Metabolomic data presents challenges for epidemiological meta-analysis: a case study of childhood body mass index from the ECHO consortium.

INTRODUCTION: Meta-analyses across diverse independent studies provide improved confidence in results. However, within the context of metabolomic epidemiology, meta-analysis investigations are complicated by differences in study design, data acquisition, and other factors that may impact reproducibility. OBJECTIVE: The objective of this study was to identify maternal blood metabolites during pregnancy (> 24 gestational weeks) related to offspring body mass index (BMI) at age two years through a meta-analysis framework. METHODS: We used adjusted linear regression summary statistics from three cohorts (total N = 1012 mother-child pairs) participating in the NIH Environmental influences on Child Health Outcomes (ECHO) Program. We applied a random-effects meta-analysis framework to regression results and adjusted by false discovery rate (FDR) using the Benjamini-Hochberg procedure. RESULTS: Only 20 metabolites were detected in all three cohorts, with an additional 127 metabolites detected in two of three cohorts. Of these 147, 6 maternal metabolites were nominally associated (P < 0.05) with offspring BMI z-scores at age 2 years in a meta-analytic framework including at least two studies: arabinose (Coefmeta = 0.40 [95% CI 0.10,0.70], Pmeta = 9.7 × 10-3), guanidinoacetate (Coefmeta = - 0.28 [- 0.54, - 0.02], Pmeta = 0.033), 3-ureidopropionate (Coefmeta = 0.22 [0.017,0.41], Pmeta = 0.033), 1-methylhistidine (Coefmeta = - 0.18 [- 0.33, - 0.04], Pmeta = 0.011), serine (Coefmeta = - 0.18 [- 0.36, - 0.01], Pmeta = 0.034), and lysine (Coefmeta = - 0.16 [- 0.32, - 0.01], Pmeta = 0.044). No associations were robust to multiple testing correction. CONCLUSIONS: Despite including three cohorts with large sample sizes (N > 100), we failed to identify significant metabolite associations after FDR correction. Our investigation demonstrates difficulties in applying epidemiological meta-analysis to clinical metabolomics, emphasizes challenges to reproducibility, and highlights the need for standardized best practices in metabolomic epidemiology.

Frailty Among Sexual and Gender Minority Older Adults: The All of Us Database.

BACKGROUND: Despite known disparities in health status among older sexual and gender minority adults (OSGM), the prevalence of frailty is unknown. The aim of this study was to develop and validate a deficit-accumulation frailty index (AoU-FI) for the All of Us database to describe and compare frailty between OSGM and non-OSGM participants. METHODS: Developed using a standardized approach, the AoU-FI consists of 33 deficits from baseline survey responses of adults aged 50+. OSGM were self-reported as "not straight" or as having discordant gender and sex assigned at birth. Descriptive statistics characterized the AoU-FI. Regression was used to assess the association between frailty, age, and gender. Validation of the AoU-FI used Cox proportional hazard models to test the association between frailty categories (robust <0.15, 0.15 ≤ pre-frail ≤ 0.25, frail >0.25) and mortality. RESULTS: There were 9 110 OSGM and 67 420 non-OSGM with sufficient data to calculate AoU-FI; 41% OSGM versus 50% non-OSGM were robust, whereas 34% versus 32% were pre-frail, and 26% versus 19% were frail. Mean AoU-FI was 0.19 (95% confidence interval [CI]: 0.187, 0.191) for OSGM and 0.168 (95% CI: 0.167, 0.169) for non-OSGM. Compared to robust, odds of mortality were higher among frail OSGM (odds ratio [OR] 6.40; 95% CI: 1.84, 22.23) and non-OSGM (OR 3.96; 95% CI: 2.96, 5.29). CONCLUSIONS: The AoU-FI identified a higher burden of frailty, increased risk of mortality, and an attenuated impact of age on frailty among OSGM compared to non-OSGM. Future work is needed to understand how frailty affects the OSGM population.

Reprint of "Using multiple imputation by super learning to assign intent to nonfatal firearm injuries".

The number of nonfatal firearm injuries in the US by intent (e.g., due to assault) is not reliably known: First, although the largest surveillance system for hospital-treated events, the Healthcare Cost and Utilization Project Nationwide Emergency Department Sample (HCUP-NEDS), provides accurate data for the number of nonfatal firearm injuries, injury intent is not coded reliably. Second, the system that reliably codes intent, the CDC's National Electronic Injury Surveillance System - Firearm Injury Surveillance Study (NEISS-FISS), while large enough to produce stable estimates of the distribution of intent, is too small to produce stable estimates of the number of these events. Third, a large proportion of cases in NEISS-FISS, notably in early years of the system, are coded as of "undetermined intent." Trends in the proportion of nonfatal firearm injuries by intent in NEISS-FISS thus depend on whether these cases are treated as a distinct category, or, instead, can be re-classified through imputation. We contrast the distributions of nonfatal firearm injury by intent generated using multiple imputation with those generated using complete-case analyses and analyses that consider "undetermined intent" as a distinct category. We produce estimates of the annual number of firearm injuries by intent in a two-step process. First, we impute intent for cases coded as "undetermined" using Multiple Imputation by Super Learning (MISL). Second, we apply MISL-derived distributions to aggregate count data from HCUP-NEDS. The proportion of non-fatal firearm assaults appears to increase over time when injuries coded as undetermined are included as a category. By contrast, the proportion of assaults remains relatively constant over time in complete-case and multiply imputed analyses. Differences between complete-case and multiple imputation approaches become apparent in subgroup analyses. Trends in the number of nonfatal firearm injuries by intent, 2006-2016, derived in our two-step process, are relatively flat. Multiple imputation strategies recovered intent distribution trends that differed from trends derived using methods that are not designed to account for the multiple complex relationships of missingness present in NEISS - FISS data. When applied to NEISS - FISS, MISL imputation produces plausible distributional estimates of firearm injury by intent.

A Machine Learning Approach to Identify Predictors of Severe COVID-19 Outcome in Patients With Rheumatoid Arthritis.

BACKGROUND: Rheumatoid arthritis (RA) patients have a lowered immune response to infection, potentially due to the use of corticosteroids and immunosuppressive drugs. Predictors of severe COVID-19 outcomes within the RA population have not yet been explored in a real-world setting. OBJECTIVES: To identify the most influential predictors of severe COVID-19 within the RA population. STUDY DESIGN: Retrospective cohort study. SETTING: Research was conducted using Optum's de-identified Clinformatics® Data Mart Database (2000-2021Q1), a US commercial claims database. METHODS: We identified adult patients with index COVID-19 (ICD-10-CM diagnosis code U07.1) between March 1, 2020, and December 31, 2020. Patients were required to have continuous enrollment and have evidence of one inpatient or 2 outpatient diagnoses of RA in the 365 days prior to index. RA patients with COVID-19 were stratified by outcome (mild vs severe), with severe cases defined as having one of the following within 60 days of COVID-19 diagnosis: death, treatment in the intensive care unit (ICU), or mechanical ventilation. Baseline demographics and clinical characteristics were extracted during the 365 days prior to index COVID-19 diagnosis. To control for improving treatment options, the month of index date was included as a potential independent variable in all models. Data were partitioned (80% train and 20% test), and a variety of machine learning algorithms (logistic regression, random forest, support vector machine [SVM], and XGBoost) were constructed to predict severe COVID-19, with model covariates ranked according to importance. RESULTS: Of 4,295 RA patients with COVID-19 included in the study, 990 (23.1%) were classified as severe. RA patients with severe COVID-19 had a higher mean age (mean [SD] = 71.6 [10.3] vs 63.4 [13.7] years, P < 0.001) and Charlson Comorbidity Index (CCI) (3.8 [2.4] vs 2.4 [1.8], P < 0.001) than those with mild cases. Males were more likely to be a severe case than mild (29.1% vs 18.5%, P < 0.001). The top 15 predictors from the best performing model (XGBoost, AUC = 75.64) were identified. While female gender, commercial insurance, and physical therapy were inversely associated with severe COVID-19 outcomes, top predictors included a March index date, older age, more inpatient visits at baseline, corticosteroid or gamma-aminobutyric acid analog (GABA) use at baseline or the need for durable medical equipment (i.e., wheelchairs), as well as comorbidities such as congestive heart failure, hypertension, fluid and electrolyte disorders, lower respiratory disease, chronic pulmonary disease, and diabetes with complication. LIMITATIONS: The cohort meeting our eligibility criteria is a relatively small sample in the context of machine learning. Additionally, diagnoses definitions rely solely on ICD-10-CM codes, and there may be unmeasured variables (such as labs and vitals) due to the nature of the data. These limitations were carefully considered when interpreting the results. CONCLUSIONS: Predictive baseline comorbidities and risk factors can be leveraged for early detection of RA patients at risk of severe COVID-19 outcomes. Further research should be conducted on modifiable factors in the RA population, such as physical therapy.

Using multiple imputation by super learning to assign intent to nonfatal firearm injuries.

The number of nonfatal firearm injuries in the US by intent (e.g., due to assault) is not reliably known: First, although the largest surveillance system for hospital-treated events, the Healthcare Cost and Utilization Project Nationwide Emergency Department Sample (HCUP-NEDS), provides accurate data for the number of nonfatal firearm injuries, injury intent is not coded reliably. Second, the system that reliably codes intent, the CDC's National Electronic Injury Surveillance System - Firearm Injury Surveillance Study (NEISS-FISS), while large enough to produce stable estimates of the distribution of intent, is too small to produce stable estimates of the number of these events. Third, a large proportion of cases in NEISS-FISS, notably in early years of the system, are coded as of "undetermined intent." Trends in the proportion of nonfatal firearm injuries by intent in NEISS-FISS thus depend on whether these cases are treated as a distinct category, or, instead, can be re-classified through imputation. We contrast the distributions of nonfatal firearm injury by intent generated using multiple imputation with those generated using complete-case analyses and analyses that consider "undetermined intent" as a distinct category. We produce estimates of the annual number of firearm injuries by intent in a two-step process. First, we impute intent for cases coded as "undetermined" using Multiple Imputation by Super Learning (MISL). Second, we apply MISL-derived distributions to aggregate count data from HCUP-NEDS. The proportion of non-fatal firearm assaults appears to increase over time when injuries coded as undetermined are included as a category. By contrast, the proportion of assaults remains relatively constant over time in complete-case and multiply imputed analyses. Differences between complete-case and multiple imputation approaches become apparent in subgroup analyses. Trends in the number of nonfatal firearm injuries by intent, 2006-2016, derived in our two-step process, are relatively flat. Multiple imputation strategies recovered intent distribution trends that differed from trends derived using methods that are not designed to account for the multiple complex relationships of missingness present in NEISS - FISS data. When applied to NEISS - FISS, MISL imputation produces plausible distributional estimates of firearm injury by intent.

MISL: Multiple imputation by super learning.

Multiple imputation techniques are commonly used when data are missing, however, there are many options one can consider. Multivariate imputation by chained equations is a popular method for generating imputations but relies on specifying models when imputing missing values. In this work, we introduce multiple imputation by super learning, an update to the multivariate imputation by chained equations method to generate imputations with ensemble learning. Ensemble methodologies have recently gained attention for use in inference and prediction as they optimally combine a variety of user-specified parametric and non-parametric models and perform well when estimating complex functions, including those with interaction terms. Through two simulations we compare inferences made using the multiple imputation by super learning approach to those made with other commonly used multiple imputation methods and demonstrate multiple imputation by super learning as a superior option when considering characteristics such as bias, confidence interval coverage rate, and confidence interval width.

Categorizing Comorbid Risk for People Living With HIV: A Latent Profile Analysis.

BACKGROUND: Categorizing clinical risk amidst heterogeneous multimorbidity in older people living with HIV/AIDS (PLWH) may help prioritize and optimize health care engagements. METHODS: PLWH and their prevalent conditions in 8 health domains diagnosed before January 1, 2015 were identified using 2014-2016 Medicare claims and the Chronic Conditions Data Warehouse. Latent profile analysis identified 4 distinct clinical subgroups based on the likelihood of conditions occurring together [G1: healthy, G2: substance use (SU), G3: pulmonary (PULM), G4: cardiovascular conditions (CV)]. Restricted mean survival time regression estimated the association of each subgroup with the 365 day mean event-free days until death, first hospitalization, and nursing home admission. Zero-inflated Poisson regression estimated hospitalization frequency in 2-year follow-up. RESULTS: Of 11,196 older PLWH, 71% were male, and the average age was 61 (SD 9.2) years. Compared with healthy group, SU group had a mean of 30 [95% confidence interval: (19.0 to 40.5)], PULM group had a mean of 28 (22.1 to 34.5), and CV group had a mean of 22 (15.0 to 22.0) fewer hospitalization-free days over 1 year. Compared with healthy group (2.8 deaths/100 person-years), CV group (8.4) had a mean of 4 (3.8 to 6.8) and PULM group (7.9) had a mean of 3 (0.7 to 5.5) fewer days alive; SU group (6.0) was not different. There was no difference in restricted mean survival time for nursing home admission. Compared with healthy group, SU group had 1.42-fold [95% confidence interval: (1.32 to 1.54)], PULM group had 1.71-fold (1.61 to 1.81), and CV group had 1.28-fold (1.20 to 1.37) higher rates of hospitalization. CONCLUSION: Identifying clinically distinct subgroups with latent profile analysis may be useful to identify targets for interventions and health care optimization in older PLWH.

Long-term nitrogen dioxide exposure and cause-specific mortality in the U.S. Medicare population.

BACKGROUND: Since 1971, the annual National Ambient Air Quality Standard (NAAQS) for nitrogen dioxide (NO2) has remained at 53 ppb, the impact of long-term NO2 exposure on mortality is poorly understood. OBJECTIVES: We examined associations between long-term NO2 exposure (12-month moving average of NO2) below the annual NAAQS and cause-specific mortality among the older adults in the U.S. METHODS: Cox proportional-hazard models were used to estimate Hazard Ratio (HR) for cause-specific mortality associated with long-term NO2 exposures among about 50 million Medicare beneficiaries living within the conterminous U.S. from 2001 to 2008. RESULTS: A 10 ppb increase in NO2 was associated with increased mortality from all-cause (HR: 1.06; 95% CI: 1.05-1.06), cardiovascular (HR: 1.10; 95% CI: 1.10-1.11), respiratory disease (HR: 1.09; 95% CI: 1.08-1.11), and cancer (HR: 1.01; 95% CI: 1.00-1.02) adjusting for age, sex, race, ZIP code as strata ZIP code- and state-level socio-economic status (SES) as covariates, and PM2.5 exposure using a 2-stage approach. NO2 was also associated with elevated mortality from ischemic heart disease, cerebrovascular disease, congestive heart failure, chronic obstructive pulmonary disease, pneumonia, and lung cancer. We found no evidence of a threshold, with positive and significant HRs across the range of NO2 exposures for all causes of death examined. Exposure-response curves were linear for all-cause, supra-linear for cardiovascular-, and sub-linear for respiratory-related mortality. HRs were highest consistently among Black beneficiaries. CONCLUSIONS: Long-term NO2 exposure is associated with elevated risks of death by multiple causes, without evidence of a threshold response. Our findings raise concerns about the sufficiency of the annual NAAQS for NO2.

The impact of Long-Term PM2.5 constituents and their sources on specific causes of death in a US Medicare cohort.

BACKGROUND: Our understanding of the impact of long-term exposures to PM2.5 constituents and sources on mortality is limited. OBJECTIVES: To examine associations between long-term exposures to PM2.5 constituents and sources and cause-specific mortality in US older adults. METHODS: We obtained demographic and mortality data for 15.4 million Medicare beneficiaries living within the conterminous United States (US) between 2000 and 2008. We assessed PM2.5 constituents exposures for each beneficiary and used factor analysis and residual-based methods to characterize PM2.5 sources and mixtures, respectively. In age-, sex-, race- and site- stratified Cox proportional hazard models adjusted for neighborhood socio-economic status (SES), we assessed associations of individual PM2.5 constituents, sources, and mixtures and cause-specific mortality and examined modification of these associations by participant demographics and location of residence. We assessed the robustness of our findings to additional adjustment for behavioral risk factors and to alternate exposure definitions and exposure windows. RESULTS: Hazard ratios (HR) were highest for all causes of death, except COPD, for PM2.5 constituents and the coal combustion-related PM2.5 components, with no evidence of confounding by behavioral covariates. We further found Pb and metal-related PM2.5 components to be significantly associated with increased HR of all causes of death, except COPD and lung cancer mortality, and nitrate (NO3-) and silicon (Si) and associated source-related PM2.5 components (traffic and soil, respectively) to be significantly associated with increased all-cause, CVD, respiratory and all cancer-related mortality HR. Associations for other examined constituents and mortality were inconsistent or largely null. Our analyses of mixtures were generally consistent with these findings. Mortality HRs were greatest for minority, especially Black, low-income urban, younger, and male beneficiaries. DISCUSSION: PM2.5 components related to coal combustion, traffic, and to a lesser extent, soil were strongly associated with mortality from CVD, respiratory disease, and cancer.

Associations Between Work-Related Factors and Psychological Distress Among Construction Workers.

OBJECTIVE: Identify work-related factors associated with the mental health and well-being of construction workers. METHODS: We completed eight key informant interviews, six worker focus groups, and a survey, informed by the interviews and focus groups, of 259 construction workers on five construction sites. Negative binomial regressions examined associations between psychological distress and work-related factors including safety climate, work-to-family conflict, psychological demands, social support, harassment, and job security. RESULTS: Three themes emerged from the interviews and focus groups, job demands and structure, social support and workplace relations, and job precarity. From the survey higher psychological demands, higher work-to-family conflict, lower supervisor support, higher discrimination, and higher likelihood of losing a job were associated with higher psychological distress. When combined into a single model job demands and work-to-family conflict remained significant. CONCLUSIONS: Work-related factors were associated with high levels of distress.

Preterm birth and PM2.5 in Puerto Rico: evidence from the PROTECT birth cohort.

BACKGROUND: Preterm birth (PTB, birth before 37 weeks of gestation) has been associated with adverse health outcomes across the lifespan. Evidence on the association between PTB and prenatal exposure to air pollutants is inconsistent, and is especially lacking for ethnic/racial minority populations. METHODS: We obtained data on maternal characteristics and behaviors and PTB and other birth outcomes for women participating in the Puerto Rico Testsite for Exploring Contamination Threats (PROTECT) cohort, who lived in municipalities located along the North Coast of Puerto Rico. We assessed pre-natal PM2.5 exposures for each infant based on the nearest US Environmental Protection Agency monitor. We estimated prenatal phthalate exposures as the geometric mean of urinary measurements obtained during pregnancy. We then examined the association between PM2.5 and PTB using modified Poisson regression and assessed modification of the association by phthalate exposure levels and sociodemographic factors such as maternal age and infant gender. RESULTS: Among 1092 singleton births, 9.1% of infants were born preterm and 92.9% of mothers had at least a high school education. Mothers had a mean (standard deviation) age of 26.9 (5.5) years and a median (range) of 2.0 (1.0-8.0) pregnancies. Nearly all women were Hispanic white, black, or mixed race. Median (range) prenatal PM2.5 concentrations were 6.0 (3.1-19.8) µ g/m3. Median (interquartile range) prenatal phthalate levels were 14.9 (8.9-26.0) and 14.5 (8.4-26.0), respectively, for di-n-butyl phthalate (DBP) and di-isobutyl phthalate (DiBP). An interquartile range increase in PM2.5 was associated with a 1.2% (95% CI 0.4, 2.1%) higher risk of PTB. There was little difference in PTB risk in strata of infant sex, mother's age, family income, history of adverse birth outcome, parity, and pre-pregnancy body mass index. Pregnancy urinary phthalate metabolite levels did not modify the PM2.5-PTB association. CONCLUSION: Among ethnic minority women in Puerto Rico, prenatal PM2.5 exposure is associated with a small but significant increase in risk of PTB.

Measuring Criterion Validity of Microinteraction Ecological Momentary Assessment (Micro-EMA): Exploratory Pilot Study With Physical Activity Measurement.

BACKGROUND: Ecological momentary assessment (EMA) is an in situ method of gathering self-report on behaviors using mobile devices. In typical phone-based EMAs, participants are prompted repeatedly with multiple-choice questions, often causing participation burden. Alternatively, microinteraction EMA (micro-EMA or µEMA) is a type of EMA where all the self-report prompts are single-question surveys that can be answered using a 1-tap glanceable microinteraction conveniently on a smartwatch. Prior work suggests that µEMA may permit a substantially higher prompting rate than EMA, yielding higher response rates and lower participation burden. This is achieved by ensuring µEMA prompt questions are quick and cognitively simple to answer. However, the validity of participant responses from µEMA self-report has not yet been formally assessed. OBJECTIVE: In this pilot study, we explored the criterion validity of µEMA self-report on a smartwatch, using physical activity (PA) assessment as an example behavior of interest. METHODS: A total of 17 participants answered 72 µEMA prompts each day for 1 week using a custom-built µEMA smartwatch app. At each prompt, they self-reported whether they were doing sedentary, light/standing, moderate/walking, or vigorous activities by tapping on the smartwatch screen. Responses were compared with a research-grade activity monitor worn on the dominant ankle simultaneously (and continuously) measuring PA. RESULTS: Participants had an 87.01% (5226/6006) µEMA completion rate and a 74.00% (5226/7062) compliance rate taking an average of only 5.4 (SD 1.5) seconds to answer a prompt. When comparing µEMA responses with the activity monitor, we observed significantly higher (P<.001) momentary PA levels on the activity monitor when participants self-reported engaging in moderate+vigorous activities compared with sedentary or light/standing activities. The same comparison did not yield any significant differences in momentary PA levels as recorded by the activity monitor when the µEMA responses were randomly generated (ie, simulating careless taps on the smartwatch). CONCLUSIONS: For PA measurement, high-frequency µEMA self-report could be used to capture information that appears consistent with that of a research-grade continuous sensor for sedentary, light, and moderate+vigorous activity, suggesting criterion validity. The preliminary results show that participants were not carelessly answering µEMA prompts by randomly tapping on the smartwatch but were reporting their true behavior at that moment. However, more research is needed to examine the criterion validity of µEMA when measuring vigorous activities.

Associations of gestational phthalate exposure and non-nutritive suck among infants from the Puerto Rico Testsite for Exploring Contamination Threats (PROTECT) birth cohort study.

BACKGROUND: Infant non-nutritive suck (NNS), or sucking on a pacifier with no nutrients being delivered, has been used as in index of brain function and has been linked to subsequent neurodevelopment. Yet, no data are available connecting NNS to environmental exposures in utero. The goal of this study was to examine the relationship between gestational exposure to phthalates (a group of chemicals found in personal care products, PVC plastics, and other products) and NNS among infants in a birth cohort study in Puerto Rico. METHODS: Urinary phthalate metabolite levels were measured in women at up to three time points in pregnancy as a measure of in utero exposure to the child. We calculated the geometric mean of each metabolite for each woman as a measure of exposure across gestation. Infants had their NNS sampled using our custom research pacifier between 4-6 (± 2 weeks) weeks of age, yielding the following NNS dependent measures: cycles/burst, frequency, amplitude, bursts/min, and cycles/min. RESULTS: Two hundred and eight mother-infant dyads completed this study We used multiple linear regression to assess associations between individual phthalate metabolites and NNS measurements, adjusting for infant sex, birthweight, and urinary specific gravity. An interquartile range (IQR) increase in mono carboxyisononyl phthalate across pregnancy was associated with 3.5% (95%CI: -6.2, -0.8%) lower NNS frequency and 8.9% (0.6, 17.3%) higher NNS amplitude. Similarly, an IQR increase in mono-2-ethylhexyl phthalate was also associated with 3.4% (-6.5, -0.2%) lower NNS frequency, while an IQR increase in di-2-ethylhexyl terephthalate metabolites was associated with 11.2% (2.9, 19.5%) higher NNS amplitude. Gestational exposure to phthalates may alter NNS amplitude and frequency in full-term infants. These findings indicate that the infants may be increasing their NNS amplitude to compensate for their slower NNS frequency. These preliminary findings could have important clinical implications for earlier detection of exposure-related deficits in neurofunction as well as implications for subsequent neurodevelopment and related interventions.

A Secure and Reusable Software Architecture for Supporting Online Data Harmonization.

Retrospective data harmonization across multiple research cohorts and studies is frequently done to increase statistical power, provide comparison analysis, and create a richer data source for data mining. However, when combining disparate data sources, harmonization projects face data management and analysis challenges. These include differences in the data dictionaries and variable definitions, privacy concerns surrounding health data representing sensitive populations, and lack of properly defined data models. With the availability of mature open-source web-based database technologies, developing a complete software architecture to overcome the challenges associated with the harmonization process can alleviate many roadblocks. By leveraging state-of-the-art software engineering and database principles, we can ensure data quality and enable cross-center online access and collaboration. This paper outlines a complete software architecture developed and customized using the Django web framework, leveraged to harmonize sensitive data collected from three NIH-support birth cohorts. We describe our framework and show how we successfully overcame challenges faced when harmonizing data from these cohorts. We discuss our efforts in data cleaning, data sharing, data transformation, data visualization, and analytics, while reflecting on what we have learned to date from these harmonized datasets.

Comparison of censoring assumptions to reduce bias in tuberculosis treatment cohort analyses.

OBJECTIVE: Observational tuberculosis cohort studies are often limited by a lack of long-term data characterizing survival beyond the initial treatment outcome. Though Cox proportional hazards models are often applied to these data, differential risk of long-term survival, dependent on the initial treatment outcome, can lead to violations of model assumptions. We evaluate the performance of two alternate censoring approaches on reducing bias in treatment effect estimates. DESIGN: We simulate a typical multidrug-resistant tuberculosis cohort study and use Cox proportional hazards models to assess the relationship of an aggressive treatment regimen with hazard of death. We compare three assumptions regarding censored observations to determine which produces least biased treatment effect estimates: conventional non-informative censoring, an extension of short-term survival informed by literature, and incorporation of predicted long-term vital status. RESULTS: The treatment regimen's protective effect on death is consistently underestimated by the conventional censoring method, up to 7.6%. Models using the two alternative censoring techniques produce treatment effect estimates consistently stronger and less biased than the conventional method, underestimating the treatment effect by less than 2.4% across all scenarios. CONCLUSIONS: When model assumptions are violated, alternative censoring techniques can more accurately estimate associations between treatment and long-term survival. In multidrug-resistant tuberculosis cohort analyses, this bias reduction may yield more accurate and, larger effect estimates. This bias reduction can be achieved through use of standard statistical procedures with a simple re-coding of the censoring indicator.

Cohort profile: Center for Research on Early Childhood Exposure and Development in Puerto Rico.

PURPOSE: Puerto Rican children experience high rates of asthma and obesity. Further, infants born in Puerto Rico are more at risk for being born prematurely compared with infants on the mainland USA. Environmental exposures from multiple sources during critical periods of child development, potentially modified by psychosocial factors, may contribute to these adverse health outcomes. To date, most studies investigating the health effects of environmental factors on infant and child health have focused on single or individual exposures. PARTICIPANTS: Infants currently in gestation whose mother is enrolled in Puerto Rico Testsite for Exploring Contamination Threats (PROTECT) cohort, and infants and children already born to mothers who participated in the PROTECT study. FINDINGS TO DATE: Data collection and processing remains ongoing. Demographic data have been collected on 437 mother-child pairs. Birth outcomes are available for 420 infants, neurodevelopmental outcomes have been collected on 319 children. Concentrations of parabens and phenols in maternal spot urine samples have been measured from 386 mothers. Center for Research on Early Childhood Exposure and Development mothers have significantly higher urinary concentrations of dichlorophenols, triclosan and triclocarban, but lower levels of several parabens compared with reference values from a similar population drawn from the National Health and Nutrition Examination Survey. FUTURE PLANS: Data will continue to be collected through recruitment of new births with a target of 600 children. Seven scheduled follow-up visits with existing and new participants are planned. Further, our research team continues to work with healthcare providers, paediatricians and early intervention providers to support parent's ability to access early intervention services for participants.

Designing a Participatory Total Worker Health® Organizational Intervention for Commercial Construction Subcontractors to Improve Worker Safety, Health, and Well-Being: The "ARM for Subs" Trial.

Background: Evidence supports organizational interventions as being effective for improving worker safety, health and well-being; however, there is a paucity of evidence-based interventions for subcontracting companies in commercial construction. Methods: A theory-driven approach supplemented by formative research through key stakeholder interviews and focus groups and an iterative vetting process with stakeholders, resulted in the development of an intervention for subcontractors in the commercial construction industry. We piloted the intervention in one subcontracting commercial construction company. We used these findings to adapt and finalize the intervention design to be tested in a future large-scale trial. Results: There were several key findings from the formative research, including challenges faced by companies and assets that should be considered in the intervention design. This resulted in a communication infrastructure company-based, continual improvement, participatory intervention design, consisting of a needs assessment and report, committee-led prioritization, action planning and implementation, and worker communication/feedback cycle. The pilot contributed to the final intervention design with modifications made with respect to timing, implementation support, capacity building, adaptability and sustainability. Conclusions: The use of a theory-driven participatory approach to developing an integrated organizational intervention for commercial construction subcontracting companies was important and necessary. It allowed us to consider the empirical evidence and relevant theories and tailor these to meet the needs of our target population. This study gives pragmatic insight into the early development of a complex intervention, with practical experience of how we adapted our intervention at each stage. This intervention will be tested in a future randomized trial.

Associations between a safety prequalification survey and worker safety experiences on commercial construction sites.

BACKGROUND: While assessment of subcontractors' safety performance during project bidding processes are common in commercial construction, the validation of organizational surveys used in these processes is largely absent. METHODS: As part of a larger research project called Assessment of Contractor Safety (ACES), we designed and tested through a cross-sectional study, a 63-item organizational survey assessing subcontractors' leading indicators of safety performance. We administered the ACES Survey to 43 subcontractors on 24 construction sites. Concurrently, we captured the safety climate of 1426 workers on these sites through worker surveys, as well as injury rates, for the duration of the project. RESULTS: At the worksite level, higher average ACES scores were associated with higher worker safety climate scores (P < .01) and lower rates of injury involving days away (P < .001). Within subcontracting companies, no associations were observed between ACES and worker safety climate scores and injuries. CONCLUSIONS: These results suggest the overall and collective importance of the construction project and its worksite in mediating worker experiences, perhaps somewhat independent of the individual subcontractor level.

The impact of long-term PM2.5 exposure on specific causes of death: exposure-response curves and effect modification among 53 million U.S. Medicare beneficiaries.

BACKGROUND: The shape of the exposure-response curve for long-term ambient fine particulate (PM2.5) exposure and cause-specific mortality is poorly understood, especially for rural populations and underrepresented minorities. METHODS: We used hybrid machine learning and Cox proportional hazard models to assess the association of long-term PM2.5 exposures on specific causes of death for 53 million U.S. Medicare beneficiaries (aged ≥65) from 2000 to 2008. Models included strata for age, sex, race, and ZIP code and controlled for neighborhood socio-economic status (SES) in our main analyses, with approximately 4 billion person-months of follow-up, and additionally for warm season average of 1-h daily maximum ozone exposures in a sensitivity analysis. The impact of non-traffic PM2.5 on mortality was examined using two stage models of PM2.5 and nitrogen dioxide (NO2). RESULTS: A 10 µg /m3 increase in 12-month average PM2.5 prior to death was associated with a 5% increase in all-cause mortality, as well as an 8.8, 5.6, and 2.5% increase in all cardiovascular disease (CVD)-, all respiratory-, and all cancer deaths, respectively, in age, gender, race, ZIP code, and SES-adjusted models. PM2.5 exposures, however, were not associated with lung cancer mortality. Results were not sensitive to control for ozone exposures. PM2.5-mortality associations for CVD- and respiratory-related causes were positive and significant for beneficiaries irrespective of their sex, race, age, SES and urbanicity, with no evidence of a lower threshold for response or of lower Risk Ratios (RRs) at low PM2.5 levels. Associations between PM2.5 and CVD and respiratory mortality were linear and were higher for younger, Black and urban beneficiaries, but were largely similar by SES. Risks associated with non-traffic PM2.5 were lower than that for all PM2.5 and were null for respiratory and lung cancer-related deaths. CONCLUSIONS: PM2.5 was associated with mortality from CVD, respiratory, and all cancer, but not lung cancer. PM2.5-associated risks of CVD and respiratory mortality were similar across PM2.5 levels, with no evidence of a threshold. Blacks, urban, and younger beneficiaries were most vulnerable to the long-term impacts of PM2.5 on mortality.

Long-term ozone exposures and cause-specific mortality in a US Medicare cohort.

We examined the association of long-term, daily 1-h maximum O3 (ozone) exposures on cause-specific mortality for 22.2 million US Medicare beneficiaries between 2000-2008. We modeled the association between O3 and mortality using age-gender-race stratified log-linear regression models, adjusted for state of residence. We examined confounding by (1) adjusting for PM2.5 (particles with aerodynamic diameters <2.5 µm) and NO2 (nitrogen dioxide) exposures, temperature, and neighborhood-level characteristics and behaviors, and (2) decomposing O3 into its temporal and spatio-temporal components and comparing estimated risk ratios. We also examined sensitivity of our results to alternate exposure measures based on warm-season 8-h daily maximum and 24-h average exposures. We found increased risks from long-term O3 exposures to be strongest and most consistent for mortality from respiratory disease (1.030, 95% CI: 1.027, 1.034) (including COPD (chronic obstructive pulmonary disease)), CHF (congestive heart failure), and lung cancer (1.015, 95% CI: 1.010, 1.020), with no evidence of confounding by PM2.5, NO2, and temperature and with results similar across O3 exposure measures. While significant, associations between long-term O3 exposures and CVD (cardiovascular)-related mortality (1.005, 95% CI: 1.003, 1.007) were confounded by PM2.5 and varied with the exposure measure, with associations no longer significantly positive when warm-season 8-h maximum or 24-h average O3 was used to assess exposures. In this large study, we provide strong evidence that O3 exposure is associated with mortality from respiratory-related causes and for the first-time, lung cancer, but raise questions regarding O3-related impacts on CVD mortality. Our findings demonstrate the need to further identify potential confounders.