Public-use Data Research Articles

The effects of childhood maltreatment and social support on the trajectories of depressive symptoms from adolescence to young adulthood.

Depression is one of the most common mental health concerns among adolescents and young adults. Childhood maltreatment is a particularly potent risk factor for the subsequent development of depressive symptoms. Meanwhile, social support has been identified as a robust protective factor against depressive symptoms. However, limited research has investigated the effects of childhood maltreatment and social support on the trajectories of depressive symptoms from adolescence to young adulthood. The present study aimed to explore trajectories of depressive symptoms from adolescence to young adulthood as well as identify childhood maltreatment as a potential risk factor and social support as a protective factor to depressive symptoms over time. We conducted growth curve modeling using public-use data from Add Health. Main variables (childhood maltreatment, social support, depressive symptoms from Wave I to Wave IV) and control variables (e.g., gender, race, poverty) were included in the analysis. Depressive symptoms demonstrated a quadratic pattern, with a decline from adolescence to emerging adulthood, and a slight increase in young adulthood. Childhood maltreatment was only significantly associated with the initial level of depressive symptoms but not the trajectories of depressive symptoms. Social support was not only associated with the initial level of depressive symptoms, but also the changes of depressive symptoms over time. The present study provides an avenue towards untangling potential factors contributing to the heterogeneity of depressive symptoms. Identifying risk and protective factors with varying trajectories of depressive symptoms will potentially improve related treatments and interventions.

P-162. Estimating the Seroprevalence of HTLV-1 in Pregnant Peruvian Women Based on Blood Bank Markers: a Road to Elimination of Mother-to-Child Transmission of HTLV-1

Abstract Background The prevention of mother-to-child transmission (PMTCT) of human T-cell lymphotropic virus type 1 (HTLV-1) is a top priority for HTLV-endemic countries due to its chronicity, lack of treatment and association with the aggressive adult T cell leukemia/lymphoma in comparison to other routes of infection. As of 2024, Japan, Brazil and Chile are few of the countries with national policies regarding antenatal screening. Peru, also an endemic country with around 150,000 to 450,000 HTLV-1 carriers, has limited data in pregnant women and does not have any PMTCT policy in place. By contrast, HTLV-1/2 antibody screening of blood donations has been endorsed since 1997 and national data are available. We aim to leverage these data to estimate the prevalence of HTLV-1 in pregnant women as a first step towards development and implementation of a PMTCT policy in Peru. Figure 1 Estimation of number of HTLV-1 infected pregnant women and HTLV-1 seroprevalence in pregnant women in Peru (2017) Methods This was an ecological study using public-use data from the Peruvian Bulletin on Blood Supply 2015-2020, the 2017 National Census, the 2017 Demographic and Health Survey (DHS) and the National Registry. The unit of analysis was Peru and its subregions in the year of 2017. To estimate the number of infected women of reproductive age (WRA, 15-49 years), we used census data and a calculated percentage of HTLV-1/2 reactive blood units for this group based on a weighted factor by sex and age (Figure 1, Table 1). Using the DHS percentage of WRA currently pregnant, we estimated the number of infected pregnant women. To obtain the HTLV-1 prevalence for this group, we used the number of life births adjusted for a 9-month period (0.75) as the denominator. Calculation of weighted factor for HTLV-1/2 seroprevalence by age and sex based on local blood bank data Results A total of 359,458 blood units were screened and 2,809 were reactive for HTLV-1/2 (0.78%) (Figure 2). We estimated between 68,456 to 72,978 HTLV-1 carriers who were WRA, of whom 2,463 to 2,773 were pregnant (Figure 3). This is comparable to rates in Japan and Brazil. Our estimated HTLV-1 seroprevalence in pregnant women was 0.61% for the entire country. Number and percentage of HTLV-1/2 reactive blood units in 2017 by departments of Peru Conclusion Our findings could help guide future strategies for PMTCT, such as universal antenatal screening and assessing the feasibility of formula supplementation for infants born to HTLV-1 positive mothers. Our limitations included the use of one-year data, HTLV-1/2 false-positives in blood donors and not using inverse probability weighting. Estimated number and percentage of HTLV-1 infected pregnant women in 2017 by Departments of Peru Disclosures All Authors: No reported disclosures

Waist circumference and altered metabolic indices are associated with increased resting pulse rate in middle-aged adults.

Metabolic indices significantly impact cardiovascular health. Research on the influence of metabolic indices on resting pulse rate in younger adults is needed. Utilizing the National Longitudinal Study of Adolescent to Adult Health (Add Health) Wave 5 Public-use biological data, we performed a multiple linear regression analysis to determine the predictive factors of resting pulse rate in adults aged 32-42 years. We controlled for sex, anthropometrics, lipid indices, diabetic status, blood pressure, and antihypertensive use. The total number of participants was 1839 (weighted n=1,390,763), the mean age was 38.1 [standard deviation (SD)=2.0] and the mean pulse rate was 75.2 (SD=11.6). Notably, body mass index did not exhibit a statistically significant relationship with resting pulse rate. Conversely, females presented a statistically significant higher pulse rate than males when adjusted for other variables [p<0.001; coefficient = 4.83; 95% confidence interval (CI): 3.54-6.13], as did individuals with increased waist circumference (p=0.023; coefficient = 2.62; 95% CI: 0.39- 4.85). The data indicated a progressive rise in pulse rate correlating with elevated low-density lipoprotein and hemoglobin A1C levels, particularly among those with severe hypercholesterolemia (p=0.048, coefficient = 6.42; 95% CI: 0.12-12.71) and diabetes (p<0.001, coefficient = 7.11; 95% CI: 3.3-10.92). A significant increase was also observed in individuals with hypertension stage 1 and 2 [p<0.001 for both; coefficients = 2.98 (95% CI: 1.45-4.52) and 7.2 (95% CI: 5.28-9.12), respectively]. These findings underscore the necessity of considering metabolic indices in understanding the pathophysiology of elevated pulse rates in adults, thereby enhancing comprehension of cardiovascular tachyarrhythmias in younger adults.

Examining Lung Cancer Screening Uptake in the United States: Recent Research and Limitations of Public-Use Data.

Eligibility criteria for lung cancer screening (LCS) were updated by the Centers for Medicare & Medicaid Services in 2022 following an updated recommendation from the United States Preventive Services Task Force. Recently, research has examined LCS use in the United States following this change, which centered on lowering the age to begin screening from 55 to 50 years and the smoking history threshold from 30 to 20 pack-years. These studies, including the accompanying article from Gudina and colleagues, have used the 2022 Behavioral Risk Factors Surveillance System data, which are publicly available and nationally representative. Although increases in LCS use have been reported in all studies, screening remains low compared with other population-level cancer screening programs, highlighting the need for interventions to improve LCS uptake and adherence. Data limitations from population-based surveys may affect the ability to accuracy track LCS rates. Additional, high-quality data are needed to further assess the uptake of-and adherence to-LCS following the updated eligibility criteria. See related article by Gudina et al., p. 35.

DISPARITIES IN HEALTH OUTCOMES AMONG MIDDLE-AGED AND OLDER SEXUAL MINORITY CANCER SURVIVORS

Abstract This cross-sectional, exploratory study analyzed 2021 Behavioral Risk Factor Surveillance System public-use data for a propensity-score matched sample (N=25,545), from states that collected sexual identity and cancer survivorship information, to explore health outcomes among sexual minority (SM) and non-SM cancer survivors. In bivariate analysis, SM survivors had more chronic conditions than non-SM survivors (males 45-64 years: 1.67 vs 1.03; females 45-64 years: 1.46 vs 1.16) and were more likely to report multiple chronic conditions (males 45-64 years: 44.5% vs 29.0%), depressive disorders (males 45-64 years: 42.9 vs 23.4%; males 65+ years: 20.8% vs 11.2%; females 45-64 years: 52.9% vs 36.5%; females 65+ years: 42.7% vs 41.3%), and difficulty concentrating, remembering, or making decisions because of a physical, mental, or emotional condition (females 45-64 years: 30.1% vs 18.9%; males 65+ years: 15.9% vs 8.5%). In logistic regression models, after controlling for demographics, health, and health behaviors, SM males and females in both age groups had similar odds of being a cancer survivor as their non-SM counterparts, and greater odds of poor health, multiple chronic conditions, depressive disorder, ADL limitations, and difficulty concentrating, remembering, or making decisions. Cancer status was not always associated with these outcomes. Further research should explore treatment experiences and factors contributing to disparities within this population and among SM cancer survivors and assess whether health outcomes of SM cancer survivors are related to stage of disease at diagnosis, time since treatment, or decisions made during treatment.

Sibling and Parental Military Service and Suicidality Among Adolescents in the United States.

This study aims to examine whether having a sibling and/or a parent on active duty in the military is associated with suicidality-think about death, better off dead, think about suicide, plan suicide, and attempt suicide-among 12- to 17-year-old adolescent girls and boys in the United States. Descriptive and multivariable logistic regression analyses were conducted using pooled cross-sectional public-use data from the 2016-2019 National Survey of Drug Use and Health. Analyses were weighted and standard errors were adjusted for the complex sampling design. Having a sibling on active duty increases the likelihood of reporting all five suicidality outcomes. Having a parent on active duty increases the likelihood of each of the suicidality outcomes except think about death for both adolescent boys and girls relative to adolescents without a parent in the military. Adolescent girls who have a parent serving in the military are more likely to report suicidal ideation (better off dead, think about suicide) than adolescent boys who have a parent currently serving in the military but are equally likely to report suicide planning and attempt. Those working clinically with military families should be mindful of the association between suicidality and sibling military service, as well as parental military service. Programs aiming to reduce the negative impact of sibling deployment need to be developed and tested empirically. The current findings suggest the need for targeted family-centered approaches to suicide prevention among youth with siblings, parents, and potentially other relatives currently serving in the armed forces.

The Prevalence of Obesity is Increased in Adolescents with Amblyopia: An Analysis of National Health and Nutrition Examination Survey Data.

Background: Amblyopia is the most common cause of vision loss in children. Amblyopia has been associated with impaired depth perception but little attention has been paid to the extent to which amblyopia increases the risk of obesity. Methods: Public-use data from the 1999-2008 National Health and Nutrition Examination Survey were used. Analyses were limited to children aged 12-18, who had a visual examination, and a best corrected visual acuity (BCVA) of at least 20/40 in the better-seeing eye. Amblyopia was defined as two or more-line interocular difference in BCVA. Obesity was defined as Body Mass Index (BMI) or body fat percentage (BFP) ≥95th percentile for age and gender. Sedentary lifestyle was defined as cardiovascular fitness level (CFL) rating of "low." We used Mantel-Haenszel odds ratios (ORs) to examine the relative prevalence of obesity in children with/without amblyopia. Results: Adolescents with amblyopia (n = 360) were more likely than those without (n = 7935) to have a high BMI [OR = 1.56; 95% confidence interval (CI): 1.24-1.98; p < 0.001]. The associations with either high BFP (OR = 1.20; 95% CI: 0.86-1.56, p = 0.167) or low CFL (OR = 1.15; 95% CI: 0.83-1.57; p = 0.267) were not statistically significant but in the direction of a priori hypotheses. Conclusions: This analysis of population-based data suggests that adolescents with amblyopia may be at higher risk of having obesity. Given the high prevalence of amblyopia and the range of morbidities associated with childhood obesity, targeted interventions to reduce the risk of obesity among children with amblyopia could be warranted.

Are the Effects of Poor Survey Measurement of Sexual Identity on Estimated Associations between Sexual Identity and Health-Related Measures Moderated by the State-Level Policy Environment?

ABSTRACT Including a “something else” response option for survey questions about sexual identity significantly moderates estimated differences between sexual identity subgroups in terms of behaviors and attitudes reported in U.S. health surveys. We hypothesize that these moderation effects will be larger in states with negative policies that fail to protect sexual minorities from discrimination. We tested this hypothesis by linking public-use data from the National Survey of Family Growth (2015–2019), which randomly assigned respondents to receive either a four-category measure of sexual identity, including “something else” as a response option, or a three-category measure omitting “something else,” to year-specific state-level policy data from the Movement Advancement Project, which tracks state policies related to sexual minorities. In multivariable models for measures of substance use, reproductive health, family formation, and other correlates of health, we tested three-way interactions involving sexual identity, measurement type, and state-level policy classification. As hypothesized, we find repeated evidence (10 of the 32 health-related measures analyzed) of changes in the moderating effects of question type on sexual identity differences, with more prominent effects in negative policy environments. Suboptimal measurement of sexual identity may therefore have the largest negative effects on estimated health disparities in states with fewer protective policies.

Advanced Practice Providers in Burn Care, 2013-2022.

This study evaluated the prevalence characteristics of advanced practice providers, including nurse practitioners and physician assistants, who provide care related to burns in the United States from 2013 to 2022 using national claims data. Our analysis was a retrospective, large cohort study evaluating debridement, wound care, and office-based services provided by advanced practice providers from 2013 to 2022 using the Medicare Provider Utilization and Payment Data Public Use Files from the CMS. The reported provider type and billing codes were used to identify healthcare professionals providing burn care. Trends over the study period and available data about care provided were analyzed. From 2013 to 2022, burn care providers included 6,918 (43.3%) surgeons, 4,264 (26.7%) other physicians, and 4,783 (30.0%) advanced practice providers. Overall, the number of advanced practice providers who billed for burn care increased by 250.4% from 714 in 2013 to 2,502 in 2022. Compared with physicians, advanced practice providers were more likely to be female and provide care in micropolitan areas with less than 50,000 people. Based on these trends, we project that the number of APPs providing services related to burns will increase by 56.7% from 2,502 in 2022 to 3,921 (95% CI, 3,541-4,303; R2=0.97) in 2030. With a growing need for burn care and predicted shortages in the surgeon workforce, the scope of practice and integration of advanced practice providers merits further discussion and evaluation.

Prevalence and Practice Characteristics of Advanced Practice Providers in Upper-Extremity Surgery, 2013–2021

This study evaluated the prevalence, characteristics, and reimbursement of advanced practice providers, including nurse practitioners and physician assistants, who provide care related to the diagnosis and treatment of diseases and conditions of the hand, wrist, and upper extremities in the United States from 2013 to2021. Our analysis was a retrospective cohort study evaluating the diagnostic, procedural, and therapeutic services provided by advanced practice providers from 2013 to 2021 using the Medicare Provider Utilization and Payment Data Public Use Files from the Centers for Medicare and Medicare Services. The reported provider type and billing codes were used to identify health care professionals providing upper-extremity care such as ordering radiographs, applying casts and splints, and performing procedures on the hand, wrist, or other anatomic regions of the upper extremity. Trends over the study period and available data about services provided were analyzed. From 2013 to 2021, providers of upper-extremity care included 19,525 (64.7%) doctor of medicine or doctor of osteopathic medicine upper-extremity surgeons, 7,612 (25.2%) physician assistants, and 3,042 (10.1%) nurse practitioners. The nonsurgeon providers were more likely to be women and provide care in micropolitan areas with less than 50,000 people compared with upper-extremity surgeons. Overall, the number of advanced practice providers who billed for upper-extremity care increased by 170.9% from 1,965 in 2013 to 5,324 in 2021. Based on these trends, the growth of APPs providing upper-extremity care is expected to continue. There is a growing prevalence of advanced practice providers in upper-extremity care, and this trend is expected to continue. With a growing need for upper-extremity care and predicted shortages in the surgeon workforce, the scope of practice and integration of advanced practice providers merits further discussion and evaluation.

Health Care Access Among Children in Latinx Families Across U.S. Destinations

Latinx children now live in a wider array of U.S. geographic areas than in the past, including both established and new areas of Latinx settlement. This geographic heterogeneity could be consequential for Latinx children’s health care access, with prior research suggesting increased health access barriers for Latinx children in new versus established areas of settlement. Merging public-use county-level data with restricted individual-level health data from the National Health Interview Survey (2010–2014), we quantitatively examine how three health access indicators—health insurance coverage, delayed care, and usual place of care—differ among children (ages 4–17) in Latinx immigrant, Latinx U.S.-born, White U.S.-born, and Black U.S.-born families (n = 89,994) across established, fast-growing hub, new, and minor Latinx destination counties. We also examine the potential roles of local immigrant hostilities and health care resources in contributing to health access differences across destinations. In fully adjusted models, children in new destinations are less likely to have health insurance than peers in established destinations, and this disparity is even wider for Latinx children of immigrants. Adjusted model results also show that children in new destinations are more likely to have delayed care than those in established destinations, and children in these four groups in new destinations, fast-growing hubs, and minor destinations are more likely to have no usual place of care than peers in established destinations. Our results are consistent with prior work suggesting that more health care access barriers exist for children, particularly Latinx children of immigrants, in new versus established Latinx destinations.

Charcot neuroarthropathy in diabetic patients in Texas

AimsCharcot neuroarthropathy (CN) is a complex disease of the bone and joints that can lead to serious and life-threatening complications. This study investigates epidemiologic trends in diabetic CN in Texas and the impact of age on these values. MethodsA retrospective analysis was conducted using the Texas Department of State Health Services Hospital Discharge Data Public Use Data File. Using International Classification of Diseases, Ninth (ICD-9) and Tenth (ICD-10) Revision codes, we identified patients with diabetes and Charcot ankle or foot. Data extracted included diagnoses, race, and gender. Population rates were estimated using census data, calculated per 1000 population and standardized by age. ResultsOverall and age-standardized rates of CN increased each year from 2006 to 2016, except for a downward trend from 2014 to 2016. Poisson regression revealed significant increases in the incidence rate ratio compared to 2006 for each year from 2008 to 2016. When age group is included, all years except 2007 show a significant increase relative to 2006, and all age groups have increased rates relative to ages 18–44. Major and minor amputations in patients with CN have increased. ConclusionsThe increasing rates of CN and amputations highlight the need for further research and standardized strategies for diagnosis and management.

Association between influenza vaccination and one-year all-cause and cardiovascular mortality risk: A self-controlled case series and matched case-control study.

Debates surrounding the efficacy of influenza vaccination for survival benefits persist, and there is a lack of data regarding its duration of protection. A self-controlled case series (SCCS) and a 1:4 matched case-control study were conducted using the National Health Interview Survey (NHIS) and public-use mortality data from 2005 to 2018 in the United States. The SCCS study identified participants who received influenza vaccination within 12 months before the survey and subsequently died within 1 year of postvaccination. The matched case-control study paired participants who died during the influenza season at the time of survey with four survivors. Among 1167 participants in the SCCS study, there was a 46% reduction in all-cause mortality and a 43% reduction in cardiovascular mortality within 29-196 days of postvaccination. The greatest protection was observed during days 29-56 (all-cause mortality: RI: 0.19; 95% CI: 0.12-0.29; cardiovascular mortality: RI: 0.28; 95% CI: 0.14-0.56). Among 626 cases and 2504 controls included in the matched case-control study, influenza vaccination was associated with a reduction in all-cause mortality (OR: 0.74, 95% CI: 0.60-0.92) and cardiovascular mortality (OR: 0.64, 95% CI: 0.44-0.93) during the influenza season. This study highlights the importance of influenza vaccination in reducing the risks of all-cause and cardiovascular mortality, with effects lasting for approximately 6 months.

Association between C-reactive protein/albumin ratio and all-cause mortality in patients with stroke: Evidence from NHANES cohort study

Background and aimsThe inflammatory nutritional status is widely associated with the long-term prognosis of non-fatal stroke. The objective of this study is to examine the correlation between the C-reactive protein to albumin ratio (CAR), a new marker indicating both inflammatory and nutritional status, and the overall mortality rate among stroke patients. Methods and resultsData were obtained from the National Health and Nutrition Examination Survey (NHANES) database and corresponding public-use mortality data from the linked National Death Index (NDI). The study utilized maximally selected rank statistics to determine the optimal cutoff points for the CAR. Subsequently, participants were stratified into higher- and lower-CAR groups based on these cutoff points. The Kaplan-Meier survival method was used to study overall survival probability. Multivariable Cox proportional regression models were employed to calculate the Hazard Ratio (HR) and corresponding confidence interval (CI). Restricted cubic spline (RCS) model was applied to detect potential non-linear relationship between CAR and mortality risk. Furthermore, stratified and sensitive analyses were performed to examine the robustness and reliability of the results. The study, encompassing 1043 participants with an average age of 64.61 years, identified a cutoff value of 0.32 for CAR, with notable variances observed across gender and age cohorts. Over an average follow-up period of 116 months, 679 instances of all-cause mortality were documented. Kaplan-Meier survival analysis unveiled noteworthy disparities in survival probabilities between groups categorized by high and low CAR levels (p = 0.00081). Continuous CAR analysis consistently demonstrated a positive correlation between elevated CAR values and heightened risk (HR = 1.78 (1.36, 2.33)) of all-cause mortality among stroke patients. Similarly, individuals in the high CAR group exhibited adjusted HR of 1.34 (0.96, 1.89) for all-cause mortality compared to their low CAR counterparts. Subgroup and sensitive analysis consistently reinforced these findings. Smoothing curve fitting further validated CAR's significance as a prognostic indicator of all-cause mortality, indicating a linear relationship. ConclusionElevated CAR is associated with increased long-term risk of mortality for individuals who have experienced a stroke, suggesting that CAR could serve as a survival indicator.

Health Disparities: Differences in Veteran and Non-Veteran Populations using Generative AI

Introduction: This study investigated self-reported health status, health screenings, vision problems, and vaccination rates among veteran and non-veteran groups to uncover health disparities that are critical for informed health system planning for veteran populations. Methods: Using public-use data from the National Health Interview Survey (2015-2018), this study adopts an ecologic cross-sectional approach to conduct an in-depth analysis and visualization of the data assisted by Generative AI, specifically ChatGPT-4. This integration of advanced AI tools with traditional epidemiological principles enables systematic data management, analysis, and visualization, offering a nuanced understanding of health dynamics across demographic segments and highlighting disparities essential for veteran health system planning. Findings: Disparities in self-reports of health outcomes, health screenings, vision problems, and vaccination rates were identified, emphasizing the need for targeted interventions and policy adjustments. Conclusion: Insights from this study could inform health system planning, using epidemiological data assessment to suggest enhancements for veteran healthcare delivery. These findings highlight the value of integrating Generative AI with epidemiological analysis in shaping public health policy and health planning.

A model for supporting biomedical and public health researcher use of publicly available All of Us data at Historically Black Colleges and Universities.

The aim of this study was to describe opportunities and challenges associated with the development and implementation of a program for supporting researchers underrepresented in biomedical research. We describe a case study of the All of Us Researcher Academy supported by the National Institutes of Health (NIH), including feedback from participants, instructors, and coaches. Lessons include the importance of inviting role models into learning networks, establishing and maintaining trusted relationships, and making coaches available for technical questions from researcher participants. Although research has focused on learning outcomes in science, technology, engineering, and mathematics at Minority Serving Institutions in the United States, literature tends to lack models for initiatives to improve everyday research experiences of faculty and researchers at such institutions or to encourage researcher use of public-use data such as that available through NIH's All of Us Research Program. The All of Us Researcher Academy offers a model that addresses these needs.

Assessing the prevalence of Beers medication utilization in the Medicare Part D population in 2020.

Medication utilization has been increasing in the U.S. year-over-year for several decades. As older adults take more medications, there is a higher risk of them being exposed to drug-drug or drug-disease interactions. The American Geriatrics Society in 2019 updated their Beers Criteria for Potentially Inappropriate Medication (PIM) Use in Older Adults. The objective of this study was to assess the prevalence of utilization of medications included in the 2019 Beers Criteria. An analysis was conducted using the Medicare Part D Provider Utilization and Payment Data Public Use File for calendar-year 2020. Medications identified in the 2019 Beers Criteria were applied to the analysis. Two categories of medications were assessed: (1) "Avoid" and (2) "Use With Caution." In 2020, 56 million prescriptions were dispensed to Medicare patients 65 years and older that are recommended to be avoided without exception (4.7% of all prescriptions) totaling $957 million in medication costs. The most utilized medication classes in this category were benzodiazepines (25,949,994 prescriptions), "Z-drugs" (6,204,909 prescriptions), long-acting sulfonylureas (5,306,577 prescriptions), 1st-generation antihistamines (5,049,289 prescriptions), and tricyclic antidepressants (4,190,062 prescriptions). Additionally, 135 million prescriptions were dispensed to Medicare beneficiaries 65 years and older for medications which the Beers Criteria states to use caution (11.3% of all prescriptions) exceeding $2.85 billion in medication costs. The most utilized medications for this category were diuretics (74,599,126 prescriptions), selective serotonin reuptake inhibitors (30,033,121 prescriptions), serotonin and norepinephrine reuptake inhibitors (11,858,968 prescriptions), tramadol (11,450,878 prescriptions), and mirtazapine (5,737,304 prescriptions). Even with the existence of the AGS Beers Criteria for PIM Use in Older Adults and its continued updated versions, 16% of medications dispensed to Medicare Part D were potentially inappropriate. Future studies are needed to assess if this has led to worsened outcomes among older adults who utilized these PIM.

Daily spiritual experiences and allostatic load trajectories: a longitudinal study of midlife African American women.

This study aimed to evaluate the association between daily spiritual experiences and allostatic load (AL) trajectories in midlife African American women. A longitudinal analysis of public-use data from 727 African American women in the Study of Women's Health Across the Nation (SWAN) was performed. We included African American women who completed the Daily Spiritual Experiences Scale at SWAN visit 4 (2000-2001) and had AL data at three or more study visits over 7 years. AL was calculated at each visit using 10 biomarkers: systolic and diastolic blood pressure, body mass index, C-reactive protein, high-density lipoprotein cholesterol, total cholesterol, waist-to-hip ratio, fasting serum glucose, triglycerides, and dehydroepiandrosterone. Group-based trajectory modeling identified women with similar patterns of AL. We used multinomial logistic regression to estimate associations between daily spiritual experiences (some days or less, most days, daily, many times a day) and AL trajectories. Our sample had a mean ± SD age of 49.9 ± 2.66 years, 47% were early perimenopausal, and 17% earned <$19,999 annually. The mean ± SD AL score was 2.52 ± 1.68. Three AL trajectories were identified: low (35.1%), moderate (44.7%), and high (20.2%). In age-adjusted models, women who reported daily comfort in religion and spirituality were less likely to follow a high AL trajectory (odds ratio, 0.41; 95% CI, 0.18-0.93); the association was attenuated when controlling for depressive symptoms (odds ratio, 0.48; 95% CI, 0.19-1.21). Findings from this study do not support an independent association between spirituality in African American women and AL trajectories in midlife. Studies with a larger sample and additional measures of spirituality are warranted in this population.

Nearly collinear robust procedures for 2SLS estimation

Stata’s two-stage least-squares (2SLS) computation procedures are sensitive to near collinearity among regressors, allowing situations in which reported results depend upon factors as irrelevant as the order of the data and variables. This article illustrates this claim with the public-use data of Oreopoulos (2006, American Economic Review 96: 152–175), where the instrumented coefficient estimate can be made to vary between 0.012 and 30.0 in one specification by permuting the order of the variables. Different methods for improving the accuracy of 2SLS estimates are reviewed, and a Stata command for collinearity-robust 2SLS estimation is provided.

Longitudinal association between e-cigarette use and respiratory symptoms among US adults: Findings from the Population Assessment of Tobacco and Health Study Waves 4-5.

We assessed longitudinal effects of e-cigarette use on respiratory symptoms in a nationally representative sample of US adults by combustible tobacco smoking status. We analyzed Waves 4-5 public-use data from the Population Assessment of Tobacco and Health Study. Study sample included adult respondents who reported no diagnosis of respiratory diseases at Wave 4, and completed Waves 4-5 surveys with no missing data on analytic variables (N = 15,291). Outcome was a validated index of functionally important respiratory symptoms based on 7 wheezing/cough questions (range 0-9). An index score of ≥2 was defined as having important respiratory symptoms. Weighted lagged logistic regression models were performed to examine the association between e-cigarette use status at Wave 4 (former/current vs. never use) and important respiratory symptoms at Wave 5 by combustible tobacco smoking status (i.e., never/former/current smokers), adjusting for Wave 4 respiratory symptom index, sociodemographic characteristics, secondhand smoke exposure, body mass index, and chronic disease. Among current combustible tobacco smokers, e-cigarette use was associated with increased odds of reporting important respiratory symptoms (former e-cigarette use: adjusted odds ratio [AOR] = 1.39, 95% confidence interval [CI]: 1.07-1.81; current e-cigarette use: AOR = 1.55, 95% CI: 1.17-2.06). Among former combustible tobacco smokers, former e-cigarette use (AOR = 1.51, 95% CI: 1.06-2.15)-but not current e-cigarette use (AOR = 1.59, 95% CI: 0.91-2.78)-was associated with increased odds of important respiratory symptoms. Among never combustible tobacco smokers, no significant association was detected between e-cigarette use and important respiratory symptoms (former e-cigarette use: AOR = 1.62, 95% CI: 0.76-3.46; current e-cigarette use: AOR = 0.82, 95% CI: 0.27-2.56). The association between e-cigarette use and respiratory symptoms varied by combustible tobacco smoking status. Current combustible tobacco smokers who use e-cigarettes have an elevated risk of respiratory impairments.