Adult Income Research Articles

Artificial Intelligence Life Cycle: The Detection and Mitigation of Bias

The rapid expansion of Artificial Intelligence(AI) has outpaced the development of ethical guidelines and regulations, raising concerns about the potential for bias in AI systems. These biases in AI can manifest in real-world applications leading to unfair or discriminatory outcomes in areas like job hiring, loan approvals or criminal justice predictions. For example, a biased AI model used for loan prediction may deny loans to qualified applicants based on demographic factors such as race or gender. This paper investigates the presence and mitigation of bias in Machine Learning(ML) models trained on the Adult Census Income dataset, known to have limitations in gender and race. Through comprehensive data analysis, focusing on sensitive attributes like gender, race and relationship status, this research sheds light on complex relationships between societal biases and algorithmic outcomes and how societal biases can be rooted and amplified by ML algorithms. Utilising fairness metrics like demographic parity(DP) and equalised odds(EO), this paper quantifies the impact of bias on model predictions. The results demonstrated that biased datasets often lead to biased models even after applying pre-processing techniques. The effectiveness of mitigation techniques such as reweighting(Exponential Gradient(EG)) to reduce disparities was examined, resulting in a measurable reduction in bias disparities. However, these improvements came with trade-offs in accuracy and sometimes in other fairness metrics, identified the complex nature of bias mitigation and the need for precise consideration of ethical implications. The findings of this research highlight the critical importance of addressing bias at all stages of the AI life cycle, from data collection to model deployment. The limitation of this research, especially the use of EG, demonstrates the need for further development of bias mitigation techniques that can address complex relationships while maintaining accuracy. This paper concludes with recommendations for best practices in Artificial Intelligence development, emphasising the need for ongoing research and collaboration to mitigate bias by prioritising ethical considerations, transparency, explainability, and accountability to ensure fairness in AI systems.

Family experiences with supplemental social security income and legal guardianship for autistic adults: A mixed-methods study

BackgroundFamily interactions with the Supplemental Social Security Income (SSI) program and the decision to pursue legal guardianship are poorly understood in services research for autistic adults. MethodsWe conducted a mixed-methods study and incorporated quantitative survey data from 122 autistic adults in an existing longitudinal cohort with qualitative interviews with 12 autistic adults and/or legal guardians. We explored sociodemographic, developmental, and behavioral features associated with the likelihood of having SSI and a legal guardian at 25 years old. Spoken interviews were transcribed verbatim, and interviews were analyzed using a rapid qualitative analytic approach. ResultsThere were quantitative differences in whether families were able to obtain SSI and guardianship based on intelligence quotient (IQ) scores, adaptive behavior, and autistic characteristics, but not by race, ethnicity, or maternal education. Qualitative data analysis revealed six themes that highlighted the challenges associated with obtaining and maintaining SSI, along with the complex, nuanced decisions associated with legal guardianship. Families noted many challenges in navigating these procedures and some potential benefits for each unique circumstance. ConclusionsThese findings offer new perspectives on experiences associated with pursuing SSI benefits and legal guardianship for autistic adults, including similarities and key differences in these procedures. Findings also provide suggestions for future research to improve coordination and supports for families throughout adulthood.

Low Income and Mortality in Young Adults Living With Type 2 Diabetes

Low Income and Mortality in Young Adults Living With Type 2 Diabetes

Bias and Its Consequences : A Study of Machine Learning Performance

This paper addresses the concern about bias affecting the results of machine learning models. For this purpose, it uses the Adult Income dataset from OpenML for income classification. The conditions for bias are induced by underrepresenting people that earn <= $50K in training data, thus checking the behavior of different models when encountering such a skewed distribution. Key metrics, namely accuracy and specificity (True Negative Rate), were analyzed for unbiased and biased training scenarios. The results show that Naive Bayes and Random Forest models were resistant to bias, but others, including SVM and Logistic Regression, suffered major performance drops. This study throws light on the robustness of different classifiers when exposed to biased data, requiring further bias mitigation strategies in real-world applications. This paper actually examines critically how bias in training data can significantly affect the performance of prediction, fairness, and model selection in income classification tasks.

Community and Patient Features and Health Care Point of Entry for Pediatric Concussion

Many recent advances in pediatric concussion care are implemented by specialists; however, children with concussion receive care across varied locations. Thus, it is critical to identify which children have access to the most up-to-date treatment strategies. To evaluate differences in the sociodemographic and community characteristics of pediatric patients who sought care for concussion across various points of entry into a regional health care network. This cross-sectional study included children seen for concussions across a regional US health care network from January 1, 2017, to August 4, 2023. Pediatric patients aged 0 to 18 years who received an International Statistical Classification of Diseases and Related Health Problems, Tenth Revision, Clinical Modification code for concussion were included. The study took place at emergency department (ED) and outpatient (primary care [PC] and specialty care [SC]) settings. Age at visit, biological sex, parent-identified race and ethnicity, payer type, median income and percentage of adults with a bachelor's degree for home zip code, and overall and subdomain Child Opportunity Index (COI) score based on patient address. The association of exposures with point of entry of ED, PC, and SC were examined in both bivariate analysis and a multinomial logistic regression. Overall, 15 631 patients were included in the study (median [IQR] age, 13 [11-15] years; 7879 [50.4%] male; 1055 [6.7%] Hispanic, 2865 [18.3%] non-Hispanic Black, and 9887 [63.7%] non-Hispanic White individuals). Race and ethnicity were significantly different across settings (1485 patients [50.0%] seen in the ED were non-Hispanic Black vs 1012 [12.0%] in PC and 368 [8.7%] in SC; P < .001) as was insurance status (1562 patients [52.6%] seen in the ED possessed public insurance vs 1624 [19.3%] in PC and 683 [16.1%] in SC; P < .001). Overall and individual COI subdomain scores were also significantly different between settings (overall COI median [IQR]: ED, 30 [9-71]; PC, 87 [68-95]; SC, 87 [69-95]; P < .001). Race, insurance status, and overall COI had the strongest associations with point of entry in the multivariable model (eg, non-Hispanic Black patients seen in the ED compared with non-Hispanic White patients: odds ratio, 2.03; 95% CI, 1.69-2.45). In this cross-sectional study, children with concussion seen in the ED setting were more likely to be non-Hispanic Black, have public insurance, and have a lower Child Opportunity Index compared with children cared for in the PC or SC setting. This highlights the importance of providing education and training for ED clinicians as well as establishing up-to-date community-level resources to optimize care delivery for pediatric patients with concussion at high risk of care inequities.

Poverty Dynamics and Caries Status in Young Adolescents.

To investigate whether timing, accumulation and trajectories of poverty are associated with dental caries in young adolescents. The study was conducted within the Generation R Study, which is an ongoing population-based prospective cohort study conducted in Rotterdam, the Netherlands. This study included 2653 children. Information about household income and number of children and adults living in a household at six time points from pregnancy to 13 years old was retrieved from parental questionnaires to construct the poverty variable. Dental caries was assessed with the decayed, missing and filled teeth index through intraoral photographs at the age of 13 years. Sociodemographic and oral health-related characteristics were included as possible confounders. The association between poverty and dental caries was analysed on the basis of the three lifecourse theories, that is, critical period, cumulative risk and social mobility model. For the latter, we used latent class growth analysis (LCGA) to identify poverty trajectories over time. Next, the associations were studied with Hurdle Negative Binomial Models. Poverty at birth and intermittent poverty up to the age of 13 were significantly associated with dental caries at 13 years of age (OR 1.41, 95% CI 1.01-1.99; OR 1.36, 95% CI 1.01-1.83 respectively) and with an increased mean number of decayed teeth by 34% (95% CI 1.02-1.76; 95% CI 1.05-1.71, respectively). LCGA showed four trajectories for the probabilities of poverty. All trajectories were significantly associated with dental caries at 13 years of age, with the 'downward mobility' trajectory showing the strongest association with dental caries (OR 1.55, 95% CI 1.05-2.29) and an increasing mean number of decayed teeth by 58% (95% CI 1.18-2.12) than the 'stable absent' trajectory. Poverty at birth, intermittent poverty and downward poverty trajectory were associated with higher odds and higher mean number of decayed teeth at 13 years of age. The three lifecourse models influenced dental caries status during adolescence, hence strategies and policies targeted to improve socioeconomic conditions on deprived children should be implemented.

Abstract A114: Inequities in advanced stage at diagnosis of non-small cell lung cancer among people with and without HIV in the US

Abstract Background: Lung cancer is the leading cause of non-AIDS death among people with HIV (PWH). PWH are also more likely to die due to lung cancer compared to people with HIV. Late presentation at diagnosis due to barriers to lung cancer screening may contribute to disproportionate deaths among PWH. Our objective was to evaluate associations of HIV status with advanced stage at diagnosis and evaluate social determinants of health (SDoH) associated with advanced stage at diagnosis of lung cancer among PWH. Methods: We used data from the 2004-2020 National Cancer Database, including patients aged 18-89 years diagnosed with lung cancer (ICD-0-3 codes: C340-C349). PWH were identified using ICD9 (04200 to 04499, 07953, and V08) and ICD10 (B20 to B24, R75, Z21, B97.35).The primary outcome was advanced stage at diagnosis (stage 3/4). Our main SDoH exposures were quantified as quartiles (Q1-Q4) and included measures of area-level SDoH based on a patient’s zip code: (1) % of adults without a high school degree, or area-level education; (2) median household income of adults, or area-level income. We used hierarchical multivariable logistic regression models clustered by facility ID to estimate adjusted odds ratios(aOR) with 95% confidence intervals(95% CI), and evaluated each exposure separately adjusting for age, sex, year of diagnosis, and Charlson comorbidity score, and stratified by race/ethnicity and insurance status. Results: Overall, we identified 6,585 PWH and 1,665,030 without HIV. Seventy one percent of PWH were diagnosed with stage 3 or 4 lung cancer compared to 64.9% of people without HIV (p&amp;lt;0.001). PWH were more likely to be under the age of 60 (43.5% vs.15.1%, p&amp;lt;0.001), After adjustment for sex, age, and year of diagnosis, PWH had 15% higher odds of advanced stage at diagnosis compared to those without HIV (aOR: 1.15; 95% CI: 1.09-1.22). Residing in areas with larger proportions of the population without a HS diploma (Q4 vs. Q1: aOR:1.19; 95% CI: 1.01-1.41) and larger proportions with lower household income (Q1 vs. Q4: aOR: 1.15; 95% CI: 1.01-1.31) were positively associated with advanced stage-at-diagnosis of lung cancer. When we stratified by insurance status, only PWH who were Medicaid insured residing in the least educated areas (aOR: 1.84; 95% CI: 1.11-3.06) and in the lowest income areas (aOR: 1.86; 95% CI: 1.30-2.65) had higher odds of advanced stage at diagnosis with lung cancer. Conclusions: Living with HIV is independently associated with presenting at a later stage of lung cancer at diagnosis. Area-level measures of lower education and lower household income are significantly associated with advanced stage-at- diagnosis of lung cancer regardless, particularly among those who are Medicaid insured. Citation Format: Jessica Y Islam, Yi Guo, Kea Turner, Amir Alishahi Tabriz, Marlene Camacho-Rivera, Gita Suneja. Inequities in advanced stage at diagnosis of non-small cell lung cancer among people with and without HIV in the US [abstract]. In: Proceedings of the 17th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2024 Sep 21-24; Los Angeles, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2024;33(9 Suppl):Abstract nr A114.

The Affordable Care Act Dependent Coverage Provision and Unintended Pregnancy.

Background: Nearly half of all pregnancies in the United States are considered unintended (mistimed or unwanted), and this rate is even higher among younger and lower income women. The Affordable Care Act (ACA) dependent coverage provision may have influenced the frequency of unintended pregnancies by increasing accessibility to and affordability of family planning services among young adults. Furthermore, the impact of this provision may differ by young adult income level as those with lower income are less likely to be insured and thus more likely to benefit from this provision. Our objective was to estimate the association between the ACA dependent coverage provision and unintended pregnancy, overall, and by young adult income level. Methods: We applied a difference-in-differences approach to data from multiple cycles of the National Survey of Family Growth (n = 10,104) and compared trends in unintended pregnancy between those who were eligible to benefit (ages 18-25 years) and those who were ineligible to benefit (ages 26-33 years) from the provision, overall, and among income subgroups. Results: We found evidence that the dependent coverage provision was associated with a -7.4 percentage point reduction (95% CI: -13.5, -1.3) in the prevalence of unintended pregnancy among young adults with lower income (<100% of the federal poverty level). There was limited evidence, however, that the provision was associated with unintended pregnancy among young adults with higher income levels. Conclusions: These findings suggest the ACA dependent coverage provision may have reduced unintended pregnancy among a particularly high-risk group (i.e., young adults with lower income).

Association between family economic situation and serum PFAS concentration in American adults with hypertension and hyperlipemia

Although there is an association between income status and concentration of perfluoroalkyl and polyfluoroalkyl substance (PFAS), the association remains uncertain in patients with hypertension, hyperlipidemia, and comorbidities. Data from the 2013–2016 National Health and Nutrition Examination Survey were analyzed. A total of 2665 adults were included, and the data included participants' serum PFAS (perfluorooctanoic acid [PFOA], perfluorononaic acid, perfluorodecanoic acid, perfluoroundecanoic acid, perfluorohexane sulfonic acid, and perfluorooctane sulfonic acid) levels and selected covariates. Multivariate linear regression models were used to examine the association between the ratio of family income to poverty (PIR) and individual serum PFAS concentrations in the hypertensive and/or hyperlipidemia groups after adjusting for covariates. The potential effects of sex and age on the results were explored using stratified analysis. A mediating effect model was used to explore the mediating effects of body mass index (BMI) and waist circumference on the association results. After adjusting for potential confounders, for hyperlipidemia and comorbidities (hypertension and hyperlipidemia), serum levels of multiple common PFAS increased by 0.09% (95%Confidence interval [CI] 0.02–0.15%) to 0.13% (95%CI 0.08–0.19%) and 0.10% (95%CI 0.02–0.17%) to 0.12% (95%CI 0.06–0.18%), respectively, with each 1% increase in PIR. The covariate model and stratified analyses results suggested the potential effects of different covariates such as age and sex, leading to changes in the statistical significance of the association results. BMI significantly mediated the effect of PIR on PFOA in hyperlipidemia (13%, P < 0.001). Household income in adults with hyperlipidemia and comorbidities positively correlated with serum PFAS concentration in the United States. Obesity played an indispensable mediating role in the association between economic income and PFAS concentration.

Sports Attitudes in Childhood and Income in Adulthood

Research shows that sports participation in youth is associated with earnings in adulthood. However, studies have often relied on self-reported earnings and on single indicators of sports participation. Using large-scale data linked to administrative records, we investigate: (a) whether sports attitudes at age 13 are uniquely related to income at ages 30–36 and (b) whether educational attainment and mental health in early adulthood mediate this association. We find that a one SD increase in sports attitudes is related to a 10% increase in income. This association is not entirely confounded by preexisting and co-occurring risk factors. Educational attainment mediates 22% of the association between sports attitudes and income. Sports attitudes are potentially a form of capital deployed in the pursuit of socioeconomic advantages.

Household income mobility and adolescent subjective well-being in China: Analyzing the mechanisms of influence

BackgroundThe relationship between adult happiness and income has been a topic of extensive discussion. However, among children and adolescents, who are not directly engaged in economic activities, the relationship between household income over time and happiness is more complex and remains less explored. MethodsWe used nationally representative data from the China Family Panel Study (CFPS) from 2012 to 2018. The analytic sample consisted of 3,607 Chinese adolescents aged 10 to 19 years in 2018. Using an ordinary least squares model, we investigated the associations of 2-year (2016–2018), 4-year (2014–2018), and 6-year (2012–2018) household income mobility (i.e., upward mobility and downward mobility) with adolescents’ subjective well-being (i.e., happiness and depression) in 2018. We further explored the potential mediating effects of parental care on these associations and whether they were moderated by adolescent age, adolescent sex, and initial household economic status. ResultsWe found that upward mobility in household income was associated with decreased subjective well-being and parental care mediated this association. Furthermore, we found that the effects of upward mobility on adolescent subjective well-being varied by age, sex, and initial household economic status. Additionally, downward income mobility did not significantly impact adolescents’ subjective well-being. ConclusionOur findings emphasized that economic progress and upward mobility should not come at the expense of parental care, as this would decrease adolescents’ subjective well-being. Collaboration between the government and families is crucial to address the conflicts between work and family responsibilities, ensuring that parents are available and supportive to their children.

Optimizing fairness and accuracy: a Pareto optimal approach for decision-making

AbstractIn the era of data-driven decision-making, ensuring fairness and equality in machine learning models has become increasingly crucial. Multiple fairness definitions have been brought forward to evaluate and mitigate unintended fairness-related harms in real-world applications, with little research on addressing their interactions with each other. This paper explores the application of a Minimax Pareto-optimized solution to optimize individual and group fairness at individual and group levels on the Adult Census Income dataset as well as on the German Credit dataset. The objective of training a classification model with a multi-objective loss function is to achieve fair outcomes without compromising utility objectives. We investigate the interplay of different fairness definitions, including definitions of performance consistency and traditional group and individual fairness measures, amongst each other coupled with performance. The results presented in this paper highlight the feasibility of incorporating several fairness considerations into machine learning models, which can be applied to use cases with multiple sensitive features and attributes that characterize real-world applications. This research is a valuable step toward building responsible and transparent machine learning systems that can be incorporated into critical decision-making processes.

Preterm Birth, Family Income, and Intergenerational Income Mobility

Preterm birth (PTB) has been associated with lower income in adulthood, but associations with intergenerational income mobility and the role of family socioeconomic status (SES) as modifying factor are unclear. To assess whether the association between PTB and income differs according to family SES at birth and to assess the association between PTB and intergenerational income mobility. This study comprised a matched cohort of live births in Canada between January 1, 1990, and December 31, 1996, with follow-up until December 31, 2018. Statistical analysis was performed between May 2023 and March 2024. Preterm birth, defined as birth between 24 and 37 weeks' gestational age (with gestational age subcategories of 34-36, 32-33, 28-31, and 24-27 weeks) vs early and full term births (gestational age, 37-41 weeks). Associations between PTB and annual adulthood income in 2018 Canadian dollars were assessed overall (current exhange rate: $1 = CAD $1.37) and stratified by family income quintiles, using generalized estimating equation regression models. Associations between PTB and percentile rank change (ie, difference between the rank of individuals and their parents in the income distribution within their respective generations) and upward or downward mobility (based on income quintile) were assessed using linear and multinomial logistic regressions, respectively. Of 1.6 million included births (51.1% boys and 48.9% girls), 6.9% infants were born preterm (5.4% born at 34-36 weeks, 0.7% born at 32-33 weeks, 0.5% born at 28-31 weeks, and 0.2% born at 24-27 weeks). After matching on baseline characteristics (eg, sex, province of birth, and parental demographics) and adjusting for age and period effects, PTB was associated with lower annual income (mean difference, CAD -$687 [95% CI, -$788 to -$586]; 3% lower per year), and the differences were greater among those belonging to families in the lowest family SES quintile (mean difference, CAD -$807 [95% CI, -$998 to -$617]; 5% lower per year). Preterm birth was also associated with lower upward mobility and higher downward mobility, particularly for those born earlier than 31 weeks' gestational age (24-27 weeks: mean difference in percentile rank change, -8.7 percentile points [95% CI, -10.5 to -6.8 percentile points]). In this population-based matched cohort study, PTB was associated with lower adulthood income, lower upward social mobility, and higher downward mobility, with greater differences among those belonging to economically disadvantaged families. Interventions to optimize socioeconomic outcomes of preterm-born individuals would need to define target population considering SES.

Comprehensive Validation on Reweighting Samples for Bias Mitigation via AIF360

Fairness Artificial Intelligence (AI) aims to identify and mitigate bias throughout the AI development process, spanning data collection, modeling, assessment, and deployment—a critical facet of establishing trustworthy AI systems. Tackling data bias through techniques like reweighting samples proves effective for promoting fairness. This paper undertakes a systematic exploration of reweighting samples for conventional Machine-Learning (ML) models, utilizing five models for binary classification on datasets such as Adult Income and COMPAS, incorporating various protected attributes. In particular, AI Fairness 360 (AIF360) from IBM, a versatile open-source library aimed at identifying and mitigating bias in machine-learning models throughout the entire AI application lifecycle, is employed as the foundation for conducting this systematic exploration. The evaluation of prediction outcomes employs five fairness metrics from AIF360, elucidating the nuanced and model-specific efficacy of reweighting samples in fostering fairness within traditional ML frameworks. Experimental results illustrate that reweighting samples effectively reduces bias in traditional ML methods for classification tasks. For instance, after reweighting samples, the balanced accuracy of Decision Tree (DT) improves to 100%, and its bias, as measured by fairness metrics such as Average Odds Difference (AOD), Equal Opportunity Difference (EOD), and Theil Index (TI), is mitigated to 0. However, reweighting samples does not effectively enhance the fairness performance of K Nearest Neighbor (KNN). This sheds light on the intricate dynamics of bias, underscoring the complexity involved in achieving fairness across different models and scenarios.

Fairness for Deep Learning Predictions Using Bias Parity Score Based Loss Function Regularization

Rising acceptance of machine learning driven decision support systems underscores the need for ensuring fairness for all stakeholders. This work proposes a novel approach to increase a Neural Network model’s fairness during the training phase. We offer a frame-work to create a family of diverse fairness enhancing regularization components that can be used in tandem with the widely accepted binary-cross-entropy based accuracy loss. We use Bias Parity Score (BPS), a metric that quantifies model bias with a single value, to build loss functions pertaining to different statistical measures — even for those that may not be developed yet. We analyze behavior and impact of the newly minted regularization components on bias. We explore their impact in the realm of recidivism and census-based adult income prediction. The results illustrate that apt fairness loss functions can mitigate bias without forsaking accuracy even for imbalanced datasets.

Which aspects of education are health protective? a life course examination of early education and adulthood cardiometabolic health in the 30-year study of early child care and Youth Development (SECCYD)

BackgroundPast research describes robust associations between education and health, yet findings have generally been limited to the examination of education as the number of years of education or educational attainment. Little is known about the specific features or processes underpinning education that are health protective. The objective of the current study was to address this gap by examining specific aspects of early education pertaining to student characteristics and experiences, as well as features of the classroom environment, in predicting cardiometabolic health in adulthood.MethodsSubjects were 1364 participants in the NICHD Study of Early Child Care and Youth Development (SECCYD, 1991–2009) and recent SECCYD 30-year follow-up, the Study of Health in Early and Adult Life (SHINE, 2018–2022). Models examined individual education indicators (student social skills, student-teacher relationship quality, and classroom emotional and instructional quality in the period of elementary school and student academic performance between ages 54 months and 15 years) in relation to a composite of cardiometabolic risk in adulthood (ages 26–31), reflecting central adiposity, blood pressure, insulin resistance, inflammation, and dyslipidemia. Models were adjusted for key explanatory factors including socio-demographics, infant characteristics, parental socioeconomic status (SES), and child health status. Follow-up analyses were performed to test potential mediators of early education effects on adult health, including adult SES (educational attainment, household income) and health behaviors (diet quality, activity level, sleep duration, smoking).ResultsIn adjusted models, results showed greater student social skills, indexed by a mean of annual teacher ratings between kindergarten and 6th grade, predicted lower cardiometabolic risk in adulthood (β=-0.009, p <.05). In follow-up analyses, results showed the protective effect of student social skills on cardiometabolic risk may be mediated by adult income (β=-0.0014, p <.05) and diet quality (β=-0.0031, p <.05). Effects of the other early education indicators were non-significant (ps > 0.05).ConclusionsFindings point to the potential significance of early student social competence as a link to long-term health, possibly via the acquisition of resources needed for the maintenance of health, as well as through engagement in health behaviors supporting healthy eating. However, more research is needed to replicate these findings and to elaborate on the role of early student social competence and the pathways explaining its effects on cardiometabolic health in adulthood.

CACTUS: A Comprehensive Abstraction and Classification Tool for Uncovering Structures

The availability of large datasets is providing the impetus for driving many current artificial intelligent developments. However, specific challenges arise in developing solutions that exploit small datasets, mainly due to practical and cost-effective deployment issues, as well as the opacity of deep learning models. To address this, the Comprehensive Abstraction and Classification Tool for Uncovering Structures (CACTUS) is presented as a means of improving secure analytics by effectively employing explainable artificial intelligence. CACTUS achieves this by providing additional support for categorical attributes, preserving their original meaning, optimising memory usage, and speeding up the computation through parallelisation. It exposes to the user the frequency of the attributes in each class and ranks them by their discriminative power. Performance is assessed by applying it to various domains, including Wisconsin Diagnostic Breast Cancer, Thyroid0387, Mushroom, Cleveland Heart Disease, and Adult Income datasets.

The association of adverse childhood experiences with household income, educational attainment and partnered status among adults aged 30-39

BackgroundThe association of adverse childhood experiences (ACEs) with young adult social outcomes is poorly understood. ObjectiveTo examine the relationship between ACEs and young adult household income, education, and partnered status. Participants13,767 respondents (population-weighted estimate N = 13,191,291) aged 30–39 completed the optional ACES module in the 2019 Behavioral Risk Factor Surveillance System (BRFSS) from 17 states. MethodsThe 2019 BRFSS was used to analyze the significance of ACE scores for the likelihood of reporting household income greater than $75,000, a college or postgraduate degree, and living with a partner among respondents aged 30–39. Poisson regression analyses controlled for sociodemographic, health status, and behavioral risk characteristics. ResultsAs compared to those with zero ACEs, respondents reporting four or more ACE exposures (20.2% of respondents) were marginally less likely to report high income (IRR: 0.87, 95% CI: 0.79–0.97) and one-third less likely to report a college degree (IRR: 0.68, 95% CI: 0.61–0.76), with no significant difference in partnered status. ConclusionsYoung adult social status may be shaped by social-emotional effects of ACEs that go beyond demographic and health status differences. Enhancing resilience to childhood adversity can benefit from a trauma-informed approach in health care, education, and employment.

Dust to Feed, Dust to Gray: The Effect of in Utero Exposure to the Dust Bowl on Old-Age Longevity

Intensive agriculture and deep plowing caused topsoil erosion and dust storms during the 1930s, affecting agricultural income and land values for years. Given the growing literature on the relevance of in utero and early-life exposures, it is surprising that studies focusing on links between the Dust Bowl and later-life health have produced inconclusive and mixed results. We reevaluate this literature and study the long-term effects of in utero and early-life exposure to topsoil erosion caused by the 1930s Dust Bowl on old-age longevity. Specifically, using Social Security Administration death records linked with the full-count 1940 census, we conduct event studies with difference-in-differences designs to compare the longevity of individuals in high- versus low-topsoil-erosion counties before versus after 1930. We find intent-to-treat reductions in longevity of approximately 0.85 months for those born in high-erosion counties after 1930. We show that these effects are not an artifact of preexisting trends in longevity. Additional analyses suggest that the effects are more pronounced among children raised in farm households, females, and those whose mothers had lower education. We also provide suggestive evidence that reductions in adulthood income are a likely mechanism for the effects we document.

Evidence for the Association Between Adverse Childhood Family Environment, Child Abuse, and Caregiver Warmth and Cardiovascular Health Across the Lifespan: The Coronary Artery Risk Development in Young Adults (CARDIA) Study.

This study aimed to quantify the association between childhood family environment and longitudinal cardiovascular health (CVH) in adult CARDIA (Coronary Artery Risk Development in Young Adults) Study participants. We further investigated whether the association differs by adult income. We applied the CVH framework from the American Heart Association including metrics for smoking, cholesterol, blood pressure, glucose, body mass index, physical activity, and diet. CVH scores (range, 0-14) were calculated at years 0, 7, and 20 of the study. Risky Family environment (range, 7-28) was assessed at year 15 retrospectively, for childhood experiences of abuse, caregiver warmth, and family or household challenges. Complete case ordinal logistic regression and mixed models associated risky family (exposure) with CVH (outcome), adjusting for age, sex, race, and alcohol use. The sample (n=2074) had a mean age of 25.3 (±3.5) years and 56% females at baseline. The median risky family was 10 with ideal CVH (≥12) met by 288 individuals at baseline (28.4%) and 165 (16.3%) at year 20. Longitudinally, for every 1-unit greater risky family, the odds of attaining high CVH (≥10) decreased by 3.6% (OR, 0.9645 [95% CI, 0.94-0.98]). Each unit greater child abuse and caregiver warmth score corresponded to 12.8% lower and 11.7% higher odds of ideal CVH (≥10), respectively (OR, 0.872 [95% CI, 0.77-0.99]; OR, 1.1165 [95% CI, 1.01-1.24]), across all 20 years of follow-up. Stratified analyses by income in adulthood demonstrated associations between risky family environment and CVH remained significant for those of the highest adult income (>$74k), but not the lowest (<$35k). Although risky family environmental factors in childhood increase the odds of poor longitudinal adult CVH, caregiver warmth may increase the odds of CVH, and socioeconomic attainment in adulthood may contextualize the level of risk. Toward a paradigm of primordial prevention of cardiovascular disease, childhood exposures and economic opportunity may play a crucial role in CVH across the life course.