Improving prediction of linear regression models by integrating external information from heterogeneous populations: James-Stein estimators.

We consider the setting where (1) an internal study builds a linear regression model for prediction based on individual-level data, (2) some external studies have fitted similar linear regression models that use only subsets of the covariates and provide coefficient estimates for the reduced models without individual-level data, and (3) there is heterogeneity across these study populations. The goal is to integrate the external model summary information into fitting the internal model to improve prediction accuracy. We adapt the James-Stein shrinkage method to propose estimators that are no worse and are oftentimes better in the prediction mean squared error after information integration, regardless of the degree of study population heterogeneity. We conduct comprehensive simulation studies to investigate the numerical performance of the proposed estimators. We also apply the method to enhance a prediction model for patella bone lead level in terms of blood lead level and other covariates by integrating summary information from published literature.

Multiple metal exposures associate with higher amyotrophic lateral sclerosis risk and mortality independent of genetic risk and correlate to self-reported exposures: a case-control study.

BACKGROUND: The pathogenesis of amyotrophic lateral sclerosis (ALS) involves both genetic and environmental factors. This study investigates associations between metal measures in plasma and urine, ALS risk and survival and exposure sources. METHODS: Participants with and without ALS from Michigan provided plasma and urine samples for metal measurement via inductively coupled plasma mass spectrometry. ORs and HRs for each metal were computed using risk and survival models. Environmental risk scores (ERS) were created to evaluate the association between exposure mixtures and ALS risk and survival and exposure source. ALS (ALS-PGS) and metal (metal-PGS) polygenic risk scores were constructed from an independent genome-wide association study and relevant literature-selected single-nucleotide polymorphisms. RESULTS: Plasma and urine samples from 454 ALS and 294 control participants were analysed. Elevated levels of individual metals, including copper, selenium and zinc, significantly associated with ALS risk and survival. ERS representing metal mixtures strongly associated with ALS risk (plasma, OR=2.95, CI=2.38-3.62, p<0.001; urine, OR=3.10, CI=2.43-3.97, p<0.001) and poorer ALS survival (plasma, HR=1.37, CI=1.20-1.58, p<0.001; urine, HR=1.44, CI=1.23-1.67, p<0.001). Addition of the ALS-PGS or metal-PGS did not alter the significance of metals with ALS risk and survival. Occupations with high potential of metal exposure associated with elevated ERS. Additionally, occupational and non-occupational metal exposures were associated with measured plasma and urine metals. CONCLUSION: Metals in plasma and urine associated with increased ALS risk and reduced survival, independent of genetic risk, and correlated with occupational and non-occupational metal exposures. These data underscore the significance of metal exposure in ALS risk and progression.

PATIENT RECRUITMENT USING ELECTRONIC HEALTH RECORDS UNDER SELECTION BIAS: A TWO-PHASE SAMPLING FRAMEWORK.

Electronic health records (EHRs) are increasingly recognized as a cost-effective resource for patient recruitment in clinical research. However, how to optimally select a cohort from millions of individuals to answer a scientific question of interest remains unclear. Consider a study to estimate the mean or mean difference of an expensive outcome. Inexpensive auxiliary covariates predictive of the outcome may often be available in patients' health records, presenting an opportunity to recruit patients selectively, which may improve efficiency in downstream analyses. In this paper we propose a two-phase sampling design that leverages available information on auxiliary covariates in EHR data. A key challenge in using EHR data for multiphase sampling is the potential selection bias, because EHR data are not necessarily representative of the target population. Extending existing literature on two-phase sampling design, we derive an optimal two-phase sampling method that improves efficiency over random sampling while accounting for the potential selection bias in EHR data. We demonstrate the efficiency gain from our sampling design via simulation studies and an application evaluating the prevalence of hypertension among U.S. adults leveraging data from the Michigan Genomics Initiative, a longitudinal biorepository in Michigan Medicine.

Decomposing the association of psychosocial wellbeing with all-cause mortality: the mediating role of physical health and lifestyle factors.

PURPOSE: This study quantified the effect of 48 psychosocial constructs on all-cause mortality using data from 7,698 individuals in the U.S. Health and Retirement Study. METHODS: Latent class analysis was used to divide participants into mutually exclusive psychosocial wellbeing groups (good, average, or poor) which was subsequently considered as the exposure. Mediation analysis was then conducted to determine the direct effect of the psychosocial wellbeing groups and the indirect (mediating) effects of physical health (functional status and comorbid conditions) and lifestyle factors (physical activity, smoking, and alcohol consumption) on overall survival. We also created a composite health index measure representing the summative effect of the mediators. RESULTS: We observed a strong and statistically significant total effect (TE) between survival time and psychosocial wellbeing group (survival time ratio (SR) = 1.73, 95% confidence interval (CI):1.50,2.01 when comparing good to poor). Mediation analysis revealed that the direct effect via psychosocial wellbeing group accounted for more than half of the TE (SR = 1.46, 95% CI:1.27,1.67). The composite health index measure mediated 36.2% of the TE with the natural indirect effect SR of 1.18 (95% CI:1.13,1.22). CONCLUSION: Our findings demonstrate the interconnectedness between psychosocial wellbeing and physical health and lifestyle factors on survival.

The mediating role of systemic inflammation and moderating role of racialization in disparities in incident dementia.

BACKGROUND: Exposure to systemic racism is linked to increased dementia burden. To assess systemic inflammation as a potential pathway linking exposure to racism and dementia disparities, we investigated the mediating role of C-reactive protein (CRP), a systemic inflammation marker, and the moderating role of the racialization process in incident dementia. METHODS: In the US Health and Retirement Study (n = 6,908), serum CRP was measured at baseline (2006, 2008 waves). Incident dementia was classified by cognitive tests over a six-year follow-up. Self-reported racialized categories were a proxy for exposure to the racialization process. We decomposed racialized disparities in dementia incidence (non-Hispanic Black and/or Hispanic vs. non-Hispanic white) into 1) the mediated effect of CRP, 2) the moderated portion attributable to the interaction between racialized group membership and CRP, and 3) the controlled direct effect (other pathways through which racism operates). RESULTS: The 6-year cumulative incidence of dementia is 12%. Among minoritized participants (i.e., non-Hispanic Black and/or Hispanic), high CRP levels ( ≥ 75th percentile or 4.73µg/mL) are associated with 1.26 (95%CI: 0.98, 1.62) times greater risk of incident dementia than low CRP ( < 4.73µg/mL). Decomposition analysis comparing minoritized versus non-Hispanic white participants shows that the mediating effect of CRP accounts for 3% (95% CI: 0%, 6%) of the racial disparity, while the interaction effect between minoritized group status and high CRP accounts for 14% (95% CI: 1%, 27%) of the disparity. Findings are robust to potential violations of causal mediation assumptions. CONCLUSIONS: Minoritized group membership modifies the relationship between systemic inflammation and incident dementia.

Higher levels of inflammation in blood are linked to greater dementia risk in older adults. Non-Hispanic Black and Hispanic Americans have higher inflammation levels compared to non-Hispanic white Americans. We conducted a study to examine whether high levels of inflammation could explain differences in dementia risk among these racial groups. We found that differences in inflammation levels in non-Hispanic Black or Hispanic adults modestly explain their higher risk of dementia compared to non-Hispanic white adults. These findings suggest that interventions aimed at reducing high levels of inflammation in minoritized US adults could ameliorate racial differences in dementia risk.

Text Messages to Promote Physical Activity in Patients With Cardiovascular Disease: A Micro-Randomized Trial of a Just-In-Time Adaptive Intervention.

BACKGROUND: Text messages may enhance physical activity levels in patients with cardiovascular disease, including those enrolled in cardiac rehabilitation. However, the independent and long-term effects of text messages remain uncertain. METHODS: The VALENTINE study (Virtual Application-supported Environment to Increase Exercise) was a micro-randomized trial that delivered text messages through a smartwatch (Apple Watch or Fitbit Versa) to participants initiating cardiac rehabilitation. Participants were randomized 4× per day over 6-months to receive no text message or a message encouraging low-level physical activity. Text messages were tailored on contextual factors (eg, weather). Our primary outcome was step count 60 minutes following a text message, and we used a centered and weighted least squares mean method to estimate causal effects. Given potential measurement differences between devices determined a priori, data were assessed separately for Apple Watch and Fitbit Versa users over 3 time periods corresponding to the initiation (0-30 days), maintenance (31-120 days), and completion (121-182 days) of cardiac rehabilitation. RESULTS: One hundred eight participants were included with 70 552 randomizations over 6 months; mean age was 59.5 (SD, 10.7) years with 36 (32.4%) female and 68 (63.0%) Apple Watch participants. For Apple Watch participants, text messages led to a trend in increased step count by 10% in the 60-minutes following a message during days 1 to 30 (95% CI, -1% to +20%), with no effect from days 31 to 120 (+1% [95% CI, -4% to +5%]), and a significant 6% increase during days 121 to 182 (95% CI, +0% to +11%). For Fitbit users, text messages significantly increased step count by 17% (95% CI, +7% to +28%) in the 60-minutes following a message in the first 30 days of the study with no effect subsequently. CONCLUSIONS: In patients undergoing cardiac rehabilitation, contextually tailored text messages may increase physical activity, but this effect varies over time and by device. REGISTRATION: URL: https://www.clinicaltrials.gov; Unique identifier: NCT04587882.

Correction: The importance of investing in data, models, experiments, team science, and public trust to help policymakers prepare for the next pandemic.

[This corrects the article DOI: 10.1371/journal.pgph.0002601.].

To weight or not to weight? The effect of selection bias in 3 large electronic health record-linked biobanks and recommendations for practice.

OBJECTIVES: To develop recommendations regarding the use of weights to reduce selection bias for commonly performed analyses using electronic health record (EHR)-linked biobank data. MATERIALS AND METHODS: We mapped diagnosis (ICD code) data to standardized phecodes from 3 EHR-linked biobanks with varying recruitment strategies: All of Us (AOU; n = 244 071), Michigan Genomics Initiative (MGI; n = 81 243), and UK Biobank (UKB; n = 401 167). Using 2019 National Health Interview Survey data, we constructed selection weights for AOU and MGI to represent the US adult population more. We used weights previously developed for UKB to represent the UKB-eligible population. We conducted 4 common analyses comparing unweighted and weighted results. RESULTS: For AOU and MGI, estimated phecode prevalences decreased after weighting (weighted-unweighted median phecode prevalence ratio [MPR]: 0.82 and 0.61), while UKB estimates increased (MPR: 1.06). Weighting minimally impacted latent phenome dimensionality estimation. Comparing weighted versus unweighted phenome-wide association study for colorectal cancer, the strongest associations remained unaltered, with considerable overlap in significant hits. Weighting affected the estimated log-odds ratio for sex and colorectal cancer to align more closely with national registry-based estimates. DISCUSSION: Weighting had a limited impact on dimensionality estimation and large-scale hypothesis testing but impacted prevalence and association estimation. When interested in estimating effect size, specific signals from untargeted association analyses should be followed up by weighted analysis. CONCLUSION: EHR-linked biobanks should report recruitment and selection mechanisms and provide selection weights with defined target populations. Researchers should consider their intended estimands, specify source and target populations, and weight EHR-linked biobank analyses accordingly.

Associations of urinary polycyclic aromatic hydrocarbon (PAH) metabolites and their mixture with thyroid hormone concentration during pregnancy in the LIFECODES cohort: A repeated measures study.

BACKGROUND: Polycyclic aromatic hydrocarbons (PAHs) are endocrine disruptors resulting from incomplete combustion. Pregnancy represents a particularly vulnerable period to such exposures, given the significant influence of hormone physiology on fetal growth and pregnancy outcomes. Maternal thyroid hormones play crucial roles in fetal development and pregnancy outcomes. However, limited studies have examined gestational PAH exposure and maternal thyroid hormones during pregnancy. METHODS: Our study included 439 women enrolled in the LIFECODES birth cohort in Boston, aiming to explore the relationship between urinary PAH metabolites and thyroid hormones throughout pregnancy. Urine samples for PAH metabolite analysis and plasma samples for thyroid hormone were measured up to four visits throughout gestation. Single pollutant analyses employed linear mixed effect models to investigate individual associations between each PAH metabolite and thyroid hormone concentration. Sensitivity analyses were conducted to assess potential susceptibility windows and fetal-sex-specific effects of PAH exposure. Mixture analyses utilized quantile g-computation to evaluate the collective impact of eight PAH metabolites on thyroid hormone concentrations. Additionally, Bayesian kernel machine regression (BKMR) was employed to explore potential non-linear associations and interactions between PAH metabolites. Subject-specific random intercepts were incorporated to address intra-individual correlation of serial measurements over time in both single pollutant and mixture analyses. RESULTS: Our findings revealed positive trends in associations between PAH metabolites and thyroid hormones, both individually and collectively as a mixture. Sensitivity analyses indicated that these associations were influenced by the study visit and fetal sex. Mixture analyses suggested non-linear relationships and interactions between different PAH exposures. CONCLUSIONS: This comprehensive investigation underscores the critical importance of understanding the impact of PAH exposures on thyroid hormone physiology during pregnancy. The findings highlight the intricate interplay between environmental pollutants and human pregnancy physiology, emphasizing the need for targeted interventions and public health policies to mitigate adverse outcomes associated with prenatal PAH exposure.

Exploring the Big Data Paradox for various estimands using vaccination data from the global COVID-19 Trends and Impact Survey (CTIS).

Selection bias poses a substantial challenge to valid statistical inference in nonprobability samples. This study compared estimates of the first-dose COVID-19 vaccination rates among Indian adults in 2021 from a large nonprobability sample, the COVID-19 Trends and Impact Survey (CTIS), and a small probability survey, the Center for Voting Options and Trends in Election Research (CVoter), against national benchmark data from the COVID Vaccine Intelligence Network. Notably, CTIS exhibits a larger estimation error on average (0.37) compared to CVoter (0.14). Additionally, we explored the accuracy (regarding mean squared error) of CTIS in estimating successive differences (over time) and subgroup differences (for females versus males) in mean vaccine uptakes. Compared to the overall vaccination rates, targeting these alternative estimands comparing differences or relative differences in two means increased the effective sample size. These results suggest that the Big Data Paradox can manifest in countries beyond the United States and may not apply equally to every estimand of interest.

Toward Realizing the Promise of AI in Precision Health Across the Spectrum of Care.

Significant progress has been made in augmenting clinical decision-making using artificial intelligence (AI) in the context of secondary and tertiary care at large academic medical centers. For such innovations to have an impact across the spectrum of care, additional challenges must be addressed, including inconsistent use of preventative care and gaps in chronic care management. The integration of additional data, including genomics and data from wearables, could prove critical in addressing these gaps, but technical, legal, and ethical challenges arise. On the technical side, approaches for integrating complex and messy data are needed. Data and design imperfections like selection bias, missing data, and confounding must be addressed. In terms of legal and ethical challenges, while AI has the potential to aid in leveraging patient data to make clinical care decisions, we also risk exacerbating existing disparities. Organizations implementing AI solutions must carefully consider how they can improve care for all and reduce inequities.

Cross-shift changes in pulmonary function and occupational exposure to particulate matter among e-waste workers in Ghana.

Introduction: Little is known on the association between cross-shift changes in pulmonary function and personal inhalation exposure to particulate matter (PM) among informal electronic-waste (e-waste) recovery workers who have substantial occupational exposure to airborne pollutants from burning e-waste. Methods: Using a cross-shift design, pre- and post-shift pulmonary function assessments and accompanying personal inhalation exposure to PM (sizes <1, <2.5 µm, and the coarse fraction, 2.5-10 µm in aerodynamic diameter) were measured among e-waste workers (n = 142) at the Agbogbloshie e-waste site and a comparison population (n = 65) in Accra, Ghana during 2017 and 2018. Linear mixed models estimated associations between percent changes in pulmonary function and personal PM. Results: Declines in forced expiratory volume in one second (FEV1) and forced vital capacity (FVC) per hour were not significantly associated with increases in PM (all sizes) among either study population, despite breathing zone concentrations of PM (all sizes) that exceeded health-based guidelines in both populations. E-waste workers who worked "yesterday" did, however, have larger cross-shift declines in FVC [-2.4% (95%CI: -4.04%, -0.81%)] in comparison to those who did not work "yesterday," suggesting a possible role of cumulative exposure. Discussion: Overall, short-term respiratory-related health effects related to PM exposure among e-waste workers were not seen in this sample. Selection bias due to the "healthy worker" effect, short shift duration, and inability to capture a true "pre-shift" pulmonary function test among workers who live at the worksite may explain results and suggest the need to adapt cross-shift studies for informal settings.

Prenatal per- and polyfluoroalkyl substances (PFAS) and maternal oxidative stress: Evidence from the LIFECODES study.

BACKGROUND: Per- and polyfluoroalkyl substances (PFAS) are synthetic chemicals linked to adverse pregnancy outcomes. Although their underlying biological mechanisms are not fully understood, evidence suggests PFAS may disrupt endocrine functions and contribute to oxidative stress (OS) and inflammation. OBJECTIVE: We examined associations between early pregnancy PFAS exposure and OS biomarkers, exploring potential effect modifications by fetal sex and maternal race. METHODS: We used data from 469 LIFECODES participants with measured plasma PFAS (median 10 weeks gestation) and repeated measures (median 10, 18, 26, and 35 weeks gestation) of urinary OS biomarkers [8-iso-prostaglandin-F2α (8-isoprostane) and 8-hydroxydeoxyguanosine (8-OHdG)]. Protein damage biomarkers (chlorotyrosine, dityrosine, and nitrotyrosine) were additionally measured in plasma from a subset (N = 167) during the third visit. Associations between each PFAS and OS biomarkers were examined using linear mixed-effects models and multivariable linear regressions, adjusting for potential confounders, including maternal age, race, education level, pre-pregnancy BMI, insurance status, and parity. Effect modifications were evaluated by including an interaction term between each PFAS and fetal sex or maternal race in the models. RESULTS: We observed significant positive associations between PFOS and 8-isoprostane, with a 9.68% increase in 8-isoprostane levels (95% CI: 0.10%, 20.18%) per interquartile range increase in PFOS. In contrast, PFUA was negatively associated [9.32% (95% CI: -17.68%, -0.11%)], while there were suggestive positive associations for MPAH and PFOA with 8-isoprostane. The associations of several PFAS with 8-OHdG varied by fetal sex, showing generally positive trends in women who delivered females, but negative or null in those who delivered males. No significant effect modification by maternal race was observed. CONCLUSIONS: This study provides evidence linking PFAS exposure to OS during pregnancy, with potential sex-specific effects of certain PFAS on 8-OHdG. Further research should explore additional OS/inflammatory biomarkers and assess the modifying effects of dietary and behavioral patterns across diverse populations.

Cross-Sectional Associations between Prenatal Per- and Poly-Fluoroalkyl Substances and Bioactive Lipids in Three Environmental Influences on Child Health Outcomes (ECHO) Cohorts.

Prenatal per- and poly-fluoroalkyl substances (PFAS) exposure may influence gestational outcomes through bioactive lipids─metabolic and inflammation pathway indicators. We estimated associations between prenatal PFAS exposure and bioactive lipids, measuring 12 serum PFAS and 50 plasma bioactive lipids in 414 pregnant women (median 17.4 weeks' gestation) from three Environmental influences on Child Health Outcomes Program cohorts. Pairwise association estimates across cohorts were obtained through linear mixed models and meta-analysis, adjusting the former for false discovery rates. Associations between the PFAS mixture and bioactive lipids were estimated using quantile g-computation. Pairwise analyses revealed bioactive lipid levels associated with PFDeA, PFNA, PFOA, and PFUdA (p < 0.05) across three enzymatic pathways (cyclooxygenase, cytochrome p450, lipoxygenase) in at least one combined cohort analysis, and PFOA and PFUdA (q < 0.2) in one linear mixed model. The strongest signature revealed doubling in PFOA corresponding with PGD2 (cyclooxygenase pathway; +24.3%, 95% CI: 7.3-43.9%) in the combined cohort. Mixture analysis revealed nine positive associations across all pathways with the PFAS mixture, the strongest signature indicating a quartile increase in the PFAS mixture associated with PGD2 (+34%, 95% CI: 8-66%), primarily driven by PFOS. Bioactive lipids emerged as prenatal PFAS exposure biomarkers, deepening insights into PFAS' influence on pregnancy outcomes.

Residential exposure associations with ALS risk, survival, and phenotype: a Michigan-based case-control study.

BACKGROUND: Environmental exposures impact amyotrophic lateral sclerosis (ALS) risk and progression, a fatal and progressive neurodegenerative disease. Better characterization of these exposures is needed to decrease disease burden. OBJECTIVE: To identify exposures in the residential setting that associate with ALS risk, survival, and onset segment. METHODS: ALS and control participants recruited from University of Michigan completed a survey that ascertained exposure risks in the residential setting. ALS risk was assessed using logistic regression models followed by latent profile analysis to consider exposure profiles. A case-only analysis considered the contribution of the residential exposure variables via a Cox proportional hazards model for survival outcomes and multinomial logistic regression for onset segment, a polytomous outcome. RESULTS: This study included 367 ALS and 255 control participants. Twelve residential variables were associated with ALS risk after correcting for multiple comparison testing, with storage in an attached garage of chemical products including gasoline or kerosene (odds ratio (OR) = 1.14, padjusted < 0.001), gasoline-powered equipment (OR = 1.16, padjusted < 0.001), and lawn care products (OR = 1.15, padjusted < 0.001) representing the top three risk factors sorted by padjusted. Latent profile analysis indicated that storage of these chemical products in both attached and detached garages increased ALS risk. Although residential variables were not associated with poorer ALS survival following multiple testing corrections, storing pesticides, lawn care products, and woodworking supplies in the home were associated with shorter ALS survival using nominal p values. No exposures were associated with ALS onset segment. CONCLUSION: Residential exposures may be important modifiable components of the ALS susceptibility and prognosis exposome.

Improving prediction models of amyotrophic lateral sclerosis (ALS) using polygenic, pre-existing conditions, and survey-based risk scores in the UK Biobank.

Background and Objectives: Amyotrophic lateral sclerosis (ALS) causes profound impairments in neurological function and a cure for this devastating disease remains elusive. Early detection and risk stratification are crucial for timely intervention and improving patient outcomes. This study aimed to identify predisposing genetic, phenotypic, and exposure-related factors for Amyotrophic lateral sclerosis using multi-modal data and assess their joint predictive potential. Methods: Utilizing data from the UK Biobank, we analyzed an unrelated set of 292 ALS cases and 408,831 controls of European descent. Two polygenic risk scores (PRS) are constructed: "GWAS Hits PRS" and "PRS-CS," reflecting oligogenic and polygenic ALS risk profiles, respectively. Time-restricted phenome-wide association studies (PheWAS) were performed to identify pre-existing conditions increasing ALS risk, integrated into phenotypic risk scores (PheRS). A poly-exposure score ("PXS") captures the influence of environmental exposures measured through survey questionnaires. We evaluate the performance of these scores for predicting ALS incidence and stratifying risk, adjusting for baseline demographic covariates. Results: Both PRSs modestly predicted ALS diagnosis, but with increased predictive power when combined (covariate-adjusted receiver operating characteristic [AAUC] = 0.584 [0.525, 0.639]). PheRS incorporated diagnoses 1 year before ALS onset (PheRS1) modestly discriminated cases from controls (AAUC = 0.515 [0.472, 0.564]). The "PXS" did not significantly predict ALS. However, a model incorporating PRSs and PheRS1 improved prediction of ALS (AAUC = 0.604 [0.547, 0.667]), outperforming a model combining all risk scores. This combined risk score identified the top 10% of risk score distribution with a 4-fold higher ALS risk (95% CI: [2.04, 7.73]) versus those in the 40%-60% range. Discussions: By leveraging UK Biobank data, our study uncovers predisposing ALS factors, highlighting the improved effectiveness of multi-factorial prediction models to identify individuals at highest risk for ALS.

Associations of maternal blood metal concentrations with plasma eicosanoids among pregnant women in Puerto Rico.

BACKGROUND/AIM: Heavy metals are known to induce oxidative stress and inflammation, and the association between metal exposure and adverse birth outcomes is well established. However, there lacks research on biomarker profiles linking metal exposures and adverse birth outcomes. Eicosanoids are lipid molecules that regulate inflammation in the body, and there is growing evidence that suggests associations between plasma eicosanoids and pregnancy outcomes. Eicosanoids may aid our understanding of etiologic birth pathways. Here, we assessed associations between maternal blood metal concentrations with eicosanoid profiles among 654 pregnant women in the Puerto Rico PROTECT birth cohort. METHODS: We measured concentrations of 11 metals in whole blood collected at median 18 and 26 weeks of pregnancy, and eicosanoid profiles measured in plasma collected at median 26 weeks. Multivariable linear models were used to regress eicosanoids on metals concentrations. Effect modification by infant sex was explored using interaction terms. RESULTS: A total of 55 eicosanoids were profiled. Notably, 12-oxoeicosatetraenoic acid (12-oxoETE) and 15-oxoeicosatetraenoic acid (15-oxoETE), both of which exert inflammatory activities, had the greatest number of significant associations with metal concentrations. These eicosanoids were associated with increased concentrations of Cu, Mn, and Zn, and decreased concentrations of Cd, Co, Ni, and Pb, with the strongest effect sizes observed for 12-oxoETE and Pb (ß:-33.5,95 %CI:-42.9,-22.6) and 15-oxoETE and Sn (ß:43.2,95 %CI:11.4,84.1). Also, we observed differences in metals-eicosanoid associations by infant sex. Particularly, Cs and Mn had the most infant sex-specific significant associations with eicosanoids, which were primarily driven by female fetuses. All significant sex-specific associations with Cs were inverse among females, while significant sex-specific associations with Mn among females were positive within the cyclooxygenase group but inverse among the lipoxygenase group. CONCLUSION: Certain metals were significantly associated with eicosanoids that are responsible for regulating inflammatory responses. Eicosanoid-metal associations may suggest a role for eicosanoids in mediating metal-induced adverse birth outcomes.

Metal mixtures associate with higher amyotrophic lateral sclerosis risk and mortality independent of genetic risk and correlate to self-reported exposures: a case-control study.

Background: The pathogenesis of amyotrophic lateral sclerosis (ALS) involves both genetic and environmental factors. This study investigates associations between metal measures in plasma and urine, ALS risk and survival, and exposure sources. Methods: Participants with and without ALS from Michigan provided plasma and urine samples for metal measurement via inductively coupled plasma mass spectrometry. Odds and hazard ratios for each metal were computed using risk and survival models. Environmental risk scores (ERS) were created to evaluate the association between exposure mixtures and ALS risk and survival and exposure source. ALS (ALS-PGS) and metal (metal-PGS) polygenic risk scores were constructed from an independent genome-wide association study and relevant literature-selected SNPs. Results: Plasma and urine samples from 454 ALS and 294 control participants were analyzed. Elevated levels of individual metals, including copper, selenium, and zinc, significantly associated with ALS risk and survival. ERS representing metal mixtures strongly associated with ALS risk (plasma, OR=2.95, CI=2.38-3.62, p<0.001; urine, OR=3.10, CI=2.43-3.97, p<0.001) and poorer ALS survival (plasma, HR=1.42, CI=1.24-1.63, p<0.001; urine, HR=1.52, CI=1.31-1.76, p<0.001). Addition of the ALS-PGS or metal-PGS did not alter the significance of metals with ALS risk and survival. Occupations with high potential of metal exposure associated with elevated ERS. Additionally, occupational and non-occupational metal exposures associated with measured plasma and urine metals. Conclusion: Metals in plasma and urine associated with increased ALS risk and reduced survival, independent of genetic risk, and correlated with occupational and non-occupational metal exposures. These data underscore the significance of metal exposure in ALS risk and progression.

Statistical methods for chemical mixtures: a roadmap for practitioners.

Quantitative characterization of the health impacts associated with exposure to chemical mixtures has received considerable attention in current environmental and epidemiological studies. With many existing statistical methods and emerging approaches, it is important for practitioners to understand when each method is best suited for their inferential goals. In this study, we conduct a review and comparison of 11 analytical methods available for use in mixtures research, through extensive simulation studies for continuous and binary outcomes. These methods fall in three different classes: identifying important components of a mixture, identifying interactions and creating a summary score for risk stratification and prediction. We carry out an illustrative data analysis in the PROTECT birth cohort from Puerto Rico. Most importantly we develop an integrated package "CompMix" that provides a platform for mixtures analysis where the practitioner can implement a pipeline for several types of mixtures analysis. Our simulation results suggest that the choice of methods depends on the goal of analysis and there is no clear winner across the board. For selection of important toxicants in the mixture and for identifying interactions, Elastic net by Zou et al. (Enet), Lasso for Hierarchical Interactions by Bien et al (HierNet), Selection of nonlinear interactions by a forward stepwise algorithm by Narisetty et al. (SNIF) have the most stable performance across simulation settings. Additionally, the predictive performance of the Super Learner ensembling method by Van de Laan et al. and HierNet are found to be superior to the rest of the methods. For overall summary or a cumulative measure, we find that using the Super Learner to combine multiple Environmental Risk Scores can lead to improved risk stratification properties. We have developed an R package "CompMix: A comprehensive toolkit for environmental mixtures analysis", allowing users to implement a variety of tasks under different settings and compare the findings. In summary, our study offers guidelines for selecting appropriate statistical methods for addressing specific scientific questions related to mixtures research. We identify critical gaps where new and better methods are needed.

To weight or not to weight? Studying the effect of selection bias in three large EHR-linked biobanks.

Objective: To explore the role of selection bias adjustment by weighting electronic health record (EHR)-linked biobank data for commonly performed analyses. Materials and methods: We mapped diagnosis (ICD code) data to standardized phecodes from three EHR-linked biobanks with varying recruitment strategies: All of Us (AOU; n=244,071), Michigan Genomics Initiative (MGI; n=81,243), and UK Biobank (UKB; n=401,167). Using 2019 National Health Interview Survey data, we constructed selection weights for AOU and MGI to be more representative of the US adult population. We used weights previously developed for UKB to represent the UKB-eligible population. We conducted four common descriptive and analytic tasks comparing unweighted and weighted results. Results: For AOU and MGI, estimated phecode prevalences decreased after weighting (weighted-unweighted median phecode prevalence ratio [MPR]: 0.82 and 0.61), while UKB's estimates increased (MPR: 1.06). Weighting minimally impacted latent phenome dimensionality estimation. Comparing weighted versus unweighted PheWAS for colorectal cancer, the strongest associations remained unaltered and there was large overlap in significant hits. Weighting affected the estimated log-odds ratio for sex and colorectal cancer to align more closely with national registry-based estimates. Discussion: Weighting had limited impact on dimensionality estimation and large-scale hypothesis testing but impacted prevalence and association estimation more. Results from untargeted association analyses should be followed by weighted analysis when effect size estimation is of interest for specific signals. Conclusion: EHR-linked biobanks should report recruitment and selection mechanisms and provide selection weights with defined target populations. Researchers should consider their intended estimands, specify source and target populations, and weight EHR-linked biobank analyses accordingly.