Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Adverse childhood experiences and health-related quality of life in adulthood: revelations from a community needs assessment

Adverse childhood experiences and health-related quality of life in adulthood: revelations from a... Background: Adverse childhood experiences (ACE) have been previously linked to quality of life, health conditions, and life expectancy in adulthood. Less is known about the potential mechanisms which mediate these associations. This study examined how ACE influences adult health-related quality of life (HRQoL) in a low-income community in Florida. Methods: A community-based participatory needs assessment was conducted from November 2013 to March 2014 with 201 residents of Tampa, Florida, USA. HRQoL was measured by an excessive number of unhealthy days experienced during the previous 30-day window. Mediation analyses for dichotomous outcomes were conducted with logistic regression. Bootstrapped confidence intervals were generated for both total and specific indirect effects. Results: Most participants reported ‘good to excellent health’ (76 %) and about a fourth reported ‘fair to poor health’ (24 %). The mean of total unhealthy days was 9 days per month (SD ±10.5). Controlling for demographic and neighborhood covariates, excessive unhealthy days was associated with ACE (AOR = 1.23; 95 % CI: 1.06, 1.43), perceived stress (AOR = 1.07; 95 % CI: 1.03, 1.10), and sleep disturbance (AOR = 8.86; 3.61, 21.77). Mediated effects were significant for stress (β = 0.08) and sleep disturbances (β = 0.11) as they related to the relationship between ACE and excessive unhealthy days. Conclusion: ACE is linked to adult HRQoL. Stress and sleep disturbances may represent later consequences of childhood adversity that modulate adult quality of life. Background during childhood can also have a long-term impact on Racial/ethnic minorities in the United States suffer a dis- health [3], and consequently, quality of life. proportionately high burden of morbidity and mortality According to the Centers for Disease Control and compared to their non-Hispanic white counterparts [1], Prevention (CDC), more than half of adults in the U.S. particularly with respect to adverse pregnancy outcomes, have suffered from adverse childhood experiences childhood illnesses, and adult chronic diseases [1]. These (ACE), such as verbal, physical abuse and family dys- health differences are likely the product of complex re- function [4]. These exposures have been linked to a variety lationships across social, economic, environmental, of health conditions in adults, including depression, healthcare, bio-behavioral, structural factors such as cardiovascular disease, cancer, diabetes [5, 6], as well as pre- discrimination and racism, as well as literacy and legis- mature mortality [7]. Moreover, research suggests that ACE lative policies [2]. In addition to stressors experienced is associated with long-term changes in the nervous, endo- during adulthood, stressors and adverse experiences crine, and immune systems [3, 8]. This damage to the body’s stress response system, coupled with the adoption of * Correspondence: [email protected] poor health behaviors to help cope with stress [9], appears Department of Family and Community Medicine, Baylor College of to contribute to a deterioration of adult health. Indeed, the Medicine, 3701 Kirby Drive, Suite 600, Houston, Texas 77098, USA enduring impact of ACE on health suggests that the overall Full list of author information is available at the end of the article © 2015 Salinas-Miranda et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 2 of 12 health-related quality of life (HRQoL) for these adults may Design also be impacted [10]. Although evidence exists linking ad- The community needs assessment survey was a cross- versity during the childhood years (i.e., adverse childhood sectional instrument designed to assess factors associ- experiences or ACE) to impaired HRQoL and shortened ated with HRQoL and was administered between life expectancy in adulthood [7, 11–15], little is known November 2013 and March 2014. Two-hundred and one about the potential mechanisms that mediate the relation- adult participants were recruited from approximately ship. Further, there have been even fewer empirical studies 110,451 residents of the target population [18]. The tar- conducted in socioeconomically disadvantaged popula- get community was predominantly African American tions assessing a wide range of interconnected risk and (60 % black, 18.3 % white, 12.1 % Hispanic, and 9.6 % protective factors that influence the HRQoL of individuals other) and tended to be economically disadvantaged, within a community context. To improve the health and with half the income and double the unemployment rate qualityoflifeofminoritypopulations,anenhancedun- of the rest of the county [19, 20]. The survey was admin- derstanding of the factors that contribute to health dis- istered through intercept interviews across a five zip parities is needed. Accordingly, we conducted this study code area [21, 22]. Intercept interviews are a type of so- to examine how ACE influences adult-onset HRQoL in cial marketing research in which respondents are a community setting. We also explored the roles played approached by culturally and linguistically matched re- by key mediating socio-demographic and other neigh- search staff and “intercepted” in various places where borhood level factors and assessed plausible mediating community members gather often [23]. This method is pathways. used to collect more representative information than con- venience samples regarding particular issues of concern to the community, and intercept interviews are more feasible Methods and less expensive than door-to-door needs assessments. Context Intercept interviews are also useful when the target popula- The project “Toward Eliminating Disparities in Mater- tion is widely dispersed and harder to reach (e.g., several nal and Child Health Populations” is a 3-year CBPR zip code-level areas with economically disadvantaged urban (community-based participatory research) initiative that communities). Accordingly, trained CAB members sur- partners researchers at the University of South Florida veyed people “in the streets” at frequently used community (USF), a non-profit advocacy and empowerment orga- locations (e.g., cafes, churches, libraries, local schools gath- nization, REACHUP, Inc., and a group of community erings, shopping centers) and at times when community representatives from selected zip codes in Central residents were accessible (e.g., weekends, after hours during Tampa, Hillsborough County, Florida. The project has weekdays). In a similar manner to 30 by 7 cluster sampling three sequential phases: 1) needs assessment to identify [24], the CAB nominated 30 community locations or “clus- priority maternal and child health issues, 2) planning ters” (23 were actually surveyed due to redundancy of par- an evidence-based intervention that addresses the ticipants in 7 locations) based on socio-demographic and priorities identified, and 3) implementation of a pilot economic characteristics using zip code level census data, community-driven intervention. The project is funded and then randomly sampled at least seven individuals from by the National Institute on Minority Health and each location. To participate in the survey, a respondent Health Disparities (5R24MD8056-02). must have been a community resident for at least the previ- This paper focuses exclusively on the needs assess- ous 12 months. Flyers, social media, and “word-of- ment survey, which was designed to explore health de- mouth” transmission of information were used for re- terminants and quality of life indicators in the target cruitment. Written informed consent was obtained and community, implemented using a CBPR approach. This a modest but appropriate monetary incentive was given CBPR study builds extensively upon a strong, existing to study participants. Approval for the study was obtained community-academic partnership. At the study’s outset, from the Institutional Review Board of the University of a Community Advisory Board (CAB) was created, which South Florida. consisted of eight adult residents from the target com- munity who were recruited based upon their knowledge, Measures participation, advocacy, and leadership in previous The development of survey questions was guided by community projects [16, 17]. In addition to the required the Life Course Perspective (LCP) with extensive input Human Subjects Protection Certification course, all from the CAB. The LCP framework is used to evaluate CAB members completed skill-building workshops in the cumulative influences of risk and protective factors research methods. Active participation of community during critical periods of human development and the members occurred in the design, data collection, and effects those events have on the health trajectories of analysis phases of the study. individuals [25–27]. First, community members, in Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 3 of 12 partnership with academic researchers, reviewed local Primary exposure: adverse childhood experiences data and voiced their concerns for critical topics. Next, To assess cumulative risk from childhood to adulthood, the academic researchers suggested questionnaires or scales main predictors included participant recall of adverse that had been validated in previous research studies for events that occurred during the first 18 years of life. We local adaptation. Subsequently, the CAB provided feedback used the Brief Family History Questionnaire from the ACE on wording and readability, question addition/deletion, study [6], a 10-item questionnaire collecting self-reported assessed acceptability of questions and technology usability physical, emotional, and sexual abuse, family dysfunction, (see Technology section), and pilot tested questions before and economic hardship. All items are dichotomous, yes/no community-wide implementation. The final survey ques- questions. An overall ACE score, which ranges from 1–10, tionnaire contained 63 questions (Likert-type and multiple is then calculated by adding the number of “yes” responses. choice questions), which covered the following inquiry Higher scores have been found to be associated with a wide domains and specific sequencing into the survey: life in the range of adverse health outcomes, impaired quality of life, neighborhood (neighborhood assets; community-wide and higher mortality [7, 11, 34]. issues), social connections (social support), health and qual- ity of life (general health, HRQoL, self-reported health Potential mediators: smoking, alcohol use, diet, physical problems, sleepless days, perceived social standing), stress activity, and sleep disturbances and unfair treatment (stress appraisal, perceived experi- We hypothesized that contemporary lifestyle risk and pro- ences of discrimination), lifestyle (smoking, alcohol use, tective factors of participants could mediate the association recreational drugs, diet, exercise, perceived HIV risk), between ACE and adult HRQoL. Lifestyle and behavioral childhood experiences (ACE), and socio-demographic questions were obtained from multiple validated instru- questions (age, education, marital status, race/ethnicity, ments that are frequently used in health services research, household income, employment). Academic researchers which were adapted using input from community assisted with the development of hypotheses, analytic members. Smoking was assessed with a question from the strategies, and statistical analyses. In the subsections 2013 Behavioral Risk Factor Surveillance System (BRFSS) below, we describe only the specific instruments and [35]: “Have you smoked at least 100 cigarettes in your en- measures used to test our main hypotheses; however, tire life?” This question was chosen by CAB members the complete survey is available as a supplemental as being easier to respond to than a detailed frequency file (Additional file 1). question [36]. The frequency of alcohol use per month was measured with a numerical question, also adapted from the 2013 BRFSS: “In a typical month, how often Primary outcome: health-related quality of life do you drink alcoholic beverages?” HRQoL was the primary study outcome and was mea- Although self-reported measures of dietary patterns sured using the CDC’s “Healthy Days Measure” instru- are vulnerable to measurement bias [37, 38], the CAB ment, which is a validated scale used frequently in members felt that asking a few questions about diet was national health surveillance surveys [28, 29]. Specific- still an important aspect to include in the survey as a ally, we used the brief version, referred to as Healthy general, non-sensitive behavioral item. Thus, the survey Days Core Module (4-items questionnaire), which cap- included a short series of questions about intake of fruits tures the self-reported number of days in the past and vegetables, fatty, sugary, salty foods, and caffeinated 30 days that individuals rated their physical or mental drinks. Participants were asked to indicate how often health as not good [30]. It includes the following com- they consumed fruits and vegetables in a typical month, ponents: 1) self-rated health, from poor to excellent (or- and we grouped responses as: “once a month or less” dinal); 2) number of days when physical health was not coded as 1, “2–3 times a month” coded as 2.5, “once a good during the past 30 days; 3) number of days when week” coded as 4, “2–3 times a week” equivalent to 10, mental health was not good during the past 30 days; and “4–6 times a week or more” coded as 20. For this and 4) number of activity limitations due to either phys- study, we focused on the consumption of fruits and ical or mental health illness (combined). Total number vegetables as key protective factor because of stronger of unhealthy days is then obtained by adding the re- evidence regarding fruits and vegetable self-reported sponses to questions 2 and 3. Any sum greater than 30 measures. Specifically, brief instruments for fruit and was capped at 30, with a maximum of 30 unhealthy days vegetable intake assessments have been found to be ad- [30]. The outcome was operationalized through dichotomi- equate for estimating relative risks in the relationship zation of the total number of unhealthy days, with poor between fruit and vegetable intake versus disease [39]. HRQoL reflected by 14 or more total days. This cut-point Because the survey included questions not previously has been used by other authors as a discriminate measure validated, we consider our approach a conservative one. to assess excessive unhealthiness [31–33]. Moreover, because we did not find that any of our Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 4 of 12 self-reported dietary measures were significantly associ- Neighborhood confounders ated with HRQoL in bivariate preliminary analyses, we Community resources, community-wide issues, and did not explore dietary items in greater detail in media- neighborhood cohesion levels were considered as tional analyses. Sleep disturbances were also measured potential confounders in our analyses. First, resources with a question we derived from the BRFSS [28]: “During available in the community were measured with the fol- the past 30 days, for about how many days have you felt lowing question: “Which of the following resources are you did not get enough rest or sleep?” Physical activity available to you in your community?” A list of resources was measured with one question that assessed physical ac- was compiled using CAB input and provided with the tivity, also from the BRFSS [28]: “During the past month, survey. Each item listed was weighted equally and the about how many days per week did you exercise for recre- final variable used in analyses was the total number of ation or to keep in shape (activities that make you assets or resources reported. Similarly, perceived sweat)?” community-wide issues were measured with one ques- tion on neighborhood problems [43]: “Which of the fol- Stress appraisal lowing is a problem in the neighborhood?” Again, a list was Cognitive appraisal of stress was measured with the 4-item developed by the CAB, with the allowance for write-in en- Perceived Stress Scale, which is a validated instrument used tries. The final variable for analysis consisted of the total to make comparisons of subjects’ perceived stress related to number of different issues reported. Lastly, neighborhood current events [40]. Questions are 5-point Likert type social cohesion was assessed by measuring the respondent’s scaled (i.e., strongly disagree = 1, to strongly agree = 5). A level of agreement (on a 5-point Likert type scale) with a composite score was derived by adding the original set of questions proposed by Cagney and colleagues [44]: 1) scores and multiplying by a factor of 5, which results in “People around here are willing to help their neighbors”;2) a 100-point scale. The higher the score, the higher the “This is a close-knit neighborhood”;3) “People in this risk for clinical psychiatric disorder [41]. neighborhood can be trusted”; and 4) “People in this neigh- borhood generally don’t get along with each other” (reverse Social support coded). These questions were then summed to provide a Perceived social support was measured with five questions total score, where higher scores indicated higher neighbor- from the Medical Outcomes Study Social Support Survey hood social cohesion. [42]. Using 5-point Likert type scales (‘None of the time’ =1 to ‘All of the time’ = 5), individuals were asked Technology to indicate how often the following types of supports The droidSURVEY software was used to design and were available to them if needed: 1) someone to con- administer the survey [45], which was installed on ten fide in or talk about yourself or your problems; 2) Hewlett-Packard Slate 7” portable tablet computers someone to share your most private worries and fears running the Android™ 4.2.2 operating system [46]. For with; 3) someone to help you if you were confined to Spanish-speaking participants, all questions were trans- bed; 4) someone to prepare your meals if you were un- lated to Spanish by native speakers and assessed for ac- able to do it yourself; and 5) someone to get together curacy of translation using back translation and pilot with for relaxation. Question scores were added to testing. The use of tablet-administered surveys allowed yield a total social support score. To facilitate inter- for a portable, convenient means of data collection, and pretation in the community setting, the original scores pilot testing revealed that community members found ranging from 4 to 20 were converted to a 100-point touchscreen technology to be intuitive, easy to use, and scale by multiplying the original score by five. enjoyable. Supervised by the principal investigator, the study coordinator trained community members in the Sociodemographic confounders use of tablets and survey implementation, the informed Socio-demographic characteristics were collected from consent process, management of tablet computers in the participants and included: age in years (categorized as field, and how to assist participants with questions about 18–35; 36–45; 46–55; and ≥ 56 years), sex, education the study. Training consisted of three 2-hour sessions in (high school vs. less than high school graduate or the preceding month to the survey. equivalent), current marital status, race/ethnicity (non- Hispanic white, non-Hispanic black, Hispanic, other), Statistical analysis employment status (employed, unemployed but able, The study population was described using descriptive and unable to work), annual household income (categor- statistics that included frequencies and percentages for ies: US$ 0–20,000; 20,001–40,000; and ≥ 40,000), and categorical variables and means and standard deviations residential time in the five target zip codes (5 years or for numerical variables. Two-sided tests of equality for less vs. more than 5 years). column proportions (z-tests for column proportions or Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 5 of 12 t-tests for column means) were conducted to assess Results differences by outcome group. Tests assumed equal Participants’ characteristics, neighborhood factors, and variances and were adjusted for multiple comparisons health-related quality of life using a Bonferroni correction. All analyses were con- Table 1 describes the study sample, by socio-demographic ducted with the IBM SPSS Statistics for Windows, Version characteristics. In total, 201 participants were surveyed, of 22.0 (IBM Corp, Armonk, NY). Statistical significance was which the majority were female (65.7 %) and non- assessed at the 0.05 level. Hispanic black (66.5 %), with a mean age of 45 ± 14 years. Under the LCP framework, we posited that numerous Nearly one-fifth of participants did not graduate from high accumulating risks and protective factors mediate the school, most were not married (67 %), and only 40 % were association between ACE and the HRQoL proxy. As a first currently employed. Income was not reported by 13 % of step, separate logistic regression models were run to assess participants. Among those who reported income, the ma- the independent effects of life course social determinants jority had incomes lower than $20,000 (about 70 %). on the outcome (≥14 unhealthy days), adjusting for indi- Moreover, nearly all participants (97.5 %) reported receiving vidual socio-demographic and neighborhood covariates. at least one form of social assistance based on federal pov- We explored the role of the following factors: stress, sleep erty line categorizations (i.e., Section 8-Housing Choice disturbances, smoking, alcohol use, physical activity, fruits Voucher Program, food stamps, school free or reduced and vegetable intake, and social support. The purpose lunches, Medicaid, supplemental security income, Tempor- was to identify significant independent associations ary Assistance for Needy Families, or other). This was also with HRQoL that could represent potential mediating a relatively transient population, with as many residents pathways in the relationship between ACE and HRQoL. who had lived for 5 or more years in the neighborhood as After a set of factors was identified as possible mediators, those that had just moved in the last 5 years. the next step was to test for mediation or indirect effects Participants demonstrated awareness of community [47]. The following three conditions must be established to resources, reporting an average of 8 community assets determine whether mediation had occurred (i.e., indirect (range was from 0 to 13 reported assets). The most effects) [48]: 1)thatthe independentvariable(ACE) pre- frequently cited resources were churches and schools dicts the dependentvariable(HRQoL),2)thatthe inde- (Fig. 1). Other high-ranking community resources (in pendent variable (ACE) predicts the mediators (Ms), 3) that order of highest frequency to least) were public trans- the mediator (M) predicts the dependent variable (i.e., ≥14 portation, parks/recreational facilities, police/fire de- unhealthy days). Additionally, the effects of confounders partments, pharmacies, libraries, and hospitals/clinics. were considered [49]. Participants also reported community-wide issues that Accordingly, mediation analyses for the dichotomous were a perceived problem for their health and their outcome (Y: HRQoL) were conducted with logistic re- neighborhood (Fig. 2). On average, participants noted gression models to estimate the path coefficients in a about 4 community-wide issues (SD ± 2.9). Drug and two-mediator model (X: ACE, M1: stress, M2: sleep dis- alcohol related issues were the most frequently stated turbance). We controlled for socio-demographic covari- followed by abandoned property, litter, homeless issues, ates (i.e., age, gender, race/ethnicity, no high school, and high crime rate, lack of affordable shopping, poor police household income) and neighborhood level factors (i.e., response, poor quality grocery stores, and lack of parks/ neighborhood cohesion, community resources, and recreational facilities. neighborhood issues). The first step was the estimation Mean neighborhood cohesion was 65.33 (SD ±18.7), of controlled direct effects through a series of logistic re- and the majority (60.3 %) of participants perceived that gressions. This was followed by the decomposition of people in their neighborhood generally got along with effects into total, indirect, and direct effects [50]. The each other. More than half (57 %) perceived their formula used to calculate indirect effects for both mediators neighborhood as a place where people were willing to was: c=c’ +(ab);where c is total effect, c’ is the direct effect, help each other. On the other hand, only 41.5 % of and a*b is the indirect effect. The indirect effect (ab) is the participants agreed or strongly agreed that people in measure of the amount of mediation, and equals the reduc- their neighborhood shared the same values. In tion of the effect of the causal variable on the outcome or: addition, only 40.4 % of participants agreed or strongly ab = c - c’. For this purpose, we used an SPSS macro with agreed that they lived in a close-knit neighborhood. asymptotic and resampling strategies for comparing indir- Approximately three of every four participants consid- ect effects in multiple mediator models, which included ered themselves in “good to excellent health.” The mean bootstrapping to estimate confidence intervals for total and number of unhealthy days due to poor physical health or specific indirect effects [47, 51]. Missing values were han- injury in the past month was 4.9 days (SD ± 7.8). The mean dled with default option in SPSS for logistic regression pro- number of unhealthy days due to stress, depression or cedure, which is listwise deletion. problems with emotions was 5.3 days (SD ± 8.4). When Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 6 of 12 Table 1 Unhealthy days during the previous 30 days by participants’ characteristics and neighborhood factors Total 0–13 unhealthy days a month 14 or more unhealthy days a month N (%) Count (%) Count (%) Age (in years) 18–35 years 62 (31.0) 46 (31.3) 16 (30.2) 36–45 years 39 (19.5) 31 (21.1) 8 (15.1) 46–55 years 50 (25.0) 38 (25.9) 12 (22.6) 56 plus years 49 (24.5) 32 (21.8) 17 (32.1) Sex Female 132 (65.7) 93 (62.8) 39 (73.6) Male 69 (34.3) 55 (37.2) 14 (26.4) Education High school complete 162 (80.6) 124 (83.8) 38 (71.7) No high school 39 (19.4) 24 (16.2) 15 (28.3) Marital status Currently married 65 (32.7) 54 (37.0) 11 (20.8) Not married now 134 (67.3) 92 (63.0) 42 (79.2) Race/ethnicity Non-Hispanic whites 13 (6.5) 8 (5.4) 5 (9.4) Non-Hispanic blacks 133 (66.5) 98 (66.7) 35 (66.0) Hispanic or Latino 54 (27.0) 41 (27.9) 13 (24.5) Household annual income $0–20,000 114 (56.7) 77 (38.3) 37 (18.4) $20,001–40,000 36 (18.0) 29 (14.4) 7 (3.5) $40,001 or more 24 (11.9) 21 (10.4) 3 (1.5) Income not reported 27 (13.4) 21 (10.4) 6 (3.0) Residential time 5 years or less 101 (50.2) 76 (51.4) 25 (47.2) More than 5 years 100 (49.8) 72 (48.6) 28 (52.8) Neighborhood factors Mean SD Mean SD Mean SD Community assets 7.78 (3.5) 7.88 3.38 7.49 3.92 Community wide issues 4.11 (2.9) 4.03 2.93 4.36 2.97 Neighborhood cohesion 65.33 (18.7) 66.03 18.81 63.40 18.73 Column comparisons of 0–13 unhealthy days a month vs. 14 or more unhealthy days. Values were significantly different at p < .05 in the two-sided test of equality for column proportions (z-tests for column proportions or t-tests for column means). Some column numbers do not add to total sample size due to missing values combined both physically and mentally unhealthy days, higher proportion of individuals with incomes lower than the mean total unhealthy days was about 9 days a $20,000 occurred among those who reported ≥14 or more month (SD ± 10.54). About a fourth of participants unhealthy days. We found no significant differences by (26 %) reported having 14 or more total unhealthy days other socio-demographic variables. When examining the per month (Table 1). The most common self-reported potential role of neighborhood factors, we found no signifi- personal health issues included back/neck problems, cant differences by assets, perceptions of community-wide stress, vision problems, arthritis, problems walking, issues, nor neighborhood social cohesion. depression/anxiety, and weight issues (Fig. 3). Marital status was significantly associated with ≥14 or Independent adjusted effects of life course determinants more unhealthy days (Table 1). Specifically, a higher pro- on unhealthy days portion of unmarried individuals in the group reported ≥14 Table 2 presents separate models for each determinant, or more unhealthy days (79.2 %). Also, low income was as- adjusting by age, sex, race/ethnicity, education, income, sociated with ≥14 or more unhealthy days, specifically a community assets, community issues, and neighborhood Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 7 of 12 Churches Schools Public transportation Parks and recreational facilities Police/Fire Department Pharmacies Libraries Hospitals and clinics Mental health facilities Colleges, universities Local businesses Community recreation centers 0.0 20.0 40.0 60.0 80.0 100.0 Percentages (N=201) Fig. 1 Community assets reported by survey participants. Notes: Percentages based on N = 201 cohesion. ACE was associated with 23 %-increased odds significantly predicted poor HRQoL. The model correctly of poor HRQoL (≥14 unhealthy days) in adulthood predicted 73 % of cases with excessive unhealthy days. In (AOR = 1.23; 95 % CI: 1.06, 1.43). Increasing stress was Model 2, this relationship was explained by stress and the also associated with increased likelihood of reporting ≥14 model had greater explanatory power (Pseudo-R =0.32) unhealthy days (AOR = 1.07; 95 % CI: 1.03, 1.10). Sleep than the first model. In Model 3, the addition of sleep disturbances were strongly associated with reports of 14 disturbances improved the explanatory power (Pseudo- or more unhealthy days. Specifically, any additional sleep R = 0.44) and the percent of correct predictions was deprived day was associated with 9 times higher odds of increased to 82.1 %. All three models demonstrated reporting ≥14 unhealthy days (AOR = 8.86; 95 % CI: 3.61, adequate goodness of fit (Hosmer-Lemeshow non- 21.77). After adjusting for covariates, no associations were significant). On the other hand, these three initial logistic found for smoking, alcohol use, physical activity, fruits and regression models only indicate measures of associations vegetable intake, or social support. Based on these find- with controlled effects and cannot estimate the magnitude ings, we selected stress and sleep disturbance as poten- of mediated or indirect effects (i.e., effect of ACE on tial mediators that needed to be examined in the test of Stress, plus the effect of stress on unhealthy days) or if the mediation. mediation is partial or total [47]. Table 4 presents the decomposition of effects into total, Test of mediation indirect, and direct effects for the HRQoL measure. The Table 3 presents the stepwise assessment of controlled three essential conditions to determine mediation were direct effects of ACE, stress, and sleep disturbances on established, even after controlling for SES and neigh- HRQoL. In Model 1, it can be appreciated that ACE borhood level factors. First, ACE (main predictor) Drug/alcohol-related problems Abandoned houses, factories, or buildings Trash or litter in empty lots, streets, or… Homeless persons/panhandling Vandalism, high crime rate Lack of affordable shopping Poor police response Lack of quality of recreation, parks, or… Poor quality grocery Poor quality of schools, or libraries Dogs and uncontrolled animals Junkyards, gasoline stations, and other… Hazards: incinerators, chemical plants, and… 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 Percentages (N=201) Fig. 2 Community-wide issues reported by survey participants. Notes: Percentages based on N = 201 Community Issues Community Assets Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 8 of 12 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 Self-Reported Health Issues in the Previous Year Fig. 3 Self-reported health issues among survey participants. Notes: Percentages based on N = 201 independently predicted self-reports of ≥14 unhealthy of stress and sleep disturbances was 0.19 (95 % CI: 0.06, days due to physical or mental illness (outcome). Second, 0.33). By exponentiation of the later beta coefficient, we ACE significantly predicted two mediators, namely per- obtain an OR of 1.20 (95 % CI: 1.06, 1.39). In other ceived stress and sleepless days. Third, these mediators words, there is an increase of 20 % in the odds of report- (stress and sleep disturbance) significantly predicted un- ing ≥14 unhealthy days for every additional ACE, which healthy days. is mediated by stress and sleep disturbances in the adult- Mediated effects were significant for stress (β = 0.08; hood. It should be highlighted that such effect size is 95 % CI: 0.01, 0.21) and sleep disturbances (β = 0.11; very close to the independent controlled effect of ACE 95 % CI: 0.03, 0.21) on the relationship between ACE on unhealthy days that was noted in Model 1 of Table 3. and excessive unhealthy days. It can be noted that when testing for mediation the direct effect of ACE on un- Discussion healthy days (c’ path) became non-significant, which in- We found that exposure to ACE is linked to impaired dicates total mediation. The joint total mediated effect adult HRQoL, with mediation effects modulated by stress and sleep disturbances. An implied hypothesis of this study was that there was an association between ACE and Table 2 Independent unadjusted and adjusted effects of adverse childhood experiences and current social HRQoL. Our study confirms this association and is con- determinants on the odds of reporting ≥14 days of unhealthy sistent with previous findings [7, 11–15]. Unique to our days per month study, however, is the way in which ACE was measured Variable ≥14 days of unhealthy days per month using the widely established theoretical framework, the LCP, to capture cumulative life experiences over time. OR (95 % CI) AOR (95 % CI) a a The LCP provided the theoretical guidance to frame the Adverse childhood experiences 1.19 (1.05, 1.34) 1.23 (1.06, 1.43) a a inquiry, while CBPR fostered active community engage- Stress 1.05 (1.03, 1.08) 1.07 (1.03, 1.10) ment in the design of research questions that are rele- Social support .99 (.98, 1.01) .99 (.97, 1.01) vant to the community context and in the collection of Smoking 1.50 (.79, 2.86) 1.30 (.59, 2.89) reliable experiential information [52, 53]. The multi-do- Alcohol use 2.54 (1.02, 6.33) 2.78 (.93, 8.31) main conceptualization and multi-level nature of LCP Physical activity .90 (.46, 1.69) .66 (.31, 1.39) along with CBPR approaches have the potential of im- proving the assessment and deepening the understanding Fruits and vegetable intake 1.04 (.55, 1.96) 1.32 (.61, 2.83) a a of determinants of health disparities. It is precisely because Sleep disturbances 6.08 (3.03, 12.24) 8.86 (3.61, 21.77) of these notable features, namely, integration (i.e., risk/ Statistically significant 95 % CI protection), multilevel (i.e., individual, families, and com- Separate models for each determinant, adjusting by age, sex, race/ethnicity, education, income, community assets, community issues, and munities), and time-orientation (i.e., life span), that the neighborhood cohesion c LCP was able to assist in explaining why and how health With the exception of smoking (yes/no), all other predictor variables were on a continuous measurement scale disparities occur. Percentage (N=201) Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 9 of 12 Table 3 Excessive unhealthy (≥14) days predicted by adverse childhood experiences and step-wise inclusion of mediators Variables Model 1 Model 2 Model 3 a a a B (95 % CI) B (95 % CI) B (95 % CI) Main Predictor Adverse childhood experiences .21 (.05, .43) .104 (−.06, .34) .03 (−.24, .28) Mediators b b Stress .061 (.03, .158) .05 (.01, .12) Sleep disturbances .10 (.06, .23) Confounders Age .01 (−.027, .04) .03 (−.01, .08) .03 (−.01, .09) Sex −.19 (−1.33, .76) −.19 (−1.67, .95) .21 (−1.37, 1.84) b b Education .78 (−.36, 2.28) 1.35 (.05, 3.48) 1.85 (.40, 4.38) Marital status .78 (−.19, 2.40) .55 (−.72, 2.02) .87 (−.47, 3.06) Non-Hispanic black −.45 (−2.44, 1.65) −.17 (−1.97, 1.85) −.65 (−3.01, 1.41) Hispanic or Latino −.81 (−3.08, 1.40) −.72 (−2.93, 1.65) −.97 (−3.64, 1.24) US$20,001–40,000 −.44 (−1.88, .71) −.02 (−1.68, 1.34) −.42 (−2.76, 1.20) ≥US$40,001 −.29 (−19.69, 1.30) −.31 (−19.29 1.13) −.30 (−20.16, 1.85) Community assets −.05 (−.19, .08) .07 (−.08, .26) .07 (−.11, .28) Neighborhood issues −.02 (−.23, .15) −.12 (−.34, .08) −.17 (−.49, .03) Neighborhood cohesion −.01 (−.03, .01) −.01 (−.05, .02) −.01 (−.05, .02) Constant −.76 (−5.04, 2.81) −4.32 (−16.16, −.47) −4.7 (−13.87, −.05) Model Chi-square [df] 22.48 [7] 38.32 [11] 56.26 [14] Nagelkerke R .18 .32 .44 % Correct predictions 73.2 80.8 82.1 Goodness of fit p-value .44 .55 .38 Model 1: binary logistic regression with Unhealthy Days as outcome, ACE as predictor, and controlling for age, gender, education, marital status, race/ethnicity, income, community assets, neighborhood issues, neighborhood cohesion as covariates Model 2: Model 1, adding stress (mediator 1) Model 3: Model 2, adding sleep disturbance (mediator 2) N = 151 for model 1,2, and 3 Bootstrapped 95 % confidence intervals based on 1000 bootstrapped samples. All values rounded to two digits. Indicates that the coefficient is statistically significant at, at least, the .05 level Another hypothesis in this study was that certain symp- intervene immediately. However, there are circumstances toms are expressed later in life among victims of ACE, (e.g., poverty, hidden abuse, etc.) which may elude early de- resulting in a symptomatology complex that includes a tection. In such instances, the other approach would be physically and emotionally poor quality of life. Our study detection of ACE through screening measures with inter- is novel in this perspective because we conducted a detailed vention coming later in life. For example, those impacted and robust mediation analysis, which mapped out mechan- by ACE may exhibit somatization by expressing symptoms ismal pathways that could potentially explain early life ex- such as sleep and stress disturbances and may greatly periences and subsequent impaired HRQoL. Specifically, benefit from improved tertiary care. we observed that ACE victims in this study were more It is important to place our results within the context of a likely to have heightened stress levels and sleep distur- methodological limitation, namely that the survey upon bances, a finding that underscores the importance of their which our findings were based was cross-sectional. None- role as potential mediators linking ACE to HRQoL. There theless, it is noteworthy that a merit of the study in- are clinical as well as public health implications of these ob- strument is that the questions were framed within servations. In particular, they represent potential avenues specific temporal windows, and therefore, the temporal for intervention to minimize the negative impact of ACE relationships between ACE, HRQoL, and the other on HRQoL. factors included could be established. Another short- There are two main approaches of effective strategies to coming of our analysis is that while our assessment improve the HRQoL among adults impacted by ACE. The was based on a representative community sample, it first is to identify those impacted by ACE early in life and cannot be taken as directly generalizable to the entire Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 10 of 12 Table 4 Mediation of stress on the relationship between approach from the “womb to the tomb,” which starts adverse childhood experiences and excessive unhealthy days by addressing psychological well-being and behavioral a, b Path Coefficients S.E. health of expecting mothers and continue to provide support to minimize the effect of unresolved childhood IV to Mediators (a paths): ACE to Stress and Sleep traumas, as well as the linkages to professional and Perceived stress 1.64 .54 community supports throughout the life span. This Sleep disturbance 1.09 .30 offers a potential path to the improvement of HRQoL Direct Effects of Mediators on DV (b paths) for victims impacted by ACE. Perceived stress .05 .02 Sleep disturbance .10 .03 Additional file Total Effect of IV on DV (c path) Adverse childhood experiences .19 .07 Additional file 1: Survey Questions and Flow. (PDF 130 kb) Direct Effect of IV on DV (c’ path) Adverse childhood experiences .03 .09 Abbreviations ACE: Adverse childhood experiences; HRQoL: Health-related quality of life; Partial Effect of Control Variables on DV SD: Standard deviation; AOR: Adjusted odds ratio; CI: Confidence interval; Age .03 .02 CDC: Centers for Disease Control and Prevention; CBPR: Community-based participatory research; USF: University of South Florida; CAB: Community Sex .18 .57 Advisory Board; LCP: Life Course Perspective; BRFSS: Behavioral Risk Factor Education 1.85 .63 Surveillance System. Marital status .79 .58 Competing interests Race/ethnicity −.43 .49 The authors declare that they have no competing interests. Household income −.23 .39 Community assets .07 .07 Authors’ contributions ASM was involved in developing the needs assessment, participated in the Neighborhood issues −.16 .09 statistical analyses, and was responsible for the first drafts of the Intro, Neighborhood cohesion −.01 .01 Methods, and Results. JLS was responsible for developing and implementing the statistical analysis plan, interpreting the results, and assisting with Indirect Effects of IV(ACE) on DV(Unhealthy days) through Proposed drafting the Intro, Methods, and Results. LK aided with development of the Mediators (ab paths) needs assessment, assisted with data collection, and assisted with drafting Data Boot (95 % CI) the manuscript. JB made considerable contributions to the conceptualization of the study and directed development of the needs assessment. She Total .19 .22 (.06, .33) worked to provide critical appraisal and subsequent revision of the initial Perceived stress .08 .09 (.01, .21) draft. EB, DA, and KS were all instrumental to conception, design, and implementation of the study, acquisition of the data, and critical review and Sleep Deprivation .11 .13 (.03, .21) revision of the manuscript. KKS helped to analyze and interpret the data, and a b All values rounded to two digits. Linear regressions for translate findings into an initial draft of the Discussion. HS was ultimately ‘a paths’ for perceived stress and sleep deprivation. All other coefficients responsible for design of the entire study, oversaw all aspects of instrument derived with binary logistic regression. Formulas: Total Effect: c = c’ + ab; development and data collection, assisting in writing the initial draft of the amount of mediation or indirect effect: ab = c - c’. Based on 1000 Discussion, and provided critical insight that guided revisions. All authors bootstrapped samples. Path coefficient significant at p < 0.05 have provided final approval of this manuscript for submission and agree to be accountable for the work herein. US population, due to the selective context of our Acknowledgments project (e.g., mostly low income, African American This work is supported by funding from the National Institute on Minority Health women). The advantage of this limitation, however, is andHealthDisparities throughanR24 grant on “Community-Based Participatory that the study addressed an important cause of low Research” (Award#: 5R24MD8056-02). The content of this paper is solely the responsibility of the authors and does not necessarily represent the official views HRQoL in a socio-economically disadvantaged setting that of the National Institute on Minority Health and Health Disparities, or the stands to benefitmostfromappropriate andtargeted University of South Florida. interventions for victims of ACE. Author details Department of Epidemiology and Biostatistics, College of Public Health, Conclusions University of South Florida, 13201 Bruce B Downs, MDC56, Tampa, Florida In summary, our findings indicate that adversity in 33612, USA. REACHUP Inc., 2902 N, Armenia Avenue Suite 100, Tampa, Florida 33607, USA. Department of Family and Community Medicine, Baylor childhood continues to affect the mental and behav- College of Medicine, 3701 Kirby Drive, Suite 600, Houston, Texas 77098, USA. ioral health trajectory of adults. Thus, we recommend Department of Community and Family Health, College of Public Health, the implementation of community health programs University of South Florida, 13201 Bruce B. Downs Blvd. MDC 56, Tampa, Florida 33612, USA. aimed at improving psychological well-being by redu- cing high stress levels, particularly among individuals Received: 31 December 2014 Accepted: 24 June 2015 who havesuffered childhood trauma.Wesuggest an Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 11 of 12 References 23. Andreasen AR. Academy for Educational D. Marketing research that won't 1. CDC. CDC Health Disparities and Inequalities Report - United States, 2011. break the bank : a practical guide to getting the information you need. San Morb Mortal Wkly Rep. 2011;60:1–114. Francisco: Jossey-Bass; 2002. 2. Truman BI, Smith KC, Roy K, Chen Z, Moonesinghe R, Zhu J, et al. Rationale 24. Hlady WG, Quenemoen LE, Armenia-Cope RR, Hurt KJ, Malilay J, Noji EK, et for regular reporting on health disparities and inequalities - United States. al. Use of a modified cluster sampling method to perform rapid needs MMWR Surveill Summ. 2011;60(Suppl):3–10. assessment after Hurricane Andrew. Ann Emerg Med. 1994;23(4):719–25. 3. Danese A, McEwen BS. Adverse childhood experiences, allostasis, allostatic 25. Braveman P, Barclay C. Health disparities beginning in childhood: a life-course load, and age-related disease. Physiol Behav. 2012;106(1):29–39. doi:10.1016/ perspective. Pediatrics. 2009;124:163–75. j.physbeh.2011.08.019. 26. Lu MC, Chen B. Racial and ethnic disparities in preterm birth: the role of stressful life events. Am J Obstet Gynecol. 2004;191(3):691–9. 4. CDC. Adverse childhood experiences reported by adults – five states, 2009. doi:10.1016/j.ajog.2004.04.018. MMWR Morb Mortal Wkly Rep. 2010;59(49):1609–13. 5. Felitti VJ, Anda RF. The relationship of adverse childhood experiences to 27. Lu MC, Halfon N. Racial and ethnic disparities in birth outcomes: a life-course adult medical disease, psychiatric disorders, and sexual behavior: perspective. Matern Child Health J. 2003;7(1):13–30. implications for healthcare. In: Lanius RA, Vermetten E, Pain C, editors. The 28. CDC. Behavioral Risk Factor Surveillance System. Atlanta, GA: US Department Impact of Early Life Trauma on Health Child and Adolescent Psychiatry and of Health and Human Services; 2013. http://www.cdc.gov/brfss/. Mental and Disease: The Hidden Epidemic. New York: Cambridge University 29. CDC. National Health and Nutrition Examination Survey (NHANES). Press; 2010. p. 77–87. Hyattsville, MD: US Department of Health and Human Services; 2012. 6. Felitti VJ, Anda RF, Nordenberg D, Williamson DF, Spitz AM, Edwards V, et al. http://www.cdc.gov/nchs/nhanes.htm. Relationship of childhood abuse and household dysfunction to many of the 30. Taylor VR. Measuring healthy days population assessment of health-related leading causes of death in adults. The Adverse Childhood Experiences (ACE) quality of life. Atlanta, Ga.: U.S. Dept. of Health and Human Services, Centers for Study. Am J Prev Med. 1998;14(4):245–58. Disease Control and Prevention, National Center for Chronic Disease Prevention 7. Brown DW, Anda RF, Tiemeier H, Felitti VJ, Edwards VJ, Croft JB, et al. and Health Promotion, Division of Adult and Community Health; 2000. Adverse childhood experiences and the risk of premature mortality. 31. CDC. Self-reported frequent mental distress among adults–United States, Am J Prev Med. 2009;37(5):389–96. doi:10.1016/j.amepre.2009.06.021. 1993–1996. MMWR Morb Mortal Wkly Rep. 1998;47(16):326–31. 8. Danese A, Moffitt TE, Harrington H, Milne BJ, Polanczyk G, Pariante CM, et al. 32. CDC. Frequently Asked Questions: Which cutpoints can I use to look at Adverse childhood experiences and adult risk factors for age-related Frequent Mental Distress or Frequent Physical Distress? National Center for disease: depression, inflammation, and clustering of metabolic risk markers. Chronic Disease Prevention and Health Promotion. Atlanta, Georgia: Division Arch Pediatr Adolesc Med. 2009;163(12):1135–43. doi:10.1001/ of Population Health; 2011. http://www.cdc.gov/hrqol/faqs.htm#10. archpediatrics.2009.214. Accessed 07/21/2014. 9. Anda RF, Croft JB, Felitti VJ, Nordenberg D, Giles WH, Williamson DF, et al. 33. Mentoring program enhances CM training. Healthcare Benchmarks Qual Adverse childhood experiences and smoking during adolescence and Improv. 2009;16(9):99–102. adulthood. JAMA. 1999;282(17):1652–8. 34. Hillis SD, Anda RF, Dube SR, Felitti VJ, Marchbanks PA, Marks JS. The 10. Corso PS, Edwards VJ, Fang X, Mercy JA. Health-related quality of life among association between adverse childhood experiences and adolescent adults who experienced maltreatment during childhood. Am J Public pregnancy, long-term psychosocial consequences, and fetal death. Health. 2008;98(6):1094–100. doi:10.2105/AJPH.2007.119826. Pediatrics. 2004;113(2):320–7. 11. Chapman DP, Whitfield CL, Felitti VJ, Dube SR, Edwards VJ, Anda RF. 35. CDC. Behavioral Risk Factor Surveillance System Questionnaire, 2013. 2012. Adverse childhood experiences and the risk of depressive disorders in http://www.cdc.gov/brfss/questionnaires/pdf-ques/2013%20BRFSS_English.pdf. adulthood. J Affect Disord. 2004;82(2):217–25. doi:10.1016/j.jad.2003.12.013. 36. Arday DR, Tomar SL, Nelson DE, Merritt RK, Schooley MW, Mowery P. State 12. Dube SR, Anda RF, Felitti VJ, Chapman DP, Williamson DF, Giles WH. smoking prevalence estimates: a comparison of the Behavioral Risk Factor Childhood abuse, household dysfunction, and the risk of attempted suicide Surveillance System and current population surveys. Am J Public Health. throughout the life span: findings from the Adverse Childhood Experiences 1997;87(10):1665–9. Study. JAMA. 2001;286(24):3089–96. 37. Brener ND, Billy JO, Grady WR. Assessment of factors affecting the validity of 13. Dube SR, Felitti VJ, Dong M, Giles WH, Anda RF. The impact of adverse self-reported health-risk behavior among adolescents: evidence from the childhood experiences on health problems: evidence from four birth scientific literature. J Adolesc Health. 2003;33(6):436–57. cohorts dating back to 1900. Prev Med. 2003;37(3):268–77. 38. Illner AK, Nothlings U, Wagner K, Ward H, Boeing H. The assessment of 14. Dube SR, Cook ML, Edwards VJ. Health-related outcomes of adverse individual usual food intake in large-scale prospective studies. Ann Nutr childhood experiences in Texas, 2002. Prev Chronic Dis. 2010;7(3):A52. Metab. 2010;56(2):99–105. doi:10.1159/000277667. doi:A52 [pii]. 39. Thompson FE, Kipnis V, Subar AF, Krebs-Smith SM, Kahle LL, Midthune D, et 15. Loria AS, Ho DH, Pollock JS. A mechanistic look at the effects of adversity al. Evaluation of 2 brief instruments and a food-frequency questionnaire to early in life on cardiovascular disease risk during adulthood. Acta Physiol estimate daily number of servings of fruit and vegetables. Am J Clin Nutr. (Oxf). 2013. doi:10.1111/apha.12189. 2000;71(6):1503–10. 16. Salihu HM, August EM, Alio AP, Jeffers D, Austin D, Berry E. Community-academic 40. Cohen S, Williamson G. Perceived stress in a probability sample of the partnerships to reduce black-white disparities in infant mortality in Florida. Prog United States. In: Spacapan S, Oskamp S, editors. The social psychology of Community Health Partnersh. 2011;5(1):53–66. doi:10.1353/cpr.2011.0009. health: Claremont Symposium on applied social psychology. Newbury Park, 17. Salihu HM, August EM, Mbah AK, Alio AP, de Cuba 2nd R, Jaward FM, et al. CA: Sage; 1988. p. 31–67. Effectiveness of a federal healthy start program in reducing the impact of 41. Cohen S, Janicki-Deverts D, Miller GE. Psychological stress and disease. particulate air pollutants on feto-infant morbidity outcomes. Matern Child JAMA. 2007;298(14):1685–7. doi:10.1001/jama.298.14.1685. Health J. 2012;16(8):1602–11. doi:10.1007/s10995-011-0854-1. 42. Sherbourne CD, Stewart AL. The MOS social support survey. Soc Sci Med. 18. Howard KS, Brooks-Gunn J. The role of home-visiting programs in 1991;32(6):705–14. preventing child abuse and neglect. Future Child. 2009;19(2):119–46. 43. Steptoe A, Feldman PJ. Neighborhood problems as sources of chronic 19. Jevitt C, Zapata L, Harrington M, Berry E. Screening for perinatal stress: development of a measure of neighborhood problems, and depression with limited psychiatric resources. J Am Psychiatr Nurses associations with socioeconomic status and health. Ann Behav Med. Assoc. 2006;11:359–63. 2001;23(3):177–85. 20. Salihu HM, Mbah AK, Jeffers D, Alio AP, Berry L. Healthy start program and 44. Cagney KA, Browning CR. Exploring neighborhood-level variation in asthma feto-infant morbidity outcomes: evaluation of program effectiveness. Matern and other respiratory diseases: the contribution of neighborhood social Child Health J. 2009;13(1):56–65. doi:10.1007/s10995-008-0400-y. context. J Gen Intern Med. 2004;19(3):229–36. 21. Bush AJ, Hair JF. An assessment of the mall intercept as a data collection 45. Contact Software Limited. droidSURVEY. droidSURVEY Wellington, New method. J Mark Res. 1985;22(May):158–67. doi:10.2307/3151361. Zealand. 2013. https://www.harvestyourdata.com/products/droidsurvey/. 22. Hornik J, Ellis S. Makhon le-mehkar asakim b-Y. Strategies to secure 46. Hewlett-Packard Company. HP Slate 7 2800 Tablet Product Specifications. compliance for a Mall intercept interview. Tel Aviv: Israel Institute of Hewlett-Packard Development Company, L.P. 2014. http://support.hp.com/ Business Research; 1986. us-en/document/c03726652. Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 12 of 12 47. Preacher KJ, Hayes AF. SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behav Res Methods Instrum Comput. 2004;36(4):717–31. 48. Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol. 1986;51(6):1173–82. 49. Vanderweele TJ, Vansteelandt S, 12. Odds ratios for mediation analysis for a dichotomous outcome. Am J Epidemiol. 2010;172:1339–48. 50. Hayes AF, Preacher KJ. Statistical mediation analysis with a multicategorical independent variable. The British journal of mathematical and statistical psychology. 2013. doi:10.1111/bmsp.12028. 51. Preacher KJ, Hayes AF. Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behav Res Methods. 2008;40(3):879–91. 52. Blane D, Netuveli G, Stone J. The development of life course epidemiology. Rev Epidemiol Sante Publique. 2007;55(1):31–8. doi:10.1016/j.respe.2006.12.004. 53. Leung MW, Yen IH, Minkler M. Community based participatory research: a promising approach for increasing epidemiology's relevance in the 21st century. Int J Epidemiol. 2004;33(3):499–506. doi:10.1093/ije/dyh010. Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color figure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Health and Quality of Life Outcomes Springer Journals

Adverse childhood experiences and health-related quality of life in adulthood: revelations from a community needs assessment

Download PDF
12 pages

Loading next page...
 
/lp/springer-journals/adverse-childhood-experiences-and-health-related-quality-of-life-in-efdAAgDJY9

References (66)

Publisher
Springer Journals
Copyright
Copyright © 2015 by Salinas-Miranda et al.
Subject
Medicine & Public Health; Quality of Life Research; Quality of Life Research
eISSN
1477-7525
DOI
10.1186/s12955-015-0323-4
pmid
26260580
Publisher site
See Article on Publisher Site

Abstract

Background: Adverse childhood experiences (ACE) have been previously linked to quality of life, health conditions, and life expectancy in adulthood. Less is known about the potential mechanisms which mediate these associations. This study examined how ACE influences adult health-related quality of life (HRQoL) in a low-income community in Florida. Methods: A community-based participatory needs assessment was conducted from November 2013 to March 2014 with 201 residents of Tampa, Florida, USA. HRQoL was measured by an excessive number of unhealthy days experienced during the previous 30-day window. Mediation analyses for dichotomous outcomes were conducted with logistic regression. Bootstrapped confidence intervals were generated for both total and specific indirect effects. Results: Most participants reported ‘good to excellent health’ (76 %) and about a fourth reported ‘fair to poor health’ (24 %). The mean of total unhealthy days was 9 days per month (SD ±10.5). Controlling for demographic and neighborhood covariates, excessive unhealthy days was associated with ACE (AOR = 1.23; 95 % CI: 1.06, 1.43), perceived stress (AOR = 1.07; 95 % CI: 1.03, 1.10), and sleep disturbance (AOR = 8.86; 3.61, 21.77). Mediated effects were significant for stress (β = 0.08) and sleep disturbances (β = 0.11) as they related to the relationship between ACE and excessive unhealthy days. Conclusion: ACE is linked to adult HRQoL. Stress and sleep disturbances may represent later consequences of childhood adversity that modulate adult quality of life. Background during childhood can also have a long-term impact on Racial/ethnic minorities in the United States suffer a dis- health [3], and consequently, quality of life. proportionately high burden of morbidity and mortality According to the Centers for Disease Control and compared to their non-Hispanic white counterparts [1], Prevention (CDC), more than half of adults in the U.S. particularly with respect to adverse pregnancy outcomes, have suffered from adverse childhood experiences childhood illnesses, and adult chronic diseases [1]. These (ACE), such as verbal, physical abuse and family dys- health differences are likely the product of complex re- function [4]. These exposures have been linked to a variety lationships across social, economic, environmental, of health conditions in adults, including depression, healthcare, bio-behavioral, structural factors such as cardiovascular disease, cancer, diabetes [5, 6], as well as pre- discrimination and racism, as well as literacy and legis- mature mortality [7]. Moreover, research suggests that ACE lative policies [2]. In addition to stressors experienced is associated with long-term changes in the nervous, endo- during adulthood, stressors and adverse experiences crine, and immune systems [3, 8]. This damage to the body’s stress response system, coupled with the adoption of * Correspondence: [email protected] poor health behaviors to help cope with stress [9], appears Department of Family and Community Medicine, Baylor College of to contribute to a deterioration of adult health. Indeed, the Medicine, 3701 Kirby Drive, Suite 600, Houston, Texas 77098, USA enduring impact of ACE on health suggests that the overall Full list of author information is available at the end of the article © 2015 Salinas-Miranda et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 2 of 12 health-related quality of life (HRQoL) for these adults may Design also be impacted [10]. Although evidence exists linking ad- The community needs assessment survey was a cross- versity during the childhood years (i.e., adverse childhood sectional instrument designed to assess factors associ- experiences or ACE) to impaired HRQoL and shortened ated with HRQoL and was administered between life expectancy in adulthood [7, 11–15], little is known November 2013 and March 2014. Two-hundred and one about the potential mechanisms that mediate the relation- adult participants were recruited from approximately ship. Further, there have been even fewer empirical studies 110,451 residents of the target population [18]. The tar- conducted in socioeconomically disadvantaged popula- get community was predominantly African American tions assessing a wide range of interconnected risk and (60 % black, 18.3 % white, 12.1 % Hispanic, and 9.6 % protective factors that influence the HRQoL of individuals other) and tended to be economically disadvantaged, within a community context. To improve the health and with half the income and double the unemployment rate qualityoflifeofminoritypopulations,anenhancedun- of the rest of the county [19, 20]. The survey was admin- derstanding of the factors that contribute to health dis- istered through intercept interviews across a five zip parities is needed. Accordingly, we conducted this study code area [21, 22]. Intercept interviews are a type of so- to examine how ACE influences adult-onset HRQoL in cial marketing research in which respondents are a community setting. We also explored the roles played approached by culturally and linguistically matched re- by key mediating socio-demographic and other neigh- search staff and “intercepted” in various places where borhood level factors and assessed plausible mediating community members gather often [23]. This method is pathways. used to collect more representative information than con- venience samples regarding particular issues of concern to the community, and intercept interviews are more feasible Methods and less expensive than door-to-door needs assessments. Context Intercept interviews are also useful when the target popula- The project “Toward Eliminating Disparities in Mater- tion is widely dispersed and harder to reach (e.g., several nal and Child Health Populations” is a 3-year CBPR zip code-level areas with economically disadvantaged urban (community-based participatory research) initiative that communities). Accordingly, trained CAB members sur- partners researchers at the University of South Florida veyed people “in the streets” at frequently used community (USF), a non-profit advocacy and empowerment orga- locations (e.g., cafes, churches, libraries, local schools gath- nization, REACHUP, Inc., and a group of community erings, shopping centers) and at times when community representatives from selected zip codes in Central residents were accessible (e.g., weekends, after hours during Tampa, Hillsborough County, Florida. The project has weekdays). In a similar manner to 30 by 7 cluster sampling three sequential phases: 1) needs assessment to identify [24], the CAB nominated 30 community locations or “clus- priority maternal and child health issues, 2) planning ters” (23 were actually surveyed due to redundancy of par- an evidence-based intervention that addresses the ticipants in 7 locations) based on socio-demographic and priorities identified, and 3) implementation of a pilot economic characteristics using zip code level census data, community-driven intervention. The project is funded and then randomly sampled at least seven individuals from by the National Institute on Minority Health and each location. To participate in the survey, a respondent Health Disparities (5R24MD8056-02). must have been a community resident for at least the previ- This paper focuses exclusively on the needs assess- ous 12 months. Flyers, social media, and “word-of- ment survey, which was designed to explore health de- mouth” transmission of information were used for re- terminants and quality of life indicators in the target cruitment. Written informed consent was obtained and community, implemented using a CBPR approach. This a modest but appropriate monetary incentive was given CBPR study builds extensively upon a strong, existing to study participants. Approval for the study was obtained community-academic partnership. At the study’s outset, from the Institutional Review Board of the University of a Community Advisory Board (CAB) was created, which South Florida. consisted of eight adult residents from the target com- munity who were recruited based upon their knowledge, Measures participation, advocacy, and leadership in previous The development of survey questions was guided by community projects [16, 17]. In addition to the required the Life Course Perspective (LCP) with extensive input Human Subjects Protection Certification course, all from the CAB. The LCP framework is used to evaluate CAB members completed skill-building workshops in the cumulative influences of risk and protective factors research methods. Active participation of community during critical periods of human development and the members occurred in the design, data collection, and effects those events have on the health trajectories of analysis phases of the study. individuals [25–27]. First, community members, in Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 3 of 12 partnership with academic researchers, reviewed local Primary exposure: adverse childhood experiences data and voiced their concerns for critical topics. Next, To assess cumulative risk from childhood to adulthood, the academic researchers suggested questionnaires or scales main predictors included participant recall of adverse that had been validated in previous research studies for events that occurred during the first 18 years of life. We local adaptation. Subsequently, the CAB provided feedback used the Brief Family History Questionnaire from the ACE on wording and readability, question addition/deletion, study [6], a 10-item questionnaire collecting self-reported assessed acceptability of questions and technology usability physical, emotional, and sexual abuse, family dysfunction, (see Technology section), and pilot tested questions before and economic hardship. All items are dichotomous, yes/no community-wide implementation. The final survey ques- questions. An overall ACE score, which ranges from 1–10, tionnaire contained 63 questions (Likert-type and multiple is then calculated by adding the number of “yes” responses. choice questions), which covered the following inquiry Higher scores have been found to be associated with a wide domains and specific sequencing into the survey: life in the range of adverse health outcomes, impaired quality of life, neighborhood (neighborhood assets; community-wide and higher mortality [7, 11, 34]. issues), social connections (social support), health and qual- ity of life (general health, HRQoL, self-reported health Potential mediators: smoking, alcohol use, diet, physical problems, sleepless days, perceived social standing), stress activity, and sleep disturbances and unfair treatment (stress appraisal, perceived experi- We hypothesized that contemporary lifestyle risk and pro- ences of discrimination), lifestyle (smoking, alcohol use, tective factors of participants could mediate the association recreational drugs, diet, exercise, perceived HIV risk), between ACE and adult HRQoL. Lifestyle and behavioral childhood experiences (ACE), and socio-demographic questions were obtained from multiple validated instru- questions (age, education, marital status, race/ethnicity, ments that are frequently used in health services research, household income, employment). Academic researchers which were adapted using input from community assisted with the development of hypotheses, analytic members. Smoking was assessed with a question from the strategies, and statistical analyses. In the subsections 2013 Behavioral Risk Factor Surveillance System (BRFSS) below, we describe only the specific instruments and [35]: “Have you smoked at least 100 cigarettes in your en- measures used to test our main hypotheses; however, tire life?” This question was chosen by CAB members the complete survey is available as a supplemental as being easier to respond to than a detailed frequency file (Additional file 1). question [36]. The frequency of alcohol use per month was measured with a numerical question, also adapted from the 2013 BRFSS: “In a typical month, how often Primary outcome: health-related quality of life do you drink alcoholic beverages?” HRQoL was the primary study outcome and was mea- Although self-reported measures of dietary patterns sured using the CDC’s “Healthy Days Measure” instru- are vulnerable to measurement bias [37, 38], the CAB ment, which is a validated scale used frequently in members felt that asking a few questions about diet was national health surveillance surveys [28, 29]. Specific- still an important aspect to include in the survey as a ally, we used the brief version, referred to as Healthy general, non-sensitive behavioral item. Thus, the survey Days Core Module (4-items questionnaire), which cap- included a short series of questions about intake of fruits tures the self-reported number of days in the past and vegetables, fatty, sugary, salty foods, and caffeinated 30 days that individuals rated their physical or mental drinks. Participants were asked to indicate how often health as not good [30]. It includes the following com- they consumed fruits and vegetables in a typical month, ponents: 1) self-rated health, from poor to excellent (or- and we grouped responses as: “once a month or less” dinal); 2) number of days when physical health was not coded as 1, “2–3 times a month” coded as 2.5, “once a good during the past 30 days; 3) number of days when week” coded as 4, “2–3 times a week” equivalent to 10, mental health was not good during the past 30 days; and “4–6 times a week or more” coded as 20. For this and 4) number of activity limitations due to either phys- study, we focused on the consumption of fruits and ical or mental health illness (combined). Total number vegetables as key protective factor because of stronger of unhealthy days is then obtained by adding the re- evidence regarding fruits and vegetable self-reported sponses to questions 2 and 3. Any sum greater than 30 measures. Specifically, brief instruments for fruit and was capped at 30, with a maximum of 30 unhealthy days vegetable intake assessments have been found to be ad- [30]. The outcome was operationalized through dichotomi- equate for estimating relative risks in the relationship zation of the total number of unhealthy days, with poor between fruit and vegetable intake versus disease [39]. HRQoL reflected by 14 or more total days. This cut-point Because the survey included questions not previously has been used by other authors as a discriminate measure validated, we consider our approach a conservative one. to assess excessive unhealthiness [31–33]. Moreover, because we did not find that any of our Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 4 of 12 self-reported dietary measures were significantly associ- Neighborhood confounders ated with HRQoL in bivariate preliminary analyses, we Community resources, community-wide issues, and did not explore dietary items in greater detail in media- neighborhood cohesion levels were considered as tional analyses. Sleep disturbances were also measured potential confounders in our analyses. First, resources with a question we derived from the BRFSS [28]: “During available in the community were measured with the fol- the past 30 days, for about how many days have you felt lowing question: “Which of the following resources are you did not get enough rest or sleep?” Physical activity available to you in your community?” A list of resources was measured with one question that assessed physical ac- was compiled using CAB input and provided with the tivity, also from the BRFSS [28]: “During the past month, survey. Each item listed was weighted equally and the about how many days per week did you exercise for recre- final variable used in analyses was the total number of ation or to keep in shape (activities that make you assets or resources reported. Similarly, perceived sweat)?” community-wide issues were measured with one ques- tion on neighborhood problems [43]: “Which of the fol- Stress appraisal lowing is a problem in the neighborhood?” Again, a list was Cognitive appraisal of stress was measured with the 4-item developed by the CAB, with the allowance for write-in en- Perceived Stress Scale, which is a validated instrument used tries. The final variable for analysis consisted of the total to make comparisons of subjects’ perceived stress related to number of different issues reported. Lastly, neighborhood current events [40]. Questions are 5-point Likert type social cohesion was assessed by measuring the respondent’s scaled (i.e., strongly disagree = 1, to strongly agree = 5). A level of agreement (on a 5-point Likert type scale) with a composite score was derived by adding the original set of questions proposed by Cagney and colleagues [44]: 1) scores and multiplying by a factor of 5, which results in “People around here are willing to help their neighbors”;2) a 100-point scale. The higher the score, the higher the “This is a close-knit neighborhood”;3) “People in this risk for clinical psychiatric disorder [41]. neighborhood can be trusted”; and 4) “People in this neigh- borhood generally don’t get along with each other” (reverse Social support coded). These questions were then summed to provide a Perceived social support was measured with five questions total score, where higher scores indicated higher neighbor- from the Medical Outcomes Study Social Support Survey hood social cohesion. [42]. Using 5-point Likert type scales (‘None of the time’ =1 to ‘All of the time’ = 5), individuals were asked Technology to indicate how often the following types of supports The droidSURVEY software was used to design and were available to them if needed: 1) someone to con- administer the survey [45], which was installed on ten fide in or talk about yourself or your problems; 2) Hewlett-Packard Slate 7” portable tablet computers someone to share your most private worries and fears running the Android™ 4.2.2 operating system [46]. For with; 3) someone to help you if you were confined to Spanish-speaking participants, all questions were trans- bed; 4) someone to prepare your meals if you were un- lated to Spanish by native speakers and assessed for ac- able to do it yourself; and 5) someone to get together curacy of translation using back translation and pilot with for relaxation. Question scores were added to testing. The use of tablet-administered surveys allowed yield a total social support score. To facilitate inter- for a portable, convenient means of data collection, and pretation in the community setting, the original scores pilot testing revealed that community members found ranging from 4 to 20 were converted to a 100-point touchscreen technology to be intuitive, easy to use, and scale by multiplying the original score by five. enjoyable. Supervised by the principal investigator, the study coordinator trained community members in the Sociodemographic confounders use of tablets and survey implementation, the informed Socio-demographic characteristics were collected from consent process, management of tablet computers in the participants and included: age in years (categorized as field, and how to assist participants with questions about 18–35; 36–45; 46–55; and ≥ 56 years), sex, education the study. Training consisted of three 2-hour sessions in (high school vs. less than high school graduate or the preceding month to the survey. equivalent), current marital status, race/ethnicity (non- Hispanic white, non-Hispanic black, Hispanic, other), Statistical analysis employment status (employed, unemployed but able, The study population was described using descriptive and unable to work), annual household income (categor- statistics that included frequencies and percentages for ies: US$ 0–20,000; 20,001–40,000; and ≥ 40,000), and categorical variables and means and standard deviations residential time in the five target zip codes (5 years or for numerical variables. Two-sided tests of equality for less vs. more than 5 years). column proportions (z-tests for column proportions or Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 5 of 12 t-tests for column means) were conducted to assess Results differences by outcome group. Tests assumed equal Participants’ characteristics, neighborhood factors, and variances and were adjusted for multiple comparisons health-related quality of life using a Bonferroni correction. All analyses were con- Table 1 describes the study sample, by socio-demographic ducted with the IBM SPSS Statistics for Windows, Version characteristics. In total, 201 participants were surveyed, of 22.0 (IBM Corp, Armonk, NY). Statistical significance was which the majority were female (65.7 %) and non- assessed at the 0.05 level. Hispanic black (66.5 %), with a mean age of 45 ± 14 years. Under the LCP framework, we posited that numerous Nearly one-fifth of participants did not graduate from high accumulating risks and protective factors mediate the school, most were not married (67 %), and only 40 % were association between ACE and the HRQoL proxy. As a first currently employed. Income was not reported by 13 % of step, separate logistic regression models were run to assess participants. Among those who reported income, the ma- the independent effects of life course social determinants jority had incomes lower than $20,000 (about 70 %). on the outcome (≥14 unhealthy days), adjusting for indi- Moreover, nearly all participants (97.5 %) reported receiving vidual socio-demographic and neighborhood covariates. at least one form of social assistance based on federal pov- We explored the role of the following factors: stress, sleep erty line categorizations (i.e., Section 8-Housing Choice disturbances, smoking, alcohol use, physical activity, fruits Voucher Program, food stamps, school free or reduced and vegetable intake, and social support. The purpose lunches, Medicaid, supplemental security income, Tempor- was to identify significant independent associations ary Assistance for Needy Families, or other). This was also with HRQoL that could represent potential mediating a relatively transient population, with as many residents pathways in the relationship between ACE and HRQoL. who had lived for 5 or more years in the neighborhood as After a set of factors was identified as possible mediators, those that had just moved in the last 5 years. the next step was to test for mediation or indirect effects Participants demonstrated awareness of community [47]. The following three conditions must be established to resources, reporting an average of 8 community assets determine whether mediation had occurred (i.e., indirect (range was from 0 to 13 reported assets). The most effects) [48]: 1)thatthe independentvariable(ACE) pre- frequently cited resources were churches and schools dicts the dependentvariable(HRQoL),2)thatthe inde- (Fig. 1). Other high-ranking community resources (in pendent variable (ACE) predicts the mediators (Ms), 3) that order of highest frequency to least) were public trans- the mediator (M) predicts the dependent variable (i.e., ≥14 portation, parks/recreational facilities, police/fire de- unhealthy days). Additionally, the effects of confounders partments, pharmacies, libraries, and hospitals/clinics. were considered [49]. Participants also reported community-wide issues that Accordingly, mediation analyses for the dichotomous were a perceived problem for their health and their outcome (Y: HRQoL) were conducted with logistic re- neighborhood (Fig. 2). On average, participants noted gression models to estimate the path coefficients in a about 4 community-wide issues (SD ± 2.9). Drug and two-mediator model (X: ACE, M1: stress, M2: sleep dis- alcohol related issues were the most frequently stated turbance). We controlled for socio-demographic covari- followed by abandoned property, litter, homeless issues, ates (i.e., age, gender, race/ethnicity, no high school, and high crime rate, lack of affordable shopping, poor police household income) and neighborhood level factors (i.e., response, poor quality grocery stores, and lack of parks/ neighborhood cohesion, community resources, and recreational facilities. neighborhood issues). The first step was the estimation Mean neighborhood cohesion was 65.33 (SD ±18.7), of controlled direct effects through a series of logistic re- and the majority (60.3 %) of participants perceived that gressions. This was followed by the decomposition of people in their neighborhood generally got along with effects into total, indirect, and direct effects [50]. The each other. More than half (57 %) perceived their formula used to calculate indirect effects for both mediators neighborhood as a place where people were willing to was: c=c’ +(ab);where c is total effect, c’ is the direct effect, help each other. On the other hand, only 41.5 % of and a*b is the indirect effect. The indirect effect (ab) is the participants agreed or strongly agreed that people in measure of the amount of mediation, and equals the reduc- their neighborhood shared the same values. In tion of the effect of the causal variable on the outcome or: addition, only 40.4 % of participants agreed or strongly ab = c - c’. For this purpose, we used an SPSS macro with agreed that they lived in a close-knit neighborhood. asymptotic and resampling strategies for comparing indir- Approximately three of every four participants consid- ect effects in multiple mediator models, which included ered themselves in “good to excellent health.” The mean bootstrapping to estimate confidence intervals for total and number of unhealthy days due to poor physical health or specific indirect effects [47, 51]. Missing values were han- injury in the past month was 4.9 days (SD ± 7.8). The mean dled with default option in SPSS for logistic regression pro- number of unhealthy days due to stress, depression or cedure, which is listwise deletion. problems with emotions was 5.3 days (SD ± 8.4). When Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 6 of 12 Table 1 Unhealthy days during the previous 30 days by participants’ characteristics and neighborhood factors Total 0–13 unhealthy days a month 14 or more unhealthy days a month N (%) Count (%) Count (%) Age (in years) 18–35 years 62 (31.0) 46 (31.3) 16 (30.2) 36–45 years 39 (19.5) 31 (21.1) 8 (15.1) 46–55 years 50 (25.0) 38 (25.9) 12 (22.6) 56 plus years 49 (24.5) 32 (21.8) 17 (32.1) Sex Female 132 (65.7) 93 (62.8) 39 (73.6) Male 69 (34.3) 55 (37.2) 14 (26.4) Education High school complete 162 (80.6) 124 (83.8) 38 (71.7) No high school 39 (19.4) 24 (16.2) 15 (28.3) Marital status Currently married 65 (32.7) 54 (37.0) 11 (20.8) Not married now 134 (67.3) 92 (63.0) 42 (79.2) Race/ethnicity Non-Hispanic whites 13 (6.5) 8 (5.4) 5 (9.4) Non-Hispanic blacks 133 (66.5) 98 (66.7) 35 (66.0) Hispanic or Latino 54 (27.0) 41 (27.9) 13 (24.5) Household annual income $0–20,000 114 (56.7) 77 (38.3) 37 (18.4) $20,001–40,000 36 (18.0) 29 (14.4) 7 (3.5) $40,001 or more 24 (11.9) 21 (10.4) 3 (1.5) Income not reported 27 (13.4) 21 (10.4) 6 (3.0) Residential time 5 years or less 101 (50.2) 76 (51.4) 25 (47.2) More than 5 years 100 (49.8) 72 (48.6) 28 (52.8) Neighborhood factors Mean SD Mean SD Mean SD Community assets 7.78 (3.5) 7.88 3.38 7.49 3.92 Community wide issues 4.11 (2.9) 4.03 2.93 4.36 2.97 Neighborhood cohesion 65.33 (18.7) 66.03 18.81 63.40 18.73 Column comparisons of 0–13 unhealthy days a month vs. 14 or more unhealthy days. Values were significantly different at p < .05 in the two-sided test of equality for column proportions (z-tests for column proportions or t-tests for column means). Some column numbers do not add to total sample size due to missing values combined both physically and mentally unhealthy days, higher proportion of individuals with incomes lower than the mean total unhealthy days was about 9 days a $20,000 occurred among those who reported ≥14 or more month (SD ± 10.54). About a fourth of participants unhealthy days. We found no significant differences by (26 %) reported having 14 or more total unhealthy days other socio-demographic variables. When examining the per month (Table 1). The most common self-reported potential role of neighborhood factors, we found no signifi- personal health issues included back/neck problems, cant differences by assets, perceptions of community-wide stress, vision problems, arthritis, problems walking, issues, nor neighborhood social cohesion. depression/anxiety, and weight issues (Fig. 3). Marital status was significantly associated with ≥14 or Independent adjusted effects of life course determinants more unhealthy days (Table 1). Specifically, a higher pro- on unhealthy days portion of unmarried individuals in the group reported ≥14 Table 2 presents separate models for each determinant, or more unhealthy days (79.2 %). Also, low income was as- adjusting by age, sex, race/ethnicity, education, income, sociated with ≥14 or more unhealthy days, specifically a community assets, community issues, and neighborhood Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 7 of 12 Churches Schools Public transportation Parks and recreational facilities Police/Fire Department Pharmacies Libraries Hospitals and clinics Mental health facilities Colleges, universities Local businesses Community recreation centers 0.0 20.0 40.0 60.0 80.0 100.0 Percentages (N=201) Fig. 1 Community assets reported by survey participants. Notes: Percentages based on N = 201 cohesion. ACE was associated with 23 %-increased odds significantly predicted poor HRQoL. The model correctly of poor HRQoL (≥14 unhealthy days) in adulthood predicted 73 % of cases with excessive unhealthy days. In (AOR = 1.23; 95 % CI: 1.06, 1.43). Increasing stress was Model 2, this relationship was explained by stress and the also associated with increased likelihood of reporting ≥14 model had greater explanatory power (Pseudo-R =0.32) unhealthy days (AOR = 1.07; 95 % CI: 1.03, 1.10). Sleep than the first model. In Model 3, the addition of sleep disturbances were strongly associated with reports of 14 disturbances improved the explanatory power (Pseudo- or more unhealthy days. Specifically, any additional sleep R = 0.44) and the percent of correct predictions was deprived day was associated with 9 times higher odds of increased to 82.1 %. All three models demonstrated reporting ≥14 unhealthy days (AOR = 8.86; 95 % CI: 3.61, adequate goodness of fit (Hosmer-Lemeshow non- 21.77). After adjusting for covariates, no associations were significant). On the other hand, these three initial logistic found for smoking, alcohol use, physical activity, fruits and regression models only indicate measures of associations vegetable intake, or social support. Based on these find- with controlled effects and cannot estimate the magnitude ings, we selected stress and sleep disturbance as poten- of mediated or indirect effects (i.e., effect of ACE on tial mediators that needed to be examined in the test of Stress, plus the effect of stress on unhealthy days) or if the mediation. mediation is partial or total [47]. Table 4 presents the decomposition of effects into total, Test of mediation indirect, and direct effects for the HRQoL measure. The Table 3 presents the stepwise assessment of controlled three essential conditions to determine mediation were direct effects of ACE, stress, and sleep disturbances on established, even after controlling for SES and neigh- HRQoL. In Model 1, it can be appreciated that ACE borhood level factors. First, ACE (main predictor) Drug/alcohol-related problems Abandoned houses, factories, or buildings Trash or litter in empty lots, streets, or… Homeless persons/panhandling Vandalism, high crime rate Lack of affordable shopping Poor police response Lack of quality of recreation, parks, or… Poor quality grocery Poor quality of schools, or libraries Dogs and uncontrolled animals Junkyards, gasoline stations, and other… Hazards: incinerators, chemical plants, and… 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 Percentages (N=201) Fig. 2 Community-wide issues reported by survey participants. Notes: Percentages based on N = 201 Community Issues Community Assets Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 8 of 12 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 Self-Reported Health Issues in the Previous Year Fig. 3 Self-reported health issues among survey participants. Notes: Percentages based on N = 201 independently predicted self-reports of ≥14 unhealthy of stress and sleep disturbances was 0.19 (95 % CI: 0.06, days due to physical or mental illness (outcome). Second, 0.33). By exponentiation of the later beta coefficient, we ACE significantly predicted two mediators, namely per- obtain an OR of 1.20 (95 % CI: 1.06, 1.39). In other ceived stress and sleepless days. Third, these mediators words, there is an increase of 20 % in the odds of report- (stress and sleep disturbance) significantly predicted un- ing ≥14 unhealthy days for every additional ACE, which healthy days. is mediated by stress and sleep disturbances in the adult- Mediated effects were significant for stress (β = 0.08; hood. It should be highlighted that such effect size is 95 % CI: 0.01, 0.21) and sleep disturbances (β = 0.11; very close to the independent controlled effect of ACE 95 % CI: 0.03, 0.21) on the relationship between ACE on unhealthy days that was noted in Model 1 of Table 3. and excessive unhealthy days. It can be noted that when testing for mediation the direct effect of ACE on un- Discussion healthy days (c’ path) became non-significant, which in- We found that exposure to ACE is linked to impaired dicates total mediation. The joint total mediated effect adult HRQoL, with mediation effects modulated by stress and sleep disturbances. An implied hypothesis of this study was that there was an association between ACE and Table 2 Independent unadjusted and adjusted effects of adverse childhood experiences and current social HRQoL. Our study confirms this association and is con- determinants on the odds of reporting ≥14 days of unhealthy sistent with previous findings [7, 11–15]. Unique to our days per month study, however, is the way in which ACE was measured Variable ≥14 days of unhealthy days per month using the widely established theoretical framework, the LCP, to capture cumulative life experiences over time. OR (95 % CI) AOR (95 % CI) a a The LCP provided the theoretical guidance to frame the Adverse childhood experiences 1.19 (1.05, 1.34) 1.23 (1.06, 1.43) a a inquiry, while CBPR fostered active community engage- Stress 1.05 (1.03, 1.08) 1.07 (1.03, 1.10) ment in the design of research questions that are rele- Social support .99 (.98, 1.01) .99 (.97, 1.01) vant to the community context and in the collection of Smoking 1.50 (.79, 2.86) 1.30 (.59, 2.89) reliable experiential information [52, 53]. The multi-do- Alcohol use 2.54 (1.02, 6.33) 2.78 (.93, 8.31) main conceptualization and multi-level nature of LCP Physical activity .90 (.46, 1.69) .66 (.31, 1.39) along with CBPR approaches have the potential of im- proving the assessment and deepening the understanding Fruits and vegetable intake 1.04 (.55, 1.96) 1.32 (.61, 2.83) a a of determinants of health disparities. It is precisely because Sleep disturbances 6.08 (3.03, 12.24) 8.86 (3.61, 21.77) of these notable features, namely, integration (i.e., risk/ Statistically significant 95 % CI protection), multilevel (i.e., individual, families, and com- Separate models for each determinant, adjusting by age, sex, race/ethnicity, education, income, community assets, community issues, and munities), and time-orientation (i.e., life span), that the neighborhood cohesion c LCP was able to assist in explaining why and how health With the exception of smoking (yes/no), all other predictor variables were on a continuous measurement scale disparities occur. Percentage (N=201) Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 9 of 12 Table 3 Excessive unhealthy (≥14) days predicted by adverse childhood experiences and step-wise inclusion of mediators Variables Model 1 Model 2 Model 3 a a a B (95 % CI) B (95 % CI) B (95 % CI) Main Predictor Adverse childhood experiences .21 (.05, .43) .104 (−.06, .34) .03 (−.24, .28) Mediators b b Stress .061 (.03, .158) .05 (.01, .12) Sleep disturbances .10 (.06, .23) Confounders Age .01 (−.027, .04) .03 (−.01, .08) .03 (−.01, .09) Sex −.19 (−1.33, .76) −.19 (−1.67, .95) .21 (−1.37, 1.84) b b Education .78 (−.36, 2.28) 1.35 (.05, 3.48) 1.85 (.40, 4.38) Marital status .78 (−.19, 2.40) .55 (−.72, 2.02) .87 (−.47, 3.06) Non-Hispanic black −.45 (−2.44, 1.65) −.17 (−1.97, 1.85) −.65 (−3.01, 1.41) Hispanic or Latino −.81 (−3.08, 1.40) −.72 (−2.93, 1.65) −.97 (−3.64, 1.24) US$20,001–40,000 −.44 (−1.88, .71) −.02 (−1.68, 1.34) −.42 (−2.76, 1.20) ≥US$40,001 −.29 (−19.69, 1.30) −.31 (−19.29 1.13) −.30 (−20.16, 1.85) Community assets −.05 (−.19, .08) .07 (−.08, .26) .07 (−.11, .28) Neighborhood issues −.02 (−.23, .15) −.12 (−.34, .08) −.17 (−.49, .03) Neighborhood cohesion −.01 (−.03, .01) −.01 (−.05, .02) −.01 (−.05, .02) Constant −.76 (−5.04, 2.81) −4.32 (−16.16, −.47) −4.7 (−13.87, −.05) Model Chi-square [df] 22.48 [7] 38.32 [11] 56.26 [14] Nagelkerke R .18 .32 .44 % Correct predictions 73.2 80.8 82.1 Goodness of fit p-value .44 .55 .38 Model 1: binary logistic regression with Unhealthy Days as outcome, ACE as predictor, and controlling for age, gender, education, marital status, race/ethnicity, income, community assets, neighborhood issues, neighborhood cohesion as covariates Model 2: Model 1, adding stress (mediator 1) Model 3: Model 2, adding sleep disturbance (mediator 2) N = 151 for model 1,2, and 3 Bootstrapped 95 % confidence intervals based on 1000 bootstrapped samples. All values rounded to two digits. Indicates that the coefficient is statistically significant at, at least, the .05 level Another hypothesis in this study was that certain symp- intervene immediately. However, there are circumstances toms are expressed later in life among victims of ACE, (e.g., poverty, hidden abuse, etc.) which may elude early de- resulting in a symptomatology complex that includes a tection. In such instances, the other approach would be physically and emotionally poor quality of life. Our study detection of ACE through screening measures with inter- is novel in this perspective because we conducted a detailed vention coming later in life. For example, those impacted and robust mediation analysis, which mapped out mechan- by ACE may exhibit somatization by expressing symptoms ismal pathways that could potentially explain early life ex- such as sleep and stress disturbances and may greatly periences and subsequent impaired HRQoL. Specifically, benefit from improved tertiary care. we observed that ACE victims in this study were more It is important to place our results within the context of a likely to have heightened stress levels and sleep distur- methodological limitation, namely that the survey upon bances, a finding that underscores the importance of their which our findings were based was cross-sectional. None- role as potential mediators linking ACE to HRQoL. There theless, it is noteworthy that a merit of the study in- are clinical as well as public health implications of these ob- strument is that the questions were framed within servations. In particular, they represent potential avenues specific temporal windows, and therefore, the temporal for intervention to minimize the negative impact of ACE relationships between ACE, HRQoL, and the other on HRQoL. factors included could be established. Another short- There are two main approaches of effective strategies to coming of our analysis is that while our assessment improve the HRQoL among adults impacted by ACE. The was based on a representative community sample, it first is to identify those impacted by ACE early in life and cannot be taken as directly generalizable to the entire Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 10 of 12 Table 4 Mediation of stress on the relationship between approach from the “womb to the tomb,” which starts adverse childhood experiences and excessive unhealthy days by addressing psychological well-being and behavioral a, b Path Coefficients S.E. health of expecting mothers and continue to provide support to minimize the effect of unresolved childhood IV to Mediators (a paths): ACE to Stress and Sleep traumas, as well as the linkages to professional and Perceived stress 1.64 .54 community supports throughout the life span. This Sleep disturbance 1.09 .30 offers a potential path to the improvement of HRQoL Direct Effects of Mediators on DV (b paths) for victims impacted by ACE. Perceived stress .05 .02 Sleep disturbance .10 .03 Additional file Total Effect of IV on DV (c path) Adverse childhood experiences .19 .07 Additional file 1: Survey Questions and Flow. (PDF 130 kb) Direct Effect of IV on DV (c’ path) Adverse childhood experiences .03 .09 Abbreviations ACE: Adverse childhood experiences; HRQoL: Health-related quality of life; Partial Effect of Control Variables on DV SD: Standard deviation; AOR: Adjusted odds ratio; CI: Confidence interval; Age .03 .02 CDC: Centers for Disease Control and Prevention; CBPR: Community-based participatory research; USF: University of South Florida; CAB: Community Sex .18 .57 Advisory Board; LCP: Life Course Perspective; BRFSS: Behavioral Risk Factor Education 1.85 .63 Surveillance System. Marital status .79 .58 Competing interests Race/ethnicity −.43 .49 The authors declare that they have no competing interests. Household income −.23 .39 Community assets .07 .07 Authors’ contributions ASM was involved in developing the needs assessment, participated in the Neighborhood issues −.16 .09 statistical analyses, and was responsible for the first drafts of the Intro, Neighborhood cohesion −.01 .01 Methods, and Results. JLS was responsible for developing and implementing the statistical analysis plan, interpreting the results, and assisting with Indirect Effects of IV(ACE) on DV(Unhealthy days) through Proposed drafting the Intro, Methods, and Results. LK aided with development of the Mediators (ab paths) needs assessment, assisted with data collection, and assisted with drafting Data Boot (95 % CI) the manuscript. JB made considerable contributions to the conceptualization of the study and directed development of the needs assessment. She Total .19 .22 (.06, .33) worked to provide critical appraisal and subsequent revision of the initial Perceived stress .08 .09 (.01, .21) draft. EB, DA, and KS were all instrumental to conception, design, and implementation of the study, acquisition of the data, and critical review and Sleep Deprivation .11 .13 (.03, .21) revision of the manuscript. KKS helped to analyze and interpret the data, and a b All values rounded to two digits. Linear regressions for translate findings into an initial draft of the Discussion. HS was ultimately ‘a paths’ for perceived stress and sleep deprivation. All other coefficients responsible for design of the entire study, oversaw all aspects of instrument derived with binary logistic regression. Formulas: Total Effect: c = c’ + ab; development and data collection, assisting in writing the initial draft of the amount of mediation or indirect effect: ab = c - c’. Based on 1000 Discussion, and provided critical insight that guided revisions. All authors bootstrapped samples. Path coefficient significant at p < 0.05 have provided final approval of this manuscript for submission and agree to be accountable for the work herein. US population, due to the selective context of our Acknowledgments project (e.g., mostly low income, African American This work is supported by funding from the National Institute on Minority Health women). The advantage of this limitation, however, is andHealthDisparities throughanR24 grant on “Community-Based Participatory that the study addressed an important cause of low Research” (Award#: 5R24MD8056-02). The content of this paper is solely the responsibility of the authors and does not necessarily represent the official views HRQoL in a socio-economically disadvantaged setting that of the National Institute on Minority Health and Health Disparities, or the stands to benefitmostfromappropriate andtargeted University of South Florida. interventions for victims of ACE. Author details Department of Epidemiology and Biostatistics, College of Public Health, Conclusions University of South Florida, 13201 Bruce B Downs, MDC56, Tampa, Florida In summary, our findings indicate that adversity in 33612, USA. REACHUP Inc., 2902 N, Armenia Avenue Suite 100, Tampa, Florida 33607, USA. Department of Family and Community Medicine, Baylor childhood continues to affect the mental and behav- College of Medicine, 3701 Kirby Drive, Suite 600, Houston, Texas 77098, USA. ioral health trajectory of adults. Thus, we recommend Department of Community and Family Health, College of Public Health, the implementation of community health programs University of South Florida, 13201 Bruce B. Downs Blvd. MDC 56, Tampa, Florida 33612, USA. aimed at improving psychological well-being by redu- cing high stress levels, particularly among individuals Received: 31 December 2014 Accepted: 24 June 2015 who havesuffered childhood trauma.Wesuggest an Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 11 of 12 References 23. Andreasen AR. Academy for Educational D. Marketing research that won't 1. CDC. CDC Health Disparities and Inequalities Report - United States, 2011. break the bank : a practical guide to getting the information you need. San Morb Mortal Wkly Rep. 2011;60:1–114. Francisco: Jossey-Bass; 2002. 2. Truman BI, Smith KC, Roy K, Chen Z, Moonesinghe R, Zhu J, et al. Rationale 24. Hlady WG, Quenemoen LE, Armenia-Cope RR, Hurt KJ, Malilay J, Noji EK, et for regular reporting on health disparities and inequalities - United States. al. Use of a modified cluster sampling method to perform rapid needs MMWR Surveill Summ. 2011;60(Suppl):3–10. assessment after Hurricane Andrew. Ann Emerg Med. 1994;23(4):719–25. 3. Danese A, McEwen BS. Adverse childhood experiences, allostasis, allostatic 25. Braveman P, Barclay C. Health disparities beginning in childhood: a life-course load, and age-related disease. Physiol Behav. 2012;106(1):29–39. doi:10.1016/ perspective. Pediatrics. 2009;124:163–75. j.physbeh.2011.08.019. 26. Lu MC, Chen B. Racial and ethnic disparities in preterm birth: the role of stressful life events. Am J Obstet Gynecol. 2004;191(3):691–9. 4. CDC. Adverse childhood experiences reported by adults – five states, 2009. doi:10.1016/j.ajog.2004.04.018. MMWR Morb Mortal Wkly Rep. 2010;59(49):1609–13. 5. Felitti VJ, Anda RF. The relationship of adverse childhood experiences to 27. Lu MC, Halfon N. Racial and ethnic disparities in birth outcomes: a life-course adult medical disease, psychiatric disorders, and sexual behavior: perspective. Matern Child Health J. 2003;7(1):13–30. implications for healthcare. In: Lanius RA, Vermetten E, Pain C, editors. The 28. CDC. Behavioral Risk Factor Surveillance System. Atlanta, GA: US Department Impact of Early Life Trauma on Health Child and Adolescent Psychiatry and of Health and Human Services; 2013. http://www.cdc.gov/brfss/. Mental and Disease: The Hidden Epidemic. New York: Cambridge University 29. CDC. National Health and Nutrition Examination Survey (NHANES). Press; 2010. p. 77–87. Hyattsville, MD: US Department of Health and Human Services; 2012. 6. Felitti VJ, Anda RF, Nordenberg D, Williamson DF, Spitz AM, Edwards V, et al. http://www.cdc.gov/nchs/nhanes.htm. Relationship of childhood abuse and household dysfunction to many of the 30. Taylor VR. Measuring healthy days population assessment of health-related leading causes of death in adults. The Adverse Childhood Experiences (ACE) quality of life. Atlanta, Ga.: U.S. Dept. of Health and Human Services, Centers for Study. Am J Prev Med. 1998;14(4):245–58. Disease Control and Prevention, National Center for Chronic Disease Prevention 7. Brown DW, Anda RF, Tiemeier H, Felitti VJ, Edwards VJ, Croft JB, et al. and Health Promotion, Division of Adult and Community Health; 2000. Adverse childhood experiences and the risk of premature mortality. 31. CDC. Self-reported frequent mental distress among adults–United States, Am J Prev Med. 2009;37(5):389–96. doi:10.1016/j.amepre.2009.06.021. 1993–1996. MMWR Morb Mortal Wkly Rep. 1998;47(16):326–31. 8. Danese A, Moffitt TE, Harrington H, Milne BJ, Polanczyk G, Pariante CM, et al. 32. CDC. Frequently Asked Questions: Which cutpoints can I use to look at Adverse childhood experiences and adult risk factors for age-related Frequent Mental Distress or Frequent Physical Distress? National Center for disease: depression, inflammation, and clustering of metabolic risk markers. Chronic Disease Prevention and Health Promotion. Atlanta, Georgia: Division Arch Pediatr Adolesc Med. 2009;163(12):1135–43. doi:10.1001/ of Population Health; 2011. http://www.cdc.gov/hrqol/faqs.htm#10. archpediatrics.2009.214. Accessed 07/21/2014. 9. Anda RF, Croft JB, Felitti VJ, Nordenberg D, Giles WH, Williamson DF, et al. 33. Mentoring program enhances CM training. Healthcare Benchmarks Qual Adverse childhood experiences and smoking during adolescence and Improv. 2009;16(9):99–102. adulthood. JAMA. 1999;282(17):1652–8. 34. Hillis SD, Anda RF, Dube SR, Felitti VJ, Marchbanks PA, Marks JS. The 10. Corso PS, Edwards VJ, Fang X, Mercy JA. Health-related quality of life among association between adverse childhood experiences and adolescent adults who experienced maltreatment during childhood. Am J Public pregnancy, long-term psychosocial consequences, and fetal death. Health. 2008;98(6):1094–100. doi:10.2105/AJPH.2007.119826. Pediatrics. 2004;113(2):320–7. 11. Chapman DP, Whitfield CL, Felitti VJ, Dube SR, Edwards VJ, Anda RF. 35. CDC. Behavioral Risk Factor Surveillance System Questionnaire, 2013. 2012. Adverse childhood experiences and the risk of depressive disorders in http://www.cdc.gov/brfss/questionnaires/pdf-ques/2013%20BRFSS_English.pdf. adulthood. J Affect Disord. 2004;82(2):217–25. doi:10.1016/j.jad.2003.12.013. 36. Arday DR, Tomar SL, Nelson DE, Merritt RK, Schooley MW, Mowery P. State 12. Dube SR, Anda RF, Felitti VJ, Chapman DP, Williamson DF, Giles WH. smoking prevalence estimates: a comparison of the Behavioral Risk Factor Childhood abuse, household dysfunction, and the risk of attempted suicide Surveillance System and current population surveys. Am J Public Health. throughout the life span: findings from the Adverse Childhood Experiences 1997;87(10):1665–9. Study. JAMA. 2001;286(24):3089–96. 37. Brener ND, Billy JO, Grady WR. Assessment of factors affecting the validity of 13. Dube SR, Felitti VJ, Dong M, Giles WH, Anda RF. The impact of adverse self-reported health-risk behavior among adolescents: evidence from the childhood experiences on health problems: evidence from four birth scientific literature. J Adolesc Health. 2003;33(6):436–57. cohorts dating back to 1900. Prev Med. 2003;37(3):268–77. 38. Illner AK, Nothlings U, Wagner K, Ward H, Boeing H. The assessment of 14. Dube SR, Cook ML, Edwards VJ. Health-related outcomes of adverse individual usual food intake in large-scale prospective studies. Ann Nutr childhood experiences in Texas, 2002. Prev Chronic Dis. 2010;7(3):A52. Metab. 2010;56(2):99–105. doi:10.1159/000277667. doi:A52 [pii]. 39. Thompson FE, Kipnis V, Subar AF, Krebs-Smith SM, Kahle LL, Midthune D, et 15. Loria AS, Ho DH, Pollock JS. A mechanistic look at the effects of adversity al. Evaluation of 2 brief instruments and a food-frequency questionnaire to early in life on cardiovascular disease risk during adulthood. Acta Physiol estimate daily number of servings of fruit and vegetables. Am J Clin Nutr. (Oxf). 2013. doi:10.1111/apha.12189. 2000;71(6):1503–10. 16. Salihu HM, August EM, Alio AP, Jeffers D, Austin D, Berry E. Community-academic 40. Cohen S, Williamson G. Perceived stress in a probability sample of the partnerships to reduce black-white disparities in infant mortality in Florida. Prog United States. In: Spacapan S, Oskamp S, editors. The social psychology of Community Health Partnersh. 2011;5(1):53–66. doi:10.1353/cpr.2011.0009. health: Claremont Symposium on applied social psychology. Newbury Park, 17. Salihu HM, August EM, Mbah AK, Alio AP, de Cuba 2nd R, Jaward FM, et al. CA: Sage; 1988. p. 31–67. Effectiveness of a federal healthy start program in reducing the impact of 41. Cohen S, Janicki-Deverts D, Miller GE. Psychological stress and disease. particulate air pollutants on feto-infant morbidity outcomes. Matern Child JAMA. 2007;298(14):1685–7. doi:10.1001/jama.298.14.1685. Health J. 2012;16(8):1602–11. doi:10.1007/s10995-011-0854-1. 42. Sherbourne CD, Stewart AL. The MOS social support survey. Soc Sci Med. 18. Howard KS, Brooks-Gunn J. The role of home-visiting programs in 1991;32(6):705–14. preventing child abuse and neglect. Future Child. 2009;19(2):119–46. 43. Steptoe A, Feldman PJ. Neighborhood problems as sources of chronic 19. Jevitt C, Zapata L, Harrington M, Berry E. Screening for perinatal stress: development of a measure of neighborhood problems, and depression with limited psychiatric resources. J Am Psychiatr Nurses associations with socioeconomic status and health. Ann Behav Med. Assoc. 2006;11:359–63. 2001;23(3):177–85. 20. Salihu HM, Mbah AK, Jeffers D, Alio AP, Berry L. Healthy start program and 44. Cagney KA, Browning CR. Exploring neighborhood-level variation in asthma feto-infant morbidity outcomes: evaluation of program effectiveness. Matern and other respiratory diseases: the contribution of neighborhood social Child Health J. 2009;13(1):56–65. doi:10.1007/s10995-008-0400-y. context. J Gen Intern Med. 2004;19(3):229–36. 21. Bush AJ, Hair JF. An assessment of the mall intercept as a data collection 45. Contact Software Limited. droidSURVEY. droidSURVEY Wellington, New method. J Mark Res. 1985;22(May):158–67. doi:10.2307/3151361. Zealand. 2013. https://www.harvestyourdata.com/products/droidsurvey/. 22. Hornik J, Ellis S. Makhon le-mehkar asakim b-Y. Strategies to secure 46. Hewlett-Packard Company. HP Slate 7 2800 Tablet Product Specifications. compliance for a Mall intercept interview. Tel Aviv: Israel Institute of Hewlett-Packard Development Company, L.P. 2014. http://support.hp.com/ Business Research; 1986. us-en/document/c03726652. Salinas-Miranda et al. Health and Quality of Life Outcomes (2015) 13:123 Page 12 of 12 47. Preacher KJ, Hayes AF. SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behav Res Methods Instrum Comput. 2004;36(4):717–31. 48. Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol. 1986;51(6):1173–82. 49. Vanderweele TJ, Vansteelandt S, 12. Odds ratios for mediation analysis for a dichotomous outcome. Am J Epidemiol. 2010;172:1339–48. 50. Hayes AF, Preacher KJ. Statistical mediation analysis with a multicategorical independent variable. The British journal of mathematical and statistical psychology. 2013. doi:10.1111/bmsp.12028. 51. Preacher KJ, Hayes AF. Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behav Res Methods. 2008;40(3):879–91. 52. Blane D, Netuveli G, Stone J. The development of life course epidemiology. Rev Epidemiol Sante Publique. 2007;55(1):31–8. doi:10.1016/j.respe.2006.12.004. 53. Leung MW, Yen IH, Minkler M. Community based participatory research: a promising approach for increasing epidemiology's relevance in the 21st century. Int J Epidemiol. 2004;33(3):499–506. doi:10.1093/ije/dyh010. Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color figure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit

Journal

Health and Quality of Life Outcomes – Springer Journals

Published: Aug 11, 2015

There are no references for this article.