Exploring the Relationship between Food Insecurity, Chronic Health Conditions, and Serious Mental Illness in the United States: Implications for Social Work.

Food insecurity (FI) is a modifiable social determinant of health that impacts approximately 10 percent of the U.S. population. FI has been linked to poorer health outcomes and higher healthcare costs. Given the prevalence of chronic health conditions in the United States, including serious mental illness (SMI), the current study aims to better understand the relationship between FI and chronic conditions, including SMI, in a nationally representative sample. Authors analyzed data from the 2016 Medical Expenditure Panel Survey household component and food security supplement. Findings indicate the prevalence of FI among those with diabetes, lung disease, stroke, and SMI is higher than among the general population, with the prevalence for those with SMI being particularly high (43 percent of the sample). Logistic regression models indicate strong, statistically significant relationships between FI and chronic conditions, including SMI, as well as FI and two or more chronic illnesses, even when controlling for sociodemographic and health factors. There are likely bidirectional relationships between FI and chronic conditions. Findings have implications for social workers, in relation to prevention and treatment of SMI and FI through direct care, advocacy, and integrated services in health, mental health, and social services.

Serving Families Where They Live: Supporting Multigenerational Health During Infancy and Early Childhood Through Community-centered Approaches.

This article underscores the critical importance of addressing mental health during infancy and early childhood through a multigenerational, multicultural, community-centered approach. It highlights the unique vulnerability of this period to environmental factors and emphasizes the interconnectedness of caregiver and child mental health. The article advocates for interventions that extend beyond clinical settings, recognizing the value of community involvement and the need to address social determinants of health. It also discusses innovative strategies, such as mental health consultation in early childhood education centers and collaborative care models, to bridge gaps in access to care.

County-level political group density, partisan polarization, and individual-level mortality among adults in the United States: A lagged multilevel study.

Objective: To investigate the associations between county-level political group density, partisan polarization, and individual-level mortality from all causes and from coronary heart disease (CHD) in the United States. Methods: Using data from five survey waves (1998-2006) of the General Social Survey-National Death Index dataset and the County Presidential Election Return 2000 dataset, we fit weighted Cox proportional hazards models to estimate the associations between (1) political group density and (2) partisan polarization measured at the county level in 2000 (n = 313 counties) categorized into quartiles with individual-level mortality (n = 14,983 participants) from all causes and CHD, controlling for individual- and county-level factors. Maximum follow-up was from one year after the survey up until 2014. We conducted these analyses using two separate measures based on county-level vote share differences and party affiliation ideological extremes. Results: In the overall sample, we found no evidence of associations between county-level political group density and individual-level mortality from all causes. There was evidence of a 13% higher risk of dying from heart disease in the highest quartile of county-level polarization (hazards ratio, HR = 1.13; 95% CI = 0.74-1.71). We observed heterogeneity of effects based on individual-level political affiliation. Among those identifying as Democrats, residing in counties with high (vs. low) levels of polarization appeared to be protective against mortality, with an associated 18% lower risk of dying from all causes (HR = 0.82, 95% CI = 0.71-0.94). This association was strongest in areas with the highest concentrations of Democrats. Conclusions: Among all study participants, political group density and polarization at the county level in 2000 were not linked to individual-level mortality. At the same time, we found that Democratic party affiliation may be protective against the adverse effects of high polarization, particularly in counties with high concentrations of Democrats. Future research should further explore these associations to potentially identify new structural interventions to address political determinants of population health.

Public Health Challenges for Post-secondary Students During COVID-19: A Scoping Review.

Research about public health impacts of COVID-19 on post-secondary students is slowly beginning to emerge. This scoping review identified common public health challenges among post-secondary students in higher-income countries during the COVID-19 pandemic. Five databases were searched to find relevant peer-reviewed literature up to March 2022. Results were categorized according to reported public health challenges and relevant socio-economic variables. After screening, 53 articles were reviewed. Most articles were from the USA (39/53). The seven main public health challenges identified were mental health (35/53), financial instability (25/53), physical health (13/53), food insecurity (12/53), social well-being (8/53), digital access (7/53), and housing or relocation (6/53). Students with low socioeconomic status experienced heightened public health challenges. This review offers insight and opportunities for the development of longitudinal tools to support social determinants of health in post-secondary populations in high-income countries and may offer insight into similar experiences for students in other settings.

The Effect of Social Deprivation on Hospital Utilization Following Distal Radius Fracture Treatment.

Purpose: Social determinants of health disparities have been shown to adversely impact outcomes following distal radius fracture (DRF) treatment. Identifying risk factors for increased hospital use following DRF has been historically difficult; however, it is of utmost concern to orthopedic surgeons to improve outcomes and decrease the total cost of care. The effect of social deprivation following DRF has yet to be fully investigated. Methods: This is a retrospective cohort analysis of a single institution's experience in treating DRF with either an operative or nonsurgical modality between 2005 and 2020. Patient demographic information and health care utilization (hospital readmission, emergency department [ED] visitation, office visits, and telephone use) were collected from within 90 days of treatment. Area Deprivation Index (ADI) national percentiles were recorded. Patients were stratified into terciles based on their relative level of deprivation, and their outcomes were compared. Secondary analyses included stratifying patients based on treatment modality, race, and legal sex. Results: In total, 2,149 patients were included. The least, intermediate, and most deprived groups consisted of 552, 1,067, and 530 patients, respectively. Risk factors for hospital readmission included higher levels of relative deprivation. Identifying as Black or African American and nonsurgical management were risk factors for increased ED visitation. No differences in rate of hospital readmission, ED visitation, office visitation, or telephone use were seen based on deprivation level. Conclusions: High levels of social deprivation, treatment modality, race, and legal sex disparities may influence the amount of hospital resource utilization following DRF treatment. Understanding and identifying risk factors for greater resource utilization can help to mitigate inappropriate use and decrease health care costs. We hope to use these findings to guide clinical decision making, educate patient populations, and optimize outcomes following DRF treatment. Type of Study/Level of Evidence: Therapeutic III.

Integrating genetic and socioeconomic data to predict the progression of nonalcoholic fatty liver disease.

OBJECTIVES: Nonalcoholic fatty liver disease (NAFLD) is the leading cause of chronic liver disease globally, with an estimated prevalence exceeding 25%. Variants in the PNPLA3 and HSD17B13 genes have been a focus of investigations surrounding the etiology and progression of NAFLD and are believed to contribute to a greater burden of disease experienced by Hispanic Americans. However, little is known about socioeconomic factors influencing NAFLD progression or its increased prevalence among Hispanics. MATERIALS AND METHODS: We cross-sectionally analyzed 264 patients to assess the role of genetic and socioeconomic variables in the development of advanced liver fibrosis in individuals at risk for NAFLD. RESULTS: Adjusting for age, sex, body mass index, and PNPLA3 genotype, lacking a college degree was associated with 3.3 times higher odds of advanced fibrosis (95% confidence interval [CI]: 1.21-8.76, p = 0.019), an effect comparable to that of possessing the major PNPLA3 risk variant. Notably, the effect of PNPLA3 genotype on advanced fibrosis was attenuated to nonsignificance following adjustment for education and other socioeconomic markers. The effect of the protective HSD17B13 variant, moreover, diminished after adjustment for education (odds ratio [OR]: 0.39 [95% CI: 0.13-1.16, p = 0.092]), while lower education continued to predict advanced fibrosis following multivariable adjustment with an OR of 8.0 (95% CI: 1.91-33.86, p = 0.005). DISCUSSION: Adjusting for education attenuated the effects of genotype and Hispanic ethnicity on liver fibrosis, suggesting that social factors-rather than genes or ethnicity-may be driving disease severity within some populations. Findings reveal the importance of including socioenvironmental controls when considering the role of genetics or ethnicity in complex disease.

More Than Skin Deep: Patient-Provider Racial and Ethnic Concordance and Discordance in Collegiate Athletics and Concussion Management.

CONTEXT: There is strong evidence that racial and ethnic disparities exist in multiple arenas of health and wellness. The causes of racial and ethnic differences in healthcare are multi- dimensional, one factor that may impact injury/illness communication, interactions, and outcomes is patient-provider racial and ethnic concordance. At present, it is unclear what role patient-provider racial and ethnic concordance and discordance plays in facilitating concussion care for collegiate athletes. OBJECTIVE: Investigate the presence of athlete-athletic trainer (AT) racial and ethnic concordance and discordance amongst diagnosed concussion cases, and examine if racial and ethnic concordance and discordance influences time (in days) until diagnosis, symptom resolution, or return-to-sport clinical milestones in collegiate athletes. DESIGN: Retrospective cohort study. SETTING: Collegiate athletics. PATIENTS OR OTHER PARTICIPANTS: A total of 694 concussion cases [38.6% (n=268) sustained by women, 61.4% (n=426) sustained by men] that occurred within the 2015-2016 through 2019- 2020 sport seasons at 9 institutions. MAIN OUTCOME MEASURE(S): The number of days from date of injury to diagnosis, symptom resolution, and return-to-sport; and from date of diagnosis to symptom resolution and return-to- sport. RESULTS: Overall, 68.4% (n=475) of concussion cases had patient-provider racial and ethnic concordance and 31.6% (n =219) were discordant. All concordant pairs included a White athlete and White AT. Time to diagnosis differed between the concordant and discordant groups (median[IQR]=1[0,2] versus 0[0,1], respectively) only in the model adjusted for sex, sport-type, and availability of an AT (OR[CI95]=1.46[1.07, 1.85]). There were no other group differences. CONCLUSIONS: One-third of concussion cases had athlete-AT racial and ethnic discordance. While this group was diagnosed with a concussion 1-day sooner than the concordant group, no differences were observed for any concussion recovery milestones. These findings suggest that patient-provider racial and ethnic concordance may play a minor role in concussion recognition or reporting, but not necessarily in the management and recovery thereafter.

Community Engagement, Equity, and Cross-Sector Public Health Practice.

In this topical review, we integrate 3 concepts-public health practice, community engagement, and cross-sector governance-to consider the following question: What is the underlying relationship between public health and cross-sector governance according to which the field can understand the role of community engagement in achieving health equity? We begin with an overview of public health practice and the practice of community engagement. Next, we position these practices in the broader turn toward cross-sector governance. The integration of these themes reveals that common tools for community engagement fail to address questions about how services should be funded, how resources should be distributed, and which members of the community have a claim to services. We therefore suggest that distinguishing between community engagement for equity and equitable community engagement is a first step toward deepening dialogue about the role of public health in achieving health equity.

A 12-year follow up of principles learned in a pediatric dentistry Master of Public Health dual degree program applied to professional practice.

Introduction: Dental public health professionals play a critical role in preventing and controlling oral diseases. The purpose of this study was to assess the application of public health principles learned in a pediatric dentistry Master of Public Health (MPH) dual degree program to professional practice upon graduation. Methods: Semi-structured interviews were conducted with pediatric dentistry/MPH dual degree alumni who graduated from the program between 2012 and 2023. Interview questions inquired about characteristics of patient population, location of providers' clinic/organization, whether the program was worthwhile to their practice and application of principles learned in the program to their professional practice. Results: Twenty of the 22 program alumni agreed to be interviewed. All alumni thought the program was extremely worthwhile to their practice. They felt the MPH component of the program gave them the public health background and tools they needed to provide comprehensive and holistic care to their patients. Additionally, all alumni reported applying the public health principles they learned in the program to their professional practice through leadership roles, research and teaching that focuses on oral disease prevention and the promotion of dental health. Discussion: Given the importance of a dental public health professionals' role in reducing oral health disparities at the population level, more pediatric dentistry MPH dual degree programs are urgently needed. Additionally, more research is necessary to demonstrate the effectiveness of these programs, which will be critical to helping ensure the value of a dual degree in dentistry and public health is recognized and promoted worldwide.

Codesigning Online Continuing Medical Education on Social Health Integration and Social Risk-Informed Care for Primary Care Providers.

INTRODUCTION: Adapting clinical care decisions for patient-reported social risks is essential to social health integration and patient-centered care. Most research in this area focuses on awareness and assistance (social-needs-targeted care), such as screening and referral to food, financial, and other resources. Limited evidence for adjustment strategies (social risk-informed care) or adapting care for social risks made it difficult for Kaiser Permanente to implement new initiatives. This article describes a co-design process to build a novel, adjustment-focused continuing medical education course. METHODS: The authors co-developed the online continuing medical education course with patients and clinicians using user-centered design. Transcripts from co-design activities were coded and analyzed by thematic analysis to identify major themes, including perceptions of social risk-informed care and barriers to care adjustment. RESULTS: Practical hurdles for implementing social risk-informed care emerged, including clinicians' concerns about the ethics of adjustment as substandard care, particularly without robust assistance activities. However, patients expressed a desire for their care to be adapted to their social circumstances, to allow for more realistic care plans. DISCUSSION: Implementation barriers identified from the co-design were addressed through an interactive, case-study approach. Existing evidence on contextualized care and shared decision making informed a general framework for primary care providers to engage in awareness and adjustment activities, paired with 3 interactive case studies based on real-world, clinician-supplied scenarios. CONCLUSION: The authors recommend that multiple stakeholder perspectives be incorporated during the development of social health integration initiatives, particularly adjustment. Education complemented by active, nuanced, flexible implementation strategies may be necessary for the successful uptake of care-delivery-based social health integration activities.

Socioeconomic Position and Health Among Children and Adolescents With CKD Across the Life-Course.

Children and adolescents in families of lower socioeconomic position (SEP) experience an inequitable burden of reduced access to healthcare and poorer health. For children living with chronic kidney disease (CKD), disadvantaged SEP may exacerbate their considerable disease burden. Across the life-course, CKD may also compromise the SEP of families and young people, leading to accumulating health and socioeconomic disadvantage. This narrative review summarizes the current evidence on relationships of SEP with kidney care and health among children and adolescents with CKD from a life-course approach, including impacts of family SEP on kidney care and health, and bidirectional impacts of CKD on SEP. It highlights relevant conceptual models from social epidemiology, current evidence, clinical and policy implications, and provides directions for future research. Reflecting the balance of available evidence, we focus primarily on high-income countries (HICs), with an overview of key issues in low- and middle-income countries (LMICs). Overall, a growing body of evidence indicates sobering socioeconomic inequities in health and kidney care among children and adolescents with CKD, and adverse socioeconomic impacts of CKD. Dedicated efforts to tackle inequities are critical to ensuring that all young people with CKD have the opportunity to live long and flourishing lives. To prevent accumulating disadvantage, the global nephrology community must advocate for local government action on upstream social determinants of health; and adopt a life-course approach to kidney care that proactively identifies and addresses unmet social needs, targets intervening factors between SEP and health, and minimizes adverse socioeconomic outcomes across financial, educational and vocational domains.

Social determinants of health that impact the time to diagnosis and treatment of infertility in Taiwan.

BACKGROUND/PURPOSE: Taiwan is one of the countries with the lowest birth rate in the world. We investigated factors associated with the time to diagnosis and treatment of infertility in Taiwan. METHODS: The study was conducted through an online questionnaire in December 2021. The questionnaire was adapted from a previously published multinational survey, and culture-specific questions were added. 91 infertile patients and 89 partners of patients in Taiwan, aged 20- to 45- year-old, were included. RESULTS: The average duration before diagnosis was 2.9 years, followed by 1.5 years before treatment. Older age at marriage (p = 0.0024), higher education level (P = 0.0001), and a higher gender equality score (p = 0.0031) were associated with earlier diagnosis. Conversely, folk therapy use was linked to later diagnosis (p < 0.0001) and treatment (p < 0.0001). Notably, in the female (p = 0.039) and patient (p = 0.0377) subgroups, a higher gender equality score was associated with a shorter duration of folk therapy. Subjectively, the most frequent factor influencing treatment decision was affordability or lack thereof. The government subsidy for in vitro fertilization led to increased treatment willingness for 46.3% of respondents, and 47.3% reported more likely to pursue earlier treatment. CONCLUSIONS: This study highlights the influence of education, gender equality, folk therapy, and government subsidy on fertility care decisions. To improve the timeliness of infertility healthcare in Taiwan, potential strategies include promoting education, fostering gender equality, providing financial support, and raising awareness on the association between folk therapy and delayed medical care.

Patterns of Social Needs Predict Quality-of-Life and Healthcare Utilization Outcomes in Patients from a Large Hospital System.

BACKGROUND: Unmet social needs (SNs) often coexist in distinct patterns within specific population subgroups, yet these patterns are understudied. OBJECTIVE: To identify patterns of social needs (PSNs) and characterize their associations with health-related quality-of-life (HRQoL) and healthcare utilization (HCU). DESIGN: Observational study using data on SNs screening, HRQoL (i.e., low mental and physical health), and 90-day HCU (i.e., emergency visits and hospital admission). Among patients with any SNs, latent class analysis was conducted to identify unique PSNs. For all patients and by race and age subgroups, compared with no SNs, we calculated the risks of poor HRQoL and time to first HCU following SNs screening for each PSN. PATIENTS: Adult patients undergoing SNs screening at the Mass General Brigham healthcare system in Massachusetts, United States, between March 2018 and January 2023. MAIN MEASURES: SNs included: education, employment, family care, food, housing, medication, transportation, and ability to pay for household utilities. HRQoL was assessed using the Patient-Reported Outcomes Measurement Information System Global-10. KEY RESULTS: Six unique PSNs were identified: "high number of social needs," "food and utility access," "employment needs," "interested in education," "housing instability," and "transportation barriers." In 14,230 patients with HRQoL data, PSNs increased the risks of poor mental health, with risk ratios ranging from 1.07(95%CI:1.01-1.13) to 1.80(95%CI:1.74-1.86). Analysis of poor physical health yielded similar findings, except that the "interested in education" showed a mild protective effect (0.97[95%CI:0.94-1.00]). In 105,110 patients, PSNs increased the risk of 90-day HCU, with hazard ratios ranging from 1.09(95%CI:0.99-1.21) to 1.70(95%CI:1.52-1.90). Findings were generally consistent in subgroup analyses by race and age. CONCLUSIONS: Certain SNs coexist in distinct patterns and result in poorer HRQoL and more HCU. Understanding PSNs allows policymakers, public health practitioners, and social workers to identify at-risk patients and implement integrated, system-wide, and community-based interventions.

Socioeconomic Disparities and Emergency Department Visits for Diabetic Retinopathy in California.

Purpose: To investigate the association between neighborhood-level socioeconomic factors, quantified by the Distressed Communities Index, and emergency department visits for diabetic retinopathy (DR). Methods: All patients who presented to the emergency department for DR in California were analyzed using the State Emergency Department Database (2018-2020). Patients were stratified by Distressed Communities Index score and DR severity. Logistic regression was applied to explore the independent correlation between Distressed Communities Index scores and proliferative DR (PDR). Results: Of 2 725 195 emergency department visits for diabetic patients, Distressed Communities Index data were available for 2 459 577 (90.3%); 39 693 were for DR, including 13 617 (34.3%) for PDR. Hispanics (44.2%) were the largest racial/ethnic group to present for PDR, followed by non-Hispanic Whites (19.6%) and non-Hispanic Blacks (19.3%). A significant association was observed between the Distressed Communities Index and emergency department visits for PDR, with distressed neighborhoods having the highest incidence (adjusted odds ratio [aOR], 1.63; 95% CI, 1.20-2.23; P = .001). Other predictors included Hispanic ethnicity (aOR, 2.21; 95% CI, 1.97-2.48; P < .001) and Black race (aOR, 1.46; 95% CI, 1.28-1.67; P < .001) compared with White race and having Medicaid (aOR, 1.37; 95% CI, 1.13-1.65; P = .001) compared with private insurance. Conclusions: The Distressed Communities Index identified patients residing in the most distressed neighborhoods as being at the highest risk for presenting to the emergency department for PDR based on 7 socioeconomic factors. Policymakers may consider the Distressed Communities Index as a tool for targeting DR prevention strategies and improving healthcare accessibility.

Geospatial Imprecision With Constraints for Precision Public Health: Algorithm Development and Validation.

BACKGROUND: Location and environmental social determinants of health are increasingly important factors in both an individual's health and the monitoring of community-level public health issues. OBJECTIVE: We aimed to measure the extent to which location obfuscation techniques, designed to protect an individual's privacy, can unintentionally shift geographical coordinates into neighborhoods with significantly different socioeconomic demographics, which limits the precision of findings for public health stakeholders. METHODS: Point obfuscation techniques intentionally blur geographic coordinates to conceal the original location. The pinwheel obfuscation method is an existing technique in which a point is moved along a pinwheel-like path given a randomly chosen angle and a maximum radius; we evaluate the impact of this technique using 2 data sets by comparing the demographics of the original point and the resulting shifted point by cross-referencing data from the United States Census Bureau. RESULTS: Using poverty measures showed that points from regions of low poverty may be shifted to regions of high poverty; similarly, points in regions with high poverty may be shifted into regions of low poverty. We varied the maximum allowable obfuscation radius; the mean difference in poverty rate before and after obfuscation ranged from 6.5% to 11.7%. Additionally, obfuscation inadvertently caused false hot spots for deaths by suicide in Cook County, Illinois. CONCLUSIONS: Privacy concerns require patient locations to be imprecise to protect against risk of identification; precision public health requires accuracy. We propose a modified obfuscation technique that is constrained to generate a new point within a specified census-designated region to preserve both privacy and analytical accuracy by avoiding demographic shifts.

Social vulnerability index and age-adjusted mortality from all causes by county: Mississippi 2016-2020.

BACKGROUND: Social variables are correlates of mortality. A number of social variables were used by the Centers for Disease Control and Prevention (CDC) to create a Social Vulnerability Index (SVI). SVI has been used as a correlate of health status. Age-adjusted mortality rates have been higher in Mississippi than in other states. Within Mississippi, the Delta region has had higher mortality. To test the hypothesis that social vulnerability was associated with mortality rate within the state, we examined SVI of counties in Mississippi as related to mortality from all causes in 2016-2020. METHODS: The CDC/ATSDR SVI ranks each census tract on 16 social factors, including poverty, lack of vehicle access, and crowded housing, and groups them into four related themes. Using CDC Wonder, we gathered data analyzing age-adjusted rate of death from all causes (AAR) in Mississippi Counties from 2016 to 2020, combined (reporting the death rate per 100,000 persons). Descriptive statistics were computed for each variable. Pearson correlation analysis, bivariable and multivariable regression analysis was done using Microsoft Excel version 16.77. The dependent variable was AAR and independent variables were for themes from the SVI. RESULTS: AAR varied greatly amongst counties in Mississippi. Higher AAR was seen in northwestern areas of Mississippi. The county with the lowest AAR (730 per 100,000 persons) had only half the AAR of the county with the highest AAR (1313.3 per 100,000 persons). The association of SVI THEME 1 (socioeconomic status) with AAR in Mississippi was positive. Linear regression analysis showed a coefficient of 203.5, 95 % CI 111.9-295.0, p = 0. 0.0000305. R square was 0.20. The addition of the following themes added little to the variation in AAR explained: SVI THEME 2 (household characteristics), SVI THEME 3 (racial and ethnic minority status), and SVI THEME 4 (housing type/transportation). CONCLUSION: Socioeconomic status explained a fifth of the variation in AAR among Mississippi counties in 2016-2020.

Disadvantaged social status contributed to sleep disorders: An observational and genome-wide gene-environment interaction analysis.

BACKGROUND: Sleep is a natural and essential physiological need for individuals. Our study aimed to research the associations between accumulated social risks and sleep disorders. METHODS: In this study, we came up with a polysocial risk score (PsRS), which is a cumulative social risk index composed of 13 social determinants of health. This research includes 239,165 individuals with sleep disorders and social determinants of health data from the UK Biobank cohort. First, logistic regression models were performed to examine the associations of social determinants of health and sleep disorders, including chronotype, narcolepsy, insomnia, snoring, short and long sleep duration. Then, PsRS was calculated based on statistically significant social determinants of health for each sleep disorder. Third, a genome-wide gene-environment interaction study was conducted to explore the interactions between single-nucleotide polymorphisms and PsRS in relation to sleep disorders. RESULTS: Higher PsRS scores were associated with worse sleep status, with the adjusted odds ratio (OR) ranging from 1.10 (95% Confidence interval [CI]: 1.09-1.11) to 1.29 (95% CI: 1.27-1.30) for sleep disorders. Emotional stress (OR = 1.36, 95% CI: 1.28-1.43) and not in paid employment (OR = 2.62, 95% CI: 2.51-2.74) were found to have significant contributions for sleep disorders. Moreover, multiple single-nucleotide polymorphisms were discovered to have interactions with PsRS, such as FRAS1 (P = 2.57 × 10-14) and CACNA1A (P = 8.62 × 10-14) for narcolepsy, and ACKR3 (P = 1.24 × 10-8) for long sleep. CONCLUSIONS: Our findings suggested that cumulative social risks was associated with sleep disorders, while the interactions between genetic susceptibility and disadvantaged social status are risk factors for the development of sleep disorders.

Predicting first time depression onset in pregnancy: applying machine learning methods to patient-reported data.

PURPOSE: To develop a machine learning algorithm, using patient-reported data from early pregnancy, to predict later onset of first time moderate-to-severe depression. METHODS: A sample of 944 U.S. patient participants from a larger longitudinal observational cohortused a prenatal support mobile app from September 2019 to April 2022. Participants self-reported clinical and social risk factors during first trimester initiation of app use and completed voluntary depression screenings in each trimester. Several machine learning algorithms were applied to self-reported data, including a novel algorithm for causal discovery. Training and test datasets were built from a randomized 80/20 data split. Models were evaluated on their predictive accuracy and their simplicity (i.e., fewest variables required for prediction). RESULTS: Among participants, 78% identified as white with an average age of 30 [IQR 26-34]; 61% had income ≥ $50,000; 70% had a college degree or higher; and 49% were nulliparous. All models accurately predicted first time moderate-severe depression using first trimester baseline data (AUC 0.74-0.89, sensitivity 0.35-0.81, specificity 0.78-0.95). Several predictors were common across models, including anxiety history, partnered status, psychosocial factors, and pregnancy-specific stressors. The optimal model used only 14 (26%) of the possible variables and had excellent accuracy (AUC = 0.89, sensitivity = 0.81, specificity = 0.83). When food insecurity reports were included among a subset of participants, demographics, including race and income, dropped out and the model became more accurate (AUC = 0.93) and simpler (9 variables). CONCLUSION: A relatively small amount of self-report data produced a highly predictive model of first time depression among pregnant individuals.