Faculty Bibliography

BACKGROUND:The Starting Early Program (StEP) promotes healthy nutrition during early life and leads to healthier child weight, but whether dietary patterns contribute to weight or mediate StEP weight outcomes has not been studied. OBJECTIVES/OBJECTIVE:This secondary analysis identified infant dietary patterns in StEP, determined associations between dietary patterns and child weight outcomes, and examined whether dietary patterns mediated the relationship between StEP and child weight. METHODS:-score (WFAz) and likelihood of being classified as overweight (WFA ≥85th percentile) were assessed using linear and logistic multivariable regression models. Mediation was used to assess intervention effects on WFAz via impacts on infant dietary patterns. RESULTS:Four classes of infant dietary patterns were identified: Breastfed-High variety, Formula fed-High variety, Formula fed-Low variety, and Mixed fed-Low variety. Compared to the Breastfed-High variety class, infants in the Formula fed-Low variety class had higher WFAz and were more likely to be classified as overweight at 24 and 36 months. Participation in StEP increased membership in Breastfed-High variety, which mediated the association between StEP and lower WFAz at 24 months. CONCLUSIONS:Infant dietary patterns were identified, and some were associated with child overweight. StEP was associated with a dietary pattern most consistent with guidelines, which mediated intervention effects on child weight.

OBJECTIVE:To examine whether disparities in preventable adverse events (AEs) and family experience for hospitalized children of parents with and without limited comfort with English (LCE) changed following implementation of a health literacy-informed, structured family-centered rounds intervention, Patient and Family Centered I-PASS. METHODS:Seven-center prospective before and after intervention study conducted from December 2014 to January 2017. Inclusion criteria includes parents and caregivers of hospitalized children less than 18 years old. LCE defined as parent and caregiver reporting any language other than English as the language in which they were most comfortable speaking to physicians and nurses. Preventable AE rates determined via (1) daily medical record surveillance, solicited clinician reports, hospital incident reports, and family safety interviews, followed by (2) dual-physician masked review and consensus rating. Changes assessed through multivariable mixed-effects logistic regression, controlling for complex chronic conditions, length of stay, parent race and ethnicity, education, and clustering by site. Family experience assessed via predischarge survey (6 domains, 5-point scale). Mixed-effects linear regression estimated changes in mean domain scores by language comfort. RESULTS:Of 1666 patients, 8.8% (n = 147) had parents with LCE. Preimplementation, over twice as many children of parents expressing LCE experienced preventable AEs, compared with children of parents expressing comfort with English (18.8% vs 7.6%; P = .001; adjusted odds ratio, 3.29; 95% CI, 1.48-7.31). Postimplementation, there were no significant differences between groups (6.4 vs 5.3%; P = .68; adjusted odds ratio, 1.26; 95% CI, 0.42-3.75). Some domains of family experience improved for both groups, although disparities in experience persisted in nurse communication. CONCLUSIONS:Patient safety disparities related to parental comfort with English narrowed after implementing a family-centered rounds communication intervention.

RATIONALE & OBJECTIVE/OBJECTIVE:While kidney failure is associated with severe COVID-19, data on early-stage chronic kidney disease (CKD) and its relationship to acute and post-acute COVID-19, as well as post-vaccination antibody responses, are limited. We evaluated pre-pandemic CKD stage in relation to these outcomes. STUDY DESIGN/METHODS:Prospective cohort. SETTING & PARTICIPANTS/METHODS:Adults in nine US population-based studies established since 1971, with pre-pandemic CKD measurements and follow-up for COVID-19 outcomes. EXPOSURE/METHODS:CKD stages (low CKD stage [reference], moderate CKD stage, high CKD stage, and very high CKD stage) defined by creatinine-based eGFR and urinary albumin-to-creatinine ratio. OUTCOMES/RESULTS:(i) COVID-19 hospitalization or death (self-report/medical records, 2020‒2023); (ii) RECOVER Long COVID Research Index score ≥11 (LCRI-positivity by questionnaires, 2023‒2024); (iii) post-vaccination anti-Spike 1 (S1) IgG levels (dried blood spots, 2021‒2022). ANALYTICAL APPROACH/METHODS:Cause-specific hazards for severe COVID-19, logistic regression for LCRI-positivity, and generalized additive models for anti-S1 IgG. RESULTS:Among 26,039 participants, CKD stage was low in 81.7%, moderate in 13.0%, high in 3.6%, and very high in 1.7%. Over a median 550-day follow-up (P25-P75: 314‒636), 734 had severe COVID-19; 12.1% of infected (669/5,527) were classified as LCRI-positive; and 3.1% (82/2,679) were anti-S1 IgG antibody non-reactive. A more severe CKD stage was associated with a greater risk for severe COVID-19: the HR for moderate stage CKD was 1.27 (95% CI, 1.03‒1.56), for high stage CKD, 2.33 (1.77‒3.06), and for very high stage CKD, 2.72 (1.88‒3.92). Higher CKD stages were associated with lower anti-S1 IgG levels: ‒12.00% (95% CI: ‒21.14% to ‒1.78%) for moderate stage CKD, ‒27.03% (95% CI, ‒40.00% to ‒11.25%) for high stage CKD, and ‒48.9% (95% CI, ‒61.60% to ‒32.01%) for very high stage CKD. A higher prevalence of LCRI positivity was observed only among infected individuals with very high stage CKD (OR=2.20; 95% CI: 1.16‒4.18). LIMITATIONS/CONCLUSIONS:Some outcomes were self-reported. CONCLUSION/CONCLUSIONS:Higher CKD stages were associated with a greater risk for severe COVID-19 and a lower anti-S1 antibody response. No consistent association was observed between CKD stage and LCRI-positivity except in individuals with very high stage CKD. INDEX WORDS/UNASSIGNED:Chronic kidney disease, kidney dysfunction, SARS-CoV-2 infection, severe COVID-19, Long COVID, anti-S1 IgG antibody, Prospective cohort.

OBJECTIVE:Using AI algorithms can exacerbate health disparities if care or resources are allocated away from underserved populations. We evaluated an algorithm for its potential to worsen health disparities across different clinical use cases. MATERIALS AND METHODS/METHODS:This was a retrospective study of patients with heart failure (HF) at an academic health system using an algorithm that predicts pharmacy fill nonadherence to evidence-based HF medications. We compared prediction performance metrics (accuracy, false positive rate, false negative rate), using rate-ratios (RRs), between subgroups with and without known HF care disparities: below vs above median neighborhood-level socioeconomic status (nSES) and Black vs White race. Results were then applied to 3 hypothetical clinical use cases. RESULTS:Among 34 697 patients (13% Black, 10% Hispanic, 65% White), algorithm accuracy was similar across nSES and racial subgroups. The algorithm assigned more false positives for medication nonadherence among low vs high nSES (RR [95%CI] 1.50 [1.44-1.56]) and Black vs White (2.05 [1.92-2.19]) subgroups. The algorithm also assigned fewer false negatives (0.63 [0.59-0.67]) to Black vs White subgroups. When applied to 3 hypothetical use cases, worsening of existing disparities was pertinent for clinical applications where false positives could be particularly harmful (e.g, if predictions of nonadherence prompted lower treatment priority). DISCUSSION/CONCLUSIONS:Although accuracy was similar across demographic groups, differences in false positive and false negative rates revealed that the same prediction may worsen disparities in some use cases, but not others. CONCLUSION/CONCLUSIONS:Evaluation of predictions in the context of clinical use is essential to avoid unintentionally worsening inequities.

In the past decade, interest in studying psychedelic compounds as potential therapeutic agents has resurged. These studies carefully exclude individuals at risk for developing psychotic symptoms in response to psychedelic use. Given the potential for psychedelics to be established as treatments in psychiatry, it is important to more robustly understand their link with psychosis and schizophrenia spectrum disorders (SSDs). In this narrative review, we examine the historical and theoretical relationship between psychedelic drugs and SSDs, including the origins of the psychotomimetic hypothesis. For key psychedelic compounds, we review their phenomenological manifestations in relation to the experiential alterations characteristic of SSDs, revealing both areas of overlap and important qualitative differences that challenge the uniform psychotomimetic classification. We also review putative neural mechanisms underlying altered experiential states associated with psychedelic use and SSDs, with attention to serotonergic, dopaminergic, and glutamatergic contributions. Clinical evidence demonstrates that psychedelics can exacerbate pre-existing psychotic illness and may trigger psychosis in vulnerable individuals, though the magnitude of these risks remains inadequately quantified. However, phenomenological and mechanistic distinctions suggest that potential therapeutic applications may exist for carefully selected symptoms (negative symptoms, depression) in stable patients using low-dose, controlled approaches. Based on published work, we provide recommendations regarding psychosis-related risk and potential avenues for the treatment of SSDs as psychedelics gain traction as therapeutics.

Research is needed to establish optimal cutoffs to determine cocaine use via oral fluid testing while considering the possibility of unintentional exposure. We compared detection of cocaine and its metabolites to self-reported use while considering other drug use to determine optimal cutoffs to determine intentional use. In this study, 1819 adults entering randomly selected nightclubs were surveyed about cocaine and other drug use and had their oral fluid analyzed using liquid chromatography quadrupole time-of-flight mass spectrometry. We compared self-reported use to detected exposure. 13.6% of participants reported past 24-hour cocaine use, 43.7% tested positive for cocaine exposure (≥1 ng/mL), and 11.8% tested positive for at least one metabolite. 28.6% either tested positive for or reported past 24-hour use of other drugs, primarily ketamine. Among those testing positive for cocaine exposure (n = 794), only 29.1% reported past 24-hour use. Among positive cases, the optimal cocaine concentration cut-point for predicting self-report was ≥ 26 ng/mL. When focusing on the full sample and a subsample with data on past 48-hour reported use, optimal cut-points were ≥ 6 ng/mL and ≥ 5 ng/mL, respectively. However, detection of metabolites was the strongest predictor of self-reported use. Using self-report as the gold standard, metabolite detection most accurately classified reported use, over and above detection of cocaine or other drugs; however, relying solely on detection of metabolites or higher cocaine concentration thresholds often led to under-detection. In conclusion, relying on detection of metabolites or higher concentrations most accurately detects intentional use, but these lack sensitivity to detect a portion of use.

OBJECTIVE:To identify proteomic signatures underlying 3-year progression from impaired fasting glucose (IFG) to diabetes. RESEARCH DESIGN AND METHODS/METHODS:We examined IFG progression in the Atherosclerosis Risk in Communities (ARIC) study from visit 2 (1990-1992) to visit 3 (1993-1995). We tested associations of 4,955 plasma proteins (SomaScan version 4.0) with ∼3-year progression from IFG (FG 100-125 mg/dL without diabetes) to diabetes (diagnosis, medication, or FG ≥126 mg/dL) using logistic regression models adjusted for demographics, cardiometabolic risk factors, and baseline glucose with Bonferroni correction (P < 10-5). We explored biological pathways enriched among the top proteins and calculated improvements in prediction (ΔAUC and net reclassification using 3-year risk thresholds of 6% and 15% in 80% training and 20% internal validation subsamples). We validated results in the Multi-Ethnic Study of Atherosclerosis (MESA) cohort. RESULTS:There were 3,786 ARIC participants with IFG (mean [SD] age 57 [6] years, 52% female, 28% Black individuals). The 3-year cumulative incidence of diabetes was 6%. Six proteins were associated with ∼3-year progression to diabetes, namely lower receptor-type tyrosine-protein phosphatase S (PTPRS), anthrax toxin receptor 2 (ANTXR2), adiponectin (ADIPOQ), ciliary neurotrophic factor receptor subunit α (CNTFR), transmembrane protein 132C (TMEM132C), and higher ADAMTS-like protein 2 (ADAMTSL2). Altered carbohydrate metabolism and glycolysis were key pathways. Adding the six proteins to covariates improved discrimination (optimism-corrected AUC 0.81, ΔAUC 0.03, P = 0.005) and net reclassification (training 12.2%, internal validation 12.0%) with predicted diabetes risk quintiles spanning <1% to ∼20%. Two of the six proteins were validated in MESA (P < 0.008). CONCLUSIONS:We identified proteins associated with 3-year IFG progression, with improvements in diabetes risk stratification.

BACKGROUND:Preterm birth (PTB, < 37 weeks of gestation) is the leading cause of child mortality in the United States (U.S.) and worldwide, and has substantial short- and long-term health consequences for mothers and infants. Each year, > 350,000 infants in the U.S. are born preterm, and rates continue to rise in parallel with maternal risk factors such as hypertension, diabetes, anemia, asthma, and mental health conditions. Evidence-based interventions exist for many of these conditions and are associated with improved pregnancy outcomes, including low-dose aspirin for preeclampsia prevention in individuals with chronic hypertension or pregestational diabetes, inhalers for asthma, iron for anemia, and therapy or medication for mental health disorders, but fewer than half of eligible individuals receive them, reflecting persistent gaps in use. To address this, we developed the PTB Actionable Risk Index (PTB-ARIx), which leverages factors with known evidence-based interventions to identify individuals who are pregnant and are at increased risk for PTB. This study evaluates performance of the PTB-ARIx throughout pregnancy with respect to risk determination and characterization of actionable risk factors, including their combined contributions to PTB. METHODS:A retrospective cohort study was conducted using linked data for 1.9 million singleton live births in California in 2016-2020, divided into training and testing sets. Poisson regression estimated associations between 18 candidate risk factors for PTB with evidence-based interventions spanning clinical, behavioral, and social risks, including preeclampsia risk composites (≥ 1 high-risk or ≥ 2 moderate-risk factors based on U.S. Preventive Services Task Force (USPSTF) criteria), maternal conditions (e.g., gestational hypertension, asthma), substance use, and social adversity. Beta coefficients were combined to construct the PTB-ARIx, evaluated by per-unit associations with PTB and by area under the receiver operating characteristic curve (AUC) overall, by early (< 32 weeks), late (32-36 weeks), spontaneous, and medically-indicated PTB, and by PTB co-occurring with preeclampsia. FINDINGS/RESULTS:All risk factors were found to be associated with increased PTB risk. Having ≥ 1 high-risk or ≥ 2 moderate-risk factors for preeclampsia (based on composites) was most strongly related to PTB (relative risk (RR) 6.73, 95% confidence interval (CI) 6.57, 6.89). Each unit increase in PTB-ARIx was associated with > 60% higher PTB risk (RRs 1.66-1.72) across training and testing samples, with consistent findings across PTB and race/ethnicity-insurance subgroups. Model performance was modest for late PTB (AUC ≈ 0.63), stronger for early PTB (0.69-0.72), and especially high for early PTB with preeclampsia (AUCs up to 0.97). Over 70% of individuals with PTB-ARIx scores ≥ 3.00 experienced PTB or another adverse outcome such as low birth weight (< 2,500 grams). CONCLUSIONS:The PTB-ARIx is a well-performing metric for identifying individuals at increased risk for PTB and other adverse pregnancy outcomes. By centering on modifiable risks, the PTB-ARIx combines risk identification with opportunities for intervention. Demonstrating strong performance across subgroups, including for early PTB and PTB with preeclampsia, the PTB-ARIx provides a potential pathway to improve patient-provider communication and uptake of equitable, evidence-based care. Further validation, including integration with treatment data, is needed to confirm its potential to reduce PTB risk and rates.

BACKGROUND AND AIMS/OBJECTIVE:Individuals who engage in illicit or nonmedical opioid use may have elevated risk of health and social consequences, including progression to opioid use disorder (OUD). Preventive interventions to reduce this risk are lacking. This trial tested the impact of a primary care-integrated collaborative care approach for reducing risky opioid use, defined as nonmedical use of prescription opioids or any use of illicit opioids. DESIGN/METHODS:Cluster-randomized controlled trial randomized primary care providers (PCPs) and their patients into the Subthreshold Opioid Use Disorder Prevention (STOP) intervention or enhanced usual care (EUC). SETTING/METHODS:Primary care clinics at 5 U.S. sites. PARTICIPANTS/METHODS:PCPs and their patients were recruited January 2021-May 2023. A total of 119 PCP clusters (STOP = 48, EUC = 51) and 202 patients (STOP = 88, EUC = 114) enrolled. Eligible patients were adults (≥18 years) having current risky opioid use, without moderate-severe OUD. Patient participants were majority female (63.4%), white (70.8%) and non-Hispanic (96.5%), with a mean age of 55.7 [standard deviation (SD) = 12.7] years. At baseline, 63.4% of participants had moderate-severe pain (Brief Pain Inventory) and below average physical (79.2%) and mental (62.4%) health (SF-12). INTERVENTIONS/METHODS:The STOP collaborative care intervention consisted of brief advice from the PCP about reducing risky opioid use, meetings with a clinic-embedded nurse care manager over 12 months and remote health coaching (2-6 sessions). Both groups received primary care treatment as usual and overdose risk reduction materials. MEASUREMENTS/METHODS:The primary outcome was total days of risky opioid use, recorded from 6 monthly electronic surveys. A key secondary outcome was moderate-severe OUD at 6 and 12 months. FINDINGS/RESULTS:A total of 77 (87.5%) STOP and 107 (93.9%) EUC participants completed the 6-month assessment period. The primary outcome analysis used the Intention-to-Treat sample with multiple imputations of missing data. Mean days of risky opioid use at 180 days were lower in STOP than EUC [12.2 (SD = 27.73) vs. 15.5 (SD = 32.64)]; the difference between groups adjusted for baseline risky opioid use was not statistically significant (rate ratio 0.95, 95% confidence interval = 0.52-1.74). One STOP participant (1.1%) and 13 EUC participants (11.4%) developed moderate-severe OUD at 6 months, and 3 (3.4%) STOP and 6 (5.3%) EUC participants had moderate-severe OUD at 12 months (P < 0.001). CONCLUSIONS:This cluster-randomized controlled trial did not find evidence that the STOP intervention for reducing risky opioid use produced greater reductions over 6 months compared with enhanced usual care, though fewer intervention participants progressed to moderate-severe opioid use disorder. Patients had a high burden of pain and comorbidities that may present challenges to reducing opioid use.