To examine the impact of adverse childhood experiences (ACEs) and support on self-reported work inability of adults reporting disability.
Adults (ages 18–64) who participated in the Behavioral Risk Factor Surveillance System in 2009 or 2010 and who reported having a disability (n = 13,009).
Design and Main Outcome Measures
The study used a retrospective cohort design with work inability as the main outcome. ACE categories included abuse (sexual, physical, emotional) and family dysfunction (domestic violence, incarceration, mental illness, substance abuse, divorce). Support included functional (perceived emotional/social support) and structural (living with another adult) support. Logistic regression was used to adjust for potential confounders (age, sex and race) and to evaluate whether there was an independent effect of ACEs on work inability after adding other important predictors (support, education, health) to the model.
ACEs were highly prevalent with almost 75% of the sample reporting at least one ACE category and over 25% having a high ACE burden (4 or more categories). ACEs were strongly associated with functional support. Participants experiencing a high ACE burden had a higher adjusted odds ratio (OR) [95% confidence interval] of 1.9 [1.5–2.4] of work inability (reference: zero ACEs). Good functional support (adjusted OR 0.52 [0.42–0.63]) and structural support (adjusted OR 0.48 [0.41–0.56]) were protective against work inability. After adding education and health to the model, ACEs no longer appeared to have an independent effect. Structural support remained highly protective, but functional support only appeared to be protective in those with good physical health.
ACEs are highly prevalent in working-age US adults with a disability, particularly young adults. ACEs are associated with decreased support, lower educational attainment and worse adult health. Health care providers are encouraged to screen for ACEs. Addressing the effects of ACEs on health and support, in addition to education and retraining, may increase ability to work in those with a disability.
Citation: Schüssler-Fiorenza Rose SM, Eslinger JG, Zimmerman L, Scaccia J, Lai BS, Lewis C, et al. (2016) Adverse Childhood Experiences, Support, and the Perception of Ability to Work in Adults with Disability. PLoS ONE 11(7): e0157726. https://doi.org/10.1371/journal.pone.0157726
Editor: Daisuke Nishi, National Center of Neurology and Psychiatry, JAPAN
Received: July 6, 2015; Accepted: June 3, 2016; Published: July 5, 2016
This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Data Availability: Original data are available for download from the Centers for Disease Control and Prevention Behavioral Risk Factor Surveillance System website. http://www.cdc.gov/brfss/annual_data/annual_data.htm.
Funding: Supported by the Department of Veteran Affairs Office of Academic Affiliations Advanced Fellowship Program in Spinal Cord Injury Medicine (SMS-FR) http://www.va.gov/oaa/specialfellows/programs/SF_SCIMinfo.asp?p=16 and the Ruth L. Kirschstein National Research Service Award Institutional Research Training grant T32AA00745530 (LZ) http://grants.nih.gov/grants/guide/pa-files/PA-14-015.html. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
In the United States, roughly 37 to 56 million people, including 11.6% of adults aged 18–64 years, live with a disability [1, 2]. Disability is associated with work inability among some individuals, but not others . Over 70% of people with a disability have experienced childhood adversity which is higher than the general population . Adverse childhood experiences (ACEs) have been linked with increased health risk behaviors , worse education outcomes , impaired worker performance , adult psychological distress [5, 8–10], worse physical and mental health [5, 11] and increased disability rates in adulthood . Much less is known about how ACEs may affect the ability to work of people with disability. Increasing the understanding of risk and protective factors that may affect the self-perceived inability to work of individuals with disability is needed, since it can inform rehabilitation assessment and treatment.
To evaluate how ACEs might influence self-perceived ability to work, we used concepts from the International Classification of Functioning, Disability and Health (ICF). The ICF framework conceptualizes disability as comprised of one or more of the following components: impairments of body structures and functions, activity limitations, and participation restrictions. Contextual factors (personal and environmental factors) also influence disability in the ICF conceptual model [13, 14]. Not all those with impairments and/or activity limitations will have participation restrictions (such as work inability). For those with activity limitations, ability to work (participation) could be influenced by contextual factors and health condition-related impairments (Fig 1). Despite evidence that ACEs affect adult mental and physical health and personal factors such as health risk behaviors, less attention has been paid to how ACEs affect environmental factors such as support and to what extent support is a protective factor for work participation by people with health-condition related activity limitations. Receipt of adequate support can help to buffer the effects of life adversity . Deficits in support have been associated with poorer health outcomes [16, 17] and decline in functioning .
Model of how ACEs may affect inability to work based on the International Classification of Functioning, Disability and Health constructs . ACEs may lead to decreased functional and structural support (an environmental factor), worse adult health and decreased educational attainment (personal factors). All of these may impact participation restrictions (which include inability to work—our primary outcome).
Our first aim was to examine adverse childhood experiences (including category, type and number of categories) as a risk factor for work inability among participants with disability in a population-based survey. We hypothesized that as the number of categories of self-reported ACEs increased, the prevalence of self-reported work inability would increase.
Our second aim was to examine the relationship between ACEs and support which was conceptualized as having two components (structural and functional). We hypothesized that high ACEs exposure is a risk factor for poor support in adulthood. Specifically we hypothesized that as the number of ACE categories experienced increased the percentage with worse functional and structural support would also increase. We also hypothesized that better functional and structural support in adulthood would be protective against work inability in adulthood and that ACEs would have an effect of work inability that is independent of their effect on support.
For our third aim we wished to understand the effects of ACEs on work inability while accounting for current health status and educational attainment. ACEs are associated with lower educational attainment [9, 19] and worse physical and mental health in adulthood [5, 20–22]. In addition, both lower education and poor health are associated with work inability [23–25]. Thus we wished to evaluate whether ACEs have an effect on work inability independent of their effect on health status and education. We hypothesized that ACEs would have an independent effect.
For our fourth aim, we evaluated the effect of support on work inability while controlling for ACEs, education and health. We hypothesized that support would have an independent effect on work inability. Furthermore we hypothesized that good support may buffer the effects of poor health and be protective against work inability.
Materials & Methods
We had previously confirmed with the University of Pennsylvania Institutional Review Board that research with the Behavioral Risk Factor Surveillance System’s publicly available de-identified data was not considered human subjects research (per the US federal definition).
We used data from the 2009 and 2010 Behavioral Risk Factor Surveillance System (BRFSS). The BRFSS is a joint project between the Centers for Disease Control and Prevention (CDC) and state health departments which surveys adults (age 18 and older) about their health risk factors and health . In 2009 and 2010, the BRFSS used a disproportionate stratified sampling strategy to identify random households with landline telephones. Post-stratification adjustment of survey weights was performed to ensure a state’s sample demographics reflected its demographic distributions [27, 28]. The survey contained core component questions (demographics, employment, days of poor health, disability status, perceived emotional/social support) used by all U.S. states and territories. It also used optional modules which were incorporated into the survey at the discretion of each state/territory. The ACE module was an optional module used by fourteen states and the District of Columbia in 2009 and/or 2010 . The ACE Module questions (Table 1) were adapted from a prior study on the effects of ACEs .
The state survey cooperation rates (percentage of those contacted who participated) ranged from 68.9 to 82.4. In addition, the Council of American Survey Research Organizations (CASRO) response rates (percentage of the estimated eligible who participated) ranged from 47.0 to 68.7 [27, 30]. Of the 89,810 survey respondents, we limited our sample to respondents younger than age 65 who affirmed one or both of the BRFSS disability questions (n = 14,983): “Are you limited in any way in any activities because of physical, mental or emotional health problems?” and “Do you now have any health problem which requires you to use special equipment, such as a cane, a wheelchair, a special bed, or a special telephone?” . Of these, 652 were missing all ACE questions (97% of these had only partially completed the BRFSS survey). Of the remaining 14,331, an additional 1,322 cases were removed due to missing data (ACE: 536 cases; functional support: 124 cases; employment: 42 cases; physical and/or mental health: 458 cases; race: 139 cases; education: 10 cases; household member: 13 cases). Our final sample size was 13,009.
The presence of ACEs was measured by questions regarding adverse childhood experiences occurring prior to age 18. The module included questions about sexual, physical and emotional abuse. It also contained questions about family dysfunction which included domestic violence, substance abuse, mental illness, and family member incarceration (Table 1). A total ACE score was created by summing the number of endorsed ACE categories (range 0–8). ACEs were also categorized by type: family dysfunction only, abuse only, both abuse and family dysfunction and none. Participants who chose, “Unable to work” as an answer to the question on employment status  were considered to have work inability.
Functional support was determined by the question, “How often do you get the emotional or social support that you need?” The variable was dichotomized into good (“Always,” “Usually”) versus poor (“Sometimes,” “Rarely,” or “Never”) functional support . Good structural support was defined as living with another adult in the household. Educational attainment was divided into 4 categories (less than high school, high school graduate, some college/technical school, college/technical school graduate). Poor physical and mental health were measured by the self-reported number of days of poor physical and mental health in the past thirty days (range 0–30) respectively .
Descriptive and logistic regression analyses were conducted with SAS® 9.3 (SAS Institute, Inc., Cary, NC, 2011). The complex survey design specifications and survey weights were used in all analyses.
We first performed the following descriptive analyses: 1. Prevalence of ACEs. 2. Demographics characteristics and prevalence of ACEs, and work inability in relation to these characteristics. 3. Inability to work, functional and structural support by ACE category, type and score. (We added logistic regression to this analysis to evaluate the relationship while controlling for demographic variables). 4. The prevalence of four or more ACEs and work inability by support, educational attainment and past month physical and mental health.
Demographic characteristics included age, sex and race. These variables were considered potential confounders. Age, in particular, was known to be a negative confounder from prior work . Based on our conceptual model (Fig 1), education and health were considered potential intermediate variables on the causal pathway between ACEs and inability to work. They were treated as other important predictors. Although support was considered to be influenced by ACEs, we also evaluated whether the effect of support varied according to health status. All variables were chosen a priori based on our conceptual model. No additional variables were evaluated.
We then evaluated whether ACEs had an independent effect on work inability after adjusting for contextual factors and health individually. This involved 5 separate logistic regression models: ACEs and functional support, ACEs and structural support, ACEs and education, ACEs and physical health, ACEs and mental health with all models also including the demographic variables (age, sex, and race).
After confirming ACEs had an independent effect in these bivariate models, we then performed a series of logistic regression analysis starting with ACEs only (1) then adding the support variables (2), followed by education (3), and then health (4). Age, gender, and race/ethnicity were included as covariates in all of the multivariate analyses. An odds ratio (OR) whose 95% confidence interval (CI) did not cross one was considered statistically significant (α = 0.05). Although, in descriptive analyses, age was reported categorically, it was entered as a continuous predictor in our logistic regression analyses.
We initially evaluated the ACE Score as a categorical variable, since model fit seemed better using ACE as a categorical variable compared to a continuous variable. However after finding that there was not a significant difference between ACEs 0–3 and that there were wide and overlapping confidence intervals (due to smaller numbers) at higher levels of ACEs, we dichotomized ACE Score into 4 or more/less than 4 categories for our multivariate analyses examining the effect of health and contextual factors.
Days of poor mental and physical health were initially modeled continuously. However, these variables had a non-uniform distribution with peaks at 30 days and 0 days with sparse data in between. There was violation of the logistic regression assumption of a linear relationship between a continuous predictor and the log odds. Thus, we also modeled these variables as categorical variables, using three categories: good (3 days or fewer), intermediate (4–15 days) and poor (16 or greater days of poor health). We also evaluated potential nonlinear transformations of the health variables using spline functions. The choice of how to model days of poor health did not affect the odds ratios (ORs) of ACEs, support or education. Therefore, we chose to present the categorical results, as these are more interpretable. In particular, the health categories were particularly important for the analysis of whether the effect support on work inability varied by level of health. The choice of categories was made prior to analysis. Although days of poor health are dichotomized in other studies , having at least three categories made sense in terms of the outcome measure of work inability. We included up to 3 days of poor health in the good health category with the rationale that even those in good health may have a few days poor health due to a minor illness and that these may not interfere with employment. Having poor health for over half the month was considered likely to have a strong effect on ability to work.
We evaluated possible differential effects of support on inability to work by health status by stratifying by physical and mental health categories separately. We also tested whether there were statistically significant differences between groups by adding an interaction term to our final model. Where there appeared to be a substantial variation of effect by strata, we also modeled the joint effect against a common reference . We assessed multiplicative interaction, which was considered most appropriate for our data [33, 34].
ACEs were common, with 72.4% of participants having experienced at least one ACE category and 27.9% having experienced four or more categories. The most commonly reported ACEs were emotional abuse (42.8%) and living with a substance abusing parent/caregiver (38.5%). The least common ACE was having an incarcerated caregiver (11.1%) Over forty percent of participants had experienced both abuse and family dysfunction categories. (Table 1).
Table 2 shows the demographic distribution of the sample along with the prevalence of ACEs and work inability in each group. Of the 13,009 respondents, 51.9% were female, and the majority of respondents (81.9%) identified as non-Hispanic white (Table 2). Over forty percent of the youngest age group (ages 18–29) reported experiencing at least four ACE categories which was over twice the rate of the oldest group (ages 55–64). The oldest two groups (45–64) had almost twice the rate of being unable to work than the youngest group. (Table 2) Although in the overall prevalence of work inability was 15.8% in youngest age group; in young adults reporting a high ACE burden, the prevalence was 27.4% (95% CI 15.7–39.1) versus 6.7% (3.2–10.2) for those with lower ACE exposure.
Table 3 explores the relationship between ACEs and work inability and support. For all ACE categories, the prevalence of work inability was higher for those having experienced the category than for those who had not. The prevalence of work inability for those who experienced a particular ACE category was close to 30% for most categories. The prevalence of work inability was over 10% higher in those experiencing the family dysfunction ACE categories (over 27%) compared to those experiencing abuse only ACE categories (16.6%) Participants experiencing 4 or more ACE categories had a work inability prevalence of 31.4% which was over 10% higher than the prevalence of those who had not experienced any ACEs. Compared to those reporting no ACEs, the adjusted odds (95% confidence interval) of work inability of those with an ACE score ≥ 4 was [OR 1.91 (1.55–2.44), p <0.0001] (Table 3).
The association between ACEs and functional support appeared stronger than between ACEs and structural support (Table 3). For each ACE category, around 40% of those who had experienced the category reported poor functional support (range 38–45%). This percentage was universally higher than the percentage of those who had not experienced the category. The percentage of those reporting poor functional support increased from 17.7% (0 ACEs) to 44.7% (4 or more ACE categories). In contrast, there was a smaller increase in percentage reporting poor structural support as ACEs increased (from 15.7% (0 ACEs) to 21.6% (4 or more ACE categories) (Table 3).
Overall reports of poor functional support (30.8%) were more common than reports of poor structural support (18.0%) (Table 4). Both poor structural support and poor functional support were associated with work inability. Almost 40% of those with poor functional support and 34.5% of those with poor structural support reported work inability compared to 21.7% and 20.8% of those reporting good structural and functional support respectively (Table 4).
Education and health were strongly associated with ACEs and work inability. The prevalence of ACEs decreased with increasing educational attainment going from 40% in those with less than a high school education to 19.7% of college graduates. The decline in work inability prevalence was even greater. Almost half of those with less than a high school education reported work inability compared to only 10% of college graduates (Table 4).
Almost half the sample reported 3 or fewer days of poor physical health and over half reported 3 or fewer days of poor mental health. The prevalence of work inability was lower in those reporting good physical health (11%) than those reporting good mental health (17%). The percentage reporting work inability was substantially higher in those reporting poor health over half the month. It was close to 50% for those with poor physical health and 44% of those reporting poor mental health (Table 4).
Logistic Regression Analyses
Logistic regression results indicated that those with an ACE score ≥ 4 had a higher adjusted odds ratio (95% confidence interval) of work inability [OR 1.74 (1.43–2.10), p <0.0001] compared to those with ACE score less than four, adjusting for age, sex and race. ACE burden was an independent predictor of work inability in the following regression models (all adjusted for age, sex and race): ACE and functional support, ACE and structural support, ACE and education, ACE and physical health and ACE and mental health.
The results of the sequential logistic regression analyses are shown in Table 5. ACEs appeared to have an independent effect on inability to work after adjusting for support (Model 2) and support and education (Model 3). After health was added (Model 4), the OR of four or more ACEs (compared to fewer than four) was closer to one and no longer statistically significant. A similar result was seen with functional support. Structural support, however, still had a substantial protective association with work inability (OR 0.50 (0.41–0.60)). Higher education was highly protective against work inability and the odds of work inability decreased as educational attainment increased. Physical and mental health were independent predictors of work inability (Table 5).
In the stratified logistic regression analysis by physical health status (adjusting for ACEs, demographics, education and mental health), functional support was protective in those with good physical health (OR 0.66 (0.46–0.95)), but not for those with intermediate (OR 1.15 (0.76–1.75)) or poor physical health (OR 0.92 (0.67–1.26)). Since there was no evidence of any protective effect at intermediate and poor physical health, we combined these levels for the logistic regression analysis using a common reference and the separate analysis evaluating an interaction term (Table 6). These analyses also supported that functional support was only protective for those with good physical health. In contrast, when performing logistic regression stratified by mental health, functional support was not protective at any level of mental health.
In contrast, while structural support was protective at all three levels of physical health, the odds ratios did not differ significantly between levels of physical health. The protective effects of structural support, however, were stronger for those with good mental health: OR (95% CI) 0.39 (0.30–0.51) for good; OR 0.58 (0.41–0.83) for intermediate and OR 0.63 (0.44–0.89) for poor mental health. This was confirmed again through the evaluation of an interaction term and logistic regression using a common reference (Table 6).
The prevalence of ACEs was strikingly high in our sample and strongly associated with decreased functional support, lower educational attainment and worse physical and mental health. As hypothesized, participants experiencing high levels of childhood adversity had a higher odds of work inability compared to those that had not. The effect, however, appeared to be more of a threshold effect as opposed to a graded effect and was strongest in the youngest age group. Functional and structural support were protective against work inability in adults with self-reported disability. After adding education and recent health to the model, there no longer was an independent effect of ACEs or functional support on work inability. Structural support, however, remained strongly protective. ACEs did appear to have an independent effect on work inability in the youngest age group. Functional support appeared to be protective in those with good physical health.
The finding of increased rates of work inability in those reporting ACEs is in line with previous research that has connected ACEs with worse worker performance , functional impairment in role performance (including work, social life, intimate relationships and household) [6, 35], lower levels of employment [9, 36, 37], and increased rates of employment disability . Childhood abuse has been associated with lower self-efficacy , an important predictor of returning to work after a long-term absence due to illness . It also has been associated with decreased executive functioning in middle adulthood , which is important for obtaining and maintaining employment. Furthermore, ACEs are associated with increased risk of experiencing physical or sexual violence in adulthood . Prior research about women with disabilities has found that women who have experienced prior physical or sexual abuse or intimate partner violence have higher unemployment rates [43, 44]. Finally, childhood adversity is strongly associated with increased engagement in health risk behaviors such as drug abuse [11, 20, 45, 46], which clearly can interfere with ability to work.
Our finding that the percentage of adults reporting good functional support decreased steadily as the number of self-reported ACE categories increased is also consistent with the literature. Previous research has suggested that childhood maltreatment, such as physical/sexual abuse and neglect, may negatively impact an adult’s ability to create and maintain healthy relationships . Further, childhood adversity can have negative effects on intimate relationships  and social network size .
Our hypotheses regarding the independent effect of ACEs and support on work inability, after inclusion of other important predictors, were only partially supported. This finding was not unexpected given the strong associations between childhood adversity and worse physical and mental health in adulthood [5, 10, 11, 20, 22, 49]. Childhood adversity is also associated with increased disorder-specific functional impairment  and disability . The effects of poor physical and mental health are intertwined and comorbid mental health conditions are an important contributor to role impairment in those with physical health conditions [45, 50, 51].
Our hypotheses regarding the relationship between support and inability to work were only partially supported. Only structural support had an independent effect after controlling for education and health. In contrast to our original hypothesis, functional support appeared to have a protective association with work inability only in those reporting good physical health but not in those reporting intermediate or poor physical health.
The survey relied on landlines only and primary cellphone users were not included in the sample, which may affect generalizability.
All measures were obtained at a single time-point and thus our findings represent associations. The cross-sectional nature also precluded us from testing formal theories of mediation (such as evaluating health as a mediator and looking at the direct and indirect effects of ACEs on work inability). Although evaluating for the presence of alternative effect pathways by controlling for an intermediate variable is a common epidemiological method , because of the correlations between of our major predictors, it is possible that the change in coefficient and loss of significance in Model 4 is due to multicollinearity as opposed to evidence of lack of an independent effect of ACEs (other than their effect on education and health). Also while structural support appeared to have a strong independent effect, it is possible this is due to residual uncontrolled confounding. While we treated support as a predictor of inability to work, is also possible that inability to work affects support. For our underlying predictor, ACEs, however, the possibility of reverse causation is less of an issue, since present day work inability cannot cause ACEs.
The study relied on retrospective self-reports of ACEs which have reasonable validity and tend to be relatively stable over time, but are prone to underreporting bias [47, 48, 53–57]. The number of questions used for each measure was limited, precluding a more in depth understanding of the answers. To address the possibility that those reporting poor functional support may not have sought or required support, as opposed to not having the support available, we included a measure of structural support .
The BRFSS definition of disability is broad  and information about the details and severity of the disability is not available. A more detailed measure would allow for more in-depth exploration of the association between specific ACEs and specific components of disability.
We do not know the reasons why people felt they were “unable to work.” However, the self-reported prevalence of work inability among respondents under age 65 without self-reported disability was 1% in the full BRFSS sample versus 25% in our disability sub-sample. Hence, it is likely that disability plays a large role in ability to work. Similarly, we may have underestimated the percentage of people whose disability prevents them from working by focusing solely on those who responded they are “unable to work.” Previous research has demonstrated that ACEs are associated with early disability-related retirement , but we could not assess for this possibility with these data.
This study helps expand the literature of risk and protective factors that may affect work inability in adults with disability. The impact of ACEs on multiple adverse adult outcomes including health risk behaviors, health, education, disability and work inability highlights the importance of interventions to prevent and/or mitigate the effects of ACEs as early as possible [4, 5, 9, 61]. Mitigation efforts should include research on ways to improve educational attainment in children, adolescents and young adults with known ACE exposures.
However, given high disability rates in the United States, an understanding of the factors that affect the ability to work of adults with disability is critical to designing effective interventions. Enabling greater numbers of adults with disability to work not only has clear societal economic advantages, but also provides individual level benefits such as increased self-esteem, greater independence and social inclusion.
Our study has implications for practice and future research. It is important to realize that a greater proportion of adults with a disability experienced ACEs and over twice as many have a high ACE burden (four or more ACEs) compared to adults without a disability . In our study, ACEs appeared to be more strongly associated with work inability in the youngest age group which also had the highest prevalence of ACEs. Effectively addressing the effects of ACEs on work ability in this group may be particularly beneficial given that this age group could have as many as 30–40 productive work years ahead of them.
Even if the effects of ACEs on work inability is more indirect such as through the effect of ACEs on health, it may still be important to address the impact of ACEs in these individuals in order to improve health. There is evidence that one needs to take ACE exposure into account to effectively address health risk behaviors and certain health problems in adults who have experienced ACEs and that ACE status may have implications for choice of treatment [62–65].
Assessing for ACEs and their effect on current health and support can guide avenues for treatment (e.g. trauma therapy, social skill development). It is reasonable to suggest that a clinician should be aware of these factors and work towards preventative measures to decrease future risk of work inability. Preventive measures can include exploring a patient’s perception of available social support, assistance with identifying potential sources of support (e.g. family, friends, available support groups), and help for the patient to increase his or her social ability and currency. The strong protective effect of education suggests that policies supporting vocational rehabilitation, and research on the most effective educational interventions for those with disability are vital .
Continued research that refines our understanding of risk and protective factors associated with the ability of adults with a disability to work is needed . Longitudinal studies in adults with disability that evaluate the effect of ACEs on outcomes such as participation are particularly important. As the field advances, specific knowledge related to the relationship between ACEs, mental and physical health, external factors, and genetic predispositions will facilitate the development of targeted treatments to enable all adults with disability reach their full employment and participation potential.
This paper originates from the ‘Paper in a Day’ workshop at the annual meeting of the International Society for Traumatic Stress Studies in Philadelphia, Pennsylvania, USA, November 2013. With thanks to Khic-Houy Prang for her valuable comments on the draft manuscript.
Conceived and designed the experiments: SMS-FR JGE LZ JS BSL CL EA. Performed the experiments: SMS-FR. Analyzed the data: SMS-FR. Contributed reagents/materials/analysis tools: SMS-FR. Wrote the paper: SMS-FR JGE LZ JS BSL CL. Writing - major manuscript revisions: SMS-FR. Writing - reviewing and editing: SMS-FR JGE LZ JS BSL CL EA. Project administration: SMS-FR JGE. Organized the Paper in A Day Workshop: EA.
- 1. Brault M. Americans With Disability: 2010. Washington, DC: US Census Bureau, 2012.
- 2. Carroll D, Courtney-Long E, Stevens A, Sloan ML, Lullo C, Visser S, et al. Vital Signs: Disability and Physical activity—United States, 2009–2012. MMWR. 2014;63:1–7.
- 3. Barnow BS. Employment rate of people with disabilities. Monthly Labor Review. 2008;131.
- 4. Schüssler-Fiorenza Rose SM. Adverse Childhood Experiences, Disability and Health-Risk Behaviors. Population Health Matters (Formerly Health Policy Newsletter) [Internet]. 2013 October 24, 2014; 26(3). Available from: http://jdc.jefferson.edu/hpn/vol26/iss3/12.
- 5. Felitti VJ, Anda RF, Nordenberg D, Williamson DF, Spitz AM, Edwards V, et al. 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. 1998;14(4):245–58. Epub 1998/06/23. pmid:9635069.
- 6. Allwood MA, Widom CS. Child Abuse and Neglect, Developmental Role Attainment, and Adult Arrests. Journal of Research in Crime and Delinquency. 2013;50(4):551–78.
- 7. Anda R, Felitti V, Fleisher V, Edwards V, Whitfield C, Dube S, et al. Childhood abuse, household dysfunction and indicators of impaired worker performance in adulthood. The Permanente Journal 2004;8(1):30–8. pmid:26704603
- 8. Hadi F, Lai BS, Llabre MM. Life outcomes influenced by war-related experiences during the Gulf crisis. Anxiety, stress, and coping. 2014;27(2):156–75. Epub 2013/09/06. pmid:24003829; PubMed Central PMCID: PMC3877742.
- 9. Currie J, Widom CS. Long-term consequences of child abuse and neglect on adult economic well-being. Child Maltreatment. 2010;15(2):111–20. pmid:20425881
- 10. Gilbert R, Widom CS, Browne K, Fergusson D, Webb E, Janson S. Burden and consequences of child maltreatment in high-income countries. Lancet. 2009;373(9657):68–81. pmid:19056114.
- 11. Anda RF, Felitti VJ, Bremner JD, Walker JD, Whitfield C, Perry BD, et al. 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. 2006;256(3):174–86. Epub 2005/11/29. pmid:16311898.
- 12. Schüssler-Fiorenza Rose SM, Xie D, Stineman M. Adverse Childhood Experiences and Disability in U.S. Adults. PM&R. 2014;6(8):670–80.
- 13. World Health Organization. Towards a Common Language for Functioning, Disability and Health ICF2002 May 26, 2014. Available from: http://www.who.int/entity/classifications/icf/training/icfbeginnersguide.pdf.
- 14. World Health Organization. International Classification of Functioning, Disability and Health: ICF. Geneva, Switzerland: World Health Organization; 2001. p. 299 p.
- 15. Moak ZB, Agrawal A. The association between perceived interpersonal social support and physical and mental health: results from the National Epidemiological Survey on Alcohol and Related Conditions. Journal of public health. 2010;32(2):191–201. Epub 2009/10/30. pmid:19864495; PubMed Central PMCID: PMC2875984.
- 16. Case RB, Moss AJ, Case N, McDermott M, Eberly S. Living alone after myocardial infarction. Impact on prognosis. JAMA: the journal of the American Medical Association. 1992;267(4):515–9. Epub 1992/01/22. pmid:1729573.
- 17. Eaker ED, Sullivan LM, Kelly-Hayes M, D'Agostino RB Sr, Benjamin EJ. Marital status, marital strain, and risk of coronary heart disease or total mortality: the Framingham Offspring Study. Psychosomatic medicine. 2007;69(6):509–13. Epub 2007/07/20. pmid:17634565.
- 18. Avlund K, Lund R, Holstein BE, Due P, Sakari-Rantala R, Heikkinen RL. The impact of structural and functional characteristics of social relations as determinants of functional decline. The journals of gerontology Series B, Psychological sciences and social sciences. 2004;59(1):S44–51. Epub 2004/01/15. pmid:14722343.
- 19. Mersky JP, Topitzes J. Comparing early adult outcomes of maltreated and non-maltreated children: A prospective longitudinal investigation. Children and Youth Services Review. 2010;32(8):1086–96. http://dx.doi.org/10.1016/j.childyouth.2009.10.018.
- 20. Norman RE, Byambaa M, De R, Butchart A, Scott J, Vos T. The long-term health consequences of child physical abuse, emotional abuse, and neglect: a systematic review and meta-analysis. PLoS medicine. 2012;9(11):e1001349. pmid:23209385; PubMed Central PMCID: PMC3507962.
- 21. Anda RF, Butchart A, Felitti VJ, Brown DW. Building a framework for global surveillance of the public health implications of adverse childhood experiences. American journal of preventive medicine. 2010;39(1):93–8. Epub 2010/06/16. pmid:20547282.
- 22. Kessler RC, Davis CG, Kendler KS. Childhood adversity and adult psychiatric disorder in the US National Comorbidity Survey. Psychological medicine. 1997;27(5):1101–19. Epub 1997/09/23. pmid:9300515.
- 23. Abraido-Lanza AF, White K, Armbrister AN, Link BG. Health status, activity limitations, and disability in work and housework among Latinos and non-Latinos with arthritis: an analysis of national data. Arthritis and rheumatism. 2006;55(3):442–50. Epub 2006/06/02. pmid:16739212; PubMed Central PMCID: PMC3606810.
- 24. Clarke P, Latham K. Life course health and socioeconomic profiles of Americans aging with disability. Disability and health journal. 2014;7(1 Suppl):S15–23. Epub 2014/01/25. pmid:24456680; PubMed Central PMCID: PMC3901952.
- 25. Okoro C, Strine T, McGuire L, Balluz L, Mokdad A. Employment status and frequent mental distress among adults with disabilities. Occup Med (Lond). 2007;57(3):217–20. Epub 2007/01/19. pmid:17229716.
- 26. Centers for Disease Control and Prevention (CDC). Behavioral Risk Factor Surveillance System Survey Data 2009–2010. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention.
- 27. Centers for Disease Control and Prevention. Behavioral Risk Factor Surveillance System 2009 Summary Data Quality Report 2009 [Nov 2, 2014]. Available from: ftp://ftp.cdc.gov/pub/Data/Brfss/2009_Summary_Data_Quality_Report.pdf.
- 28. Centers for Disease Control and Prevention (CDC). Overview: BRFSS 2009Nov 2, 2014. Available from: http://www.cdc.gov/brfss/annual_data/2009/overview_09.rtf.
- 29. Ford ES, Anda RF, Edwards VJ, Perry GS, Zhao G, Li C, et al. Adverse childhood experiences and smoking status in five states. Preventive medicine. 2011;53(3):188–93. Epub 2011/07/06. pmid:21726575.
- 30. Centers for Disease Control and Prevention. Behavioral Risk Factor Surveillance System 2010 Summary Data Quality Report 2010 [Nov 2, 2014]. Available from: ftp://ftp.cdc.gov/pub/Data/Brfss/2010_Summary_Data_Quality_Report.pdf.
- 31. Behavioral Risk Factor Surveillance System Survey Questionnaire [Internet]. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. 2009–2010.
- 32. Vandenbroucke JP, von Elm E, Altman DG, Gøtzsche PC, Mulrow CD, Pocock SJ, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): Explanation and Elaboration. PLoS medicine. 2007;4(10):e297. PMC2020496. pmid:17941715
- 33. Weinberg CR. Interaction and Exposure Modification: Are We Asking the Right Questions? American Journal of Epidemiology. 2012;175(7):602–5. PMC3324436. pmid:22306562
- 34. Wacholder S, Han SS, Weinberg CR. Inference From a Multiplicative Model of Joint Genetic Effects or Ovarian Cancer Risk. JNCI Journal of the National Cancer Institute. 2011;103(2):82–3. PMC3022621. pmid:21169538
- 35. McLaughlin KA, Green JG, Gruber MJ, Sampson NA, Zaslavsky AM, Kessler RC. Childhood adversities and adult psychopathology in the National Comorbidity Survey Replication (NCS-R) III: associations with functional impairment related to DSM-IV disorders. Psychological medicine. 2010;40(5):847–59. Epub 2009/09/08. pmid:19732483; PubMed Central PMCID: PMC2847368.
- 36. Tam TW, Zlotnick C, Robertson MJ. Longitudinal perspective: adverse childhood events, substance use, and labor force participation among homeless adults. The American journal of drug and alcohol abuse. 2003;29(4):829–46. Epub 2004/01/10. pmid:14713142.
- 37. Lund T, Andersen JH, Winding TN, Biering K, Labriola M. Negative life events in childhood as risk indicators of labour market participation in young adulthood: a prospective birth cohort study. PloS one. 2013;8(9):e75860. Epub 2013/09/17. pmid:24040421; PubMed Central PMCID: PMC3770657.
- 38. Sansone RA, Dakroub H, Pole M, Butler M. Childhood trauma and employment disability. International journal of psychiatry in medicine. 2005;35(4):395–404. pmid:16673839.
- 39. Sachs-Ericsson N, Medley AN, Kendall-Tackett K, Taylor J. Childhood Abuse and Current Health Problems among Older Adults: The Mediating Role of Self-Efficacy. Psychology of violence. 2011;1(2):106–20. pmid:21922052; PubMed Central PMCID: PMC3171165.
- 40. Brouwer S, Reneman MF, Bültmann U, van der Klink JJ, Groothoff JW. A prospective study of return to work across health conditions: perceived work attitude, self-efficacy and perceived social support. Journal of occupational rehabilitation. 2010;20(1):104–12. pmid:19894106
- 41. Nikulina V, Widom CS. Child maltreatment and executive functioning in middle adulthood: a prospective examination. Neuropsychology. 2013;27(4):417–27. pmid:23876115; PubMed Central PMCID: PMC3855039.
- 42. Widom CS, Czaja SJ, Dutton MA. Childhood victimization and lifetime revictimization. Child abuse & neglect. 2008;32(8):785–96. Epub 2008/09/02. pmid:18760474; PubMed Central PMCID: PMC2572709.
- 43. Smith DL. Disability, Gender and Intimate Partner Violence: Relationships from the Behavioral Risk Factor Surveillance System. Sexuality and Disability. 2007;26(1):15–28.
- 44. Smith DL, Strauser DR. Examining the impact of physical and sexual abuse on the employment of women with disabilities in the United States: An exploratory analysis. Disability & Rehabilitation. 2008;30(14):1039–46.
- 45. Widom CS, Marmorstein NR, White HR. Childhood victimization and illicit drug use in middle adulthood. Psychology of Addictive Behaviors. 2006;20(4):394–403. WOS:000242952600004. pmid:17176174
- 46. Anda RF, Brown DW, Dube SR, Bremner JD, Felitti VJ, Giles WH. Adverse childhood experiences and chronic obstructive pulmonary disease in adults. American journal of preventive medicine. 2008;34(5):396–403. Epub 2008/04/15. pmid:18407006.
- 47. Colman RA, Widom CS. Childhood abuse and neglect and adult intimate relationships: a prospective study. Child abuse & neglect. 2004;28(11):1133–51. Epub 2004/11/30. pmid:15567020.
- 48. Ford E, Clark C, Stansfeld SA. The influence of childhood adversity on social relations and mental health at mid-life. Journal of affective disorders. 2011;133(1–2):320–7. Epub 2011/04/13. pmid:21482436.
- 49. Green JG, McLaughlin KA, Berglund PA, Gruber MJ, Sampson NA, Zaslavsky AM, et al. Childhood adversities and adult psychiatric disorders in the national comorbidity survey replication I: associations with first onset of DSM-IV disorders. Archives of general psychiatry. 2010;67(2):113–23. Epub 2010/02/04. pmid:20124111; PubMed Central PMCID: PMC2822662.
- 50. Kessler RC, Ormel J, Demler O, Stang PE. Comorbid mental disorders account for the role impairment of commonly occurring chronic physical disorders: results from the National Comorbidity Survey. Journal of occupational and environmental medicine / American College of Occupational and Environmental Medicine. 2003;45(12):1257–66. Epub 2003/12/11. pmid:14665811.
- 51. Druss BG, Hwang I, Petukhova M, Sampson NA, Wang PS, Kessler RC. Impairment in role functioning in mental and chronic medical disorders in the United States: results from the National Comorbidity Survey Replication. Molecular psychiatry. 2009;14(7):728–37. Epub 2008/02/20. pmid:18283278; PubMed Central PMCID: PMC2700857.
- 52. Szklo M, Nieto FJ. Epidemiology Beyond the Basics 1st ed. Sudbury, MA: Jones & Bartlett Learning 2003 2004.
- 53. Dube SR, Williamson DF, Thompson T, Felitti VJ, Anda RF. Assessing the reliability of retrospective reports of adverse childhood experiences among adult HMO members attending a primary care clinic. Child abuse & neglect. 2004;28(7):729–37. Epub 2004/07/21. pmid:15261468.
- 54. Fergusson DM, Horwood LJ, Woodward LJ. The stability of child abuse reports: a longitudinal study of the reporting behaviour of young adults. Psychological medicine. 2000;30(03):529–44. null.
- 55. Hardt J, Rutter M. Validity of adult retrospective reports of adverse childhood experiences: review of the evidence. Journal of child psychology and psychiatry, and allied disciplines. 2004;45(2):260–73. Epub 2004/02/26. pmid:14982240.
- 56. Widom CS, Morris S. Accuracy of adult recollections of childhood victimization .2. Childhood sexual abuse. Psychol Assess. 1997;9(1):34–46. WOS:A1997WN06500005.
- 57. Widom CS, Shepard RL. Accuracy of adult recollections of childhood victimization .1. Childhood physical abuse. Psychol Assess. 1996;8(4):412–21. WOS:A1996WE52700014.
- 58. Holt-Lunstad J, Smith TB, Layton JB. Social relationships and mortality risk: a meta-analytic review. PLoS medicine. 2010;7(7):e1000316. Epub 2010/07/30. pmid:20668659; PubMed Central PMCID: PMC2910600.
- 59. Erickson WA, Dumoulin-Smith A. A Guide to Disability Statistics from the Behavioral Risk Factors Surveillance System. Ithaca: Cornell University; 2009.
- 60. Harkonmaki K, Korkeila K, Vahtera J, Kivimaki M, Suominen S, Sillanmaki L, et al. Childhood adversities as a predictor of disability retirement. Journal of epidemiology and community health. 2007;61(6):479–84. Epub 2007/05/15. pmid:17496255; PubMed Central PMCID: PMC2465717.
- 61. Garner AS, Shonkoff JP, Siegel BS, Dobbins MI, Earls MF, Garner AS, et al. Early Childhood Adversity, Toxic Stress, and the Role of the Pediatrician: Translating Developmental Science Into Lifelong Health. Pediatrics. 2012;129(1):e224–e31. pmid:22201148
- 62. Felitti VJ, Jakstis K, Pepper V, Ray A. Obesity: problem, solution, or both? Perm J. 2010;14(1):24–30. pmid:20740128; PubMed Central PMCID: PMCPMC2912711.
- 63. Teicher MH, Samson JA. Childhood maltreatment and psychopathology: A case for ecophenotypic variants as clinically and neurobiologically distinct subtypes. Am J Psychiatry. 2013;170(10):1114–33. pmid:23982148; PubMed Central PMCID: PMCPMC3928064.
- 64. Nanni V, Uher R, Danese A. Childhood maltreatment predicts unfavorable course of illness and treatment outcome in depression: a meta-analysis. Am J Psychiatry. 2012;169(2):141–51. pmid:22420036.
- 65. Nemeroff CB, Heim CM, Thase ME, Klein DN, Rush AJ, Schatzberg AF, et al. Differential responses to psychotherapy versus pharmacotherapy in patients with chronic forms of major depression and childhood trauma. Proc Natl Acad Sci U S A. 2003;100(24):14293–6. pmid:14615578; PubMed Central PMCID: PMCPMC283585.
- 66. Ottomanelli L, Barnett SD, Goetz LL. Effectiveness of supported employment for veterans with spinal cord injury: 2-year results. Archives of physical medicine and rehabilitation. 2014;95(4):784–90. Epub 2013/12/10. pmid:24316325.
- 67. Cornelius L, Van der Klink J, Groothoff J, Brouwer S. Prognostic factors of long term disability due to mental disorders: a systematic review. Journal of occupational rehabilitation. 2011;21(2):259–74. pmid:21057974