An Enduring Health Risk of Childhood Adversity: Earlier, More Severe, and Longer Lasting Work Disability in Adult Life

An Enduring Health Risk of Childhood Adversity: Earlier, More Severe, and Longer Lasting Work... Abstract Objectives Childhood adversity has been linked with adult health problems. We hypothesized that childhood adversity would also be associated with work limitations due to physical or nervous health problems, known as work disability. Method With data from the Panel Study of Income Dynamics (PSID) (1968–2013; n = 6,045; 82,374 transitions; 129,107 person-years) and the 2014 PSID Childhood Retrospective Circumstances Study, we estimated work disability transition probabilities with multinomial logistic Markov models. Four or more adversities defined a high level. Microsimulations quantified adult work disability patterns for African American and non-Hispanic white women and men, accounting for age, education, race, sex, diabetes, heart disease, obesity, and sedentary behavior. Results Childhood adversity was significantly associated with work disability. Of African American women with high adversity, 10.2% had moderate work disability at age 30 versus 4.1% with no reported adversities; comparable results for severe work disability were 5.6% versus 1.9% (both p < .01). Comparable results for whites were 11.3% versus 4.7%, and 3.5% versus 1.1% (p < .01). The association of childhood adversity with work disability remained significant after adjusting for diabetes, heart disease, obesity, and sedentary behavior (p < .05). Conclusions Childhood adversity may increase work disability throughout adult life. Cumulative advantage/disadvantage, Early origins of health, Epidemiology, Life course analysis Adverse circumstances during childhood (childhood adversity) may be associated with educational attainment and occupational choice, health-related behaviors, adult physical and mental health, adult socioeconomic status, and relationships, all of which have been linked to functional status and disability (e.g., Bowen & González, 2010; Haas, Krueger, & Rohlfsen, 2012; Laditka & Laditka, 2017; Luo & Waite, 2005; Montez & Hayward, 2014; Smith et al., 2016; Turner, Thomas, & Brown, 2016). Associations of childhood adversity with educational attainment are especially important because people with less education are more likely to work in jobs with greater occupational health risks such as machine operations, construction, repair or moving operations, helping professions, and cleaning (Hatch, 2005). Childhood adversity may be linked with personal control and efficacy, affecting workplace communications, stress, perceptions about work, and confidence to request accommodations that enable work for people who develop functional impairments (Hatch, 2005; Silverstein, 2008). Yet, few studies have examined associations of childhood adversity with the ability to work. We address this knowledge gap. The Concept of Work Disability Disability covered by Social Security is a “medical condition [that] must significantly limit your ability to do basic work activities—such as lifting, standing, walking, sitting, and remembering—for at least 12 months” (Social Security Administration, 2017). However, work disability is often defined more broadly, by self-reports of physical or nervous conditions that limit work or make it impossible (Burkhauser, Daly, Houtenville, & Nargis, 2002; Jette & Badley, 2000; Rank & Hirschl, 2014). We adopt the conceptual framework of work disability by Jette and Badley (2000), Nagi (1991) and the World Health Organization (WHO, 2001), which emphasizes the role of social arrangements in disability. Factors contributing to work disability include job characteristics that worsen health, and social contexts that impede work for people with functional limitations (Jette & Badley, 2000; Nagi, 1991; Pransky et al., 2016; Silverstein, 2008). Work disability measures a practical outcome of such factors. In a recent study, nearly 55% of people in the United States reported a work disability at least once between ages 25 and 60, nearly one-quarter a severe work disability (Rank & Hirschl, 2014). Another recent study found that a majority of Americans have a work disability between ages 20 and 65 (Laditka & Laditka, 2018). Even when a person recovers from work disability, career advancement may be limited permanently (Breslin et al., 2007). Shuey and Willson (2017) found that people with work disabilities have double the risk of others of experiencing poverty. Workforce aging makes work disability increasingly important: one-quarter of the United States’ labor force will be age 55 or older by 2020 (Toossi, 2012). Of working age adults in the United States, more than 40% have at least one chronic health condition, and 20% two or more, increasing the risk of work disability (Pransky et al., 2016; Silverstein, 2008; Ward, 2015). Childhood Adversity—A Life Course Perspective Researchers use four theoretical models to characterize the association of childhood adversity with adult health. The accumulation model emphasizes risk from persistent, recurring, or multiple childhood adversities (e.g., Cohen, Janicki-Deverts, Chen, & Matthews, 2010; O’Rand & Hamil-Luker, 2005). A second model focuses on pathways, stressing the relationship of childhood adversity with biological risks and health behaviors that increase adult chronic disease and functional impairment (e.g., Pudrovska & Anikputa, 2014). A third model focuses on timing, emphasizing risks in limited periods of susceptibility, especially in utero and in early childhood (e.g., Cohen et al., 2010). Finally, the change model, also called the social mobility model, posits that adult socioeconomic status can modify effects of childhood adversity (e.g., Ben-Shlomo & Kuh, 2002). The mechanisms of these models may combine to influence adult health, so it may not be possible to identify their separate effects (Pudrovska & Anikputa, 2014). With regard to race, African American children are more likely to live in poor families and neighborhoods than whites (e.g., Smith et al., 2016). They have less educational opportunity and are more likely to experience violence (e.g., Haas et al., 2012). With regard to gender, the economic benefits of education may be lower for women than for men (Luo & Waite, 2005); therefore, if education moderates lifetime effects of childhood adversity on health then women may be less likely to benefit from that moderating effect. However, little research has examined variation in the association of childhood adversity with adult health by gender or race. Researchers have either not found significant differences (Luo & Waite, 2005; Smith et al., 2016) or that differences did not persist in adjusted results (Haas et al., 2012). One recent study examined associations of childhood adversity with relationship strain by race, and how that strain may affect adult health (Umberson, Williams, Thomas, Liu, & Thomeer, 2014). The researchers concluded that relationship strain associated with childhood adversity had a greater negative effect on adult health for African American men than for white men. Study Contributions and Hypotheses We examined the association of childhood adversity with work disability, using longitudinal data that were nationally representative of African Americans and non-Hispanic whites (hereafter whites) ages 20 and over living in the community. Consistent with the expectation that risks to health can accumulate in childhood, our first hypothesis was that people reporting high levels of childhood adversity would have significantly more work disability throughout adult life than people reporting none of the adversities. To contribute to knowledge about the pathway model, we examined the association of childhood adversity with adult diabetes, heart disease, obesity, and sedentary behavior, all of which are established causes of functional impairment (Andrade, 2010; Laditka & Laditka, 2015; Laditka & Laditka, 2017). We also examined how those factors may mediate the association of adversity with work disability. Although health behaviors are likely to be powerful links between early life adversity and adult health, other pathways may include education, biological factors, and mental health. Adversity may also affect adult health directly, through a pathway that is not mediated by such factors, or through other mediators that are not currently known. Consistent with these possibilities, our second hypothesis was that childhood adversity would be associated with work disability even after controlling for adult diabetes, heart disease, obesity, and sedentary behavior. We also examined educational and biological pathways. It is well established that African Americans have more functional impairment than whites (e.g., Laditka & Laditka, 2009; Montez & Hayward, 2014), more morbidity (e.g., Laditka & Laditka, 2009), and greater exposure to social and economic disparities throughout life (e.g., Haas et al., 2012). Little research has specifically studied how the association of childhood adversity with adult health may vary with race. However, given recent evidence of such variation (Umberson et al., 2014), we hypothesized that childhood adversity would be associated with more work disability among African Americans than among whites. Methods Data We used data from the Panel Study of Income Dynamics (PSID), the longest running household panel study in the world, and the 2014 PSID Childhood Retrospective Circumstances Study (CRCS). The PSID surveyed participants every year from 1968 through 1997, and thereafter every 2 years, typically with response rates from 96% to 98% (Schoeni, Stafford, McGonagle, & Andreski, 2013). The CRCS attempted to interview all PSID household heads and their spouses or partners, who represent adults living in the community in the United States. We included CRCS respondents who also provided information about work disability in at least two PSID survey waves. We excluded participants other than African Americans and whites due to their small numbers. Dependent Variable—Measuring Work Disability Beginning in 1968, the PSID asked household heads, “Do you have any physical or nervous condition that limits the type of work or the amount of work you can do?” The PSID then asked those who responded “yes,” “Does this condition keep you from doing some types of work?” Available responses were, “yes,” “no,” or “can do nothing.” We considered participants to have a severe work disability if they could “do nothing.” Those who responded “yes” were asked if the condition limited work “a lot,” “somewhat,” “just a little,” or “not at all.” We also considered those who said “a lot” to have a severe work disability, and those who said “somewhat” to have a moderate work disability. Related questions in 1969, 1970, and 1971 did not measure severity; we excluded those years. In some years before 1980 the PSID asked only new participants about work disability; we used their responses but did not use responses that were simply repeated from previous waves. The PSID asked about work disability for spouses and partners in 1976, 1978, and in all waves beginning with 1981; we used all of that data. Childhood Adversities and Other Measures We measured adversities using an index of nine childhood measures of: health status (e.g., Turner et al., 2016), family characteristics (e.g., Haas, 2007), socioeconomic status (e.g., Montez & Hayward, 2014), and victimization (e.g., Wolke, Copeland, Angold, & Costello, 2013). Adverse health was represented by self-reports of fair or poor childhood health, compared with good, very good, or excellent. Family adversities were: parents divorced when the participant was less than age 17, or the participant was raised by a single parent. Socioeconomic adversities were: one or both parents with education no more than grade 8; a parent experienced unemployment; financial “struggle”; neighborhood unsafe at night; and neighbors either not “close knit” or not reliable if help was needed. Participant reports of having been bullied “a lot” or “sometimes” in or out of school indicated victimization. Preliminary analyses associated each of these circumstances with more work disability throughout life. Few participants reported more than four adversities; we used a single categorical variable representing 0, 1, 2, 3, or 4 or more. We refer to four or more as high childhood adversity. We also examined the risks associated with childhood adversity of having diabetes, heart disease, obesity, or sedentary behavior at ages 40 and 65. The PSID asked, “Has a doctor or other health professional ever told you that you had diabetes or high blood sugar?” An analogous question asked about heart disease. Beginning in 1999, participants were asked how long they had the disease, or in some waves the year or age when diagnosed. We used this information to identify diabetes and heart disease status in previous years. The PSID asked participants their height and weight in 1986 and in all survey waves beginning with 1999; we used body mass index guidelines from the Centers for Disease Control and Prevention to measure obesity. The PSID asked participants how often they had light, moderate, or vigorous physical activity, in 1986 and in all waves beginning with 1999. We considered those who responded “never” to all three questions to be sedentary. We used the information about obesity and sedentary behavior that most closely corresponded to the year that began each measured work disability transition. We controlled for adult health and socioeconomic status with a three-level measure of education: less than high school, high school graduation (including the General Educational Development credential), or postsecondary education. We controlled for age in years, age-squared, and sex. A dummy variable indicated whether observations represented African Americans or whites. Interaction terms provided separate probabilities for each combination of childhood adversity, sex, and race. Analytical Approach We used t tests and standard logistic regression to compare PSID participants who completed the CRCS and those who did not. Our principal methods were a multinomial logistic Markov chain regression model, estimated by maximum likelihood, and dynamic microsimulation (e.g., Andrade, 2010; Laditka, 1998; Laditka & Laditka, 2009, 2014b, 2015; Laditka & Laditka, 2014b, 2016a, 2016b, 2018; Laditka & Wolf, 1998; Yong, & Saito, 2012). The model estimated work disability transition probabilities specific to each age beginning at 20, adjusting the probability of each transition type for the time between interviews and accommodating any pattern or number of unrecorded work disability transitions between interviews (e.g., Laditka & Laditka, 2018; Laditka & Wolf, 1998). We first estimated a separate model for each childhood adversity, adjusted for age, sex, and race. In the principal model, the probability of a given transition at a given age was conditional on the value of the adversity index, current work disability status, age, education, race, and sex. Another model estimated transition probabilities beginning at age 40, when adult chronic diseases become prevalent, adding adjustments for diabetes, heart disease, sedentary behavior, and obesity. To provide adequate statistical power, this second model did not stratify by race. With the probabilities estimated by the Markov models, we used microsimulation to create large simulated populations in which each individual had a complete annual history of work disability: none, moderate, or severe. Details of the method are published (Laditka, 1998; Laditka & Laditka, 2009, 2014a; Laditka & Wolf, 1998). All participants responded to the CRCS, so we could not estimate mortality risks. We simulated each individual life through death using mortality data from the National Center for Health Statistics. In a microsimulation for each population, such as African American women with high childhood adversity, we created 100,000 lives. For both moderate and severe work disability, we calculated the population prevalence at each age, the average onset age of the first spell, the number of spells for the average individual, and the percentage of life at ages 30–65 with work disability. Controlling for education may adjust for a pathway linking childhood adversity with adult health. We therefore also estimated the model without a control for education. To contribute to knowledge about the pathway model, we first examined associations of childhood adversity with diabetes, heart disease, obesity, or sedentary behavior at ages 40 and 65. For each outcome, we estimated an unadjusted standard logistic model, then adjusted for education. Given the expectation that poor childhood health would be associated with those outcomes, and to control for a biological pathway, we also estimated models representing only participants reporting good to excellent childhood health; all other models included childhood health in the adversity index. We then conducted microsimulations in which all individuals separately had diabetes, heart disease, sedentary behavior, or obesity beginning at age 40, or none of those conditions, or all of them, focused on results for childhood adversity. We analyzed the consistency across survey waves of reports of the ages when participants said they were diagnosed with diabetes or heart disease. When a participant responded inconsistently, we used the most common response. When a household head and the head’s spouse or partner both reported onset ages for the spouse or partner, which can occur when the head provides this information and then dies or is divorced, we used the responses from the spouse or partner for the analytic data for that individual. We examined the reliability of the onset age responses by calculating the intra-class correlation coefficient (ICC) using the INTRACC SAS macro (Hamer, 2007). An ICC ≥ 0.70 indicates adequate reliability (Special Advisory Committee of the Medical Outcomes Trust, 2002). As the PSID follows families for decades, it is useful to examine whether characteristics shared within families might be associated with childhood adversity and work disability, and therefore might affect the results. To examine this issue, we estimated a multinomial logistic hazard model predicting the three levels of work disability with both fixed-effects and random-effects (Allison, 2005). Fixed effects controlled for all characteristics shared within families; random effects did not. We tested whether these effects differed, which would indicate that the association of childhood adversity with work disability varied systematically among families. We estimated variation in the results with bootstrapping, repeating each microsimulation 1,000 times, each with parameters randomly selected from their 95% confidence intervals (CIs). The point estimates for the outcomes were the means; the CIs ranged from the 2.5 to the 97.5 percentiles. We created the software for estimating the Markov models and conducting the microsimulations using SAS IML (Cary, NC). For all other analyses, we used SAS 9.4 (Cary, NC). The Institutional Review Board (IRB) at the University of North Carolina at Charlotte determined that this research did not require IRB review. Results Sample Characteristics The CRCS response rate was 67%; 6,045 participants fulfilled the inclusion criteria. Participants and nonparticipants did not differ significantly in age, sex, race, or reports of work disability. Participants had 82,374 recorded work disability status transitions through 129,107 person-years, each participant averaging 13.6 (SD 10.1) work disability measurements (data not shown). The mean baseline age was 29.5 years (SD 8.2 years, not shown), where the baseline was the first recorded work disability status for each individual; weighted for national representativeness and accounting for the survey design the mean baseline age was 32.8 years (CI 32.2–33.4). The mean age when participants responded to the CRCS was 48.7 (SD 15.0; 50.5 years weighted, CI 49.9–51.1). Women were 57.8% of the sample (54.1 weighted, CI 50.7–57.5). The PSID oversampled African Americans, 29.9% of the sample (9.1 weighted, CI 6.7–11.6). As for childhood adversities, 30.4% reported none, 28.9% one, 21.2% two, 11.9% three, and 7.7% four or more. Table 1 describes the sample. White women and men had the smallest percentages with high adversity, 6.1% and 5.7%, respectively; comparable results were 13.2% and 9.3% for African Americans (all p < .0001). Not shown, obesity and sedentary behavior were measured in the same year as work disability in 60.1% of the analytic observations; the average period between the obesity and sedentary measurements and the work disability measurement was 1.8 years (SD 1.0). Also not shown, the sample used to test our second hypothesis (ages 40 and over with complete information on diabetes, heart disease, obesity, and sedentary behavior) represented 4,135 participants with 28,084 functional status transitions. Other characteristics of the second sample were similar to those reported in Table 1. Table 1. Associations of Childhood Adversity With Work Disability, Characteristics of the Sample, by Sex and Racea   Women  Men    African American  White  African American  White  Sample size, nb  1,175  2,320  634  1,916  Work disability status transitions, nb  14,022  31,072  7,304  29,976  Person-years, nb  22,867  50,467  11,709  44,064  Age at baseline, mean (CI)c  31.2 (30.2–32.3)  32.2 (31.6–32.8)  35.9 (33.9–38.0)  33.4 (32.5–34.4)  Education years, mean (CI)c  12.8 (12.0–13.6)  13.6 (13.3–13.9)  12.7 (12.3–13.1)  13.5 (13.2–13.8)  Adversity index (0–4), mean (CI)c  1.7 (1.5–1.9)  1.3 (1.2–1.4)  1.6 (1.3–1.8)  1.2 (1.1–1.3)   Index=0, %d  20.0  33.2  22.9  35.1   Index=1, %d  27.1  29.1  27.9  30.0   Index=2, %d  22.5  21.0  23.8  19.7   Index=3, %d  17.3  10.5  16.1  8.9   Index=4+, %d  13.2  6.1  9.3  5.7    Women  Men    African American  White  African American  White  Sample size, nb  1,175  2,320  634  1,916  Work disability status transitions, nb  14,022  31,072  7,304  29,976  Person-years, nb  22,867  50,467  11,709  44,064  Age at baseline, mean (CI)c  31.2 (30.2–32.3)  32.2 (31.6–32.8)  35.9 (33.9–38.0)  33.4 (32.5–34.4)  Education years, mean (CI)c  12.8 (12.0–13.6)  13.6 (13.3–13.9)  12.7 (12.3–13.1)  13.5 (13.2–13.8)  Adversity index (0–4), mean (CI)c  1.7 (1.5–1.9)  1.3 (1.2–1.4)  1.6 (1.3–1.8)  1.2 (1.1–1.3)   Index=0, %d  20.0  33.2  22.9  35.1   Index=1, %d  27.1  29.1  27.9  30.0   Index=2, %d  22.5  21.0  23.8  19.7   Index=3, %d  17.3  10.5  16.1  8.9   Index=4+, %d  13.2  6.1  9.3  5.7  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study, n = 6,045. bUnweighted sample characteristics. cResults weighted to be nationally representative and accounting for the survey design; CI = 95% confidence interval. dIndex level comparisons across race, p < .0001 for both women and men. View Large Table 1. Associations of Childhood Adversity With Work Disability, Characteristics of the Sample, by Sex and Racea   Women  Men    African American  White  African American  White  Sample size, nb  1,175  2,320  634  1,916  Work disability status transitions, nb  14,022  31,072  7,304  29,976  Person-years, nb  22,867  50,467  11,709  44,064  Age at baseline, mean (CI)c  31.2 (30.2–32.3)  32.2 (31.6–32.8)  35.9 (33.9–38.0)  33.4 (32.5–34.4)  Education years, mean (CI)c  12.8 (12.0–13.6)  13.6 (13.3–13.9)  12.7 (12.3–13.1)  13.5 (13.2–13.8)  Adversity index (0–4), mean (CI)c  1.7 (1.5–1.9)  1.3 (1.2–1.4)  1.6 (1.3–1.8)  1.2 (1.1–1.3)   Index=0, %d  20.0  33.2  22.9  35.1   Index=1, %d  27.1  29.1  27.9  30.0   Index=2, %d  22.5  21.0  23.8  19.7   Index=3, %d  17.3  10.5  16.1  8.9   Index=4+, %d  13.2  6.1  9.3  5.7    Women  Men    African American  White  African American  White  Sample size, nb  1,175  2,320  634  1,916  Work disability status transitions, nb  14,022  31,072  7,304  29,976  Person-years, nb  22,867  50,467  11,709  44,064  Age at baseline, mean (CI)c  31.2 (30.2–32.3)  32.2 (31.6–32.8)  35.9 (33.9–38.0)  33.4 (32.5–34.4)  Education years, mean (CI)c  12.8 (12.0–13.6)  13.6 (13.3–13.9)  12.7 (12.3–13.1)  13.5 (13.2–13.8)  Adversity index (0–4), mean (CI)c  1.7 (1.5–1.9)  1.3 (1.2–1.4)  1.6 (1.3–1.8)  1.2 (1.1–1.3)   Index=0, %d  20.0  33.2  22.9  35.1   Index=1, %d  27.1  29.1  27.9  30.0   Index=2, %d  22.5  21.0  23.8  19.7   Index=3, %d  17.3  10.5  16.1  8.9   Index=4+, %d  13.2  6.1  9.3  5.7  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study, n = 6,045. bUnweighted sample characteristics. cResults weighted to be nationally representative and accounting for the survey design; CI = 95% confidence interval. dIndex level comparisons across race, p < .0001 for both women and men. View Large Transition Probabilities from the Markov Model In results not shown in a table, all of the Markov model parameter estimates associated childhood adversity with a higher a likelihood of becoming disabled, or more severely disabled, and a lower likelihood of recovering from disability; five of the six parameters were statistically significant (p < .001). The parameters governed the microsimulations. Work Disability in Working Life Figure 1 shows the increase in the percentage of years at ages 30–65 with work disability, separately for each of the nine adversities, each compared to no adversities (all p < .01). Combining results for moderate and severe work disability, as shown in Figure 1, the adversities associated with the largest differences in work disability were bullying, being raised by a single parent, and having a parent with education no more than grade 8. Not shown are comparable separate results by gender and race, in which the largest differences in work disability were associated with: fair/poor childhood health, neighborhood characteristics, and bullying for African American women and men; bullying, parent education, and childhood health for white women; and being raised by a single parent, neighborhood safety, and bullying for white men. Figure 1. View largeDownload slide Increase in the percentage of years ages 30–65 with moderate and severe work disability for individuals reporting four or more childhood adversities, compared to individuals reporting none of the adversities. Data source: Panel Study of Income Dynamics. Childhood adversity codes: 1=bullied; 2=single parent; 3=parent education ≤ grade 8; 4=fair or poor childhood health; 5=unsafe neighborhood; 6=neighbors not reliable for help; 7=parents divorced; 8=parent unemployment; 9=financial struggle. Figure 1. View largeDownload slide Increase in the percentage of years ages 30–65 with moderate and severe work disability for individuals reporting four or more childhood adversities, compared to individuals reporting none of the adversities. Data source: Panel Study of Income Dynamics. Childhood adversity codes: 1=bullied; 2=single parent; 3=parent education ≤ grade 8; 4=fair or poor childhood health; 5=unsafe neighborhood; 6=neighbors not reliable for help; 7=parents divorced; 8=parent unemployment; 9=financial struggle. Table 2 shows the principal microsimulation results. We controlled for education through stratified analysis; the results shown represent individuals with high school education. Women and men with high adversity had significantly more moderate and severe work disability than others (all p < .01). For example, among African American women at age 30, of those with high childhood adversity 10.2% reported moderate work disability and 5.6% reported severe. Comparable results for no adversities were 4.1% and 1.9%. Table 2. Association of Childhood Adversity With Adult Work Disabilitya   Women  Men    African American  White  African American  White  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  Moderate work disability   Age 1st spell, mean  40.6  34.2*  40.6  32.1*  42.1  38.6*  42.1  37.1*   Spells, n, mean  2.1  4.2*  2.0  4.9*  1.6  2.7*  1.5  2.9*   Population %, age 30  4.1  10.2*  4.7  11.3*  3.3  6.1*  3.7  7.2*   Population %, age 65  12.1  23.0*  16.0  28.9*  10.1  16.3*  11.5  18.9*   % of life ages 30–65  8.8  16.7*  10.2  19.9*  8.0  11.7*  9.4  13.3*  Severe work disability   Age 1st spell, mean  45.7  40.6*  47.5  41.6*  46.7  43.1*  47.9  46.6*   Spells, n, mean  1.3  2.8*  0.7  2.5*  1.0  2.0*  0.6  1.0*   Population %, age 30  1.9  5.6*  1.1  3.5*  1.8  4.1*  1.0  2.0*   Population %, age 65  10.9  21.4*  8.4  17.8*  10.7  19.0*  6.2  11.5*   % of life ages 30–65  5.4  12.5*  3.2  10.1*  5.2  10.3*  3.0  5.8*    Women  Men    African American  White  African American  White  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  Moderate work disability   Age 1st spell, mean  40.6  34.2*  40.6  32.1*  42.1  38.6*  42.1  37.1*   Spells, n, mean  2.1  4.2*  2.0  4.9*  1.6  2.7*  1.5  2.9*   Population %, age 30  4.1  10.2*  4.7  11.3*  3.3  6.1*  3.7  7.2*   Population %, age 65  12.1  23.0*  16.0  28.9*  10.1  16.3*  11.5  18.9*   % of life ages 30–65  8.8  16.7*  10.2  19.9*  8.0  11.7*  9.4  13.3*  Severe work disability   Age 1st spell, mean  45.7  40.6*  47.5  41.6*  46.7  43.1*  47.9  46.6*   Spells, n, mean  1.3  2.8*  0.7  2.5*  1.0  2.0*  0.6  1.0*   Population %, age 30  1.9  5.6*  1.1  3.5*  1.8  4.1*  1.0  2.0*   Population %, age 65  10.9  21.4*  8.4  17.8*  10.7  19.0*  6.2  11.5*   % of life ages 30–65  5.4  12.5*  3.2  10.1*  5.2  10.3*  3.0  5.8*  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study, n = 6,045. *p < .01. View Large Table 2. Association of Childhood Adversity With Adult Work Disabilitya   Women  Men    African American  White  African American  White  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  Moderate work disability   Age 1st spell, mean  40.6  34.2*  40.6  32.1*  42.1  38.6*  42.1  37.1*   Spells, n, mean  2.1  4.2*  2.0  4.9*  1.6  2.7*  1.5  2.9*   Population %, age 30  4.1  10.2*  4.7  11.3*  3.3  6.1*  3.7  7.2*   Population %, age 65  12.1  23.0*  16.0  28.9*  10.1  16.3*  11.5  18.9*   % of life ages 30–65  8.8  16.7*  10.2  19.9*  8.0  11.7*  9.4  13.3*  Severe work disability   Age 1st spell, mean  45.7  40.6*  47.5  41.6*  46.7  43.1*  47.9  46.6*   Spells, n, mean  1.3  2.8*  0.7  2.5*  1.0  2.0*  0.6  1.0*   Population %, age 30  1.9  5.6*  1.1  3.5*  1.8  4.1*  1.0  2.0*   Population %, age 65  10.9  21.4*  8.4  17.8*  10.7  19.0*  6.2  11.5*   % of life ages 30–65  5.4  12.5*  3.2  10.1*  5.2  10.3*  3.0  5.8*    Women  Men    African American  White  African American  White  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  Moderate work disability   Age 1st spell, mean  40.6  34.2*  40.6  32.1*  42.1  38.6*  42.1  37.1*   Spells, n, mean  2.1  4.2*  2.0  4.9*  1.6  2.7*  1.5  2.9*   Population %, age 30  4.1  10.2*  4.7  11.3*  3.3  6.1*  3.7  7.2*   Population %, age 65  12.1  23.0*  16.0  28.9*  10.1  16.3*  11.5  18.9*   % of life ages 30–65  8.8  16.7*  10.2  19.9*  8.0  11.7*  9.4  13.3*  Severe work disability   Age 1st spell, mean  45.7  40.6*  47.5  41.6*  46.7  43.1*  47.9  46.6*   Spells, n, mean  1.3  2.8*  0.7  2.5*  1.0  2.0*  0.6  1.0*   Population %, age 30  1.9  5.6*  1.1  3.5*  1.8  4.1*  1.0  2.0*   Population %, age 65  10.9  21.4*  8.4  17.8*  10.7  19.0*  6.2  11.5*   % of life ages 30–65  5.4  12.5*  3.2  10.1*  5.2  10.3*  3.0  5.8*  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study, n = 6,045. *p < .01. View Large Table 2 also shows analogous results for the percentage of years from ages 30 through 65 with moderate and severe work disability. For all groups, the percentages were significantly larger with high adversity than with none (all p < .01). For example, white women with high adversity could expect moderate work disability for 19.9% of those years, severe for 10.1%; comparable results for white women reporting no adversities were 10.2% with moderate work disability and 3.2% with severe. Women and men with high adversity also had work disability at younger ages than those reporting no adversities, and more spells (all p < .01). African American women and men were more likely to have a severe work disability at ages 30 and 65 than whites. However, there was no evidence of notable differences between African Americans and whites in the association of childhood adversity with work disability. Women were more likely to have moderate and severe work disability than men, and also had more spells of both moderate and severe work disability. There was modest evidence that associations of adversity with work disability may have been larger for women than for men; for example among whites at age 65, the prevalence of moderate work disability was 80.6% greater with high adversity than with no adversity for women, 64.3% greater for men (percent differences not shown). Results of the model that did not control for education also associated adversity with significantly more work disability (results not shown), although the size of the effect was substantially smaller. For example, in that model white women with high adversity could expect moderate work disability for 9.7% of the years from ages 30–65, severe for 3.0%, compared with 9.1% and 2.7% with no adversities (both p < .05). Associations of Childhood Adversity with Chronic Conditions and Sedentary Behavior Table 3 shows odds ratios estimating the association of childhood adversity with diabetes, heart disease, obesity, and sedentary behavior at ages 40 and 65 for three models: unadjusted, adjusted for education, and adjusted for education and childhood health. In all models, the results suggest that high adversity was associated with greater risks of having diabetes or being obese at ages 40 and 65, and greater risk of having heart disease at age 65. For example, in the unadjusted model, compared with participants reporting no adversities, the odds of having diabetes at age 40 were twice as great with high adversity (p < .001). The bottom row of Table 3 shows analogous associations of childhood adversity with having at least one of the four conditions at ages 40 and 65, all of which were significant. In all models, participants reporting high adversity had at least 80% higher odds of having one or more of the conditions at age 65 than those reporting none. Table 3. Odds Ratios Estimating the Risk of Having Diabetes, Heart Disease, Obesity, or Sedentary Behavior at Ages 40 and 65, Comparing People With High Childhood Adversity to Those With Low Childhood Adversitya   Unadjusted  Adjusted for years of education  Participants reporting good to excellent childhood health, adjusted for years of education    Age 40  Age 65  Age 40  Age 65  Age 40  Age 65  Diabetes  2.0***  1.6*  1.9**  1.6*  1.8**  1.4  Heart Disease  1.5  1.8*  1.4  1.8*  1.5  1.8*  Obesity  1.5**  1.7*  1.4*  1.6*  1.4*  1.5  Sedentary  1.3  1.3  1.2  1.1  1.1  1.1  Any condition  1.5***  1.9*  1.4**  1.8*  1.3*  1.8*    Unadjusted  Adjusted for years of education  Participants reporting good to excellent childhood health, adjusted for years of education    Age 40  Age 65  Age 40  Age 65  Age 40  Age 65  Diabetes  2.0***  1.6*  1.9**  1.6*  1.8**  1.4  Heart Disease  1.5  1.8*  1.4  1.8*  1.5  1.8*  Obesity  1.5**  1.7*  1.4*  1.6*  1.4*  1.5  Sedentary  1.3  1.3  1.2  1.1  1.1  1.1  Any condition  1.5***  1.9*  1.4**  1.8*  1.3*  1.8*  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study; low childhood adversity = values 0 and 1 on a five-point adversity scale (values 0 through 4); high childhood adversity = values 3 and 4 on the same scale; Any condition = any of the other four listed conditions. *p < .05, **p < .01, ***p < .001. View Large Table 3. Odds Ratios Estimating the Risk of Having Diabetes, Heart Disease, Obesity, or Sedentary Behavior at Ages 40 and 65, Comparing People With High Childhood Adversity to Those With Low Childhood Adversitya   Unadjusted  Adjusted for years of education  Participants reporting good to excellent childhood health, adjusted for years of education    Age 40  Age 65  Age 40  Age 65  Age 40  Age 65  Diabetes  2.0***  1.6*  1.9**  1.6*  1.8**  1.4  Heart Disease  1.5  1.8*  1.4  1.8*  1.5  1.8*  Obesity  1.5**  1.7*  1.4*  1.6*  1.4*  1.5  Sedentary  1.3  1.3  1.2  1.1  1.1  1.1  Any condition  1.5***  1.9*  1.4**  1.8*  1.3*  1.8*    Unadjusted  Adjusted for years of education  Participants reporting good to excellent childhood health, adjusted for years of education    Age 40  Age 65  Age 40  Age 65  Age 40  Age 65  Diabetes  2.0***  1.6*  1.9**  1.6*  1.8**  1.4  Heart Disease  1.5  1.8*  1.4  1.8*  1.5  1.8*  Obesity  1.5**  1.7*  1.4*  1.6*  1.4*  1.5  Sedentary  1.3  1.3  1.2  1.1  1.1  1.1  Any condition  1.5***  1.9*  1.4**  1.8*  1.3*  1.8*  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study; low childhood adversity = values 0 and 1 on a five-point adversity scale (values 0 through 4); high childhood adversity = values 3 and 4 on the same scale; Any condition = any of the other four listed conditions. *p < .05, **p < .01, ***p < .001. View Large Work Disability With and Without Chronic Conditions and Sedentary Behavior Table 4 shows results comparing high adversity to none, for women and men with and without adult diabetes, heart disease, obesity, or sedentary behavior, and for those with all four conditions, ages 40 through 65. For example, at age 65, among women with none of the four conditions 8.4% of those reporting no adversity had moderate work disability, compared with 14.2% with high adversity (p < .01). In analogous results for severe work disability, among women with none of the four conditions 2.3% of those reporting no adversity had severe work disability, compared with 6.7% with high adversity (p < .01). Most results in Table 4 also show that individuals with high adversity: experienced their first moderate or severe work disability at significantly younger ages than those with no reported adversities; had significantly more spells of moderate and severe work disability; and had moderate and severe work disability for significantly greater proportions of life from ages 40 through 65. Table 4. Association of Childhood Adversity, and Adult Diabetes, Heart Disease, Obesity, and Sedentary Behavior, With Work Disability, Ages 40–65a   No adult conditionsb  Diabetes  Heart disease  Obesity  Sedentary  All adult conditions  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  0  4+  0  4+  Women                          Moderate work disability                           Age 1st spell, mean  59.2  54.7*  57.7  52.8*  55.8  51.0*  55.7  50.9*  53.8  49.3*  46.5  44.8*   Spells, n, mean  2.2  3.7*  2.6  4.4*  3.1  5.0*  3.3  5.3*  4.0  6.1*  6.8  8.0   Population %, age 65  8.4  14.2*  10.7  17.2*  14.7  22.5*  13.7  20.5*  17.1  24.5*  37.0  37.0   % of life, ages 40–65  6.5  11.2*  8.3  13.8*  11.6  18.5*  10.8  16.9*  13.6  20.5*  33.6  36.3  Severe work disability   Age 1st spell, mean  63.9  61.0*  63.0  58.7*  62.7  58.7*  61.4  55.7*  61.4  55.6*  51.7  46.9*   Spells, n, mean  0.7  1.7*  1.0  2.3*  1.0  2.2*  1.5  3.3*  1.6  3.3*  4.0  5.9*   Population %, age 65  2.3  6.7*  3.7  9.9*  3.2  8.2*  6.6  16.5*  6.3  15.4*  21.8  36.1*   % of life, ages 40–65  1.8  5.0*  2.8  7.7*  2.5  6.5*  5.0  13.0*  4.7  12.1*  18.7  33.0*  Men  Moderate work disability   Age 1st spell, mean  60.6  56.5*  59.2  54.8*  57.5  52.8*  57.6  53.0*  55.7  51.1*  48.0  46.3*   Spells, n, mean  1.7  3.0*  2.1  3.5*  2.5  4.0*  2.6  4.1*  3.2  4.8*  5.5  6.2   Population %, age 65  7.1  12.1*  9.1  14.7*  12.8  19.9*  11.6  17.4*  14.7  21.3*  33.3  32.4   % of life ages 40–65  5.5  9.4*  7.0  11.7*  9.9  16.1*  9.0  14.1*  11.6  17.6*  29.7  31.7  Severe work disability   Age 1st spell, mean  64.1  62.4*  63.6  60.6*  63.4  60.5*  62.5  58.0*  62.7  58.0*  53.8  48.5*   Spells, n, mean  0.5  1.2*  0.7  1.6*  0.7  1.5*  1.1  2.3*  1.1  2.3*  2.9  4.2*   Population %, age 65  2.1  6.4*  3.5  10.1*  3.0  8.1*  6.7  17.6*  6.1  16.1*  24.8  41.8 +   % of life ages 40–65  1.7  4.9*  2.7  7.7*  2.3  6.3*  5.1  13.5*  4.6  12.3*  20.7  37.2*    No adult conditionsb  Diabetes  Heart disease  Obesity  Sedentary  All adult conditions  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  0  4+  0  4+  Women                          Moderate work disability                           Age 1st spell, mean  59.2  54.7*  57.7  52.8*  55.8  51.0*  55.7  50.9*  53.8  49.3*  46.5  44.8*   Spells, n, mean  2.2  3.7*  2.6  4.4*  3.1  5.0*  3.3  5.3*  4.0  6.1*  6.8  8.0   Population %, age 65  8.4  14.2*  10.7  17.2*  14.7  22.5*  13.7  20.5*  17.1  24.5*  37.0  37.0   % of life, ages 40–65  6.5  11.2*  8.3  13.8*  11.6  18.5*  10.8  16.9*  13.6  20.5*  33.6  36.3  Severe work disability   Age 1st spell, mean  63.9  61.0*  63.0  58.7*  62.7  58.7*  61.4  55.7*  61.4  55.6*  51.7  46.9*   Spells, n, mean  0.7  1.7*  1.0  2.3*  1.0  2.2*  1.5  3.3*  1.6  3.3*  4.0  5.9*   Population %, age 65  2.3  6.7*  3.7  9.9*  3.2  8.2*  6.6  16.5*  6.3  15.4*  21.8  36.1*   % of life, ages 40–65  1.8  5.0*  2.8  7.7*  2.5  6.5*  5.0  13.0*  4.7  12.1*  18.7  33.0*  Men  Moderate work disability   Age 1st spell, mean  60.6  56.5*  59.2  54.8*  57.5  52.8*  57.6  53.0*  55.7  51.1*  48.0  46.3*   Spells, n, mean  1.7  3.0*  2.1  3.5*  2.5  4.0*  2.6  4.1*  3.2  4.8*  5.5  6.2   Population %, age 65  7.1  12.1*  9.1  14.7*  12.8  19.9*  11.6  17.4*  14.7  21.3*  33.3  32.4   % of life ages 40–65  5.5  9.4*  7.0  11.7*  9.9  16.1*  9.0  14.1*  11.6  17.6*  29.7  31.7  Severe work disability   Age 1st spell, mean  64.1  62.4*  63.6  60.6*  63.4  60.5*  62.5  58.0*  62.7  58.0*  53.8  48.5*   Spells, n, mean  0.5  1.2*  0.7  1.6*  0.7  1.5*  1.1  2.3*  1.1  2.3*  2.9  4.2*   Population %, age 65  2.1  6.4*  3.5  10.1*  3.0  8.1*  6.7  17.6*  6.1  16.1*  24.8  41.8 +   % of life ages 40–65  1.7  4.9*  2.7  7.7*  2.3  6.3*  5.1  13.5*  4.6  12.3*  20.7  37.2*  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study, n = 4,135, 28,084 functional status transitions. +p < .05, *p < .01. View Large Table 4. Association of Childhood Adversity, and Adult Diabetes, Heart Disease, Obesity, and Sedentary Behavior, With Work Disability, Ages 40–65a   No adult conditionsb  Diabetes  Heart disease  Obesity  Sedentary  All adult conditions  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  0  4+  0  4+  Women                          Moderate work disability                           Age 1st spell, mean  59.2  54.7*  57.7  52.8*  55.8  51.0*  55.7  50.9*  53.8  49.3*  46.5  44.8*   Spells, n, mean  2.2  3.7*  2.6  4.4*  3.1  5.0*  3.3  5.3*  4.0  6.1*  6.8  8.0   Population %, age 65  8.4  14.2*  10.7  17.2*  14.7  22.5*  13.7  20.5*  17.1  24.5*  37.0  37.0   % of life, ages 40–65  6.5  11.2*  8.3  13.8*  11.6  18.5*  10.8  16.9*  13.6  20.5*  33.6  36.3  Severe work disability   Age 1st spell, mean  63.9  61.0*  63.0  58.7*  62.7  58.7*  61.4  55.7*  61.4  55.6*  51.7  46.9*   Spells, n, mean  0.7  1.7*  1.0  2.3*  1.0  2.2*  1.5  3.3*  1.6  3.3*  4.0  5.9*   Population %, age 65  2.3  6.7*  3.7  9.9*  3.2  8.2*  6.6  16.5*  6.3  15.4*  21.8  36.1*   % of life, ages 40–65  1.8  5.0*  2.8  7.7*  2.5  6.5*  5.0  13.0*  4.7  12.1*  18.7  33.0*  Men  Moderate work disability   Age 1st spell, mean  60.6  56.5*  59.2  54.8*  57.5  52.8*  57.6  53.0*  55.7  51.1*  48.0  46.3*   Spells, n, mean  1.7  3.0*  2.1  3.5*  2.5  4.0*  2.6  4.1*  3.2  4.8*  5.5  6.2   Population %, age 65  7.1  12.1*  9.1  14.7*  12.8  19.9*  11.6  17.4*  14.7  21.3*  33.3  32.4   % of life ages 40–65  5.5  9.4*  7.0  11.7*  9.9  16.1*  9.0  14.1*  11.6  17.6*  29.7  31.7  Severe work disability   Age 1st spell, mean  64.1  62.4*  63.6  60.6*  63.4  60.5*  62.5  58.0*  62.7  58.0*  53.8  48.5*   Spells, n, mean  0.5  1.2*  0.7  1.6*  0.7  1.5*  1.1  2.3*  1.1  2.3*  2.9  4.2*   Population %, age 65  2.1  6.4*  3.5  10.1*  3.0  8.1*  6.7  17.6*  6.1  16.1*  24.8  41.8 +   % of life ages 40–65  1.7  4.9*  2.7  7.7*  2.3  6.3*  5.1  13.5*  4.6  12.3*  20.7  37.2*    No adult conditionsb  Diabetes  Heart disease  Obesity  Sedentary  All adult conditions  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  0  4+  0  4+  Women                          Moderate work disability                           Age 1st spell, mean  59.2  54.7*  57.7  52.8*  55.8  51.0*  55.7  50.9*  53.8  49.3*  46.5  44.8*   Spells, n, mean  2.2  3.7*  2.6  4.4*  3.1  5.0*  3.3  5.3*  4.0  6.1*  6.8  8.0   Population %, age 65  8.4  14.2*  10.7  17.2*  14.7  22.5*  13.7  20.5*  17.1  24.5*  37.0  37.0   % of life, ages 40–65  6.5  11.2*  8.3  13.8*  11.6  18.5*  10.8  16.9*  13.6  20.5*  33.6  36.3  Severe work disability   Age 1st spell, mean  63.9  61.0*  63.0  58.7*  62.7  58.7*  61.4  55.7*  61.4  55.6*  51.7  46.9*   Spells, n, mean  0.7  1.7*  1.0  2.3*  1.0  2.2*  1.5  3.3*  1.6  3.3*  4.0  5.9*   Population %, age 65  2.3  6.7*  3.7  9.9*  3.2  8.2*  6.6  16.5*  6.3  15.4*  21.8  36.1*   % of life, ages 40–65  1.8  5.0*  2.8  7.7*  2.5  6.5*  5.0  13.0*  4.7  12.1*  18.7  33.0*  Men  Moderate work disability   Age 1st spell, mean  60.6  56.5*  59.2  54.8*  57.5  52.8*  57.6  53.0*  55.7  51.1*  48.0  46.3*   Spells, n, mean  1.7  3.0*  2.1  3.5*  2.5  4.0*  2.6  4.1*  3.2  4.8*  5.5  6.2   Population %, age 65  7.1  12.1*  9.1  14.7*  12.8  19.9*  11.6  17.4*  14.7  21.3*  33.3  32.4   % of life ages 40–65  5.5  9.4*  7.0  11.7*  9.9  16.1*  9.0  14.1*  11.6  17.6*  29.7  31.7  Severe work disability   Age 1st spell, mean  64.1  62.4*  63.6  60.6*  63.4  60.5*  62.5  58.0*  62.7  58.0*  53.8  48.5*   Spells, n, mean  0.5  1.2*  0.7  1.6*  0.7  1.5*  1.1  2.3*  1.1  2.3*  2.9  4.2*   Population %, age 65  2.1  6.4*  3.5  10.1*  3.0  8.1*  6.7  17.6*  6.1  16.1*  24.8  41.8 +   % of life ages 40–65  1.7  4.9*  2.7  7.7*  2.3  6.3*  5.1  13.5*  4.6  12.3*  20.7  37.2*  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study, n = 4,135, 28,084 functional status transitions. +p < .05, *p < .01. View Large Comparisons of high adversity with none were statistically significant for women and men with none of the adult conditions, and with each of the individual conditions. In several instances, the comparisons were not significant for individuals with all four conditions; this result may be due to substantial disability risks for individuals with all of these conditions, regardless of their childhood adversity status. In all models, the association of childhood adversity with work disability was generally linear. Thus, for example, results for individuals who reported any two adversities were approximately mid-way between the results for high adversity and none. Diagnosis Reports and Family Effects Regarding reports of diabetes or heart disease diagnosis ages, for participants who reported diagnosis ages in 2, 3, 4, or 5 waves (2005–2013), respectively, the ICCs were: 0.99, 0.79, 0.78, and 0.74 for diabetes and 0.89, 0.84, 079, and 0.76 for heart disease (results not shown in tables). For those who reported diagnosis ages in five waves, where variation was greatest, the average deviation from each individual’s mean reported age was 3.3 years for diabetes (SD 4.2), 2.8 years for heart disease (SD 3.4). However, 92.5% of participants reported in two or fewer waves; 97.8% were within 1 year for diabetes, 48.7% for heart disease. Results were similar when the 1999, 2001, and 2003 reports of the time since diagnosis were included. In the analysis of family effects, the average participant had 2.6 other family members who also participated (data not shown). The comparison of the fixed and random effects approaches was not significant (p = .128). Thus, there was no evidence that family membership affected the results. Discussion We examined associations of childhood adversity with work disability throughout working life. Consistent with our first hypothesis, African American and white women and men with high adversity were more likely to have moderate and severe work disability throughout working life than those reporting no adversities. Although these results are consistent with studies linking childhood adversity with functional status (e.g., Laditka & Laditka, 2017; Montez & Hayward, 2014; Smith et al., 2016), only one previous study examined this association with work disability: Shuey and Willson (2017) found that poor health and poverty in childhood were associated with work disability. Our results underscore the burden of childhood adversity for individuals and society. Consistent with the pathway model (e.g., Pudrovska & Anikputa, 2014), adults with high childhood adversity were more likely than those with no adversity to have the four causes of functional impairment that we studied. The results were also consistent with our second hypothesis, that childhood adversity would be associated with more work disability after controlling for education, health conditions, and a health behavior that are established causes of work disability. It would be useful to examine whether other factors such as mental health may account for this association. There was modest evidence that the association of childhood adversity with work disability may be larger for women than for men. Comparing African Americans and whites, there were no notable differences in the association of adversity with work disability. Thus, the results did not support our hypothesis regarding race. Our results regarding race were also not consistent with those of the one previous study in this area (Umberson et al., 2014). That analysis focused on adults ages 60 years and older, with substantially fewer African American participants and many fewer outcome measurements. That study also found that when socioeconomic status was controlled the association with race did not persist among men, and was greatly attenuated for women. However, African Americans represented in our data were much more likely than whites to have experienced high levels of childhood adversity, in contrast to the finding of no difference by Umberson et al. (2014). Thus, the overall health burden of childhood adversity is likely to be larger for African Americans than for whites. The results for gender and race should be interpreted with caution as they may be sensitive to the measures of childhood adversity in the index. Taken as a whole, the results were consistent with the accumulation of risk model. We found that work disability increased with childhood adversity and adult chronic conditions. Our study provided evidence that childhood adversity may contribute to work disability separately from the pathway through adult diabetes, heart disease, and obesity, although the specific mechanisms of that contribution are not clear. Our models did not control for workplace conditions or exposures, or adult socioeconomic circumstances, although the control for education is likely to have accounted for some of the variation associated with those factors. Limitations Self-reports of work disabilities may be subject to measurement error (see, e.g., Burkhauser et al., 2002; Mathiowetz, 2000). However, work disability is not limited to the presence of functional limitations that objectively limit work (Jette & Badley, 2000; Nagi, 1991; WHO, 2001). Aside from disability determinations for public or private insurance benefits, if an individual believes that a health limitation makes it difficult or impossible to work it may be reasonable to accept that judgment (Pransky et al., 2016; Silverstein, 2008). Consistent with most related studies, participants reported adversities retrospectively (e.g., Haas et al., 2012; O’Rand & Hamil-Luker, 2005). Although recall bias may have affected the results, recollections of childhood socioeconomic status and health are typically valid (Batty, Lawlor, Macintyre, Clark, & Leon, 2005; Haas, 2007; Smith, 2009). Consistent with previous research, the index assumed that the adversities were equally associated with work disability (e.g., Montez & Hayward, 2014). However, the results suggested that the associations of individual childhood adversities with work disability varied within and between groups. Adversities may also interact to modify those associations. Results indicated a linear association of childhood adversity with work disability. This result should be interpreted with caution. It may depend on the proportion of individuals who experience each type of adversity, variation among the adversities in the strength of the association, the specific combinations of the adversities to which individuals are exposed, the periods of childhood when the adversities are experienced and the durations of the exposures, exposures to adversities that we did not measure, and other childhood risks to adult health and factors that may help to protect adult health that were not represented in our models. We did not examine abuse during childhood, mental illness, violence, environmental risks, substance abuse, or involvement with the criminal justice system. The model did not consider that some adversity in childhood, or some types of adversity may promote skills or traits associated with better health in adulthood (Schafer, Ferraro, & Mustillo, 2011). Results of the model that did not control for education also associated adversity with more work disability, although the size of the effect was smaller. The latter analysis averaged risks across all education levels, and the majority of adults have at least some postsecondary education. Thus, the average individual represented in the model that did not control for education was more highly educated than those who were the focus of our principal microsimulation results, people with high school education. Consistent with expectations that education is negatively associated with childhood adversity and predicts positive health outcomes, the results suggest that education moderates the association of childhood adversity with work disability. We focused the presentation of the microsimulation results on high school education because people with high school education are the largest educational attainment group. For example, nearly 30% of people aged 55 years and older have high school education, compared with 18% of the same age group with bachelor’s degrees. We also examined whether characteristics shared within families might have affected the results. Using a PSID subsample focused on siblings, Hass (2006) estimated a fixed-effects model examining the association of childhood health with socioeconomic attainment; results suggested that family characteristics may affect the association. However, with our much larger PSID sample and the requirement of CRCS participation, we found no evidence that such characteristics affected the association of childhood adversity with work disability. For observations from before 1999, we assigned diabetes and heart disease status based on later reports of disease onset. A study of PSID cancer diagnosis ages found variation across individuals’ reports over time, due in part to household heads responding for their spouses or partners, and possibly also to stigma associated with cancer (Zajacova, Dowd, Beam, Schoeni, & Wallace, 2010). In the present study, when spouses or partners became household heads in one or more waves we used their own reports of their diagnosis ages. The ICCs for the onset ages suggested acceptable reliability. Conclusions and Implications Participants with childhood adversity had significantly more work disability than those who did not have that experience. The results suggest that reducing socioeconomic, family, and health adversities in childhood, increasing neighborhood cohesion and safety, and reducing bullying may help to reduce work disability. Whether or not an individual with functional impairments has a work disability may often depend on workplace and community cultures, policies, and accommodations (Pransky et al., 2016). Addressing those factors is also desirable because people who experience work disability are significantly more likely than others to live in poverty as they approach retirement, and less likely to own a home or to have a pension (Shuey & Willson, 2017). The fact that African Americans were significantly more likely to have experienced childhood adversity suggests that they may be particularly affected by this financial disparity. Given our aging workforce, greater racial and ethnic diversity, and the increasing number of employees with functional limitations and multiple chronic health conditions, a better understanding of the social production of disability would be useful. It would also be useful to reduce childhood adversity, and to promote resilience and positive health behaviors for people who have that enduring risk to health. Funding This research was supported by the Panel Study of Income Dynamics with a grant from the National Institute on Aging under Grant P01AG029409. The collection of data used in this study was partly supported by the National Institutes of Health under Grant R01 HD069609, and the National Science Foundation under award number 1157698. Conflict of Interest None reported. Acknowledgments The authors are grateful to Cheryl Elman, PhD, Philippa Clarke, PhD, and two anonymous reviewers for valuable comments about this research. Author contributions: S. B. Laditka and J. N. Laditka participated in conceptualizing and designing this study, analyzing the data and interpreting the results, drafting the manuscript, and approving the final manuscript. The data used for this study are available from the Panel Study of Income Dynamics, http://psidonline.isr.umich.edu/default.aspx. References Allison, P. D. ( 2005). Fixed effects regression methods for longitudinal data using SAS . Cary, NC: SAS Institute Inc. Andrade, F. C. ( 2010). Measuring the impact of diabetes on life expectancy and disability-free life expectancy among older adults in Mexico. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 65B, 381– 389. doi: 10.1093/geronb/gbp119 Google Scholar CrossRef Search ADS   Batty, G. D., Lawlor, D. A., Macintyre, S., Clark, H., & Leon, D. A. ( 2005). Accuracy of adults’ recall of childhood social class: Findings from the Aberdeen children of the 1950s study. Journal of Epidemiology and Community Health , 59, 898– 903. doi: 10.1136/jech.2004.030932 Google Scholar CrossRef Search ADS PubMed  Ben-Shlomo, Y., & Kuh, D. ( 2002). A life course approach to chronic disease epidemiology: Conceptual models, empirical challenges and interdisciplinary perspectives. International Journal of Epidemiology , 31, 285– 293. doi: 10.1093/acprof:oso/9780198578154.001.0001 Google Scholar CrossRef Search ADS PubMed  Bowen, M. E., & González, H. M. ( 2010). Childhood socioeconomic position and disability in later life: Results of the Health and Retirement Study. American Journal of Public Health , 100( Suppl.1), S197– S203. doi: 10.2105/AJPH.2009.160986 Google Scholar CrossRef Search ADS PubMed  Breslin, F. C., Tompa, E., Zhao, R., Amick, B. C. 3rd, Pole, J. D., Smith, P., & Hogg-Johnson, S. ( 2007). Work disability absence among young workers with respect to earnings losses in the following year. Scandinavian Journal of Work, Environment & Health , 33, 192– 197. Google Scholar CrossRef Search ADS PubMed  Burkhauser, R. V., Daly, M. C., Houtenville, A. J., & Nargis, N. ( 2002). Self-reported work-limitation data: What they can and cannot tell us. Demography , 39, 541– 555. doi: 10.1353/dem.2002.0025 Google Scholar CrossRef Search ADS PubMed  Cohen, S., Janicki-Deverts, D., Chen, E., & Matthews, K. A. ( 2010). Childhood socioeconomic status and adult health. Annals of the New York Academy of Sciences , 1186, 37– 55. doi: 10.1111/j.1749-6632.2009.05334.x Google Scholar CrossRef Search ADS PubMed  Hass, S. A. ( 2006). Health selection and the process of social stratification: The effect of childhood health on socioeconomic attainment. Journal of Health and Social Behavior , 47, 339– 354. doi: 10.1177/002214650604700403 Google Scholar CrossRef Search ADS PubMed  Haas, S. A. ( 2007). The long-term effects of poor childhood health: An assessment and application of retrospective reports. Demography , 44, 113– 135. Google Scholar CrossRef Search ADS PubMed  Haas, S. A., Krueger, P. M., & Rohlfsen, L. ( 2012). Race/ethnic and nativity disparities in later life physical performance: The role of health and socioeconomic status over the life course. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 67, 238– 248. doi: 10.1093/geronb/gbr155 Google Scholar CrossRef Search ADS   Hamer, R. M. ( 2007). The %INTRACC macro. Retrieved from: http://support.sas.com/kb/25/031.html. Accessed October 23, 2017. Hatch, S. L. ( 2005). Conceptualizing and identifying cumulative adversity and protective resources: Implications for understanding health inequalities. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 60, S130– S134. doi: 10.1093/geronb/60.Special_Issue_2.S130 Google Scholar CrossRef Search ADS   Jette, A. M., & Badley, E. ( 2000). Conceptual issues in the measurement of work disability. In N. Mathiowetz & G.S. Wunderlich (Eds.), Survey measurement of work disability: Summary of a workshop  (pp. 4– 27). Washington, D.C.: National Academy Press. Laditka, S. B. ( 1998). Modeling lifetime nursing home use under assumptions of better health. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 53, S177– S187. Google Scholar CrossRef Search ADS   Laditka, S. B., & Laditka, J. N. ( 2009). Active life expectancy: A central measure of population health. In P. Uhlenberg (Ed.). International handbook of population aging  (pp. 543– 565). Netherlands: Springer-Verlag. Google Scholar CrossRef Search ADS   Laditka, S. B., & Laditka, J. N. ( 2014a). More education may limit disability and extend life for people with cognitive impairment. American Journal of Alzheimer’s Disease and Other Dementias , 29, 436– 447. doi: 10.1177/1533317513518648 Google Scholar CrossRef Search ADS   Laditka, J. N., & Laditka, S. B. ( 2014b). Stroke and active life expectancy in the United States, 1999-2009. Disability and Health Journal , 7, 472– 477. doi: 10.1016/j.dhjo.2014.06.005 Google Scholar CrossRef Search ADS   Laditka, S. B., & Laditka, J. N. ( 2015). Active life expectancy of Americans with diabetes: Risks of heart disease, obesity, and inactivity. Diabetes Research and Clinical Practice , 107, 37– 45. doi: 10.1016/j.diabres.2014.10.008 Google Scholar CrossRef Search ADS PubMed  Laditka, J. N., & Laditka, S. B. ( 2016a). Associations of educational attainment with disability and life expectancy by race and gender in the United States: A longitudinal analysis of the Panel Study of Income Dynamics. Journal of Aging and Health , 28, 1403– 1425. doi: 10.1177/0898264315620590 Google Scholar CrossRef Search ADS   Laditka, J. N., & Laditka, S. B. ( 2016b). Associations of multiple chronic health conditions with active life expectancy in the United States. Disability and Rehabilitation , 38, 273– 289. doi: 10.3109/09638288.2015.1041614 Google Scholar CrossRef Search ADS   Laditka, J. N., & Laditka, S. B. ( 2017). Adverse childhood circumstances and functional status throughout life. Journal of Aging and Health , Advance online publication. doi: 10.1177/ 0898264317715976 Laditka, J. N., & Laditka, S. B. ( 2018). Work disability in the United States, 1968-2015: Prevalence, duration, recovery, and trends. SSM - Population Health , 4, 126– 134. doi: 10.1016/j.ssmph.2017.12.006 Google Scholar CrossRef Search ADS PubMed  Laditka, S. B., & Wolf, D. A. ( 1998). New methods for analyzing active life expectancy. Journal of Aging and Health , 10, 214– 241. Google Scholar CrossRef Search ADS   Luo, Y., & Waite, L. J. ( 2005). The impact of childhood and adult SES on physical, mental, and cognitive well-being in later life. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 60, S93– S101. Google Scholar CrossRef Search ADS   Mathiowetz, N. ( 2000). Methodological issues in the measurement of work disability. In N. Mathiowetz & G.S. Wunderlich (Eds.), Survey measurement of work disability: Summary of a workshop  (pp. 28– 52). Washington, D.C.: National Academy Press. Montez, J. K., & Hayward, M. D. ( 2014). Cumulative childhood adversity, educational attainment, and active life expectancy among U.S. adults. Demography , 51, 413– 435. doi: 10.1007/s13524-013-0261-x Google Scholar CrossRef Search ADS PubMed  Nagi, S. ( 1991). Disability concepts revisited: Implications for prevention. Institute of Medicine, Disability in America: Toward a national agenda for prevention  (pp. 309– 327). Washington, D.C.: National Academy Press. doi:doi.org/10.17226/1579 O’Rand, A. M., & Hamil-Luker, J. ( 2005). Processes of cumulative adversity: Childhood disadvantage and increased risk of heart attack across the life course. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 60B, 117– 124. Google Scholar CrossRef Search ADS   Pransky, G. S., Fassier, J-B., Besen, E., Blanck, P., Ekberg, K., Feuerstein, M., … The Hopkinton Conference Working Group on Workplace Disability Prevention. ( 2016). Sustaining work participation across the life course. Journal of Occupational Rehabilitation , 26, 465– 479. doi: 10.1007/s10926-016- 9670-1 Google Scholar CrossRef Search ADS PubMed  Pudrovska, T., & Anikputa, B. ( 2014). Early-life socioeconomic status and mortality in later life: An integration of four life-course mechanisms. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 69, 451– 460. doi: 10.1093/geronb/gbt122 Google Scholar CrossRef Search ADS   Rank, M. R., & Hirschl, T. A. ( 2014). The risk of developing a work disability across the adulthood years. Disability and Health Journal , 7, 189– 195. doi: 10.1016/j.dhjo.2013.12.001 Google Scholar CrossRef Search ADS PubMed  Schafer, M. H., Ferraro, K. F., & Mustillo, S. A. ( 2011). Children of misfortune: Early adversity and cumulative inequality in perceived life trajectories. American Journal of Sociology , 116, 1053– 1091. Google Scholar CrossRef Search ADS   Schoeni, R., Stafford, F., McGonagle, K. A., & Andreski, P. ( 2013). Response rates in national surveys. Annals of the American Academy of Political and Social Science , 645, 60– 87. doi: 10.1177/0002716212456363 Google Scholar CrossRef Search ADS PubMed  Shuey, K. M., & Willson, A. E. ( 2017). Trajectories of work disability and economic insecurity approaching retirement. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , Advance Access publication July 8, 2017, 1– 11. doi: 10.1093/geronb/gbx096 Silverstein, M. ( 2008). Meeting the challenges of an aging workforce. American Journal of Industrial Medicine , 51, 269– 280. doi: 10.1002/ajim.20569 Google Scholar CrossRef Search ADS PubMed  Smith, J. P. ( 2009). Reconstructing childhood health histories. Demography , 46, 387– 403. doi: 10.1353/dem.0.0058 Google Scholar CrossRef Search ADS PubMed  Smith, N. R., Ferraro, K. F., Kemp, B. R., Morton, P. M., Mustillo, S. A., & Angel, J. L. ( 2016). Childhood misfortune and handgrip strength among Black, White, and Hispanic Americans. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , Advance access December 6, 2016. doi: 10.1093/geronb/gbw147 Social Security Administration( 2017). Disability benefits. Publication No. 05-10029. ICN 456000. Available at: https://www.ssa.gov/pubs/EN-05-10029.pdf. Special Advisory Committee of the Medical Outcomes Trust.( 2002). Assessing health status and quality-of-life instruments: Attributes and review criteria. Quality of Life Research , 11, 193– 205. CrossRef Search ADS PubMed  Toossi, M. ( 2012). Labor force projections to 2020: A more slowly growing workforce. Monthly Labor Review , 43– 64. Turner, R. J., Thomas, C. S., & Brown, T. H. ( 2016). Childhood adversity and adult health: Evaluating intervening mechanisms. Social Science & Medicine , 156, 114– 124. doi: org/10106/j.socscimed.2016.02.026 Google Scholar CrossRef Search ADS   Umberson, D., Williams, K., Thomas, P. A., Liu, H., & Thomeer, M. B. ( 2014). Race, gender, and chains of disadvantage: Childhood adversity, social relationships, and health. Journal of Health and Social Behavior , 55, 20– 38. doi: 10.1177/ 0022146514521426 Google Scholar CrossRef Search ADS PubMed  Ward, B. W. ( 2015). Multiple chronic conditions and labor force outcomes: A population study of U.S. adults. American Journal of Industrial Medicine , 58, 943– 954. doi: 10.1002/ajim.22439 Google Scholar CrossRef Search ADS PubMed  Wolke, D., Copeland, W. E., Angold, A., & Costello, E. J. ( 2013). Impact of bullying in childhood on adult health, wealth, crime, and social outcomes. Psychological Science , 24, 1958– 1970. doi: 10.1177/0956797613481608 Google Scholar CrossRef Search ADS PubMed  World Health Organization (WHO). ( 2001). ICIDH-2: International classification of impairments, activities, and participation. A manual of dimensions of disablement and functioning. Final draft . Geneva: World Health Organization. Yong, V., & Saito, Y. ( 2012). Are there education differentials in disability and mortality transitions and active life expectancy among Japanese older adults? Findings from a 10-year prospective cohort study. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 67, 343– 353. doi: 10.1093/geronb/gbs029 Google Scholar CrossRef Search ADS   Zajacova, A., Dowd, J. B., Schoeni, R. F., & Wallace, R. B. ( 2010). Consistency and precision of cancer reporting in a multiwave national panel survey. Population Health Metrics , 8, 20. doi: 10.1186/1478-7954-8-20 Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Journals of Gerontology Series B: Psychological Sciences and Social Sciences Oxford University Press

An Enduring Health Risk of Childhood Adversity: Earlier, More Severe, and Longer Lasting Work Disability in Adult Life

Loading next page...
 
/lp/ou_press/an-enduring-health-risk-of-childhood-adversity-earlier-more-severe-and-f01zn20W09
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
1079-5014
eISSN
1758-5368
D.O.I.
10.1093/geronb/gby018
Publisher site
See Article on Publisher Site

Abstract

Abstract Objectives Childhood adversity has been linked with adult health problems. We hypothesized that childhood adversity would also be associated with work limitations due to physical or nervous health problems, known as work disability. Method With data from the Panel Study of Income Dynamics (PSID) (1968–2013; n = 6,045; 82,374 transitions; 129,107 person-years) and the 2014 PSID Childhood Retrospective Circumstances Study, we estimated work disability transition probabilities with multinomial logistic Markov models. Four or more adversities defined a high level. Microsimulations quantified adult work disability patterns for African American and non-Hispanic white women and men, accounting for age, education, race, sex, diabetes, heart disease, obesity, and sedentary behavior. Results Childhood adversity was significantly associated with work disability. Of African American women with high adversity, 10.2% had moderate work disability at age 30 versus 4.1% with no reported adversities; comparable results for severe work disability were 5.6% versus 1.9% (both p < .01). Comparable results for whites were 11.3% versus 4.7%, and 3.5% versus 1.1% (p < .01). The association of childhood adversity with work disability remained significant after adjusting for diabetes, heart disease, obesity, and sedentary behavior (p < .05). Conclusions Childhood adversity may increase work disability throughout adult life. Cumulative advantage/disadvantage, Early origins of health, Epidemiology, Life course analysis Adverse circumstances during childhood (childhood adversity) may be associated with educational attainment and occupational choice, health-related behaviors, adult physical and mental health, adult socioeconomic status, and relationships, all of which have been linked to functional status and disability (e.g., Bowen & González, 2010; Haas, Krueger, & Rohlfsen, 2012; Laditka & Laditka, 2017; Luo & Waite, 2005; Montez & Hayward, 2014; Smith et al., 2016; Turner, Thomas, & Brown, 2016). Associations of childhood adversity with educational attainment are especially important because people with less education are more likely to work in jobs with greater occupational health risks such as machine operations, construction, repair or moving operations, helping professions, and cleaning (Hatch, 2005). Childhood adversity may be linked with personal control and efficacy, affecting workplace communications, stress, perceptions about work, and confidence to request accommodations that enable work for people who develop functional impairments (Hatch, 2005; Silverstein, 2008). Yet, few studies have examined associations of childhood adversity with the ability to work. We address this knowledge gap. The Concept of Work Disability Disability covered by Social Security is a “medical condition [that] must significantly limit your ability to do basic work activities—such as lifting, standing, walking, sitting, and remembering—for at least 12 months” (Social Security Administration, 2017). However, work disability is often defined more broadly, by self-reports of physical or nervous conditions that limit work or make it impossible (Burkhauser, Daly, Houtenville, & Nargis, 2002; Jette & Badley, 2000; Rank & Hirschl, 2014). We adopt the conceptual framework of work disability by Jette and Badley (2000), Nagi (1991) and the World Health Organization (WHO, 2001), which emphasizes the role of social arrangements in disability. Factors contributing to work disability include job characteristics that worsen health, and social contexts that impede work for people with functional limitations (Jette & Badley, 2000; Nagi, 1991; Pransky et al., 2016; Silverstein, 2008). Work disability measures a practical outcome of such factors. In a recent study, nearly 55% of people in the United States reported a work disability at least once between ages 25 and 60, nearly one-quarter a severe work disability (Rank & Hirschl, 2014). Another recent study found that a majority of Americans have a work disability between ages 20 and 65 (Laditka & Laditka, 2018). Even when a person recovers from work disability, career advancement may be limited permanently (Breslin et al., 2007). Shuey and Willson (2017) found that people with work disabilities have double the risk of others of experiencing poverty. Workforce aging makes work disability increasingly important: one-quarter of the United States’ labor force will be age 55 or older by 2020 (Toossi, 2012). Of working age adults in the United States, more than 40% have at least one chronic health condition, and 20% two or more, increasing the risk of work disability (Pransky et al., 2016; Silverstein, 2008; Ward, 2015). Childhood Adversity—A Life Course Perspective Researchers use four theoretical models to characterize the association of childhood adversity with adult health. The accumulation model emphasizes risk from persistent, recurring, or multiple childhood adversities (e.g., Cohen, Janicki-Deverts, Chen, & Matthews, 2010; O’Rand & Hamil-Luker, 2005). A second model focuses on pathways, stressing the relationship of childhood adversity with biological risks and health behaviors that increase adult chronic disease and functional impairment (e.g., Pudrovska & Anikputa, 2014). A third model focuses on timing, emphasizing risks in limited periods of susceptibility, especially in utero and in early childhood (e.g., Cohen et al., 2010). Finally, the change model, also called the social mobility model, posits that adult socioeconomic status can modify effects of childhood adversity (e.g., Ben-Shlomo & Kuh, 2002). The mechanisms of these models may combine to influence adult health, so it may not be possible to identify their separate effects (Pudrovska & Anikputa, 2014). With regard to race, African American children are more likely to live in poor families and neighborhoods than whites (e.g., Smith et al., 2016). They have less educational opportunity and are more likely to experience violence (e.g., Haas et al., 2012). With regard to gender, the economic benefits of education may be lower for women than for men (Luo & Waite, 2005); therefore, if education moderates lifetime effects of childhood adversity on health then women may be less likely to benefit from that moderating effect. However, little research has examined variation in the association of childhood adversity with adult health by gender or race. Researchers have either not found significant differences (Luo & Waite, 2005; Smith et al., 2016) or that differences did not persist in adjusted results (Haas et al., 2012). One recent study examined associations of childhood adversity with relationship strain by race, and how that strain may affect adult health (Umberson, Williams, Thomas, Liu, & Thomeer, 2014). The researchers concluded that relationship strain associated with childhood adversity had a greater negative effect on adult health for African American men than for white men. Study Contributions and Hypotheses We examined the association of childhood adversity with work disability, using longitudinal data that were nationally representative of African Americans and non-Hispanic whites (hereafter whites) ages 20 and over living in the community. Consistent with the expectation that risks to health can accumulate in childhood, our first hypothesis was that people reporting high levels of childhood adversity would have significantly more work disability throughout adult life than people reporting none of the adversities. To contribute to knowledge about the pathway model, we examined the association of childhood adversity with adult diabetes, heart disease, obesity, and sedentary behavior, all of which are established causes of functional impairment (Andrade, 2010; Laditka & Laditka, 2015; Laditka & Laditka, 2017). We also examined how those factors may mediate the association of adversity with work disability. Although health behaviors are likely to be powerful links between early life adversity and adult health, other pathways may include education, biological factors, and mental health. Adversity may also affect adult health directly, through a pathway that is not mediated by such factors, or through other mediators that are not currently known. Consistent with these possibilities, our second hypothesis was that childhood adversity would be associated with work disability even after controlling for adult diabetes, heart disease, obesity, and sedentary behavior. We also examined educational and biological pathways. It is well established that African Americans have more functional impairment than whites (e.g., Laditka & Laditka, 2009; Montez & Hayward, 2014), more morbidity (e.g., Laditka & Laditka, 2009), and greater exposure to social and economic disparities throughout life (e.g., Haas et al., 2012). Little research has specifically studied how the association of childhood adversity with adult health may vary with race. However, given recent evidence of such variation (Umberson et al., 2014), we hypothesized that childhood adversity would be associated with more work disability among African Americans than among whites. Methods Data We used data from the Panel Study of Income Dynamics (PSID), the longest running household panel study in the world, and the 2014 PSID Childhood Retrospective Circumstances Study (CRCS). The PSID surveyed participants every year from 1968 through 1997, and thereafter every 2 years, typically with response rates from 96% to 98% (Schoeni, Stafford, McGonagle, & Andreski, 2013). The CRCS attempted to interview all PSID household heads and their spouses or partners, who represent adults living in the community in the United States. We included CRCS respondents who also provided information about work disability in at least two PSID survey waves. We excluded participants other than African Americans and whites due to their small numbers. Dependent Variable—Measuring Work Disability Beginning in 1968, the PSID asked household heads, “Do you have any physical or nervous condition that limits the type of work or the amount of work you can do?” The PSID then asked those who responded “yes,” “Does this condition keep you from doing some types of work?” Available responses were, “yes,” “no,” or “can do nothing.” We considered participants to have a severe work disability if they could “do nothing.” Those who responded “yes” were asked if the condition limited work “a lot,” “somewhat,” “just a little,” or “not at all.” We also considered those who said “a lot” to have a severe work disability, and those who said “somewhat” to have a moderate work disability. Related questions in 1969, 1970, and 1971 did not measure severity; we excluded those years. In some years before 1980 the PSID asked only new participants about work disability; we used their responses but did not use responses that were simply repeated from previous waves. The PSID asked about work disability for spouses and partners in 1976, 1978, and in all waves beginning with 1981; we used all of that data. Childhood Adversities and Other Measures We measured adversities using an index of nine childhood measures of: health status (e.g., Turner et al., 2016), family characteristics (e.g., Haas, 2007), socioeconomic status (e.g., Montez & Hayward, 2014), and victimization (e.g., Wolke, Copeland, Angold, & Costello, 2013). Adverse health was represented by self-reports of fair or poor childhood health, compared with good, very good, or excellent. Family adversities were: parents divorced when the participant was less than age 17, or the participant was raised by a single parent. Socioeconomic adversities were: one or both parents with education no more than grade 8; a parent experienced unemployment; financial “struggle”; neighborhood unsafe at night; and neighbors either not “close knit” or not reliable if help was needed. Participant reports of having been bullied “a lot” or “sometimes” in or out of school indicated victimization. Preliminary analyses associated each of these circumstances with more work disability throughout life. Few participants reported more than four adversities; we used a single categorical variable representing 0, 1, 2, 3, or 4 or more. We refer to four or more as high childhood adversity. We also examined the risks associated with childhood adversity of having diabetes, heart disease, obesity, or sedentary behavior at ages 40 and 65. The PSID asked, “Has a doctor or other health professional ever told you that you had diabetes or high blood sugar?” An analogous question asked about heart disease. Beginning in 1999, participants were asked how long they had the disease, or in some waves the year or age when diagnosed. We used this information to identify diabetes and heart disease status in previous years. The PSID asked participants their height and weight in 1986 and in all survey waves beginning with 1999; we used body mass index guidelines from the Centers for Disease Control and Prevention to measure obesity. The PSID asked participants how often they had light, moderate, or vigorous physical activity, in 1986 and in all waves beginning with 1999. We considered those who responded “never” to all three questions to be sedentary. We used the information about obesity and sedentary behavior that most closely corresponded to the year that began each measured work disability transition. We controlled for adult health and socioeconomic status with a three-level measure of education: less than high school, high school graduation (including the General Educational Development credential), or postsecondary education. We controlled for age in years, age-squared, and sex. A dummy variable indicated whether observations represented African Americans or whites. Interaction terms provided separate probabilities for each combination of childhood adversity, sex, and race. Analytical Approach We used t tests and standard logistic regression to compare PSID participants who completed the CRCS and those who did not. Our principal methods were a multinomial logistic Markov chain regression model, estimated by maximum likelihood, and dynamic microsimulation (e.g., Andrade, 2010; Laditka, 1998; Laditka & Laditka, 2009, 2014b, 2015; Laditka & Laditka, 2014b, 2016a, 2016b, 2018; Laditka & Wolf, 1998; Yong, & Saito, 2012). The model estimated work disability transition probabilities specific to each age beginning at 20, adjusting the probability of each transition type for the time between interviews and accommodating any pattern or number of unrecorded work disability transitions between interviews (e.g., Laditka & Laditka, 2018; Laditka & Wolf, 1998). We first estimated a separate model for each childhood adversity, adjusted for age, sex, and race. In the principal model, the probability of a given transition at a given age was conditional on the value of the adversity index, current work disability status, age, education, race, and sex. Another model estimated transition probabilities beginning at age 40, when adult chronic diseases become prevalent, adding adjustments for diabetes, heart disease, sedentary behavior, and obesity. To provide adequate statistical power, this second model did not stratify by race. With the probabilities estimated by the Markov models, we used microsimulation to create large simulated populations in which each individual had a complete annual history of work disability: none, moderate, or severe. Details of the method are published (Laditka, 1998; Laditka & Laditka, 2009, 2014a; Laditka & Wolf, 1998). All participants responded to the CRCS, so we could not estimate mortality risks. We simulated each individual life through death using mortality data from the National Center for Health Statistics. In a microsimulation for each population, such as African American women with high childhood adversity, we created 100,000 lives. For both moderate and severe work disability, we calculated the population prevalence at each age, the average onset age of the first spell, the number of spells for the average individual, and the percentage of life at ages 30–65 with work disability. Controlling for education may adjust for a pathway linking childhood adversity with adult health. We therefore also estimated the model without a control for education. To contribute to knowledge about the pathway model, we first examined associations of childhood adversity with diabetes, heart disease, obesity, or sedentary behavior at ages 40 and 65. For each outcome, we estimated an unadjusted standard logistic model, then adjusted for education. Given the expectation that poor childhood health would be associated with those outcomes, and to control for a biological pathway, we also estimated models representing only participants reporting good to excellent childhood health; all other models included childhood health in the adversity index. We then conducted microsimulations in which all individuals separately had diabetes, heart disease, sedentary behavior, or obesity beginning at age 40, or none of those conditions, or all of them, focused on results for childhood adversity. We analyzed the consistency across survey waves of reports of the ages when participants said they were diagnosed with diabetes or heart disease. When a participant responded inconsistently, we used the most common response. When a household head and the head’s spouse or partner both reported onset ages for the spouse or partner, which can occur when the head provides this information and then dies or is divorced, we used the responses from the spouse or partner for the analytic data for that individual. We examined the reliability of the onset age responses by calculating the intra-class correlation coefficient (ICC) using the INTRACC SAS macro (Hamer, 2007). An ICC ≥ 0.70 indicates adequate reliability (Special Advisory Committee of the Medical Outcomes Trust, 2002). As the PSID follows families for decades, it is useful to examine whether characteristics shared within families might be associated with childhood adversity and work disability, and therefore might affect the results. To examine this issue, we estimated a multinomial logistic hazard model predicting the three levels of work disability with both fixed-effects and random-effects (Allison, 2005). Fixed effects controlled for all characteristics shared within families; random effects did not. We tested whether these effects differed, which would indicate that the association of childhood adversity with work disability varied systematically among families. We estimated variation in the results with bootstrapping, repeating each microsimulation 1,000 times, each with parameters randomly selected from their 95% confidence intervals (CIs). The point estimates for the outcomes were the means; the CIs ranged from the 2.5 to the 97.5 percentiles. We created the software for estimating the Markov models and conducting the microsimulations using SAS IML (Cary, NC). For all other analyses, we used SAS 9.4 (Cary, NC). The Institutional Review Board (IRB) at the University of North Carolina at Charlotte determined that this research did not require IRB review. Results Sample Characteristics The CRCS response rate was 67%; 6,045 participants fulfilled the inclusion criteria. Participants and nonparticipants did not differ significantly in age, sex, race, or reports of work disability. Participants had 82,374 recorded work disability status transitions through 129,107 person-years, each participant averaging 13.6 (SD 10.1) work disability measurements (data not shown). The mean baseline age was 29.5 years (SD 8.2 years, not shown), where the baseline was the first recorded work disability status for each individual; weighted for national representativeness and accounting for the survey design the mean baseline age was 32.8 years (CI 32.2–33.4). The mean age when participants responded to the CRCS was 48.7 (SD 15.0; 50.5 years weighted, CI 49.9–51.1). Women were 57.8% of the sample (54.1 weighted, CI 50.7–57.5). The PSID oversampled African Americans, 29.9% of the sample (9.1 weighted, CI 6.7–11.6). As for childhood adversities, 30.4% reported none, 28.9% one, 21.2% two, 11.9% three, and 7.7% four or more. Table 1 describes the sample. White women and men had the smallest percentages with high adversity, 6.1% and 5.7%, respectively; comparable results were 13.2% and 9.3% for African Americans (all p < .0001). Not shown, obesity and sedentary behavior were measured in the same year as work disability in 60.1% of the analytic observations; the average period between the obesity and sedentary measurements and the work disability measurement was 1.8 years (SD 1.0). Also not shown, the sample used to test our second hypothesis (ages 40 and over with complete information on diabetes, heart disease, obesity, and sedentary behavior) represented 4,135 participants with 28,084 functional status transitions. Other characteristics of the second sample were similar to those reported in Table 1. Table 1. Associations of Childhood Adversity With Work Disability, Characteristics of the Sample, by Sex and Racea   Women  Men    African American  White  African American  White  Sample size, nb  1,175  2,320  634  1,916  Work disability status transitions, nb  14,022  31,072  7,304  29,976  Person-years, nb  22,867  50,467  11,709  44,064  Age at baseline, mean (CI)c  31.2 (30.2–32.3)  32.2 (31.6–32.8)  35.9 (33.9–38.0)  33.4 (32.5–34.4)  Education years, mean (CI)c  12.8 (12.0–13.6)  13.6 (13.3–13.9)  12.7 (12.3–13.1)  13.5 (13.2–13.8)  Adversity index (0–4), mean (CI)c  1.7 (1.5–1.9)  1.3 (1.2–1.4)  1.6 (1.3–1.8)  1.2 (1.1–1.3)   Index=0, %d  20.0  33.2  22.9  35.1   Index=1, %d  27.1  29.1  27.9  30.0   Index=2, %d  22.5  21.0  23.8  19.7   Index=3, %d  17.3  10.5  16.1  8.9   Index=4+, %d  13.2  6.1  9.3  5.7    Women  Men    African American  White  African American  White  Sample size, nb  1,175  2,320  634  1,916  Work disability status transitions, nb  14,022  31,072  7,304  29,976  Person-years, nb  22,867  50,467  11,709  44,064  Age at baseline, mean (CI)c  31.2 (30.2–32.3)  32.2 (31.6–32.8)  35.9 (33.9–38.0)  33.4 (32.5–34.4)  Education years, mean (CI)c  12.8 (12.0–13.6)  13.6 (13.3–13.9)  12.7 (12.3–13.1)  13.5 (13.2–13.8)  Adversity index (0–4), mean (CI)c  1.7 (1.5–1.9)  1.3 (1.2–1.4)  1.6 (1.3–1.8)  1.2 (1.1–1.3)   Index=0, %d  20.0  33.2  22.9  35.1   Index=1, %d  27.1  29.1  27.9  30.0   Index=2, %d  22.5  21.0  23.8  19.7   Index=3, %d  17.3  10.5  16.1  8.9   Index=4+, %d  13.2  6.1  9.3  5.7  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study, n = 6,045. bUnweighted sample characteristics. cResults weighted to be nationally representative and accounting for the survey design; CI = 95% confidence interval. dIndex level comparisons across race, p < .0001 for both women and men. View Large Table 1. Associations of Childhood Adversity With Work Disability, Characteristics of the Sample, by Sex and Racea   Women  Men    African American  White  African American  White  Sample size, nb  1,175  2,320  634  1,916  Work disability status transitions, nb  14,022  31,072  7,304  29,976  Person-years, nb  22,867  50,467  11,709  44,064  Age at baseline, mean (CI)c  31.2 (30.2–32.3)  32.2 (31.6–32.8)  35.9 (33.9–38.0)  33.4 (32.5–34.4)  Education years, mean (CI)c  12.8 (12.0–13.6)  13.6 (13.3–13.9)  12.7 (12.3–13.1)  13.5 (13.2–13.8)  Adversity index (0–4), mean (CI)c  1.7 (1.5–1.9)  1.3 (1.2–1.4)  1.6 (1.3–1.8)  1.2 (1.1–1.3)   Index=0, %d  20.0  33.2  22.9  35.1   Index=1, %d  27.1  29.1  27.9  30.0   Index=2, %d  22.5  21.0  23.8  19.7   Index=3, %d  17.3  10.5  16.1  8.9   Index=4+, %d  13.2  6.1  9.3  5.7    Women  Men    African American  White  African American  White  Sample size, nb  1,175  2,320  634  1,916  Work disability status transitions, nb  14,022  31,072  7,304  29,976  Person-years, nb  22,867  50,467  11,709  44,064  Age at baseline, mean (CI)c  31.2 (30.2–32.3)  32.2 (31.6–32.8)  35.9 (33.9–38.0)  33.4 (32.5–34.4)  Education years, mean (CI)c  12.8 (12.0–13.6)  13.6 (13.3–13.9)  12.7 (12.3–13.1)  13.5 (13.2–13.8)  Adversity index (0–4), mean (CI)c  1.7 (1.5–1.9)  1.3 (1.2–1.4)  1.6 (1.3–1.8)  1.2 (1.1–1.3)   Index=0, %d  20.0  33.2  22.9  35.1   Index=1, %d  27.1  29.1  27.9  30.0   Index=2, %d  22.5  21.0  23.8  19.7   Index=3, %d  17.3  10.5  16.1  8.9   Index=4+, %d  13.2  6.1  9.3  5.7  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study, n = 6,045. bUnweighted sample characteristics. cResults weighted to be nationally representative and accounting for the survey design; CI = 95% confidence interval. dIndex level comparisons across race, p < .0001 for both women and men. View Large Transition Probabilities from the Markov Model In results not shown in a table, all of the Markov model parameter estimates associated childhood adversity with a higher a likelihood of becoming disabled, or more severely disabled, and a lower likelihood of recovering from disability; five of the six parameters were statistically significant (p < .001). The parameters governed the microsimulations. Work Disability in Working Life Figure 1 shows the increase in the percentage of years at ages 30–65 with work disability, separately for each of the nine adversities, each compared to no adversities (all p < .01). Combining results for moderate and severe work disability, as shown in Figure 1, the adversities associated with the largest differences in work disability were bullying, being raised by a single parent, and having a parent with education no more than grade 8. Not shown are comparable separate results by gender and race, in which the largest differences in work disability were associated with: fair/poor childhood health, neighborhood characteristics, and bullying for African American women and men; bullying, parent education, and childhood health for white women; and being raised by a single parent, neighborhood safety, and bullying for white men. Figure 1. View largeDownload slide Increase in the percentage of years ages 30–65 with moderate and severe work disability for individuals reporting four or more childhood adversities, compared to individuals reporting none of the adversities. Data source: Panel Study of Income Dynamics. Childhood adversity codes: 1=bullied; 2=single parent; 3=parent education ≤ grade 8; 4=fair or poor childhood health; 5=unsafe neighborhood; 6=neighbors not reliable for help; 7=parents divorced; 8=parent unemployment; 9=financial struggle. Figure 1. View largeDownload slide Increase in the percentage of years ages 30–65 with moderate and severe work disability for individuals reporting four or more childhood adversities, compared to individuals reporting none of the adversities. Data source: Panel Study of Income Dynamics. Childhood adversity codes: 1=bullied; 2=single parent; 3=parent education ≤ grade 8; 4=fair or poor childhood health; 5=unsafe neighborhood; 6=neighbors not reliable for help; 7=parents divorced; 8=parent unemployment; 9=financial struggle. Table 2 shows the principal microsimulation results. We controlled for education through stratified analysis; the results shown represent individuals with high school education. Women and men with high adversity had significantly more moderate and severe work disability than others (all p < .01). For example, among African American women at age 30, of those with high childhood adversity 10.2% reported moderate work disability and 5.6% reported severe. Comparable results for no adversities were 4.1% and 1.9%. Table 2. Association of Childhood Adversity With Adult Work Disabilitya   Women  Men    African American  White  African American  White  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  Moderate work disability   Age 1st spell, mean  40.6  34.2*  40.6  32.1*  42.1  38.6*  42.1  37.1*   Spells, n, mean  2.1  4.2*  2.0  4.9*  1.6  2.7*  1.5  2.9*   Population %, age 30  4.1  10.2*  4.7  11.3*  3.3  6.1*  3.7  7.2*   Population %, age 65  12.1  23.0*  16.0  28.9*  10.1  16.3*  11.5  18.9*   % of life ages 30–65  8.8  16.7*  10.2  19.9*  8.0  11.7*  9.4  13.3*  Severe work disability   Age 1st spell, mean  45.7  40.6*  47.5  41.6*  46.7  43.1*  47.9  46.6*   Spells, n, mean  1.3  2.8*  0.7  2.5*  1.0  2.0*  0.6  1.0*   Population %, age 30  1.9  5.6*  1.1  3.5*  1.8  4.1*  1.0  2.0*   Population %, age 65  10.9  21.4*  8.4  17.8*  10.7  19.0*  6.2  11.5*   % of life ages 30–65  5.4  12.5*  3.2  10.1*  5.2  10.3*  3.0  5.8*    Women  Men    African American  White  African American  White  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  Moderate work disability   Age 1st spell, mean  40.6  34.2*  40.6  32.1*  42.1  38.6*  42.1  37.1*   Spells, n, mean  2.1  4.2*  2.0  4.9*  1.6  2.7*  1.5  2.9*   Population %, age 30  4.1  10.2*  4.7  11.3*  3.3  6.1*  3.7  7.2*   Population %, age 65  12.1  23.0*  16.0  28.9*  10.1  16.3*  11.5  18.9*   % of life ages 30–65  8.8  16.7*  10.2  19.9*  8.0  11.7*  9.4  13.3*  Severe work disability   Age 1st spell, mean  45.7  40.6*  47.5  41.6*  46.7  43.1*  47.9  46.6*   Spells, n, mean  1.3  2.8*  0.7  2.5*  1.0  2.0*  0.6  1.0*   Population %, age 30  1.9  5.6*  1.1  3.5*  1.8  4.1*  1.0  2.0*   Population %, age 65  10.9  21.4*  8.4  17.8*  10.7  19.0*  6.2  11.5*   % of life ages 30–65  5.4  12.5*  3.2  10.1*  5.2  10.3*  3.0  5.8*  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study, n = 6,045. *p < .01. View Large Table 2. Association of Childhood Adversity With Adult Work Disabilitya   Women  Men    African American  White  African American  White  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  Moderate work disability   Age 1st spell, mean  40.6  34.2*  40.6  32.1*  42.1  38.6*  42.1  37.1*   Spells, n, mean  2.1  4.2*  2.0  4.9*  1.6  2.7*  1.5  2.9*   Population %, age 30  4.1  10.2*  4.7  11.3*  3.3  6.1*  3.7  7.2*   Population %, age 65  12.1  23.0*  16.0  28.9*  10.1  16.3*  11.5  18.9*   % of life ages 30–65  8.8  16.7*  10.2  19.9*  8.0  11.7*  9.4  13.3*  Severe work disability   Age 1st spell, mean  45.7  40.6*  47.5  41.6*  46.7  43.1*  47.9  46.6*   Spells, n, mean  1.3  2.8*  0.7  2.5*  1.0  2.0*  0.6  1.0*   Population %, age 30  1.9  5.6*  1.1  3.5*  1.8  4.1*  1.0  2.0*   Population %, age 65  10.9  21.4*  8.4  17.8*  10.7  19.0*  6.2  11.5*   % of life ages 30–65  5.4  12.5*  3.2  10.1*  5.2  10.3*  3.0  5.8*    Women  Men    African American  White  African American  White  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  Moderate work disability   Age 1st spell, mean  40.6  34.2*  40.6  32.1*  42.1  38.6*  42.1  37.1*   Spells, n, mean  2.1  4.2*  2.0  4.9*  1.6  2.7*  1.5  2.9*   Population %, age 30  4.1  10.2*  4.7  11.3*  3.3  6.1*  3.7  7.2*   Population %, age 65  12.1  23.0*  16.0  28.9*  10.1  16.3*  11.5  18.9*   % of life ages 30–65  8.8  16.7*  10.2  19.9*  8.0  11.7*  9.4  13.3*  Severe work disability   Age 1st spell, mean  45.7  40.6*  47.5  41.6*  46.7  43.1*  47.9  46.6*   Spells, n, mean  1.3  2.8*  0.7  2.5*  1.0  2.0*  0.6  1.0*   Population %, age 30  1.9  5.6*  1.1  3.5*  1.8  4.1*  1.0  2.0*   Population %, age 65  10.9  21.4*  8.4  17.8*  10.7  19.0*  6.2  11.5*   % of life ages 30–65  5.4  12.5*  3.2  10.1*  5.2  10.3*  3.0  5.8*  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study, n = 6,045. *p < .01. View Large Table 2 also shows analogous results for the percentage of years from ages 30 through 65 with moderate and severe work disability. For all groups, the percentages were significantly larger with high adversity than with none (all p < .01). For example, white women with high adversity could expect moderate work disability for 19.9% of those years, severe for 10.1%; comparable results for white women reporting no adversities were 10.2% with moderate work disability and 3.2% with severe. Women and men with high adversity also had work disability at younger ages than those reporting no adversities, and more spells (all p < .01). African American women and men were more likely to have a severe work disability at ages 30 and 65 than whites. However, there was no evidence of notable differences between African Americans and whites in the association of childhood adversity with work disability. Women were more likely to have moderate and severe work disability than men, and also had more spells of both moderate and severe work disability. There was modest evidence that associations of adversity with work disability may have been larger for women than for men; for example among whites at age 65, the prevalence of moderate work disability was 80.6% greater with high adversity than with no adversity for women, 64.3% greater for men (percent differences not shown). Results of the model that did not control for education also associated adversity with significantly more work disability (results not shown), although the size of the effect was substantially smaller. For example, in that model white women with high adversity could expect moderate work disability for 9.7% of the years from ages 30–65, severe for 3.0%, compared with 9.1% and 2.7% with no adversities (both p < .05). Associations of Childhood Adversity with Chronic Conditions and Sedentary Behavior Table 3 shows odds ratios estimating the association of childhood adversity with diabetes, heart disease, obesity, and sedentary behavior at ages 40 and 65 for three models: unadjusted, adjusted for education, and adjusted for education and childhood health. In all models, the results suggest that high adversity was associated with greater risks of having diabetes or being obese at ages 40 and 65, and greater risk of having heart disease at age 65. For example, in the unadjusted model, compared with participants reporting no adversities, the odds of having diabetes at age 40 were twice as great with high adversity (p < .001). The bottom row of Table 3 shows analogous associations of childhood adversity with having at least one of the four conditions at ages 40 and 65, all of which were significant. In all models, participants reporting high adversity had at least 80% higher odds of having one or more of the conditions at age 65 than those reporting none. Table 3. Odds Ratios Estimating the Risk of Having Diabetes, Heart Disease, Obesity, or Sedentary Behavior at Ages 40 and 65, Comparing People With High Childhood Adversity to Those With Low Childhood Adversitya   Unadjusted  Adjusted for years of education  Participants reporting good to excellent childhood health, adjusted for years of education    Age 40  Age 65  Age 40  Age 65  Age 40  Age 65  Diabetes  2.0***  1.6*  1.9**  1.6*  1.8**  1.4  Heart Disease  1.5  1.8*  1.4  1.8*  1.5  1.8*  Obesity  1.5**  1.7*  1.4*  1.6*  1.4*  1.5  Sedentary  1.3  1.3  1.2  1.1  1.1  1.1  Any condition  1.5***  1.9*  1.4**  1.8*  1.3*  1.8*    Unadjusted  Adjusted for years of education  Participants reporting good to excellent childhood health, adjusted for years of education    Age 40  Age 65  Age 40  Age 65  Age 40  Age 65  Diabetes  2.0***  1.6*  1.9**  1.6*  1.8**  1.4  Heart Disease  1.5  1.8*  1.4  1.8*  1.5  1.8*  Obesity  1.5**  1.7*  1.4*  1.6*  1.4*  1.5  Sedentary  1.3  1.3  1.2  1.1  1.1  1.1  Any condition  1.5***  1.9*  1.4**  1.8*  1.3*  1.8*  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study; low childhood adversity = values 0 and 1 on a five-point adversity scale (values 0 through 4); high childhood adversity = values 3 and 4 on the same scale; Any condition = any of the other four listed conditions. *p < .05, **p < .01, ***p < .001. View Large Table 3. Odds Ratios Estimating the Risk of Having Diabetes, Heart Disease, Obesity, or Sedentary Behavior at Ages 40 and 65, Comparing People With High Childhood Adversity to Those With Low Childhood Adversitya   Unadjusted  Adjusted for years of education  Participants reporting good to excellent childhood health, adjusted for years of education    Age 40  Age 65  Age 40  Age 65  Age 40  Age 65  Diabetes  2.0***  1.6*  1.9**  1.6*  1.8**  1.4  Heart Disease  1.5  1.8*  1.4  1.8*  1.5  1.8*  Obesity  1.5**  1.7*  1.4*  1.6*  1.4*  1.5  Sedentary  1.3  1.3  1.2  1.1  1.1  1.1  Any condition  1.5***  1.9*  1.4**  1.8*  1.3*  1.8*    Unadjusted  Adjusted for years of education  Participants reporting good to excellent childhood health, adjusted for years of education    Age 40  Age 65  Age 40  Age 65  Age 40  Age 65  Diabetes  2.0***  1.6*  1.9**  1.6*  1.8**  1.4  Heart Disease  1.5  1.8*  1.4  1.8*  1.5  1.8*  Obesity  1.5**  1.7*  1.4*  1.6*  1.4*  1.5  Sedentary  1.3  1.3  1.2  1.1  1.1  1.1  Any condition  1.5***  1.9*  1.4**  1.8*  1.3*  1.8*  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study; low childhood adversity = values 0 and 1 on a five-point adversity scale (values 0 through 4); high childhood adversity = values 3 and 4 on the same scale; Any condition = any of the other four listed conditions. *p < .05, **p < .01, ***p < .001. View Large Work Disability With and Without Chronic Conditions and Sedentary Behavior Table 4 shows results comparing high adversity to none, for women and men with and without adult diabetes, heart disease, obesity, or sedentary behavior, and for those with all four conditions, ages 40 through 65. For example, at age 65, among women with none of the four conditions 8.4% of those reporting no adversity had moderate work disability, compared with 14.2% with high adversity (p < .01). In analogous results for severe work disability, among women with none of the four conditions 2.3% of those reporting no adversity had severe work disability, compared with 6.7% with high adversity (p < .01). Most results in Table 4 also show that individuals with high adversity: experienced their first moderate or severe work disability at significantly younger ages than those with no reported adversities; had significantly more spells of moderate and severe work disability; and had moderate and severe work disability for significantly greater proportions of life from ages 40 through 65. Table 4. Association of Childhood Adversity, and Adult Diabetes, Heart Disease, Obesity, and Sedentary Behavior, With Work Disability, Ages 40–65a   No adult conditionsb  Diabetes  Heart disease  Obesity  Sedentary  All adult conditions  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  0  4+  0  4+  Women                          Moderate work disability                           Age 1st spell, mean  59.2  54.7*  57.7  52.8*  55.8  51.0*  55.7  50.9*  53.8  49.3*  46.5  44.8*   Spells, n, mean  2.2  3.7*  2.6  4.4*  3.1  5.0*  3.3  5.3*  4.0  6.1*  6.8  8.0   Population %, age 65  8.4  14.2*  10.7  17.2*  14.7  22.5*  13.7  20.5*  17.1  24.5*  37.0  37.0   % of life, ages 40–65  6.5  11.2*  8.3  13.8*  11.6  18.5*  10.8  16.9*  13.6  20.5*  33.6  36.3  Severe work disability   Age 1st spell, mean  63.9  61.0*  63.0  58.7*  62.7  58.7*  61.4  55.7*  61.4  55.6*  51.7  46.9*   Spells, n, mean  0.7  1.7*  1.0  2.3*  1.0  2.2*  1.5  3.3*  1.6  3.3*  4.0  5.9*   Population %, age 65  2.3  6.7*  3.7  9.9*  3.2  8.2*  6.6  16.5*  6.3  15.4*  21.8  36.1*   % of life, ages 40–65  1.8  5.0*  2.8  7.7*  2.5  6.5*  5.0  13.0*  4.7  12.1*  18.7  33.0*  Men  Moderate work disability   Age 1st spell, mean  60.6  56.5*  59.2  54.8*  57.5  52.8*  57.6  53.0*  55.7  51.1*  48.0  46.3*   Spells, n, mean  1.7  3.0*  2.1  3.5*  2.5  4.0*  2.6  4.1*  3.2  4.8*  5.5  6.2   Population %, age 65  7.1  12.1*  9.1  14.7*  12.8  19.9*  11.6  17.4*  14.7  21.3*  33.3  32.4   % of life ages 40–65  5.5  9.4*  7.0  11.7*  9.9  16.1*  9.0  14.1*  11.6  17.6*  29.7  31.7  Severe work disability   Age 1st spell, mean  64.1  62.4*  63.6  60.6*  63.4  60.5*  62.5  58.0*  62.7  58.0*  53.8  48.5*   Spells, n, mean  0.5  1.2*  0.7  1.6*  0.7  1.5*  1.1  2.3*  1.1  2.3*  2.9  4.2*   Population %, age 65  2.1  6.4*  3.5  10.1*  3.0  8.1*  6.7  17.6*  6.1  16.1*  24.8  41.8 +   % of life ages 40–65  1.7  4.9*  2.7  7.7*  2.3  6.3*  5.1  13.5*  4.6  12.3*  20.7  37.2*    No adult conditionsb  Diabetes  Heart disease  Obesity  Sedentary  All adult conditions  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  0  4+  0  4+  Women                          Moderate work disability                           Age 1st spell, mean  59.2  54.7*  57.7  52.8*  55.8  51.0*  55.7  50.9*  53.8  49.3*  46.5  44.8*   Spells, n, mean  2.2  3.7*  2.6  4.4*  3.1  5.0*  3.3  5.3*  4.0  6.1*  6.8  8.0   Population %, age 65  8.4  14.2*  10.7  17.2*  14.7  22.5*  13.7  20.5*  17.1  24.5*  37.0  37.0   % of life, ages 40–65  6.5  11.2*  8.3  13.8*  11.6  18.5*  10.8  16.9*  13.6  20.5*  33.6  36.3  Severe work disability   Age 1st spell, mean  63.9  61.0*  63.0  58.7*  62.7  58.7*  61.4  55.7*  61.4  55.6*  51.7  46.9*   Spells, n, mean  0.7  1.7*  1.0  2.3*  1.0  2.2*  1.5  3.3*  1.6  3.3*  4.0  5.9*   Population %, age 65  2.3  6.7*  3.7  9.9*  3.2  8.2*  6.6  16.5*  6.3  15.4*  21.8  36.1*   % of life, ages 40–65  1.8  5.0*  2.8  7.7*  2.5  6.5*  5.0  13.0*  4.7  12.1*  18.7  33.0*  Men  Moderate work disability   Age 1st spell, mean  60.6  56.5*  59.2  54.8*  57.5  52.8*  57.6  53.0*  55.7  51.1*  48.0  46.3*   Spells, n, mean  1.7  3.0*  2.1  3.5*  2.5  4.0*  2.6  4.1*  3.2  4.8*  5.5  6.2   Population %, age 65  7.1  12.1*  9.1  14.7*  12.8  19.9*  11.6  17.4*  14.7  21.3*  33.3  32.4   % of life ages 40–65  5.5  9.4*  7.0  11.7*  9.9  16.1*  9.0  14.1*  11.6  17.6*  29.7  31.7  Severe work disability   Age 1st spell, mean  64.1  62.4*  63.6  60.6*  63.4  60.5*  62.5  58.0*  62.7  58.0*  53.8  48.5*   Spells, n, mean  0.5  1.2*  0.7  1.6*  0.7  1.5*  1.1  2.3*  1.1  2.3*  2.9  4.2*   Population %, age 65  2.1  6.4*  3.5  10.1*  3.0  8.1*  6.7  17.6*  6.1  16.1*  24.8  41.8 +   % of life ages 40–65  1.7  4.9*  2.7  7.7*  2.3  6.3*  5.1  13.5*  4.6  12.3*  20.7  37.2*  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study, n = 4,135, 28,084 functional status transitions. +p < .05, *p < .01. View Large Table 4. Association of Childhood Adversity, and Adult Diabetes, Heart Disease, Obesity, and Sedentary Behavior, With Work Disability, Ages 40–65a   No adult conditionsb  Diabetes  Heart disease  Obesity  Sedentary  All adult conditions  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  0  4+  0  4+  Women                          Moderate work disability                           Age 1st spell, mean  59.2  54.7*  57.7  52.8*  55.8  51.0*  55.7  50.9*  53.8  49.3*  46.5  44.8*   Spells, n, mean  2.2  3.7*  2.6  4.4*  3.1  5.0*  3.3  5.3*  4.0  6.1*  6.8  8.0   Population %, age 65  8.4  14.2*  10.7  17.2*  14.7  22.5*  13.7  20.5*  17.1  24.5*  37.0  37.0   % of life, ages 40–65  6.5  11.2*  8.3  13.8*  11.6  18.5*  10.8  16.9*  13.6  20.5*  33.6  36.3  Severe work disability   Age 1st spell, mean  63.9  61.0*  63.0  58.7*  62.7  58.7*  61.4  55.7*  61.4  55.6*  51.7  46.9*   Spells, n, mean  0.7  1.7*  1.0  2.3*  1.0  2.2*  1.5  3.3*  1.6  3.3*  4.0  5.9*   Population %, age 65  2.3  6.7*  3.7  9.9*  3.2  8.2*  6.6  16.5*  6.3  15.4*  21.8  36.1*   % of life, ages 40–65  1.8  5.0*  2.8  7.7*  2.5  6.5*  5.0  13.0*  4.7  12.1*  18.7  33.0*  Men  Moderate work disability   Age 1st spell, mean  60.6  56.5*  59.2  54.8*  57.5  52.8*  57.6  53.0*  55.7  51.1*  48.0  46.3*   Spells, n, mean  1.7  3.0*  2.1  3.5*  2.5  4.0*  2.6  4.1*  3.2  4.8*  5.5  6.2   Population %, age 65  7.1  12.1*  9.1  14.7*  12.8  19.9*  11.6  17.4*  14.7  21.3*  33.3  32.4   % of life ages 40–65  5.5  9.4*  7.0  11.7*  9.9  16.1*  9.0  14.1*  11.6  17.6*  29.7  31.7  Severe work disability   Age 1st spell, mean  64.1  62.4*  63.6  60.6*  63.4  60.5*  62.5  58.0*  62.7  58.0*  53.8  48.5*   Spells, n, mean  0.5  1.2*  0.7  1.6*  0.7  1.5*  1.1  2.3*  1.1  2.3*  2.9  4.2*   Population %, age 65  2.1  6.4*  3.5  10.1*  3.0  8.1*  6.7  17.6*  6.1  16.1*  24.8  41.8 +   % of life ages 40–65  1.7  4.9*  2.7  7.7*  2.3  6.3*  5.1  13.5*  4.6  12.3*  20.7  37.2*    No adult conditionsb  Diabetes  Heart disease  Obesity  Sedentary  All adult conditions  Childhood Adversities, n:  0  4+  0  4+  0  4+  0  4+  0  4+  0  4+  Women                          Moderate work disability                           Age 1st spell, mean  59.2  54.7*  57.7  52.8*  55.8  51.0*  55.7  50.9*  53.8  49.3*  46.5  44.8*   Spells, n, mean  2.2  3.7*  2.6  4.4*  3.1  5.0*  3.3  5.3*  4.0  6.1*  6.8  8.0   Population %, age 65  8.4  14.2*  10.7  17.2*  14.7  22.5*  13.7  20.5*  17.1  24.5*  37.0  37.0   % of life, ages 40–65  6.5  11.2*  8.3  13.8*  11.6  18.5*  10.8  16.9*  13.6  20.5*  33.6  36.3  Severe work disability   Age 1st spell, mean  63.9  61.0*  63.0  58.7*  62.7  58.7*  61.4  55.7*  61.4  55.6*  51.7  46.9*   Spells, n, mean  0.7  1.7*  1.0  2.3*  1.0  2.2*  1.5  3.3*  1.6  3.3*  4.0  5.9*   Population %, age 65  2.3  6.7*  3.7  9.9*  3.2  8.2*  6.6  16.5*  6.3  15.4*  21.8  36.1*   % of life, ages 40–65  1.8  5.0*  2.8  7.7*  2.5  6.5*  5.0  13.0*  4.7  12.1*  18.7  33.0*  Men  Moderate work disability   Age 1st spell, mean  60.6  56.5*  59.2  54.8*  57.5  52.8*  57.6  53.0*  55.7  51.1*  48.0  46.3*   Spells, n, mean  1.7  3.0*  2.1  3.5*  2.5  4.0*  2.6  4.1*  3.2  4.8*  5.5  6.2   Population %, age 65  7.1  12.1*  9.1  14.7*  12.8  19.9*  11.6  17.4*  14.7  21.3*  33.3  32.4   % of life ages 40–65  5.5  9.4*  7.0  11.7*  9.9  16.1*  9.0  14.1*  11.6  17.6*  29.7  31.7  Severe work disability   Age 1st spell, mean  64.1  62.4*  63.6  60.6*  63.4  60.5*  62.5  58.0*  62.7  58.0*  53.8  48.5*   Spells, n, mean  0.5  1.2*  0.7  1.6*  0.7  1.5*  1.1  2.3*  1.1  2.3*  2.9  4.2*   Population %, age 65  2.1  6.4*  3.5  10.1*  3.0  8.1*  6.7  17.6*  6.1  16.1*  24.8  41.8 +   % of life ages 40–65  1.7  4.9*  2.7  7.7*  2.3  6.3*  5.1  13.5*  4.6  12.3*  20.7  37.2*  Note: aData source: 1968–2013 Panel Study of Income Dynamics (PSID) with 2014 PSID Childhood Retrospective Circumstances Study, n = 4,135, 28,084 functional status transitions. +p < .05, *p < .01. View Large Comparisons of high adversity with none were statistically significant for women and men with none of the adult conditions, and with each of the individual conditions. In several instances, the comparisons were not significant for individuals with all four conditions; this result may be due to substantial disability risks for individuals with all of these conditions, regardless of their childhood adversity status. In all models, the association of childhood adversity with work disability was generally linear. Thus, for example, results for individuals who reported any two adversities were approximately mid-way between the results for high adversity and none. Diagnosis Reports and Family Effects Regarding reports of diabetes or heart disease diagnosis ages, for participants who reported diagnosis ages in 2, 3, 4, or 5 waves (2005–2013), respectively, the ICCs were: 0.99, 0.79, 0.78, and 0.74 for diabetes and 0.89, 0.84, 079, and 0.76 for heart disease (results not shown in tables). For those who reported diagnosis ages in five waves, where variation was greatest, the average deviation from each individual’s mean reported age was 3.3 years for diabetes (SD 4.2), 2.8 years for heart disease (SD 3.4). However, 92.5% of participants reported in two or fewer waves; 97.8% were within 1 year for diabetes, 48.7% for heart disease. Results were similar when the 1999, 2001, and 2003 reports of the time since diagnosis were included. In the analysis of family effects, the average participant had 2.6 other family members who also participated (data not shown). The comparison of the fixed and random effects approaches was not significant (p = .128). Thus, there was no evidence that family membership affected the results. Discussion We examined associations of childhood adversity with work disability throughout working life. Consistent with our first hypothesis, African American and white women and men with high adversity were more likely to have moderate and severe work disability throughout working life than those reporting no adversities. Although these results are consistent with studies linking childhood adversity with functional status (e.g., Laditka & Laditka, 2017; Montez & Hayward, 2014; Smith et al., 2016), only one previous study examined this association with work disability: Shuey and Willson (2017) found that poor health and poverty in childhood were associated with work disability. Our results underscore the burden of childhood adversity for individuals and society. Consistent with the pathway model (e.g., Pudrovska & Anikputa, 2014), adults with high childhood adversity were more likely than those with no adversity to have the four causes of functional impairment that we studied. The results were also consistent with our second hypothesis, that childhood adversity would be associated with more work disability after controlling for education, health conditions, and a health behavior that are established causes of work disability. It would be useful to examine whether other factors such as mental health may account for this association. There was modest evidence that the association of childhood adversity with work disability may be larger for women than for men. Comparing African Americans and whites, there were no notable differences in the association of adversity with work disability. Thus, the results did not support our hypothesis regarding race. Our results regarding race were also not consistent with those of the one previous study in this area (Umberson et al., 2014). That analysis focused on adults ages 60 years and older, with substantially fewer African American participants and many fewer outcome measurements. That study also found that when socioeconomic status was controlled the association with race did not persist among men, and was greatly attenuated for women. However, African Americans represented in our data were much more likely than whites to have experienced high levels of childhood adversity, in contrast to the finding of no difference by Umberson et al. (2014). Thus, the overall health burden of childhood adversity is likely to be larger for African Americans than for whites. The results for gender and race should be interpreted with caution as they may be sensitive to the measures of childhood adversity in the index. Taken as a whole, the results were consistent with the accumulation of risk model. We found that work disability increased with childhood adversity and adult chronic conditions. Our study provided evidence that childhood adversity may contribute to work disability separately from the pathway through adult diabetes, heart disease, and obesity, although the specific mechanisms of that contribution are not clear. Our models did not control for workplace conditions or exposures, or adult socioeconomic circumstances, although the control for education is likely to have accounted for some of the variation associated with those factors. Limitations Self-reports of work disabilities may be subject to measurement error (see, e.g., Burkhauser et al., 2002; Mathiowetz, 2000). However, work disability is not limited to the presence of functional limitations that objectively limit work (Jette & Badley, 2000; Nagi, 1991; WHO, 2001). Aside from disability determinations for public or private insurance benefits, if an individual believes that a health limitation makes it difficult or impossible to work it may be reasonable to accept that judgment (Pransky et al., 2016; Silverstein, 2008). Consistent with most related studies, participants reported adversities retrospectively (e.g., Haas et al., 2012; O’Rand & Hamil-Luker, 2005). Although recall bias may have affected the results, recollections of childhood socioeconomic status and health are typically valid (Batty, Lawlor, Macintyre, Clark, & Leon, 2005; Haas, 2007; Smith, 2009). Consistent with previous research, the index assumed that the adversities were equally associated with work disability (e.g., Montez & Hayward, 2014). However, the results suggested that the associations of individual childhood adversities with work disability varied within and between groups. Adversities may also interact to modify those associations. Results indicated a linear association of childhood adversity with work disability. This result should be interpreted with caution. It may depend on the proportion of individuals who experience each type of adversity, variation among the adversities in the strength of the association, the specific combinations of the adversities to which individuals are exposed, the periods of childhood when the adversities are experienced and the durations of the exposures, exposures to adversities that we did not measure, and other childhood risks to adult health and factors that may help to protect adult health that were not represented in our models. We did not examine abuse during childhood, mental illness, violence, environmental risks, substance abuse, or involvement with the criminal justice system. The model did not consider that some adversity in childhood, or some types of adversity may promote skills or traits associated with better health in adulthood (Schafer, Ferraro, & Mustillo, 2011). Results of the model that did not control for education also associated adversity with more work disability, although the size of the effect was smaller. The latter analysis averaged risks across all education levels, and the majority of adults have at least some postsecondary education. Thus, the average individual represented in the model that did not control for education was more highly educated than those who were the focus of our principal microsimulation results, people with high school education. Consistent with expectations that education is negatively associated with childhood adversity and predicts positive health outcomes, the results suggest that education moderates the association of childhood adversity with work disability. We focused the presentation of the microsimulation results on high school education because people with high school education are the largest educational attainment group. For example, nearly 30% of people aged 55 years and older have high school education, compared with 18% of the same age group with bachelor’s degrees. We also examined whether characteristics shared within families might have affected the results. Using a PSID subsample focused on siblings, Hass (2006) estimated a fixed-effects model examining the association of childhood health with socioeconomic attainment; results suggested that family characteristics may affect the association. However, with our much larger PSID sample and the requirement of CRCS participation, we found no evidence that such characteristics affected the association of childhood adversity with work disability. For observations from before 1999, we assigned diabetes and heart disease status based on later reports of disease onset. A study of PSID cancer diagnosis ages found variation across individuals’ reports over time, due in part to household heads responding for their spouses or partners, and possibly also to stigma associated with cancer (Zajacova, Dowd, Beam, Schoeni, & Wallace, 2010). In the present study, when spouses or partners became household heads in one or more waves we used their own reports of their diagnosis ages. The ICCs for the onset ages suggested acceptable reliability. Conclusions and Implications Participants with childhood adversity had significantly more work disability than those who did not have that experience. The results suggest that reducing socioeconomic, family, and health adversities in childhood, increasing neighborhood cohesion and safety, and reducing bullying may help to reduce work disability. Whether or not an individual with functional impairments has a work disability may often depend on workplace and community cultures, policies, and accommodations (Pransky et al., 2016). Addressing those factors is also desirable because people who experience work disability are significantly more likely than others to live in poverty as they approach retirement, and less likely to own a home or to have a pension (Shuey & Willson, 2017). The fact that African Americans were significantly more likely to have experienced childhood adversity suggests that they may be particularly affected by this financial disparity. Given our aging workforce, greater racial and ethnic diversity, and the increasing number of employees with functional limitations and multiple chronic health conditions, a better understanding of the social production of disability would be useful. It would also be useful to reduce childhood adversity, and to promote resilience and positive health behaviors for people who have that enduring risk to health. Funding This research was supported by the Panel Study of Income Dynamics with a grant from the National Institute on Aging under Grant P01AG029409. The collection of data used in this study was partly supported by the National Institutes of Health under Grant R01 HD069609, and the National Science Foundation under award number 1157698. Conflict of Interest None reported. Acknowledgments The authors are grateful to Cheryl Elman, PhD, Philippa Clarke, PhD, and two anonymous reviewers for valuable comments about this research. Author contributions: S. B. Laditka and J. N. Laditka participated in conceptualizing and designing this study, analyzing the data and interpreting the results, drafting the manuscript, and approving the final manuscript. The data used for this study are available from the Panel Study of Income Dynamics, http://psidonline.isr.umich.edu/default.aspx. References Allison, P. D. ( 2005). Fixed effects regression methods for longitudinal data using SAS . Cary, NC: SAS Institute Inc. Andrade, F. C. ( 2010). Measuring the impact of diabetes on life expectancy and disability-free life expectancy among older adults in Mexico. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 65B, 381– 389. doi: 10.1093/geronb/gbp119 Google Scholar CrossRef Search ADS   Batty, G. D., Lawlor, D. A., Macintyre, S., Clark, H., & Leon, D. A. ( 2005). Accuracy of adults’ recall of childhood social class: Findings from the Aberdeen children of the 1950s study. Journal of Epidemiology and Community Health , 59, 898– 903. doi: 10.1136/jech.2004.030932 Google Scholar CrossRef Search ADS PubMed  Ben-Shlomo, Y., & Kuh, D. ( 2002). A life course approach to chronic disease epidemiology: Conceptual models, empirical challenges and interdisciplinary perspectives. International Journal of Epidemiology , 31, 285– 293. doi: 10.1093/acprof:oso/9780198578154.001.0001 Google Scholar CrossRef Search ADS PubMed  Bowen, M. E., & González, H. M. ( 2010). Childhood socioeconomic position and disability in later life: Results of the Health and Retirement Study. American Journal of Public Health , 100( Suppl.1), S197– S203. doi: 10.2105/AJPH.2009.160986 Google Scholar CrossRef Search ADS PubMed  Breslin, F. C., Tompa, E., Zhao, R., Amick, B. C. 3rd, Pole, J. D., Smith, P., & Hogg-Johnson, S. ( 2007). Work disability absence among young workers with respect to earnings losses in the following year. Scandinavian Journal of Work, Environment & Health , 33, 192– 197. Google Scholar CrossRef Search ADS PubMed  Burkhauser, R. V., Daly, M. C., Houtenville, A. J., & Nargis, N. ( 2002). Self-reported work-limitation data: What they can and cannot tell us. Demography , 39, 541– 555. doi: 10.1353/dem.2002.0025 Google Scholar CrossRef Search ADS PubMed  Cohen, S., Janicki-Deverts, D., Chen, E., & Matthews, K. A. ( 2010). Childhood socioeconomic status and adult health. Annals of the New York Academy of Sciences , 1186, 37– 55. doi: 10.1111/j.1749-6632.2009.05334.x Google Scholar CrossRef Search ADS PubMed  Hass, S. A. ( 2006). Health selection and the process of social stratification: The effect of childhood health on socioeconomic attainment. Journal of Health and Social Behavior , 47, 339– 354. doi: 10.1177/002214650604700403 Google Scholar CrossRef Search ADS PubMed  Haas, S. A. ( 2007). The long-term effects of poor childhood health: An assessment and application of retrospective reports. Demography , 44, 113– 135. Google Scholar CrossRef Search ADS PubMed  Haas, S. A., Krueger, P. M., & Rohlfsen, L. ( 2012). Race/ethnic and nativity disparities in later life physical performance: The role of health and socioeconomic status over the life course. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 67, 238– 248. doi: 10.1093/geronb/gbr155 Google Scholar CrossRef Search ADS   Hamer, R. M. ( 2007). The %INTRACC macro. Retrieved from: http://support.sas.com/kb/25/031.html. Accessed October 23, 2017. Hatch, S. L. ( 2005). Conceptualizing and identifying cumulative adversity and protective resources: Implications for understanding health inequalities. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 60, S130– S134. doi: 10.1093/geronb/60.Special_Issue_2.S130 Google Scholar CrossRef Search ADS   Jette, A. M., & Badley, E. ( 2000). Conceptual issues in the measurement of work disability. In N. Mathiowetz & G.S. Wunderlich (Eds.), Survey measurement of work disability: Summary of a workshop  (pp. 4– 27). Washington, D.C.: National Academy Press. Laditka, S. B. ( 1998). Modeling lifetime nursing home use under assumptions of better health. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 53, S177– S187. Google Scholar CrossRef Search ADS   Laditka, S. B., & Laditka, J. N. ( 2009). Active life expectancy: A central measure of population health. In P. Uhlenberg (Ed.). International handbook of population aging  (pp. 543– 565). Netherlands: Springer-Verlag. Google Scholar CrossRef Search ADS   Laditka, S. B., & Laditka, J. N. ( 2014a). More education may limit disability and extend life for people with cognitive impairment. American Journal of Alzheimer’s Disease and Other Dementias , 29, 436– 447. doi: 10.1177/1533317513518648 Google Scholar CrossRef Search ADS   Laditka, J. N., & Laditka, S. B. ( 2014b). Stroke and active life expectancy in the United States, 1999-2009. Disability and Health Journal , 7, 472– 477. doi: 10.1016/j.dhjo.2014.06.005 Google Scholar CrossRef Search ADS   Laditka, S. B., & Laditka, J. N. ( 2015). Active life expectancy of Americans with diabetes: Risks of heart disease, obesity, and inactivity. Diabetes Research and Clinical Practice , 107, 37– 45. doi: 10.1016/j.diabres.2014.10.008 Google Scholar CrossRef Search ADS PubMed  Laditka, J. N., & Laditka, S. B. ( 2016a). Associations of educational attainment with disability and life expectancy by race and gender in the United States: A longitudinal analysis of the Panel Study of Income Dynamics. Journal of Aging and Health , 28, 1403– 1425. doi: 10.1177/0898264315620590 Google Scholar CrossRef Search ADS   Laditka, J. N., & Laditka, S. B. ( 2016b). Associations of multiple chronic health conditions with active life expectancy in the United States. Disability and Rehabilitation , 38, 273– 289. doi: 10.3109/09638288.2015.1041614 Google Scholar CrossRef Search ADS   Laditka, J. N., & Laditka, S. B. ( 2017). Adverse childhood circumstances and functional status throughout life. Journal of Aging and Health , Advance online publication. doi: 10.1177/ 0898264317715976 Laditka, J. N., & Laditka, S. B. ( 2018). Work disability in the United States, 1968-2015: Prevalence, duration, recovery, and trends. SSM - Population Health , 4, 126– 134. doi: 10.1016/j.ssmph.2017.12.006 Google Scholar CrossRef Search ADS PubMed  Laditka, S. B., & Wolf, D. A. ( 1998). New methods for analyzing active life expectancy. Journal of Aging and Health , 10, 214– 241. Google Scholar CrossRef Search ADS   Luo, Y., & Waite, L. J. ( 2005). The impact of childhood and adult SES on physical, mental, and cognitive well-being in later life. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 60, S93– S101. Google Scholar CrossRef Search ADS   Mathiowetz, N. ( 2000). Methodological issues in the measurement of work disability. In N. Mathiowetz & G.S. Wunderlich (Eds.), Survey measurement of work disability: Summary of a workshop  (pp. 28– 52). Washington, D.C.: National Academy Press. Montez, J. K., & Hayward, M. D. ( 2014). Cumulative childhood adversity, educational attainment, and active life expectancy among U.S. adults. Demography , 51, 413– 435. doi: 10.1007/s13524-013-0261-x Google Scholar CrossRef Search ADS PubMed  Nagi, S. ( 1991). Disability concepts revisited: Implications for prevention. Institute of Medicine, Disability in America: Toward a national agenda for prevention  (pp. 309– 327). Washington, D.C.: National Academy Press. doi:doi.org/10.17226/1579 O’Rand, A. M., & Hamil-Luker, J. ( 2005). Processes of cumulative adversity: Childhood disadvantage and increased risk of heart attack across the life course. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 60B, 117– 124. Google Scholar CrossRef Search ADS   Pransky, G. S., Fassier, J-B., Besen, E., Blanck, P., Ekberg, K., Feuerstein, M., … The Hopkinton Conference Working Group on Workplace Disability Prevention. ( 2016). Sustaining work participation across the life course. Journal of Occupational Rehabilitation , 26, 465– 479. doi: 10.1007/s10926-016- 9670-1 Google Scholar CrossRef Search ADS PubMed  Pudrovska, T., & Anikputa, B. ( 2014). Early-life socioeconomic status and mortality in later life: An integration of four life-course mechanisms. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 69, 451– 460. doi: 10.1093/geronb/gbt122 Google Scholar CrossRef Search ADS   Rank, M. R., & Hirschl, T. A. ( 2014). The risk of developing a work disability across the adulthood years. Disability and Health Journal , 7, 189– 195. doi: 10.1016/j.dhjo.2013.12.001 Google Scholar CrossRef Search ADS PubMed  Schafer, M. H., Ferraro, K. F., & Mustillo, S. A. ( 2011). Children of misfortune: Early adversity and cumulative inequality in perceived life trajectories. American Journal of Sociology , 116, 1053– 1091. Google Scholar CrossRef Search ADS   Schoeni, R., Stafford, F., McGonagle, K. A., & Andreski, P. ( 2013). Response rates in national surveys. Annals of the American Academy of Political and Social Science , 645, 60– 87. doi: 10.1177/0002716212456363 Google Scholar CrossRef Search ADS PubMed  Shuey, K. M., & Willson, A. E. ( 2017). Trajectories of work disability and economic insecurity approaching retirement. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , Advance Access publication July 8, 2017, 1– 11. doi: 10.1093/geronb/gbx096 Silverstein, M. ( 2008). Meeting the challenges of an aging workforce. American Journal of Industrial Medicine , 51, 269– 280. doi: 10.1002/ajim.20569 Google Scholar CrossRef Search ADS PubMed  Smith, J. P. ( 2009). Reconstructing childhood health histories. Demography , 46, 387– 403. doi: 10.1353/dem.0.0058 Google Scholar CrossRef Search ADS PubMed  Smith, N. R., Ferraro, K. F., Kemp, B. R., Morton, P. M., Mustillo, S. A., & Angel, J. L. ( 2016). Childhood misfortune and handgrip strength among Black, White, and Hispanic Americans. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , Advance access December 6, 2016. doi: 10.1093/geronb/gbw147 Social Security Administration( 2017). Disability benefits. Publication No. 05-10029. ICN 456000. Available at: https://www.ssa.gov/pubs/EN-05-10029.pdf. Special Advisory Committee of the Medical Outcomes Trust.( 2002). Assessing health status and quality-of-life instruments: Attributes and review criteria. Quality of Life Research , 11, 193– 205. CrossRef Search ADS PubMed  Toossi, M. ( 2012). Labor force projections to 2020: A more slowly growing workforce. Monthly Labor Review , 43– 64. Turner, R. J., Thomas, C. S., & Brown, T. H. ( 2016). Childhood adversity and adult health: Evaluating intervening mechanisms. Social Science & Medicine , 156, 114– 124. doi: org/10106/j.socscimed.2016.02.026 Google Scholar CrossRef Search ADS   Umberson, D., Williams, K., Thomas, P. A., Liu, H., & Thomeer, M. B. ( 2014). Race, gender, and chains of disadvantage: Childhood adversity, social relationships, and health. Journal of Health and Social Behavior , 55, 20– 38. doi: 10.1177/ 0022146514521426 Google Scholar CrossRef Search ADS PubMed  Ward, B. W. ( 2015). Multiple chronic conditions and labor force outcomes: A population study of U.S. adults. American Journal of Industrial Medicine , 58, 943– 954. doi: 10.1002/ajim.22439 Google Scholar CrossRef Search ADS PubMed  Wolke, D., Copeland, W. E., Angold, A., & Costello, E. J. ( 2013). Impact of bullying in childhood on adult health, wealth, crime, and social outcomes. Psychological Science , 24, 1958– 1970. doi: 10.1177/0956797613481608 Google Scholar CrossRef Search ADS PubMed  World Health Organization (WHO). ( 2001). ICIDH-2: International classification of impairments, activities, and participation. A manual of dimensions of disablement and functioning. Final draft . Geneva: World Health Organization. Yong, V., & Saito, Y. ( 2012). Are there education differentials in disability and mortality transitions and active life expectancy among Japanese older adults? Findings from a 10-year prospective cohort study. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 67, 343– 353. doi: 10.1093/geronb/gbs029 Google Scholar CrossRef Search ADS   Zajacova, A., Dowd, J. B., Schoeni, R. F., & Wallace, R. B. ( 2010). Consistency and precision of cancer reporting in a multiwave national panel survey. Population Health Metrics , 8, 20. doi: 10.1186/1478-7954-8-20 Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.

Journal

The Journals of Gerontology Series B: Psychological Sciences and Social SciencesOxford University Press

Published: Feb 8, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off