education, race, and age. We also controlled for sexual abuse occurring in childhood. This variable assessed whether respondents had experienced completed or attempted forced vaginal, oral, or anal intercourse before the age of 18 years.
Interaction Term. We computed an interaction term to represent the cross-product of gender and physical abuse in childhood.
First, we examined the demographic composition of the total sample, as well as for males and females separately. Second, we examined differences between males and females on the reported prevalence of physical abuse in childhood. Third, we conducted bivariate logistic regression analyses to test the main effects of physical abuse in childhood on health problems in adulthood. Fourth, we conducted multivariate logistic regression analyses that tested for the main effects of physical abuse in childhood on health problems in adulthood while controlling for the demographic variables and child sexual abuse. In this way, we could assess the unique contribution of physical abuse in childhood on health problems in adulthood while holding constant other factors (e.g., age) that have also been shown to be related to health problems. Fifth, to determine whether physical abuse in childhood had differential effects on the health measures based on the respondent's gender, we conducted moderational analyses using Baron and Kenny's criteria for testing moderation. Specifically, the interaction term between the predictor (physical abuse in childhood) and the hypothesized moderator (gender) must be significantly related to the dependent variables (health measures) after we controlled for the main effects of both the predictor and hypothesized moderator variable. We conducted the moderator analyses for all of the outcome variables. Last, we conducted post hoc probing of significant moderational effects using Holmbeck's recommended procedures.
The sample comprised 50% men aged 18 years and older and 50% women aged 18 years and older. Approximately two thirds (64.9%) of the sample were married (men=66.9%, women = 62.9%), 68.4% were employed (men=78.2%, women = 59.0%), 89.6% had at least a high school education (men=89.9%, women = 89.3%), 82.5% were White (men= 82.8%, women=82.2%), 9.2% were Black (men=8.5%, women=9.9%), and 8.3% (men=8.7%, women=7.9%) were of other races (e.g., Native American/Alaska Native; Asian/Pacific Islander). The mean age of the sample was 43.33 years (SD=15.76) (men=42.47 (SD=15.33), women=44.19 (SD=16.13)).
Table 1 presents descriptive data for males and females on each of the abuse items (these data also appear in Tjaden and Thoennes(n28(p38)). Men were significantly more likely than women to have experienced 7 of the 12 violent behaviors perpetrated by a parent, stepparent, or guardian. Specifically, men were more likely than women to have had something thrown at them that could hurt; to have been pushed, grabbed, or shoved; to have been slapped or hit; to have been kicked or bitten; to have been beaten up; to have been hit with some object; and to have been threatened with a weapon other than a gun during their childhood. No gender differences were found on the other items.
Results from bivariate analyses are presented in Table 2 and indicated that respondents who had experienced physical abuse in childhood were significantly more likely than their nonabused counterparts to have sustained a serious injury in adulthood (crude odds ratio [COR] = 1.73, 95% confidence interval [CI] = 1.54, 1.95), acquired a mental health condition in adulthood (COR=2.05, 95% CI=1.57, 2.67), used alcohol daily in the past year (COR=1.51, 95% CI=1.32, 1.73), and used the following substances in the past month: tranquilizers (COR=1.74, 95% CI=1.51, 2.01), pain killers (COR=1.45, 95% CI=1.31, 1.61), antidepressants (COR= 1.84, 95% CI=1.55, 2.17), and illegal drugs (COR=2.78, 95% CI=2.25, 3.42). There were no significant differences between abused and nonabused respondents in their likelihood of acquiring a physical health condition in adulthood or reporting perceptions of unfavorable health.
Findings from the multivariate analyses are also presented in Table 2. Results for the demographic statistical controls are presented although we only describe the main effects for gender and physical abuse in childhood.
Gender. When we controlled for the other demographic variables, childhood sexual abuse, and childhood physical abuse, men were significantly more likely than women to have sustained a serious injury in adulthood (adjusted odds ratio [AOR]=1.33, 95% CI= 1.17, 1.52), used alcohol daily in the past year (AOR=2.68, 95% CI=2.29, 3.13), and used illegal drugs in the past month (AOR = 2.80, 95% CI=2.21, 3.56). Women were significantly more likely than men to have acquired a physical health condition in adulthood (AOR=0.79, 95% CI=0.71, 0.89), acquired a mental health condition in adulthood (AOR=0.69, 95% CI=0.51, 0.94), used tranquilizers in the past month (AOR=0.61, 95% CI=0.51, 0.71), and used antidepressants in the past month (AOR = 0.38, 95% CI=0.31, 0.46).
Physical Abuse in Childhood. When we controlled for demographic variables and childhood sexual abuse, childhood physical abuse was significantly associated with all of the assessed health problems in adulthood. Specifically, respondents who had experienced physical abuse in childhood were significantly more likely than their nonabused counterparts to have sustained a serious injury in adulthood (AOR=1.74, 95% CI=1.53, 1.97); acquired a physical health condition in adulthood (AOR=1.23, 95% CI=1.10, 1.38); acquired a mental health condition in adulthood (AOR=2.36, 95% CI=1.75, 3.18); used alcohol daily in the past year (AOR=1.42, 95% CI=1.24, 1.63); used tranquilizers (AOR=2.10, 95% CI=1.80, 2.46), painkillers (AOR=1.58, 95% CI=1.42, 1.76), antidepressants (AOR=2.13, 95% CI=1.78, 2.54), and illegal drugs (AOR=2.38, 95% CI=1.91, 2.96) in the past month; and reported perceptions of unfavorable health (AOR=1.25, 95% CI=1.13, 1.39). The largest effects were found for mental health problems, tranquilizer use, antidepressant use, and illegal drug use.
Test of Gender x Childhood Physical Abuse Interaction Terms. The interaction term was statistically significant for acquiring a chronic mental health condition in adulthood (Wald=6.48, P<.01) and reporting perceptions of unfavorable health (Wald = 4.02, P<.05). These significant interaction terms indicate that the effects of physical abuse in childhood on health problems in adulthood differed for men and women. However, they did not tell us in what way the effects differed.
To interpret the nature of the interaction terms (i.e., for which group--men, women, or both--physical abuse in childhood was significantly associated with health problems in adulthood), we conducted post hoc probing of the moderational effects. This entailed computing 2 conditional moderator variables and then running 2 more regression models, 1 for each conditional moderator variable. In the first model, men were assigned a score of zero, and in the second model, women were assigned a score of zero. On the basis of these post hoc analyses, we derived adjusted odds ratios, 95% confidence intervals, and t values for the associations between physical abuse in childhood and the health measures for both men and women. In this way, we were able to determine whether physical abuse in childhood was significantly associated with acquiring a mental health condition in adulthood and reporting perceptions of unfavorable health for men only, for women only, or for both genders but in differing magnitudes.
Results from these analyses indicated that physical abuse in childhood was significantly related to acquiring a mental health condition in adulthood for women (AOR=3.10; 95% CI=2.13, 4.50; t=5.92), but not for men (AOR=1.40; 95% CI=0.86, 2.28; t= 1.36). Physical abuse in childhood also was significantly related to reporting current perceptions of unfavorable health for women (AOR=1.38; 95% CI=1.20, 1.59; t=4.44), but not for men (AOR=1.12; 95% CI=0.96, 1.30; t=1.38).
Using data from a large, nationally representative sample, we found that physical abuse in childhood was more prevalent among men than women. We also found that physical abuse in childhood was related to health problems in adulthood for the sample as a whole and adversely affected the mental health and general perceptions of health of women more than men.
The inclusion of a large sample of men and women allowed testing how the association between physical abuse in childhood and health problems in adulthood might vary by gender. Our results are consistent with other studies that have examined gender differences in the effects of child abuse. That is, child abuse is generally detrimental for both males and females. However, female abuse victims appear to be at greater risk for some health problems than their male counterparts. As with prior studies, we found this to be the case with mental health problems. However, unlike prior studies, we did not find gender differences in the effects of child abuse on alcohol or drug use.
Although this study had several strengths, including its large national sample and the statistical test of the interaction between physical abuse in childhood and gender, there were some limitations. First, even though most of the examined relations were statistically significant, the magnitudes of the odds ratios were small. This is not surprising given the amount of time that elapsed between the occurrence of physical abuse in childhood and the occurrence of health problems in adulthood. The elapsed time allowed a large and varied number of experiences that could have affected the associations between child abuse and health problems in adulthood. Second, although we took efforts to ensure that the temporal order between physical abuse in childhood and health problems in adulthood was correct, our findings do not provide firm conclusions that child abuse is causally related to health problems in adulthood. Third, data were retrospective, so recall bias was possible. For 2 reasons, recall bias may be particularly problematic when the stressor under study is child victimization. First, because the amount of time that elapsed since the childhood victimization allows a great and varied number of experiences, recall bias may be more likely as people perceive past events in light of their later and current experiences. Second, research suggests that emotional trauma might cause memory impairment.
Although our data suggest that physical abuse in childhood is related to adverse health outcomes in adulthood, they do not address why childhood physical abuse would lead to later health problems, or why some of these associations differed by gender. Future research should address variables that might explain or mediate the child abuse--adult health problems relation. Physical abuse in childhood has been found to be related to several potential mediators of the child abuse--adult health problem association, including insecure attachment patterns and more aggression (interpersonal problems), deficits in receptive and expressive language and poor academic achievement (cognitive problems), and an increased likelihood of risky sexual behavior, physical inactivity, and smoking (risky health behaviors).These potential mediators, in turn, have been found to be associated with health problems.
Our data also do not address why physical abuse in childhood was associated with poor mental health and perceived physical health in adulthood for women but not for men. Some researchers have speculated that child abuse may be a marker for other negative childhood experiences that are more common for girls than for boys. Although we were able to control for childhood sexual abuse, we were not able to control for other early childhood experiences that may account for the observed gender differences. In 1 study, abused and neglected girls were at increased risk of substance abuse and arrests for violent crimes compared with their counterparts even after researchers controlled for such family background characteristics as parental substance abuse, parental arrest, and family welfare status. For boys, however, previously significant bivariate relations between abuse and these outcomes were reduced to nonsignificance after researchers controlled for these family background variables. Some researchers have speculated that females may be more likely than males to engage in self-blame after child abuse, and that this accounts for females' increased risk for mental health problems in adulthood. Others have speculated that males and females react to stress differently, with females being more likely to internalize stress symptoms (e.g., depression) and males being more likely to externalize stress reactions (e.g., aggressive behaviors). Because our study did not include measures of externalizing behaviors, we could not test this hypothesis. It has also been suggested that females may be more likely than males to evidence problems after abuse because the abuse was more persistent or severe? Data from the current study indicate that males actually experience more forms of physical abuse in childhood than females (men had higher mean score on sum of Conflict Tactics Scale items than women did), so differences in the magnitude of abuse did not account for our finding of gender differences. However, Widom and White suggest that abuse may be tolerated more for boys than girls, and hence nonvictim comparison groups for boys may have more false negatives than nonvictim comparison groups for girls. More research is needed to determine in what ways and why males and females differ in the consequences of physical abuse in childhood.
Although this study did not directly address why there may be gender differences in the association between physical abuse in childhood and health problems in adulthood, the findings suggest the importance of considering potential long-term adverse health consequences in the development of intervention strategies to address physical abuse in childhood. Health care providers should be aware that physical abuse in childhood might be associated with health problems in adulthood, especially among females. Intervening at an early stage may reduce a child's likelihood of developing long-term health sequelae and also reduce the public health burden of child abuse by preventing future health problems. Attention also should be paid to the primary prevention of physical abuse of children. Some efforts to prevent the initial occurrence of child abuse have shown promising results.
In sum, little research has focused on differences between males and females in the consequences of physical abuse in childhood. This study helps to address this research gap by examining the moderating role of gender in the associations between physical abuse in childhood and health problems in adulthood. We found that childhood physical abuse was more prevalent among males, and although it was related to adverse health outcomes for both genders, the effect was generally greater for females. These findings can help inform intervention strategies by alerting public health and medical practitioners of the potential for physical abuse in childhood to be related to health problems in adulthood.
- Thompson, M, Kngree, J, & Desai, S. (2004). Gender differences in long-term health consequences of physical abuse of children: data from a nationally representative survey. American Journal of Public Health, 94(4).
“Personal Reflection” Journaling Activity #8
The preceding section contained information regarding gender differences in consequences of physical abuse of children. Write three case study examples regarding how you might use the content of this section of the Manual in your practice. Affix extra paper for your Journaling entries to the end of this Manual.
What seven specific forms of violent behavior were men more likely than women to have had inflicted upon them as a child?
Others who bought this Self-Esteem Course