- Email: [email protected]
Colorectal cancer screening and adverse childhood experiences: Which adversities matter?
Child Abuse & Neglect 69 (2017) 145–150
Contents lists available at ScienceDirect
Child Abuse & Neglect journal homepage: www.elsevier.com/locate/chiabuneg
Research article
Colorectal cancer screening and adverse childhood experiences: Which adversities matter?
MARK
⁎
Héctor E. Alcalá , Jessica Keim-Malpass, Emma Mitchell University of Virginia, United States
AR TI CLE I NF O
AB S T R A CT
Keywords: Adverse childhood experiences Cancer Colorectal cancer Cancer screening Child abuse
Adverse Childhood Experiences (ACEs) have been associated with an increased risk of a variety of diseases, including cancer. However, research has not paid enough attention to the association between ACEs and cancer screening. As such, the present study examined the association between ACEs and ever using colorectal cancer (CRC) screening, among adults age 50 and over. Analyses used the 2011 Behavioral Risk Factor Surveillance System (n = 24,938) to model odds of ever engaging in CRC screening from nine different adversities. Bivariate and multivariate models were fit. In bivariate models, physical abuse, having parents that were divorced or separated, and living in a household where adults treated each other violently were associated with lower odds of engaging in CRC. In multivariate models that accounted for potential confounders, emotional and sexual abuse were each associated with higher odds of engaging in CRC. Results suggest potential pathways by which early childhood experiences can impact future health behaviors. Future research should examine this association longitudinally.
1. Introduction Adverse childhood experiences (ACEs) encompass childhood conditions such as abuse and household dysfunction (Centers for Disease Control and Prevention, 2014) and have been linked to many negative health outcomes for adults. A greater number of ACEs has been positively associated with ischemic heart disease (Felitti et al., 1998), stroke (Felitti et al., 1998), chronic bronchitis or emphysema (Felitti et al., 1998), mental illness and substance use, (Anda et al., 2006; Felitti et al., 1998; Mersky, Topitzes, & Reynolds, 2013) severe obesity (Anda et al., 2006), adult compromised metabolic functioning (Lehman, Taylor, Kiefe, & Seeman, 2005) and cardiovascular disease (Batten, Asl an, Maciejewski, & Mazure, 2004). A growing body of research has linked experiences of childhood adversity to cancer (Alcalá, 2016; Brown et al., 2010; Felitti et al., 1998; Kelly-Irving et al., 2013). Despite the growing interest in the association between childhood adversity and cancer, there is a lack of research examining how childhood adversity impacts the utilization of preventative health care services, like cancer screening, which may help us understand why the association exists to begin with. One cancer that can be acted upon in early stages by utilization of health care and consistent screening is colorectal cancer (CRC). Colorectal cancer kills more than 50,000 people in the United States annually, ranking it behind only lung and bronchial cancers in terms of site-specific cancer deaths (Siegel, Naishadham, & Jemal, 2013). Fortunately, several screening options exist including fecal occult blood tests (FOBT), colonoscopy and sigmoidoscopy, that have been proven effective at disease detection (Whitlock, Lin, Liles, Beil, & Fu, 2008). Consequently, opportunities for treatment of colorectal cancer exists if screening identifies cancer in its early stages. Unfortunately, sizable proportions of the US population have not undergone recent CRC screening (Siegel, DeSantis, & Jemal, 2014).
⁎
Corresponding author. E-mail address: [email protected] (H.E. Alcalá).
http://dx.doi.org/10.1016/j.chiabu.2017.04.026 Received 22 February 2016; Received in revised form 30 March 2017; Accepted 25 April 2017 0145-2134/ © 2017 Elsevier Ltd. All rights reserved.
Child Abuse & Neglect 69 (2017) 145–150
H.E. Alcalá et al.
Current guidelines recommend that adults aged 50–75 undergo screening for CRC (U.S. Preventative Services Task Force, 2008). Given the importance of CRC screening, it is critical to examine if ACEs impact uptake of screening. Available information suggests that ACEs may be associated with utilization of health care services, but that this relationship depends on both the adversity and services under examination. Reporting a greater number of ACEs has been associated with higher odds of having more than six general practitioner visits, more than two emergency room visits and 25 or more visits to health professionals (i.e. family doctor, medical specialist, nurse, optometrist, chiropractor, physiotherapist, dentist, pharmacist, psychologist or other health professional) in the past 12 months (Chartier, Walker, & Naimark, 2010). When the same study examined specific ACE items (which included measures of physical abuse, sexual abuse and other adversities), only some ACE items were associated with use of certain services. Of note, sexual and physical abuse were not associated with use of general practitioner services but were associated with higher use of specialty care and emergency services. This suggests that abuse does not promote use of preventive care (Chartier et al., 2010). In terms of cancer screening, childhood sexual abuse is associated with lower odds of current compliance with screening recommendations, after accounting for confounders (Alcalá, Mitchell, & Keim-Malpass, 2017), while eight other ACEs (i.e. physical abuse, parental drug use, parental separation etc.) where not associated with cervical cancer screening. Findings not focusing solely on ACEs have also shown that childhood sexual abuse is associated with lower odds of cervical cancer screening (Farley, Golding, & Minkoff, 2002). Consequently, sexual abuse, and not other ACEs, may discourage use of cervical cancer screening. Discrepancies between the patterns observed more between omnibus measures of utilization of care and cancer screening may exist for a variety of reasons. First, the discrepancy between cervical cancer screening and general use of care may be due to the invasive nature of cervical cancer screenings. Women with a history of sexual abuse may be re-traumatized by this procedure in a way that women experiencing other adversities are not. Similarly, it is possible that victims of sexual abuse involving anal penetration may avoid potential re-traumatization from procedures like colonoscopies and sigmoidoscopies. Second, the discrepancy in findings may be because cancer screening is preventative in nature, while other services, like emergency department utilization are usually motivated by existing health care needs. This is important to note because people who experience ACEs have worse health overall and thus need more health care. Finally, these findings may not be discrepant at all. Specifically, because cancer screenings prevent cancer, avoiding these screening services may necessitate usage of other services like the emergency department. ACEs, broadly speaking, may discourage use of preventative health services, like cancer screening, because people who have experienced ACEs may have a broader pattern of engaging in negative health behaviors. Specifically, ACEs have been linked to riskier health behaviors such as alcohol abuse, substance abuse, tobacco use, having more sexual partners, not using contraception, earlier age at sexual debut and being physically inactive (Alcalá, von Ehrenstein, & Tomiyama, 2016; Coker, Hopenhayn, DeSimone, Bush, & Crofford, 2009; Dube, Anda, Felitti, Edwards, & Croft, 2002; Felitti et al., 1998; Hillis, Anda, Felitti, & Marchbanks, 2001). Thus, individuals who experience ACEs may eschew preventative health care because of an underlying predisposition to not engage in health promoting behaviors, which differs from use of emergency or specialty care, which will arise out of medical necessity. Finally, ACEs may negatively impact utilization of health care because of the confluence of ACEs with limited socioeconomic status. Several specific ACEs have been associated with diminished socioeconomic status in adulthood and adolescence (Astone & McLanahan, 1991; Boden, Horwood, & Fergusson, 2007; Currie & Spatz Widom, 2010; Paolucci, Genuis, & Violato, 2001; Paradise, Rose, Sleeper, & Nathanson, 1994). Lower SES, in turn, has been associated with decreased usage of several different cancer screening procedures (Achat, Close, & Taylor, 2005; Rundle et al., 2013; Segnan, 1996; Stein, Fox, & Murata, 1991), including CRC screening (Singh et al., 2004). The goal of the present study is to examine associations between ACEs and colorectal cancer screening. Given that this is an understudied area of inquiry, we are specifically interested in whether or not specific adversities occurring in childhood impact colorectal cancer screening among adults age 50 and older.
2. Methods 2.1. Data source This study utilized publically available data from the 2011 Behavioral Risk Factor Surveillance System (BRFSS). This multistage, random digit dial telephone survey is designed to be representative of non-institutionalized adults (ages 18 and over) living in U.S. states and territories. The BRFSS is an annual survey, with a core set of questions asked of all participants in all states and territories. Optional questions were asked of all or some participants in states or territories choosing to administer them. Core questions were collected using both landlines and cell phones in all states and territories. Optional questions were administered with landlines and/or cellphones (CDC, 2011a). In the 2011 BRFSS cycle, only 10 states (California, Maine, Minnesota, Montana, Nebraska, Nevada, Oregon, Vermont, Washington and Wisconsin) administered a module of questions measuring ACEs (CDC, 2011b). Of these 10 states, only two states (Nebraska and Washington) administered a module of questions about CRC screening. Nebraska and Washington had a total of 28,198 respondents who were over 50. Of these, 3,260 were excluded from the present analyses because they had responses of, “don’t know” or “refused” or because they were missing on any variable in the present study. This yielded an analytic sample of 24,938 respondents.
146
Child Abuse & Neglect 69 (2017) 145–150
H.E. Alcalá et al.
2.2. Measures The main independent variables for analyses were from the Adverse Childhood Experience (ACE) scale. The BRFSS included an 11-item version of this scale. The items measured if the following events had been experienced by the respondent prior to age 18: 1) physical abuse; 2) being touched sexually; 3) having someone attempt to make the respondent touch someone sexually; 4) being forced to have sex; 5) psychological/emotional abuse; 6) living with an adult who was depressed, mentally ill or suicidal; 7) living with anyone who was a problem drinker or alcoholic; 8) living with a drug user or abuser; 9) living with someone who had been sentenced to serve time in a prison, jail, or other correctional facility; 10) having parents who were separated or divorced and 11) living in a home where adults or parents physically harmed each other (Felitti et al., 1998). Items about mental illness, problem drinking, drug use, incarceration and separation or divorce asked whether or not the respondent had been exposed to the specific adversity. The remaining items allowed respondents to specify the frequency of occurrence of the specific adversity (i.e. never, once or more than once). For the current study, all items were dichotomized to indicate if the experience had ever happened. This coding scheme is consistent with previous studies but refrains from summing these items into a scale (Brown, Thacker, & Cohen, 2013). Because three items measured facets of sexual abuse, they were combined into a single measure of sexual abuse. This resulted in nine ACE measures. The dependent variable of interest was whether or not a respondent had ever been screened for CRC. The BRFSS asked individuals age 50 and over about their CRC screening history. Respondents were asked, “A blood stool test is a test that may use a special kit at home to determine whether the stool contains blood. Have you ever had this test using a home kit?” If they indicated not having a FOBT, they received the follow-up question, “Sigmoidoscopy and colonoscopy are exams in which a tube is inserted in the rectum to view the colon for signs of cancer or other health problems. Have you ever had either of these exams?” Respondents were considered to ever have undertaken CRC screening if they responded affirmatively to either question. Several variables were selected as potential confounders. Age was included as a continuous variable. Gender was measured as a dichotomous variable. Race and ethnicity was measured using a categorical variable representing the Office of Management and Budget’s race and ethnicity categories (i.e. non-Latino white, non-Latino Black/African American, non-Latino Asian, non-Latino other race, and Latino). State of residence was measured using a dichotomous variable measuring whether the respondent resided in Nebraska or Washington. Educational attainment was recoded from its original categories (i.e. kindergarten or less, 1st through 8th grade, 9th through 11th grade, high school graduate, 1–3 years of college and 4 or more years of college) to continuous values that represented the midpoint of the category in terms of years of education, except for the last category which was coded as 16 years. Current insurance status was measured using a dichotomous variable, indicating whether or not the respondent had any health insurance coverage. A dichotomous indicator of lifetime cancer diagnoses, excluding skin cancer, was also included as a control. 2.3. Analyses All analyses were conducted using Stata 14.1, using appropriate weights to account for survey design. Univariate statistics were run for all study variables. Next three groups of logistic regression models were fitted to estimate odds of ever receiving colorectal cancer screenings. For the first group, bivariate associations between each of the nine ACE measures and CRC screening were modeled. For the second group, confounders were added to these models. In the final group, a model including all ACE measures simultaneously, while also including confounders was modeled. 3. Results Table 1 shows the sample characteristics. Over a fourth of respondents were ever screened for CRC. Emotional abuse was the most commonly experienced ACE measure (33.14%) and living with someone who was jailed (3.41%) was the least commonly experienced. Slightly over half of respondents were female. A majority of respondents were non-Latino white. On average, respondents had completed slightly less than a 2-year college degree and were just over 63 years old. About an eighth of the sample had ever been diagnosed with cancer. The vast majority of respondents were currently insured. More than three fourths of respondents were from Washington. Model 1 in Table 2 shows the bivariate associations between individual ACE items and ever receiving colorectal cancer screening. Here, three specific adversities were associated with lower odds of colorectal cancer screenings: physical abuse (OR = 0.89; 95% CI = 0.81,0.99), having parents who were separated or divorced (OR = 0.90; 95% CI = 0.81,0.99) and living with adults who treated each other violently (OR = 0.86; 95% CI = 0.78,0.96). In Model 2, after accounting for confounders, sexual (AOR = 1.18; 95% CI = 1.06,1.33) and emotional abuse (AOR = 1.09; 95% CI = 1.01,1.18) were associated with increased odds of CRC screening. Because both of these measures were not associated with CRC screening in bivariate models, follow-up analyses were conducted to investigate which of the confounding variables were causing the change in significance, that is which variables were functioning as potential suppressors (MacKinnon, Krull, & Lockwood, 2000). Briefly, suppression requires that the suppressor be significantly associated, in opposite directions, with the independent and dependent variable. To examine this, each control variable was individually introduced into bivariate models, to yield three variable models. For sexual abuse, the introduction of state of residence and age each made the association with CRC screening significant. CRC screening was less common in Washington, relative to Nebraska (OR = 0.61; 95% CI = 0.56,0.67), while sexual abuse was more common in Washington relative to Nebraska (OR = 1.70; 95% CI = 1.44,2.02), thus, supporting the argument that state of residence functioned as a suppressor. CRC screening was more common with increasing age (OR = 1.02; 95% CI = 1.01,1.02), while sexual 147
Child Abuse & Neglect 69 (2017) 145–150
H.E. Alcalá et al.
Table 1 Sample Characteristics, Adults 50 and older, BRFSS 2011 (N = 24,938). Variable Colorectal Cancer Screening Ever screened Never screened Adverse childhood experiences Physical abuse Sexual abuse Emotional abuse Lived with some one who was mentally ill Lived with problem drinker Lived with drug user Lived with some one who was jailed Parents divorced or separated Adults in household treated each other violently Age Gender Male Female Race White Black Hispanic Asian Other Educational attainment (years) Currently insured No Yes Ever diagnosed with cancer No Yes State of residence Nebraska Washington
% or Mean
SE
27.05% 72.95%
0.35% 0.35%
16.91% 14.15% 33.14% 14.49% 24.15% 5.97% 3.41% 17.11% 15.99% 63.33
0.48% 0.43% 0.58% 0.43% 0.53% 0.33% 0.26% 0.48% 0.48% 0.12
47.21% 52.79%
0.06% 0.06%
87.29% 1.95% 3.93% 3.06% 3.77% 13.66
0.53% 0.24% 0.33% 0.31% 0.25% 0.03
8.68% 91.32%
0.42% 0.42%
87.52% 12.48%
0.36% 0.36%
21.21% 78.79%
0.31% 0.31%
abuse was less common with increasing age (OR = 0.98; 95% CI = 0.97,0.98), thus, supporting the argument that age functioned as a suppressor. For emotional abuse, no single variable altered the significance from the bivariate model, suggesting that the cumulative effect of control variables functioned as suppressors. In Model 3, after accounting for confounders and all ACEs, sexual abuse (OR = 1.18; 95% CI = 1.05,1.32) was associated with increased odds of CRC screening.
Table 2 Logistic Regression Modeling Odds for Colorectal Cancer, Adults 50 and older, BRFSS 2011 (N = 24,938). Model 1: Bivariate Variable Physical abuse Sexual abuse Emotional abuse Lived with some one who was mentally ill Lived with problem drinker Lived with drug user Lived with some one who was jailed Parents divorced or separated Adults in household treated each other violently
OR 0.89 1.08 0.98 1.00 0.97 0.85 0.86 0.90 0.86
95% CI *
(0.81,0.99) (0.97,1.20) (0.91,1.06) (0.90,1.11) (0.89,1.06) (0.71,1.02) (0.67,1.11) (0.81,0.99)* (0.78,0.96)**
Model 2: Controls
Model 3: All ACE measures and Controls
AOR
95% CI
AOR
95% CI
1.00 1.18 1.09 1.07 1.07 0.99 1.02 1.01 0.98
(0.90,1.10) (1.06,1.33)** (1.01,1.18)* (0.96,1.18) (0.98,1.16) (0.83,1.19) (0.80,1.31) (0.92,1.12) (0.88,1.09)
0.93 1.18 1.09 1.04 1.06 0.95 0.99 0.99 0.92
(0.82,1.05) (1.05,1.32)** (1.00,1.20) (0.93,1.15) (0.97,1.17) (0.79,1.13) (0.78,1.27) (0.89,1.10) (0.82,1.04)
Model 2 adds age, gender, race, years of education, cancer status, state of residence and insurance status as controls to bivariate models. Model 3 introduces all ACEs simultaneously while also controlling for age, gender, race, years of education, cancer status, state of residence and current insurance status as controls to bivariate models. * p < 0.05. ** p < 0.01.
148
Child Abuse & Neglect 69 (2017) 145–150
H.E. Alcalá et al.
4. Discussion To the authors’ knowledge, this study represents the first investigation of the association between ACEs and CRC screening. Results revealed that not all adversities impact the use of colorectal cancer screenings. This is consistent with previous research that has shown that not all adversities are related to cervical cancer screenings (Farley et al., 2002). This is also consistent with recent arguments that contend that certain adversities, like parental divorce, may be less damaging because they are more common (Finkelhor, Shattuck, Turner, & Hamby, 2015). Thus, future research investigating the health impact of ACEs should carefully consider the consequences of summing ACE items into a scale. Doing so assumes that ACEs have a cumulative impact and ignores the possibility that only specific ACEs are important in the relationship under examination. This study also showed important findings related to ACEs and CRC screening. In bivariate models, some ACEs (physical abuse, having parents that were separated or divorced and living in a household were adults treated each other violently) were associated with decreased odds of CRC screening. This suggests that at the population level, specific ACEs may be associated with reduced utilization of CRC screening. This is important for public health planning of CRC promotion efforts. Conversely, after controlling for socio-demographic characteristics all inverse associations were eliminated. As such, the initial bivariate associations may be due to underlying sociodemographic and other characteristics, which may be important risk factors independent of ACEs (i.e. age, gender and race) or may be compromised by ACEs (i.e. insurance status, cancer and educational attainment). Relatedly, because suppression analyses revealed age and location function as important suppressors, we know that difference in access to resources that could be impacted by ACEs (i.e. insurance status and educational attainment) where not obscuring the association. These findings suggest that, after accounting for confounders, individuals who experience child sexual or emotional abuse may be more likely to engage in CRC screening. While we are unable to test reasons for these findings in the present study, it may be that these individuals are sicker in general and thus consuming more health services or they are more motivated to seek preventative services given their previous adversity experiences. This contrasts to previous work in the cervical cancer screening context which showed that a history of child sexual abuse was associated with lower odds of screening (Alcalá, Mitchell et al., 2016; Farley et al., 2002). This difference may be attributable to the fact that, at the time of the study, cervical cancer screening was recommended of women age 21–65 (Salloum et al., 2014), while CRC screening is recommended for individuals age 50–70 (U.S. Preventative Services Task Force, 2008). As such, individuals with the highest burden and impact of ACEs may not survive long enough to need CRC screening, thus yielding discordant results. Thus, longitudinal work in this area is greatly needed. There are several limitations to consider when interpreting the results of the present study. First, because the data are crosssectional several problems are inherent (i.e. directionality of relationships cannot be established and recall bias cannot be accounted for). Second, the study sample is limited to two US states, which may or may not reflect the rest of the country. As a result, generalizations must be made with caution. Third, while weights and adjustments created by BRFSS attempt to make the survey representative of the underlying population, non-response bias has still been a documented problem for the BRFSS (Schneider, Clark, Rakowski, & Lapane, 2012). Consequently, the present analyses may not be fully representative of the states that were included. Finally, because of the way the BRFSS asked questions about CRC screening we cannot determine if ACEs impact the type of CRC screening an individual chooses. Future research is needed to determine if ACEs have an impact on preference for screening modalities, which could in turn have strong implications for healthcare providers attempting to address healthcare needs specific to this population. Despite limitations, the study provides a deeper understanding into the association between ACEs and health. Our results suggest that experiences of adversity are associated with lower CRC screening at the population level, but this association is fueled by other factors. Consequently, population-based efforts to promote screening may find value in promoting screening among those with specific ACEs. In the one-on-one medical encounter health care providers who screen for or become aware of a history of ACEs may find it beneficial to encourage CRC screening to those patients. Furthermore, because findings showed that some ACEs are associated with increased CRC screening, after adjusting for confounders, further research should be conducted to understand why adverse childhood experiences may result in positive health behaviors. Conflict of interest Authors have no conflicts to disclose. References Achat, H., Close, G., & Taylor, R. (2005). Who has regular mammograms? Effects of knowledge, beliefs, socioeconomic status, and health-related factors. Preventive Medicine, 41(1), 312–320. Alcalá, H. E. (2016). Making the connection between child abuse and cancer: Definitional, methodological, and theoretical issues. Social Theory & Health, 14(4), 458–474. http://dx.doi.org/10.1057/s41285-016-0016-z. Alcalá, H. E., Mitchell, E., & Keim-Malpass, J. (2017). Adverse childhood experiences and cervical cancer screening. Journal of Women's Health, 26(1), 58–63. http://dx. doi.org/10.1089/jwh.2016.5823. Alcalá, H. E., von Ehrenstein, O. S., & Tomiyama, A. J. (2016). Adverse childhood experiences and use of cigarettes and smokeless tobacco products. Journal of Community Health, 1–8. http://dx.doi.org/10.1007/s10900-016-0179-5. Anda, R. F., Felitti, V. J., Bremner, J. D., Walker, J. D., Whitfield, C., Perry, B. D., ... Giles, W. H. (2006). The enduring effects of abuse and related adverse experiences in childhood. A convergence of evidence from neurobiology and epidemiology. European Archives of Psychiatry and Clinical Neuroscience, 256(3), 174–186. http:// dx.doi.org/10.1007/s00406-005-0624-4. Astone, N. M., & McLanahan, S. S. (1991). Family structure, parental practices and high school completion. American Sociological Review, 309–320.
149
Child Abuse & Neglect 69 (2017) 145–150
H.E. Alcalá et al.
Batten, S. V., Aslan, M., Maciejewski, P. K., & Mazure, C. M. (2004). Childhood maltreatment as a risk factor for adult cardiovascular disease and depression. Journal of Clinical Psychiatry, 65(2), 249–254. Boden, J. M., Horwood, L. J., & Fergusson, D. M. (2007). Exposure to childhood sexual and physical abuse and subsequent educational achievement outcomes. Child Abuse & Neglect, 31(10), 1101–1114. http://dx.doi.org/10.1016/j.chiabu.2007.03.022. Brown, D. W., Anda, R. F., Felitti, V. J., Edwards, V. J., Malarcher, A. M., Croft, J. B., & Giles, W. H. (2010). Adverse childhood experiences are associated with the risk of lung cancer: A prospective cohort study. BMC Public Health, 10, 20. http://dx.doi.org/10.1186/1471-2458-10-20. Brown, M. J., Thacker, L. R., & Cohen, S. A. (2013). Association between adverse childhood experiences and diagnosis of cancer. PUBLIC LIBRARY OF SCIENCE, 8(6), e65524. http://dx.doi.org/10.1371/journal.pone.0065524. CDC (2011a). Overview: BRFSS 2011. from http://www.cdc.gov/brfss/annual_data/2011/overview_11.pdf. CDC (2011b). Preparing 2011 BRFSS module data for analysis. from http://www.cdc.gov/brfss/annual_data/2011/BRFSS2011_Analysis.pdf. Centers for Disease Control and Prevention (2014). ACE study: Data and statistics. Retrieved January 3, 2015, from http://www.cdc.gov/violenceprevention/acestudy/ prevalence.html. Chartier, M. J., Walker, J. R., & Naimark, B. (2010). Separate and cumulative effects of adverse childhood experiences in predicting adult health and health care utilization. Child Abuse & Neglect, 34(6), 454–464. http://dx.doi.org/10.1016/j.chiabu.2009.09.020. Coker, A. L., Hopenhayn, C., DeSimone, C. P., Bush, H. M., & Crofford, L. (2009). Violence against women raises risk of cervical cancer. Journal of Women's Health (Larchmt), 18(8), 1179–1185. http://dx.doi.org/10.1089/jwh.2008.1048. Currie, J., & Spatz Widom, C. (2010). Long-term consequences of child abuse and neglect on adult economic well-being. Child Maltreatment, 15(2), 111–120. http://dx. doi.org/10.1177/1077559509355316. Dube, S. R., Anda, R. F., Felitti, V. J., Edwards, V. J., & Croft, J. B. (2002). Adverse childhood experiences and personal alcohol abuse as an adult. Addictive Behaviors, 27(5), 713–725. http://dx.doi.org/10.1016/S0306-4603(01)00204-0. Farley, M., Golding, J. M., & Minkoff, J. R. (2002). Is a history of trauma associated with a reduced likelihood of cervical cancer screening? Journal of Family Practice, 51(10), 827–831. Felitti, V. J., Anda, R. F., Nordenberg, D., Williamson, D. F., Spitz, A. M., Edwards, V., ... Marks, J. S. (1998). Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults. The Adverse Childhood Experiences (ACE) Study. American Journal of Preventive Medicine, 14(4), 245–258. Finkelhor, D., Shattuck, A., Turner, H., & Hamby, S. (2015). A revised inventory of adverse childhood experiences. Child Abuse & Neglect, 48, 13–21. http://dx.doi.org/ 10.1016/j.chiabu.2015.07.011. Hillis, S. D., Anda, R. F., Felitti, V. J., & Marchbanks, P. A. (2001). Adverse childhood experiences and sexual risk behaviors in women: A retrospective cohort study. Family Planning Perspectives, 33(5), 206–211. http://dx.doi.org/10.2307/2673783. Kelly-Irving, M., Lepage, B., Dedieu, D., Lacey, R., Cable, N., Bartley, M., ... Delpierre, C. (2013). Childhood adversity as a risk for cancer: Findings from the 1958 British birth cohort study. BMC Public Health, 13(1), 767. http://dx.doi.org/10.1186/1471-2458-13-767. Lehman, B. J., Taylor, S. E., Kiefe, C. I., & Seeman, T. E. (2005). Relation of childhood socioeconomic status and family environment to adult metabolic functioning in the CARDIA study. Psychosomatic Medicine, 67(6), 846–854. http://dx.doi.org/10.1097/01.psy.0000188443.48405.eb. MacKinnon, D. P., Krull, J. L., & Lockwood, C. M. (2000). Equivalence of the mediation, confounding and suppression effect. Prevention Science, 1(4), 173–181. Mersky, J. P., Topitzes, J., & Reynolds, A. J. (2013). Impacts of adverse childhood experiences on health, mental health, and substance use in early adulthood: A cohort study of an urban, minority sample in the U.S. Child Abuse & Neglect, 37(11), 917–925. http://dx.doi.org/10.1016/j.chiabu.2013.07.011. Paolucci, E. O., Genuis, M. L., & Violato, C. (2001). A meta-analysis of the published research on the effects of child sexual abuse. The Journal of Psychology, 135(1), 17–36. Paradise, J. E., Rose, L., Sleeper, L. A., & Nathanson, M. (1994). Behavior, family function, school performance, and predictors of persistent disturbance in sexually abused children. Pediatrics, 93(3), 452–459. Rundle, A., Neckerman, K., Sheehan, D., Jankowski, M., Kryvenko, O., Tang, D., & Rybicki, B. (2013). A prospective study of socioeconomic status, prostate cancer screening and incidence among men at high risk for prostate cancer. Cancer Causes & Control, 24(2), 297–303. http://dx.doi.org/10.1007/s10552-012-0108-6. Salloum, R. G., Kohler, R. E., Jensen, G. A., Sheridan, S. L., Carpenter, W. R., & Biddle, A. K. (2014). U.S. preventive services task force recommendations and cancer screening among female medicare beneficiaries. Journal of Women's Health, 23(3), 211–217. http://dx.doi.org/10.1089/jwh.2013.4421. Schneider, K. L., Clark, M. A., Rakowski, W., & Lapane, K. L. (2012). Evaluating the impact of non-response bias in the behavioral risk factor surveillance system (BRFSS). Journal of Epidemiology and Community Health, 66(4), 290–295. http://dx.doi.org/10.1136/jech.2009.103861. Segnan, N. (1996). Socioeconomic status and cancer screening, 138, IARC scientific publications369–376. Siegel, R., Naishadham, D., & Jemal, A. (2013). Cancer statistics, 2013. CA : A Cancer Journal for Clinicians, 63(1), 11–30. Siegel, R., DeSantis, C., & Jemal, A. (2014). Colorectal cancer statistics, 2014. CA : A Cancer Journal for Clinicians, 64(2), 104–117. Singh, S. M., Paszat, L. F., Li, C., He, J., Vinden, C., & Rabeneck, L. (2004). Association of socioeconomic status and receipt of colorectal cancer investigations: A population-based retrospective cohort study. Canadian Medical Association Journal, 171(5), 461–465. Stein, J. A., Fox, S. A., & Murata, P. J. (1991). The influence of ethnicity, socioeconomic status, and psychological barriers on use of mammography. Journal of Health and Social Behavior, 101–113. U.S. Preventative Services Task Force (2008). Colorectal cancer: Screening. from http://www.uspreventiveservicestaskforce.org/Page/Document/ UpdateSummaryFinal/colorectal-cancer-screening. Whitlock, E. P., Lin, J. S., Liles, E., Beil, T. L., & Fu, R. (2008). Screening for colorectal cancer: A targeted, updated systematic review for the U.S. preventive services task force. Annals of Internal Medicine, 149(9), 638–658. http://dx.doi.org/10.7326/0003-4819-149-9-200811040-00245.
150
Contents lists available at ScienceDirect
Child Abuse & Neglect journal homepage: www.elsevier.com/locate/chiabuneg
Research article
Colorectal cancer screening and adverse childhood experiences: Which adversities matter?
MARK
⁎
Héctor E. Alcalá , Jessica Keim-Malpass, Emma Mitchell University of Virginia, United States
AR TI CLE I NF O
AB S T R A CT
Keywords: Adverse childhood experiences Cancer Colorectal cancer Cancer screening Child abuse
Adverse Childhood Experiences (ACEs) have been associated with an increased risk of a variety of diseases, including cancer. However, research has not paid enough attention to the association between ACEs and cancer screening. As such, the present study examined the association between ACEs and ever using colorectal cancer (CRC) screening, among adults age 50 and over. Analyses used the 2011 Behavioral Risk Factor Surveillance System (n = 24,938) to model odds of ever engaging in CRC screening from nine different adversities. Bivariate and multivariate models were fit. In bivariate models, physical abuse, having parents that were divorced or separated, and living in a household where adults treated each other violently were associated with lower odds of engaging in CRC. In multivariate models that accounted for potential confounders, emotional and sexual abuse were each associated with higher odds of engaging in CRC. Results suggest potential pathways by which early childhood experiences can impact future health behaviors. Future research should examine this association longitudinally.
1. Introduction Adverse childhood experiences (ACEs) encompass childhood conditions such as abuse and household dysfunction (Centers for Disease Control and Prevention, 2014) and have been linked to many negative health outcomes for adults. A greater number of ACEs has been positively associated with ischemic heart disease (Felitti et al., 1998), stroke (Felitti et al., 1998), chronic bronchitis or emphysema (Felitti et al., 1998), mental illness and substance use, (Anda et al., 2006; Felitti et al., 1998; Mersky, Topitzes, & Reynolds, 2013) severe obesity (Anda et al., 2006), adult compromised metabolic functioning (Lehman, Taylor, Kiefe, & Seeman, 2005) and cardiovascular disease (Batten, Asl an, Maciejewski, & Mazure, 2004). A growing body of research has linked experiences of childhood adversity to cancer (Alcalá, 2016; Brown et al., 2010; Felitti et al., 1998; Kelly-Irving et al., 2013). Despite the growing interest in the association between childhood adversity and cancer, there is a lack of research examining how childhood adversity impacts the utilization of preventative health care services, like cancer screening, which may help us understand why the association exists to begin with. One cancer that can be acted upon in early stages by utilization of health care and consistent screening is colorectal cancer (CRC). Colorectal cancer kills more than 50,000 people in the United States annually, ranking it behind only lung and bronchial cancers in terms of site-specific cancer deaths (Siegel, Naishadham, & Jemal, 2013). Fortunately, several screening options exist including fecal occult blood tests (FOBT), colonoscopy and sigmoidoscopy, that have been proven effective at disease detection (Whitlock, Lin, Liles, Beil, & Fu, 2008). Consequently, opportunities for treatment of colorectal cancer exists if screening identifies cancer in its early stages. Unfortunately, sizable proportions of the US population have not undergone recent CRC screening (Siegel, DeSantis, & Jemal, 2014).
⁎
Corresponding author. E-mail address: [email protected] (H.E. Alcalá).
http://dx.doi.org/10.1016/j.chiabu.2017.04.026 Received 22 February 2016; Received in revised form 30 March 2017; Accepted 25 April 2017 0145-2134/ © 2017 Elsevier Ltd. All rights reserved.
Child Abuse & Neglect 69 (2017) 145–150
H.E. Alcalá et al.
Current guidelines recommend that adults aged 50–75 undergo screening for CRC (U.S. Preventative Services Task Force, 2008). Given the importance of CRC screening, it is critical to examine if ACEs impact uptake of screening. Available information suggests that ACEs may be associated with utilization of health care services, but that this relationship depends on both the adversity and services under examination. Reporting a greater number of ACEs has been associated with higher odds of having more than six general practitioner visits, more than two emergency room visits and 25 or more visits to health professionals (i.e. family doctor, medical specialist, nurse, optometrist, chiropractor, physiotherapist, dentist, pharmacist, psychologist or other health professional) in the past 12 months (Chartier, Walker, & Naimark, 2010). When the same study examined specific ACE items (which included measures of physical abuse, sexual abuse and other adversities), only some ACE items were associated with use of certain services. Of note, sexual and physical abuse were not associated with use of general practitioner services but were associated with higher use of specialty care and emergency services. This suggests that abuse does not promote use of preventive care (Chartier et al., 2010). In terms of cancer screening, childhood sexual abuse is associated with lower odds of current compliance with screening recommendations, after accounting for confounders (Alcalá, Mitchell, & Keim-Malpass, 2017), while eight other ACEs (i.e. physical abuse, parental drug use, parental separation etc.) where not associated with cervical cancer screening. Findings not focusing solely on ACEs have also shown that childhood sexual abuse is associated with lower odds of cervical cancer screening (Farley, Golding, & Minkoff, 2002). Consequently, sexual abuse, and not other ACEs, may discourage use of cervical cancer screening. Discrepancies between the patterns observed more between omnibus measures of utilization of care and cancer screening may exist for a variety of reasons. First, the discrepancy between cervical cancer screening and general use of care may be due to the invasive nature of cervical cancer screenings. Women with a history of sexual abuse may be re-traumatized by this procedure in a way that women experiencing other adversities are not. Similarly, it is possible that victims of sexual abuse involving anal penetration may avoid potential re-traumatization from procedures like colonoscopies and sigmoidoscopies. Second, the discrepancy in findings may be because cancer screening is preventative in nature, while other services, like emergency department utilization are usually motivated by existing health care needs. This is important to note because people who experience ACEs have worse health overall and thus need more health care. Finally, these findings may not be discrepant at all. Specifically, because cancer screenings prevent cancer, avoiding these screening services may necessitate usage of other services like the emergency department. ACEs, broadly speaking, may discourage use of preventative health services, like cancer screening, because people who have experienced ACEs may have a broader pattern of engaging in negative health behaviors. Specifically, ACEs have been linked to riskier health behaviors such as alcohol abuse, substance abuse, tobacco use, having more sexual partners, not using contraception, earlier age at sexual debut and being physically inactive (Alcalá, von Ehrenstein, & Tomiyama, 2016; Coker, Hopenhayn, DeSimone, Bush, & Crofford, 2009; Dube, Anda, Felitti, Edwards, & Croft, 2002; Felitti et al., 1998; Hillis, Anda, Felitti, & Marchbanks, 2001). Thus, individuals who experience ACEs may eschew preventative health care because of an underlying predisposition to not engage in health promoting behaviors, which differs from use of emergency or specialty care, which will arise out of medical necessity. Finally, ACEs may negatively impact utilization of health care because of the confluence of ACEs with limited socioeconomic status. Several specific ACEs have been associated with diminished socioeconomic status in adulthood and adolescence (Astone & McLanahan, 1991; Boden, Horwood, & Fergusson, 2007; Currie & Spatz Widom, 2010; Paolucci, Genuis, & Violato, 2001; Paradise, Rose, Sleeper, & Nathanson, 1994). Lower SES, in turn, has been associated with decreased usage of several different cancer screening procedures (Achat, Close, & Taylor, 2005; Rundle et al., 2013; Segnan, 1996; Stein, Fox, & Murata, 1991), including CRC screening (Singh et al., 2004). The goal of the present study is to examine associations between ACEs and colorectal cancer screening. Given that this is an understudied area of inquiry, we are specifically interested in whether or not specific adversities occurring in childhood impact colorectal cancer screening among adults age 50 and older.
2. Methods 2.1. Data source This study utilized publically available data from the 2011 Behavioral Risk Factor Surveillance System (BRFSS). This multistage, random digit dial telephone survey is designed to be representative of non-institutionalized adults (ages 18 and over) living in U.S. states and territories. The BRFSS is an annual survey, with a core set of questions asked of all participants in all states and territories. Optional questions were asked of all or some participants in states or territories choosing to administer them. Core questions were collected using both landlines and cell phones in all states and territories. Optional questions were administered with landlines and/or cellphones (CDC, 2011a). In the 2011 BRFSS cycle, only 10 states (California, Maine, Minnesota, Montana, Nebraska, Nevada, Oregon, Vermont, Washington and Wisconsin) administered a module of questions measuring ACEs (CDC, 2011b). Of these 10 states, only two states (Nebraska and Washington) administered a module of questions about CRC screening. Nebraska and Washington had a total of 28,198 respondents who were over 50. Of these, 3,260 were excluded from the present analyses because they had responses of, “don’t know” or “refused” or because they were missing on any variable in the present study. This yielded an analytic sample of 24,938 respondents.
146
Child Abuse & Neglect 69 (2017) 145–150
H.E. Alcalá et al.
2.2. Measures The main independent variables for analyses were from the Adverse Childhood Experience (ACE) scale. The BRFSS included an 11-item version of this scale. The items measured if the following events had been experienced by the respondent prior to age 18: 1) physical abuse; 2) being touched sexually; 3) having someone attempt to make the respondent touch someone sexually; 4) being forced to have sex; 5) psychological/emotional abuse; 6) living with an adult who was depressed, mentally ill or suicidal; 7) living with anyone who was a problem drinker or alcoholic; 8) living with a drug user or abuser; 9) living with someone who had been sentenced to serve time in a prison, jail, or other correctional facility; 10) having parents who were separated or divorced and 11) living in a home where adults or parents physically harmed each other (Felitti et al., 1998). Items about mental illness, problem drinking, drug use, incarceration and separation or divorce asked whether or not the respondent had been exposed to the specific adversity. The remaining items allowed respondents to specify the frequency of occurrence of the specific adversity (i.e. never, once or more than once). For the current study, all items were dichotomized to indicate if the experience had ever happened. This coding scheme is consistent with previous studies but refrains from summing these items into a scale (Brown, Thacker, & Cohen, 2013). Because three items measured facets of sexual abuse, they were combined into a single measure of sexual abuse. This resulted in nine ACE measures. The dependent variable of interest was whether or not a respondent had ever been screened for CRC. The BRFSS asked individuals age 50 and over about their CRC screening history. Respondents were asked, “A blood stool test is a test that may use a special kit at home to determine whether the stool contains blood. Have you ever had this test using a home kit?” If they indicated not having a FOBT, they received the follow-up question, “Sigmoidoscopy and colonoscopy are exams in which a tube is inserted in the rectum to view the colon for signs of cancer or other health problems. Have you ever had either of these exams?” Respondents were considered to ever have undertaken CRC screening if they responded affirmatively to either question. Several variables were selected as potential confounders. Age was included as a continuous variable. Gender was measured as a dichotomous variable. Race and ethnicity was measured using a categorical variable representing the Office of Management and Budget’s race and ethnicity categories (i.e. non-Latino white, non-Latino Black/African American, non-Latino Asian, non-Latino other race, and Latino). State of residence was measured using a dichotomous variable measuring whether the respondent resided in Nebraska or Washington. Educational attainment was recoded from its original categories (i.e. kindergarten or less, 1st through 8th grade, 9th through 11th grade, high school graduate, 1–3 years of college and 4 or more years of college) to continuous values that represented the midpoint of the category in terms of years of education, except for the last category which was coded as 16 years. Current insurance status was measured using a dichotomous variable, indicating whether or not the respondent had any health insurance coverage. A dichotomous indicator of lifetime cancer diagnoses, excluding skin cancer, was also included as a control. 2.3. Analyses All analyses were conducted using Stata 14.1, using appropriate weights to account for survey design. Univariate statistics were run for all study variables. Next three groups of logistic regression models were fitted to estimate odds of ever receiving colorectal cancer screenings. For the first group, bivariate associations between each of the nine ACE measures and CRC screening were modeled. For the second group, confounders were added to these models. In the final group, a model including all ACE measures simultaneously, while also including confounders was modeled. 3. Results Table 1 shows the sample characteristics. Over a fourth of respondents were ever screened for CRC. Emotional abuse was the most commonly experienced ACE measure (33.14%) and living with someone who was jailed (3.41%) was the least commonly experienced. Slightly over half of respondents were female. A majority of respondents were non-Latino white. On average, respondents had completed slightly less than a 2-year college degree and were just over 63 years old. About an eighth of the sample had ever been diagnosed with cancer. The vast majority of respondents were currently insured. More than three fourths of respondents were from Washington. Model 1 in Table 2 shows the bivariate associations between individual ACE items and ever receiving colorectal cancer screening. Here, three specific adversities were associated with lower odds of colorectal cancer screenings: physical abuse (OR = 0.89; 95% CI = 0.81,0.99), having parents who were separated or divorced (OR = 0.90; 95% CI = 0.81,0.99) and living with adults who treated each other violently (OR = 0.86; 95% CI = 0.78,0.96). In Model 2, after accounting for confounders, sexual (AOR = 1.18; 95% CI = 1.06,1.33) and emotional abuse (AOR = 1.09; 95% CI = 1.01,1.18) were associated with increased odds of CRC screening. Because both of these measures were not associated with CRC screening in bivariate models, follow-up analyses were conducted to investigate which of the confounding variables were causing the change in significance, that is which variables were functioning as potential suppressors (MacKinnon, Krull, & Lockwood, 2000). Briefly, suppression requires that the suppressor be significantly associated, in opposite directions, with the independent and dependent variable. To examine this, each control variable was individually introduced into bivariate models, to yield three variable models. For sexual abuse, the introduction of state of residence and age each made the association with CRC screening significant. CRC screening was less common in Washington, relative to Nebraska (OR = 0.61; 95% CI = 0.56,0.67), while sexual abuse was more common in Washington relative to Nebraska (OR = 1.70; 95% CI = 1.44,2.02), thus, supporting the argument that state of residence functioned as a suppressor. CRC screening was more common with increasing age (OR = 1.02; 95% CI = 1.01,1.02), while sexual 147
Child Abuse & Neglect 69 (2017) 145–150
H.E. Alcalá et al.
Table 1 Sample Characteristics, Adults 50 and older, BRFSS 2011 (N = 24,938). Variable Colorectal Cancer Screening Ever screened Never screened Adverse childhood experiences Physical abuse Sexual abuse Emotional abuse Lived with some one who was mentally ill Lived with problem drinker Lived with drug user Lived with some one who was jailed Parents divorced or separated Adults in household treated each other violently Age Gender Male Female Race White Black Hispanic Asian Other Educational attainment (years) Currently insured No Yes Ever diagnosed with cancer No Yes State of residence Nebraska Washington
% or Mean
SE
27.05% 72.95%
0.35% 0.35%
16.91% 14.15% 33.14% 14.49% 24.15% 5.97% 3.41% 17.11% 15.99% 63.33
0.48% 0.43% 0.58% 0.43% 0.53% 0.33% 0.26% 0.48% 0.48% 0.12
47.21% 52.79%
0.06% 0.06%
87.29% 1.95% 3.93% 3.06% 3.77% 13.66
0.53% 0.24% 0.33% 0.31% 0.25% 0.03
8.68% 91.32%
0.42% 0.42%
87.52% 12.48%
0.36% 0.36%
21.21% 78.79%
0.31% 0.31%
abuse was less common with increasing age (OR = 0.98; 95% CI = 0.97,0.98), thus, supporting the argument that age functioned as a suppressor. For emotional abuse, no single variable altered the significance from the bivariate model, suggesting that the cumulative effect of control variables functioned as suppressors. In Model 3, after accounting for confounders and all ACEs, sexual abuse (OR = 1.18; 95% CI = 1.05,1.32) was associated with increased odds of CRC screening.
Table 2 Logistic Regression Modeling Odds for Colorectal Cancer, Adults 50 and older, BRFSS 2011 (N = 24,938). Model 1: Bivariate Variable Physical abuse Sexual abuse Emotional abuse Lived with some one who was mentally ill Lived with problem drinker Lived with drug user Lived with some one who was jailed Parents divorced or separated Adults in household treated each other violently
OR 0.89 1.08 0.98 1.00 0.97 0.85 0.86 0.90 0.86
95% CI *
(0.81,0.99) (0.97,1.20) (0.91,1.06) (0.90,1.11) (0.89,1.06) (0.71,1.02) (0.67,1.11) (0.81,0.99)* (0.78,0.96)**
Model 2: Controls
Model 3: All ACE measures and Controls
AOR
95% CI
AOR
95% CI
1.00 1.18 1.09 1.07 1.07 0.99 1.02 1.01 0.98
(0.90,1.10) (1.06,1.33)** (1.01,1.18)* (0.96,1.18) (0.98,1.16) (0.83,1.19) (0.80,1.31) (0.92,1.12) (0.88,1.09)
0.93 1.18 1.09 1.04 1.06 0.95 0.99 0.99 0.92
(0.82,1.05) (1.05,1.32)** (1.00,1.20) (0.93,1.15) (0.97,1.17) (0.79,1.13) (0.78,1.27) (0.89,1.10) (0.82,1.04)
Model 2 adds age, gender, race, years of education, cancer status, state of residence and insurance status as controls to bivariate models. Model 3 introduces all ACEs simultaneously while also controlling for age, gender, race, years of education, cancer status, state of residence and current insurance status as controls to bivariate models. * p < 0.05. ** p < 0.01.
148
Child Abuse & Neglect 69 (2017) 145–150
H.E. Alcalá et al.
4. Discussion To the authors’ knowledge, this study represents the first investigation of the association between ACEs and CRC screening. Results revealed that not all adversities impact the use of colorectal cancer screenings. This is consistent with previous research that has shown that not all adversities are related to cervical cancer screenings (Farley et al., 2002). This is also consistent with recent arguments that contend that certain adversities, like parental divorce, may be less damaging because they are more common (Finkelhor, Shattuck, Turner, & Hamby, 2015). Thus, future research investigating the health impact of ACEs should carefully consider the consequences of summing ACE items into a scale. Doing so assumes that ACEs have a cumulative impact and ignores the possibility that only specific ACEs are important in the relationship under examination. This study also showed important findings related to ACEs and CRC screening. In bivariate models, some ACEs (physical abuse, having parents that were separated or divorced and living in a household were adults treated each other violently) were associated with decreased odds of CRC screening. This suggests that at the population level, specific ACEs may be associated with reduced utilization of CRC screening. This is important for public health planning of CRC promotion efforts. Conversely, after controlling for socio-demographic characteristics all inverse associations were eliminated. As such, the initial bivariate associations may be due to underlying sociodemographic and other characteristics, which may be important risk factors independent of ACEs (i.e. age, gender and race) or may be compromised by ACEs (i.e. insurance status, cancer and educational attainment). Relatedly, because suppression analyses revealed age and location function as important suppressors, we know that difference in access to resources that could be impacted by ACEs (i.e. insurance status and educational attainment) where not obscuring the association. These findings suggest that, after accounting for confounders, individuals who experience child sexual or emotional abuse may be more likely to engage in CRC screening. While we are unable to test reasons for these findings in the present study, it may be that these individuals are sicker in general and thus consuming more health services or they are more motivated to seek preventative services given their previous adversity experiences. This contrasts to previous work in the cervical cancer screening context which showed that a history of child sexual abuse was associated with lower odds of screening (Alcalá, Mitchell et al., 2016; Farley et al., 2002). This difference may be attributable to the fact that, at the time of the study, cervical cancer screening was recommended of women age 21–65 (Salloum et al., 2014), while CRC screening is recommended for individuals age 50–70 (U.S. Preventative Services Task Force, 2008). As such, individuals with the highest burden and impact of ACEs may not survive long enough to need CRC screening, thus yielding discordant results. Thus, longitudinal work in this area is greatly needed. There are several limitations to consider when interpreting the results of the present study. First, because the data are crosssectional several problems are inherent (i.e. directionality of relationships cannot be established and recall bias cannot be accounted for). Second, the study sample is limited to two US states, which may or may not reflect the rest of the country. As a result, generalizations must be made with caution. Third, while weights and adjustments created by BRFSS attempt to make the survey representative of the underlying population, non-response bias has still been a documented problem for the BRFSS (Schneider, Clark, Rakowski, & Lapane, 2012). Consequently, the present analyses may not be fully representative of the states that were included. Finally, because of the way the BRFSS asked questions about CRC screening we cannot determine if ACEs impact the type of CRC screening an individual chooses. Future research is needed to determine if ACEs have an impact on preference for screening modalities, which could in turn have strong implications for healthcare providers attempting to address healthcare needs specific to this population. Despite limitations, the study provides a deeper understanding into the association between ACEs and health. Our results suggest that experiences of adversity are associated with lower CRC screening at the population level, but this association is fueled by other factors. Consequently, population-based efforts to promote screening may find value in promoting screening among those with specific ACEs. In the one-on-one medical encounter health care providers who screen for or become aware of a history of ACEs may find it beneficial to encourage CRC screening to those patients. Furthermore, because findings showed that some ACEs are associated with increased CRC screening, after adjusting for confounders, further research should be conducted to understand why adverse childhood experiences may result in positive health behaviors. Conflict of interest Authors have no conflicts to disclose. References Achat, H., Close, G., & Taylor, R. (2005). Who has regular mammograms? Effects of knowledge, beliefs, socioeconomic status, and health-related factors. Preventive Medicine, 41(1), 312–320. Alcalá, H. E. (2016). Making the connection between child abuse and cancer: Definitional, methodological, and theoretical issues. Social Theory & Health, 14(4), 458–474. http://dx.doi.org/10.1057/s41285-016-0016-z. Alcalá, H. E., Mitchell, E., & Keim-Malpass, J. (2017). Adverse childhood experiences and cervical cancer screening. Journal of Women's Health, 26(1), 58–63. http://dx. doi.org/10.1089/jwh.2016.5823. Alcalá, H. E., von Ehrenstein, O. S., & Tomiyama, A. J. (2016). Adverse childhood experiences and use of cigarettes and smokeless tobacco products. Journal of Community Health, 1–8. http://dx.doi.org/10.1007/s10900-016-0179-5. Anda, R. F., Felitti, V. J., Bremner, J. D., Walker, J. D., Whitfield, C., Perry, B. D., ... Giles, W. H. (2006). The enduring effects of abuse and related adverse experiences in childhood. A convergence of evidence from neurobiology and epidemiology. European Archives of Psychiatry and Clinical Neuroscience, 256(3), 174–186. http:// dx.doi.org/10.1007/s00406-005-0624-4. Astone, N. M., & McLanahan, S. S. (1991). Family structure, parental practices and high school completion. American Sociological Review, 309–320.
149
Child Abuse & Neglect 69 (2017) 145–150
H.E. Alcalá et al.
Batten, S. V., Aslan, M., Maciejewski, P. K., & Mazure, C. M. (2004). Childhood maltreatment as a risk factor for adult cardiovascular disease and depression. Journal of Clinical Psychiatry, 65(2), 249–254. Boden, J. M., Horwood, L. J., & Fergusson, D. M. (2007). Exposure to childhood sexual and physical abuse and subsequent educational achievement outcomes. Child Abuse & Neglect, 31(10), 1101–1114. http://dx.doi.org/10.1016/j.chiabu.2007.03.022. Brown, D. W., Anda, R. F., Felitti, V. J., Edwards, V. J., Malarcher, A. M., Croft, J. B., & Giles, W. H. (2010). Adverse childhood experiences are associated with the risk of lung cancer: A prospective cohort study. BMC Public Health, 10, 20. http://dx.doi.org/10.1186/1471-2458-10-20. Brown, M. J., Thacker, L. R., & Cohen, S. A. (2013). Association between adverse childhood experiences and diagnosis of cancer. PUBLIC LIBRARY OF SCIENCE, 8(6), e65524. http://dx.doi.org/10.1371/journal.pone.0065524. CDC (2011a). Overview: BRFSS 2011. from http://www.cdc.gov/brfss/annual_data/2011/overview_11.pdf. CDC (2011b). Preparing 2011 BRFSS module data for analysis. from http://www.cdc.gov/brfss/annual_data/2011/BRFSS2011_Analysis.pdf. Centers for Disease Control and Prevention (2014). ACE study: Data and statistics. Retrieved January 3, 2015, from http://www.cdc.gov/violenceprevention/acestudy/ prevalence.html. Chartier, M. J., Walker, J. R., & Naimark, B. (2010). Separate and cumulative effects of adverse childhood experiences in predicting adult health and health care utilization. Child Abuse & Neglect, 34(6), 454–464. http://dx.doi.org/10.1016/j.chiabu.2009.09.020. Coker, A. L., Hopenhayn, C., DeSimone, C. P., Bush, H. M., & Crofford, L. (2009). Violence against women raises risk of cervical cancer. Journal of Women's Health (Larchmt), 18(8), 1179–1185. http://dx.doi.org/10.1089/jwh.2008.1048. Currie, J., & Spatz Widom, C. (2010). Long-term consequences of child abuse and neglect on adult economic well-being. Child Maltreatment, 15(2), 111–120. http://dx. doi.org/10.1177/1077559509355316. Dube, S. R., Anda, R. F., Felitti, V. J., Edwards, V. J., & Croft, J. B. (2002). Adverse childhood experiences and personal alcohol abuse as an adult. Addictive Behaviors, 27(5), 713–725. http://dx.doi.org/10.1016/S0306-4603(01)00204-0. Farley, M., Golding, J. M., & Minkoff, J. R. (2002). Is a history of trauma associated with a reduced likelihood of cervical cancer screening? Journal of Family Practice, 51(10), 827–831. Felitti, V. J., Anda, R. F., Nordenberg, D., Williamson, D. F., Spitz, A. M., Edwards, V., ... Marks, J. S. (1998). Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults. The Adverse Childhood Experiences (ACE) Study. American Journal of Preventive Medicine, 14(4), 245–258. Finkelhor, D., Shattuck, A., Turner, H., & Hamby, S. (2015). A revised inventory of adverse childhood experiences. Child Abuse & Neglect, 48, 13–21. http://dx.doi.org/ 10.1016/j.chiabu.2015.07.011. Hillis, S. D., Anda, R. F., Felitti, V. J., & Marchbanks, P. A. (2001). Adverse childhood experiences and sexual risk behaviors in women: A retrospective cohort study. Family Planning Perspectives, 33(5), 206–211. http://dx.doi.org/10.2307/2673783. Kelly-Irving, M., Lepage, B., Dedieu, D., Lacey, R., Cable, N., Bartley, M., ... Delpierre, C. (2013). Childhood adversity as a risk for cancer: Findings from the 1958 British birth cohort study. BMC Public Health, 13(1), 767. http://dx.doi.org/10.1186/1471-2458-13-767. Lehman, B. J., Taylor, S. E., Kiefe, C. I., & Seeman, T. E. (2005). Relation of childhood socioeconomic status and family environment to adult metabolic functioning in the CARDIA study. Psychosomatic Medicine, 67(6), 846–854. http://dx.doi.org/10.1097/01.psy.0000188443.48405.eb. MacKinnon, D. P., Krull, J. L., & Lockwood, C. M. (2000). Equivalence of the mediation, confounding and suppression effect. Prevention Science, 1(4), 173–181. Mersky, J. P., Topitzes, J., & Reynolds, A. J. (2013). Impacts of adverse childhood experiences on health, mental health, and substance use in early adulthood: A cohort study of an urban, minority sample in the U.S. Child Abuse & Neglect, 37(11), 917–925. http://dx.doi.org/10.1016/j.chiabu.2013.07.011. Paolucci, E. O., Genuis, M. L., & Violato, C. (2001). A meta-analysis of the published research on the effects of child sexual abuse. The Journal of Psychology, 135(1), 17–36. Paradise, J. E., Rose, L., Sleeper, L. A., & Nathanson, M. (1994). Behavior, family function, school performance, and predictors of persistent disturbance in sexually abused children. Pediatrics, 93(3), 452–459. Rundle, A., Neckerman, K., Sheehan, D., Jankowski, M., Kryvenko, O., Tang, D., & Rybicki, B. (2013). A prospective study of socioeconomic status, prostate cancer screening and incidence among men at high risk for prostate cancer. Cancer Causes & Control, 24(2), 297–303. http://dx.doi.org/10.1007/s10552-012-0108-6. Salloum, R. G., Kohler, R. E., Jensen, G. A., Sheridan, S. L., Carpenter, W. R., & Biddle, A. K. (2014). U.S. preventive services task force recommendations and cancer screening among female medicare beneficiaries. Journal of Women's Health, 23(3), 211–217. http://dx.doi.org/10.1089/jwh.2013.4421. Schneider, K. L., Clark, M. A., Rakowski, W., & Lapane, K. L. (2012). Evaluating the impact of non-response bias in the behavioral risk factor surveillance system (BRFSS). Journal of Epidemiology and Community Health, 66(4), 290–295. http://dx.doi.org/10.1136/jech.2009.103861. Segnan, N. (1996). Socioeconomic status and cancer screening, 138, IARC scientific publications369–376. Siegel, R., Naishadham, D., & Jemal, A. (2013). Cancer statistics, 2013. CA : A Cancer Journal for Clinicians, 63(1), 11–30. Siegel, R., DeSantis, C., & Jemal, A. (2014). Colorectal cancer statistics, 2014. CA : A Cancer Journal for Clinicians, 64(2), 104–117. Singh, S. M., Paszat, L. F., Li, C., He, J., Vinden, C., & Rabeneck, L. (2004). Association of socioeconomic status and receipt of colorectal cancer investigations: A population-based retrospective cohort study. Canadian Medical Association Journal, 171(5), 461–465. Stein, J. A., Fox, S. A., & Murata, P. J. (1991). The influence of ethnicity, socioeconomic status, and psychological barriers on use of mammography. Journal of Health and Social Behavior, 101–113. U.S. Preventative Services Task Force (2008). Colorectal cancer: Screening. from http://www.uspreventiveservicestaskforce.org/Page/Document/ UpdateSummaryFinal/colorectal-cancer-screening. Whitlock, E. P., Lin, J. S., Liles, E., Beil, T. L., & Fu, R. (2008). Screening for colorectal cancer: A targeted, updated systematic review for the U.S. preventive services task force. Annals of Internal Medicine, 149(9), 638–658. http://dx.doi.org/10.7326/0003-4819-149-9-200811040-00245.
150