Supervised contrastive learning enhances graph convolutional networks for predicting neurodevelopmental deficits in very preterm infants using brain structural connectome.

Very preterm (VPT) infants (born at less than 32 weeks gestational age) are at high risk for various adverse neurodevelopmental deficits. Unfortunately, most of these deficits cannot be accurately diagnosed until the age of 2-5 years old. Given the benefits of early interventions, accurate diagnosis and prediction soon after birth are urgently needed for VPT infants. Previous studies have applied deep learning models to learn the brain structural connectome (SC) to predict neurodevelopmental deficits in the preterm population. However, none of these models are specifically designed for graph-structured data, and thus may potentially miss certain topological information conveyed in the brain SC. In this study, we aim to develop deep learning models to learn the SC acquired at term-equivalent age for early prediction of neurodevelopmental deficits at 2 years corrected age in VPT infants. We directly treated the brain SC as a graph, and applied graph convolutional network (GCN) models to capture complex topological information of the SC. In addition, we applied the supervised contrastive learning (SCL) technique to mitigate the effects of the data scarcity problem, and enable robust training of GCN models. We hypothesize that SCL will enhance GCN models for early prediction of neurodevelopmental deficits in VPT infants using the SC. We used a regional prospective cohort of ∼280 VPT infants who underwent MRI examinations at term-equivalent age from the Cincinnati Infant Neurodevelopment Early Prediction Study (CINEPS). These VPT infants completed neurodevelopmental assessment at 2 years corrected age to evaluate cognition, language, and motor skills. Using the SCL technique, the GCN model achieved mean areas under the receiver operating characteristic curve (AUCs) in the range of 0.72∼0.75 for predicting three neurodevelopmental deficits, outperforming several competing models. Our results support our hypothesis that the SCL technique is able to enhance the GCN model in our prediction tasks.

Estimating effects of longitudinal and cumulative exposure to PFAS mixtures on early adolescent body composition.

Few methods have been used to characterize repeatedly measured biomarkers of chemical mixtures. We applied latent profile analysis (LPA) to serum concentrations of 4 perfluoroalkyl and polyfluoroalkyl substances (PFAS) measured at 4 time points from gestation to age 12 years. We evaluated the relationships between profiles and z scores of height, body mass index, fat mass index, and lean body mass index at age 12 years (n = 218). We compared LPA findings with an alternative approach for cumulative PFAS mixtures using g-computation to estimate the effect of simultaneously increasing the area under the receiver operating characteristic curve (AUC) for all PFAS. We identified 2 profiles: a higher PFAS profile (35% of sample) and a lower PFAS profile (relative to each other), based on their average PFAS concentrations at all time points. The higher PFAS profile had generally lower z scores for all outcomes, with somewhat larger effects for males, though all 95% CIs crossed the null. For example, the higher PFAS profile was associated with a 0.50-unit lower (ß = -0.50; 95% CI, -1.07 to 0.08) BMI z score among males but not among females (ß = 0.04; 95% CI, -0.45 to 0.54). We observed similar patterns with AUCs. We found that a higher childhood PFAS profile and higher cumulative PFAS mixtures may be associated with altered growth in early adolescence. This article is part of a Special Collection on Environmental Epidemiology.

Associations Between Brain Metabolites Measured With MR Spectroscopy and Head Impacts in High School American Football Athletes.

BACKGROUND: While changes in brain metabolites after injury have been reported, relationships between metabolite changes and head impacts are less characterized. PURPOSE: To investigate alterations in neurochemistry in high school athletes as a function of head impacts, concussion, and the use of a jugular vein compression (JVC) collar. STUDY TYPE: Prospective controlled trial. SUBJECTS: A total of 284 male American football players, divided into JVC collar and noncollar groups; 215 included in final analysis (age = 15.9 ± 1.0 years; 114 in collar group). FIELD STRENGTH/SEQUENCE: 3 Tesla/T1-weighted gradient echo, 1H point resolved spectroscopy, acquired between August and November 2018. ASSESSMENT: Head impacts were quantified using accelerometers. Concussion was diagnosed by medical professionals for each team. Pre- to postseason differences in total N-acetylaspartate (tNAA), total choline (tCho), myo-inositol (myoI), and glutamate + glutamine (Glx), in primary motor cortex (M1) and anterior cingulate cortex (ACC), relative to total creatine (tCr), were determined. STATISTICAL TESTS: Group-wise comparisons were performed using Wilcoxon signed-rank, Friedman's, and Mann-Whitney U tests. Relationships between ∆metabolite/tCr and mean g-force were analyzed using linear regressions accounting for concussion and JVC collar. Significance was set at P ≤ 0.05. RESULTS: In participants without concussion, a significant decrease in tCho/tCr (0.233 ± 1.40 × 10-3 to 0.227 ± 1.47 × 10-7) and increase in Glx/tCr (1.60 ± 8.75 × 10-3 to 1.63 ± 1.08 × 10-2) in ACC were observed pre- to postseason. The relationship between ∆tCho/tCr in M1 and ACC and mean g-force from >80 g to >140 g differed significantly between participants with and without concussion (M1 ß ranged from 3.9 × 10-3 to 2.1 × 10-3; ACC ß ranged from 2.7 × 10-3 to 2.1 × 10-3). Posthoc analyses revealed increased tCho/tCr in M1 was positively associated with mean g-force >100 g (ß = 3.6 × 10-3) and >110 g (ß = 2.9 × 10-3) in participants with concussion. Significant associations between ∆ myoI / tCr $$ \Delta \mathrm{myoI}/\mathrm{tCr} $$ in ACC and mean g-force >110 g (ß = -1.1 × 10-3) and >120 g (ß = -1.1 × 10-3) were observed in the collar group only. DATA CONCLUSION: Diagnosed concussion and the use of a JVC collar result in distinct neurochemical trends after repeated head impacts. LEVEL OF EVIDENCE: 2 TECHNICAL EFFICACY: Stage 3.

From urges to tics in children with Tourette syndrome: associations with supplementary motor area GABA and right motor cortex physiology.

Tourette syndrome (TS) is a childhood-onset disorder in which tics are often preceded by premonitory sensory urges. More severe urges correlate with worse tics and can render behavioral therapies less effective. The supplementary motor area (SMA) is a prefrontal region believed to influence tic performance. To determine whether cortical physiological properties correlate with urges and tics, we evaluated, in 8-12-year-old right-handed TS children (n = 17), correlations of urge and tic severity scores and compared both to cortical excitability (CE) and short- and long-interval cortical inhibition (SICI and LICI) in both left and right M1. We also modeled these M1 transcranial magnetic stimulation measures with SMA gamma-amino butyric acid (GABA) levels in TS and typically developing control children (n = 16). Urge intensity correlated strongly with tic scores. More severe urges correlated with lower CE and less LICI in both right and left M1. Unexpectedly, in right M1, lower CE and less LICI correlated with less severe tics. We found that SMA GABA modulation of right, but not left, M1 CE and LICI differed in TS. We conclude that in young children with TS, lower right M1 CE and LICI, modulated by SMA GABA, may reflect compensatory mechanisms to diminish tics in response to premonitory urges.

Gestational PBDE concentrations, persistent externalizing, and emerging internalizing behaviors in adolescents: The HOME study.

BACKGROUND: Polybrominated diphenyl ethers (PBDEs) are ubiquitous environmental chemicals used as flame retardants in commercial and consumer products. Gestational PBDE concentrations are associated with adverse behaviors in children; however, the persistence of these associations into adolescence remains understudied. OBJECTIVE: We estimated the association of gestational PBDE serum concentrations with early adolescent self- and caregiver-reported behaviors at age 12 years and determined the consistency with previously observed associations in childhood with caregiver-reported behaviors in a prospective pregnancy and birth cohort. METHODS: We measured maternal serum concentrations of five individual PBDE congeners and created a summary exposure variable (∑5BDE: 28, -47, -99, -100 and -153) during pregnancy. At age 12 years, we assessed behaviors for 237 adolescents using self- and caregiver-reports with the Behavioral Assessment System for Children-3 (BASC3). We used multivariable linear regression models to estimate covariate-adjusted associations of lipid standardized, log10-transformed gestational PBDE concentrations with BASC3 scores. We obtained estimates and 95% confidence intervals through a bootstrapping approach. We evaluated potential effect measure modification (EMM) of adolescent sex by examining sex-stratified regression models and estimating the EMM p-values. RESULTS: Gestational PBDE concentrations were positively associated with adolescent-reported BASC3 composite indices for inattention & hyperactivity (BDE-28, -47, -99, -100, ∑5BDE), internalizing problems (BDE-28, -47, -99), functional impairment (BDE-28, ∑5BDE), and emotional symptoms (BDE-28). Gestational PBDE concentrations were positively associated with caregiver-reported BASC3 composite indices for externalizing problems (BDE-28, -47, -99, -100, -153, ∑5BDE) and behavioral symptoms (BDE-99). For caregiver reported behaviors, we observed stronger associations with gestational BDE concentrations among males, especially for executive functioning (BDE-28, -47, -99, -100, ∑5BDE). DISCUSSION: Gestational PBDE serum concentrations were associated with self-reported internalizing and externalizing behavior problems in early adolescence. Caregiver-reported externalizing behaviors recognized during childhood remain associated with gestational PBDE concentrations and persist into early adolescence. Internalizing behaviors were less recognized by caregivers.

Per- and polyfluoroalkyl substances and bone mineral content in early adolescence: Modification by diet and physical activity.

BACKGROUND: Per- and polyfluoroalkyl substance (PFAS) exposures may negatively impact bone mineral accrual, but little is known about potential mitigators of this relation. We assessed whether associations of PFAS and their mixture with bone mineral content (BMC) in adolescence were modified by diet and physical activity. METHODS: We included 197 adolescents enrolled in a prospective pregnancy and birth cohort in Cincinnati, Ohio (2003-2006). At age 12 years, we collected serum for PFAS measurements and used dual-energy x-ray absorptiometry to measure BMC. We calculated dietary calcium intake and Health Eating Index (HEI) scores from repeated 24-h dietary recalls, physical activity scores using the Physical Activity Questionnaire for Older Children (PAQ-C), and average moderate to vigorous physical activity (MVPA) based on accelerometry. We estimated covariate-adjusted differences in BMC z-scores per interquartile range (IQR) increase of individual PFAS concentrations using linear regression and per simultaneous IQR increase in all four PFAS using g-computation. We evaluated effect measure modification (EMM) using interaction terms between each modifier and PFAS. RESULTS: Higher serum perfluorooctanoic acid, perfluorooctanesulfonic acid, and perfluorononanoic acid concentrations and the PFAS mixture were associated with lower BMC z-scores. An IQR increase in all PFAS was associated with a 0.27 (-0.54, 0.01) lower distal radius BMC z-score. Associations with lower BMC were generally stronger among adolescents classified as < median for calcium intake, HEI scores, or MVPA compared to those ≥ median. The difference in distal radius BMC z-score per IQR increase in all PFAS was -0.38 (-0.72, -0.04) for those with

Early life exposure to secondhand tobacco smoke and eating behaviors at age 12 years.

BACKGROUND: Prenatal or early childhood secondhand tobacco smoke (SHS) exposure increases obesity risk. However, the potential mechanisms underlying this association are unclear, but obesogenic eating behaviors are one pathway that components of SHS could perturb. Our aim was to assess associations of prenatal and early childhood SHS exposure with adolescent eating behaviors. METHODS: Data came from a prospective pregnancy and birth cohort (N = 207, Cincinnati, OH). With multiple informant models, we estimated associations of prenatal (mean of 16 and 26 weeks of gestation maternal serum cotinine concentrations) and early childhood cotinine (average concentration across ages 12, 24, 36, and 48 months) with eating behaviors at age 12 years (Child Eating Behaviors Questionnaire). We tested whether associations differed by exposure periods and adolescent's sex. Models adjusted for maternal and child covariates. RESULTS: We found no statistically significant associations between cotinine measures and adolescent's eating behaviors. Yet, in females, prenatal cotinine was associated with greater food responsiveness (ß: 0.23; 95% CI: 0.08, 0.38) and lower satiety responsiveness (ß: -0.14; 95% CI: -0.26, -0.02); in males, prenatal and postnatal cotinine was related to lower food responsiveness (prenatal: ß: -0.25; 95% CI: -0.04, -0.06; postnatal: ß: -0.36; 95% CI: -0.06, -0.11). No significant effect modification by sex or exposure window was found for other eating behaviors. CONCLUSION: Prenatal and early childhood SHS exposures were not related to adolescent's eating behavior in this cohort; however, biological sex may modify these associations.

Associations of a Prenatal Serum Per- and Polyfluoroalkyl Substance Mixture with the Cord Serum Metabolome in the HOME Study.

Per- and polyfluoroalkyl substances (PFAS) are ubiquitous and persistent chemicals associated with multiple adverse health outcomes; however, the biological pathways affected by these chemicals are unknown. To address this knowledge gap, we used data from 264 mother-infant dyads in the Health Outcomes and Measures of the Environment (HOME) Study and employed quantile-based g-computation to estimate covariate-adjusted associations between a prenatal (∼16 weeks' gestation) serum PFAS mixture [perfluorooctanesulfonic acid (PFOS), perfluorooctanoic acid (PFOA), perfluorohexanesulfonic acid (PFHxS), and perfluorononanoic acid (PFNA)] and 14,402 features measured in cord serum. The PFAS mixture was associated with four features: PFOS, PFHxS, a putatively identified metabolite (3-monoiodo-l-thyronine 4-O-sulfate), and an unidentified feature (590.0020 m/z and 441.4 s retention time; false discovery rate <0.20). Using pathway enrichment analysis coupled with quantile-based g-computation, the PFAS mixture was associated with 49 metabolic pathways, most notably amino acid, carbohydrate, lipid and cofactor and vitamin metabolism, as well as glycan biosynthesis and metabolism (P(Gamma) <0.05). Future studies should assess if these pathways mediate associations of prenatal PFAS exposure with infant or child health outcomes, such as birthweight or vaccine response.

Pre- and postnatal exposure to secondhand tobacco smoke and cardiometabolic risk at 12 years: Periods of susceptibility.

BACKGROUND: To identify periods of heightened susceptibility to the association of secondhand tobacco smoke (SHS) exposure with cardiometabolic (CM) risk at age 12 years. METHODS: We used data from 212 adolescents from the HOME Study, a prospective pregnancy and birth cohort in Cincinnati, OH. Using multiple informant models, we estimated associations of maternal serum cotinine (mean of concentrations at 16 and 26 weeks of pregnancy) and children's serum cotinine concentrations (mean of concentrations at ages 1, 2, 3, and 4 years) with a CM risk summary score constructed of five risk components measured at age 12 years. We determined if these associations differed for pre- and postnatal exposure periods, and adolescent's sex. RESULTS: We found some evidence that the cotinine-outcome associations differed by exposure period and sex. Postnatal, but not prenatal, cotinine was associated with higher CM risk scores and individual CM risk component values (interaction p-values = 0.04 to 0.35). Each 10-fold increase in postnatal cotinine was associated with 0.57 (95% CI: 0.32, 1.45), 0.09 (95% CI: 0.13, 0.31), 0.14 (-0.08, 0.35), 0.07 (95% CI: 0.34, 0.48), and 0.11 (95% CI: 0.04, 0.27) higher CM risk, HOMA-IR, TG to HDL-C ratio, leptin to adiponectin ratio, and visceral fat area. Postnatal cotinine was associated with higher visceral fat area among females but not males (sex × period × cotinine interaction p-value = 0.01). CONCLUSIONS: Serum cotinine concentrations during the postnatal period had greater influence on adolescent's CM risk compared to the prenatal period, and these associations may be sex-specific. This study reinforces the need for ongoing public health interventions to minimize children's exposure to SHS.

Dietary per- and polyfluoroalkyl substance (PFAS) exposure in adolescents: The HOME study.

BACKGROUND: Diet is the primary exposure pathway for per- and polyfluoroalkyl substances (PFAS) in non-occupationally exposed populations. Few studies have examined associations of dietary quality and macronutrient intake with PFAS exposure among US adolescents. OBJECTIVE: To assess relationships of self-reported dietary quality and macronutrient intake with serum PFAS concentrations in adolescents. METHODS: We used cross-sectional data from 193 Cincinnati, Ohio area adolescents (median age 12.3 years) collected from 2016 to 2019. Using 24-h food recalls completed by adolescents on three separate days, we derived Healthy Eating Index (HEI) scores, HEI components, and macronutrient intake. We measured perfluorooctanoic acid (PFOA), perfluorooctane sulfonic acid (PFOS), perfluorohexane sulfonic acid (PFHxS), and perfluorononanoic acid (PFNA) concentrations in fasting serum samples. Using linear regression, we estimated covariate-adjusted associations between dietary variables and serum PFAS concentrations. RESULTS: The median HEI score was 44 and median serum PFOA, PFOS, PFHxS, and PFNA concentrations were 1.3, 2.4, 0.7, and 0.3 ng/mL respectively. In adjusted models, higher total HEI scores, whole fruit and total fruit HEI component scores, and total dietary fiber intake were associated with lower concentrations of all four PFAS. For example, serum PFOA concentrations were 7% lower (95% CI: -15, 2) per standard deviation increase in total HEI score and 9% lower (95% CI: -18, 1) per standard deviation increase in dietary fiber. SIGNIFICANCE: Given adverse health effects associated with PFAS exposure, it is crucial to understand modifiable exposure pathways. Findings from this study may inform future policy decisions aiming to limit human exposure to PFAS.

Frequency and Intensity of Premonitory Urges-to-Tic in Tourette Syndrome Is Associated With Supplementary Motor Area GABA+ Levels.

BACKGROUND: Individuals with Tourette syndrome (TS) often report that they express tics as a means of alleviating the experience of unpleasant sensations. These sensations are perceived as an urge to act and are referred to as premonitory urges. Premonitory urges have been the focus of recent efforts to develop interventions to reduce tic expression in those with TS. OBJECTIVE: The aim of this study was to examine the contribution of brain γ-aminobutyric acid (GABA) and glutamate levels of the right primary sensorimotor cortex (SM1), supplementary motor area (SMA), and insular cortex (insula) to tic and urge severity in children with TS. METHODS: Edited magnetic resonance spectroscopy was used to assess GABA+ (GABA + macromolecules) and Glx (glutamate + glutamine) of the right SM1, SMA, and insula in 68 children with TS (MAge = 10.59, SDAge = 1.33) and 41 typically developing control subjects (MAge = 10.26, SDAge = 2.21). We first compared GABA+ and Glx levels of these brain regions between groups. We then explored the association between regional GABA+ and Glx levels with urge and tic severity. RESULTS: GABA+ and Glx of the right SM1, SMA, and insula were comparable between the children with TS and typically developing control subjects. In children with TS, lower levels of SMA GABA+ were associated with more severe and more frequent premonitory urges. Neither GABA+ nor Glx levels were associated with tic severity. CONCLUSIONS: These results broadly support the role of GABAergic neurotransmission within the SMA in the experience of premonitory urges in children with TS. © 2021 International Parkinson and Movement Disorder Society.

Magnetic resonance spectroscopy brain metabolites at term and 3-year neurodevelopmental outcomes in very preterm infants.

BACKGROUND: Noninvasive advanced neuroimaging and neurochemical assessment can identify subtle abnormalities and predict neurodevelopmental impairments. Our objective was to quantify white matter metabolite levels and evaluate their relationship with neurodevelopmental outcomes at age 3 years. METHODS: Our study evaluated a longitudinal prospective cohort of very premature infants (<32 weeks gestational age) with single-voxel proton magnetic resonance spectroscopy from the centrum semiovale performed at term-equivalent age and standardized cognitive, verbal, and motor assessments at 3 years corrected age. We separately examined metabolite ratios in the left and right centrum semiovale. We also conducted an exploratory interaction analysis for high/low socioeconomic status (SES) to evaluate the relationship between metabolites and neurodevelopmental outcomes, after adjusting for confounders. RESULTS: We found significant relationships between choline/creatine levels in the left and right centrum semiovale and motor development scores. Exploratory interaction analyses revealed that, for infants with low SES, there was a negative association between choline/creatine in the left centrum semiovale and motor assessment scores at age 3 years. CONCLUSIONS: Brain metabolites from the centrum semiovale at term-equivalent age were associated with motor outcomes for very preterm infants at 3 years corrected age. This effect may be most pronounced for infants with low SES. IMPACT: Motor development at 3 years corrected age for very preterm infants is inversely associated with choline neurochemistry within the centrum semiovale on magnetic resonance spectroscopy at term-equivalent age, especially in infants with low socioeconomic status. No prior studies have studied metabolites in the centrum semiovale to predict neurodevelopmental outcomes at 3 years corrected age based on high/low socioeconomic status. For very preterm infants with lower socioeconomic status, higher choline-to-creatine ratio in central white matter is associated with worse neurodevelopmental outcomes.

Early exposure to flame retardants is prospectively associated with anxiety symptoms in adolescents: A prospective birth cohort study.

BACKGROUND: Anxiety disorders emerge during childhood and adolescence and are frequently preceded by subsyndromal anxiety symptoms. Environmental toxicants, including gestational polybrominated diphenyl ether (PBDE) exposure, are associated with neuropsychiatric sequelae; however, the role of PBDEs as risk factors for anxiety in adolescence is unclear. METHODS: Using data from the Health Outcomes and Measures of the Environment (HOME) Study, a prospective pregnancy and birth cohort enrolled from 2003 to 2006, we investigated the relationship between gestational serum PBDE concentrations and anxiety symptoms in adolescents (N = 236). We measured five PBDE congeners (PBDE-28, -47, -99, -100, and -153) at 16 ± 3 weeks of gestation and calculated their sum (∑PBDE). We assessed self-reported anxiety symptoms using the Screen for Child Anxiety Related Emotional Disorders (SCARED) and depressive symptoms using the Children's Depression Inventory (CDI-2) at age 12 years. We estimated the associations of maternal PBDE concentrations with child anxiety and depressive symptoms using multivariable linear regression and modified Poisson regression. Covariates included child sex, maternal race, maternal age at delivery, maternal marital status, maternal education, and household income at the 12-year study visit as well as maternal depressive and anxiety symptoms. Sensitivity analyses were performed to control for maternal lead and mercury at delivery. RESULTS: After adjusting for predetermined covariates, each doubling in maternal PBDE concentrations was associated with increased SCARED scores (e.g., for ∑PBDE, SCARED total score, ß = 1.6 95% confidence interval [CI]: 0.3-2.9, p = .019) and a nonsignificant increase in depressive symptoms (e.g., for CDI total score, ß = .8, 95% CI: -0.2-1.8, p = .11). CONCLUSIONS: Gestational serum PBDE concentrations just before mid-pregnancy and during a period of active cortical and limbic neurogenesis, synaptogenesis and myelogenesis may be a risk factor for developing anxiety symptoms in early adolescence.

Childhood urinary organophosphate esters and cognitive abilities in a longitudinal cohort study.

The use of organophosphate esters (OPEs) as flame retardants, which has increased over the past two decades, raises concerns that OPEs may be harmful to humans, especially children. Animal studies and some human studies have reported that OPEs may adversely impact brain development, but few human studies evaluated OPE exposure during early childhood and neurodevelopmental outcomes. We aimed to fill this knowledge gap with the present study on urinary OPE metabolite concentrations at ages 1-5 years and cognitive abilities at 8 years. We used data of 223 children from the Health Outcomes and Measures of the Environment (HOME) Study, a prospective pregnancy and birth cohort in Cincinnati, Ohio. The point estimates for bis-2-chloroethyl-phosphate (BCEP) and bis(1,3-dichloro-2-propyl)-phosphate (BDCIPP) in association with IQ tended to be small and positive, while the point estimates for diphenyl-phosphate (DPHP) were small and negative, with 95% CIs including the null. However, we did find that socioeconomic status (SES) variables modified associations between OPEs and child IQ, with adverse OPE-IQ associations being stronger in socioeconomically disadvantaged children than in others. We identified an additional 1- to 2-point decrease in Full Scale IQ for every log-unit increase in BDCIPP, BCEP, and DPHP among those with lower maternal education, non-white race, lower income, or living in more deprived neighborhoods. We observed similar results for the Perceptual Reasoning, Verbal Comprehension, and Working Memory Index Scores. We suspect that there is residual confounding related to socioeconomic disadvantage, which was not captured with the available SES variables typically used in epidemiologic studies.

Physical activity modifies the relation between gestational perfluorooctanoic acid exposure and adolescent cardiometabolic risk.

OBJECTIVE: Exposure to per- and polyfluoroalkyl substances (PFAS) - endocrine disrupting chemicals - may increase cardiometabolic risk. We evaluated whether adolescent lifestyle factors modified associations between gestational PFAS exposure and cardiometabolic risk using a prospective cohort study. METHODS: In 166 mother-child pairs (HOME Study), we measured concentrations of four PFAS in maternal serum collected during pregnancy. When children were age 12 years, we calculated cardiometabolic risk scores from visceral adiposity area, blood pressure, and fasting serum biomarkers. We assessed adolescent physical activity and Healthy Eating Index scores using the Physical Activity Questionnaire for Older Children (PAQ-C), actigraphy, and 24-h diet recalls. Using multivariable linear regression and weighted quantile sum regression, we examined whether physical activity or diet modified covariate-adjusted associations of PFAS and their mixture with cardiometabolic risk scores. RESULTS: Physical activity modified associations between perfluorooctanoic acid (PFOA) and cardiometabolic risk scores. Each doubling of PFOA was associated with worse cardiometabolic risk scores among children with PAQ-C scores < median (ß:1.4; 95% CI:0.5, 2.2, n = 82), but not among those with PAQ-C scores ≥ median (ß: 0.2; 95% CI: 1.2, 0.7, n = 84) (interaction p-value = 0.01). Associations were most prominent for insulin resistance, leptin-adiponectin ratio, and visceral fat area. We observed results suggesting that physical activity modified the association of PFAS mixture with cardiometabolic risk scores, insulin resistance, and visceral fat area (interaction p-values = 0.17, 0.07, and 0.10, respectively); however, the 95% CIs of the interaction terms included the null value. We observed similar, but attenuated patterns for PFOA and actigraphy-based measures of physical activity. Diet did not modify any associations. Physical activity or diet did not modify associations for other PFAS. CONCLUSIONS: Childhood physical activity modified associations of prenatal serum PFOA concentrations with children's cardiometabolic risk in this cohort, indicating that lifestyle interventions may ameliorate the adverse effects of PFOA exposure.

Gestational and childhood phthalate exposures and adolescent body composition: The HOME study.

BACKGROUND: Early life phthalate exposures may disrupt metabolism but results from human studies are inconsistent and few have examined body composition during adolescence. We investigated associations of gestational and childhood urinary phthalate biomarker concentrations with body composition at age 12 years. METHODS: We used data from 206 mother-child pairs in a prospective pregnancy and birth cohort enrolled in Cincinnati, OH from 2003 to 2006. We measured nine phthalate metabolites in spot urine samples collected twice from mothers during pregnancy and up to seven times from children at 1, 2, 3, 4, 5, 8, and 12 years. At age 12 years, we assessed fat and lean mass of the whole body and android and gynoid subregions, and visceral fat area with dual x-ray absorptiometry, and calculated android to gynoid %fat ratio and age- and sex-standardized fat and lean mass index z-scores. Using a multiple informant model, we estimated covariate-adjusted associations between urinary phthalate biomarker concentrations at each time period and outcomes at age 12 years. We assessed effect measure modification by child sex using stratified models. RESULTS: Generally, urinary mono-benzyl phthalate (MBzP) concentrations were modestly associated with lower fat and lean mass. Each 10-fold increase in urinary MBzP concentrations during gestation and at ages 5 and 8 years was associated with a -0.34 (95%CI: -0.72, 0.05), -0.44 (95% CI: -0.83, -0.05), and -0.35 (95% CI: -0.71, 0.00) z-score difference in lean body mass index, respectively. Urinary monoethyl phthalate, mono-(3-carboxypropyl) phthalate, and summed di(2-ethylhexyl) phthalate metabolites were associated with greater lean mass at some exposure periods. Slightly weaker but similar patterns of association were found with other body composition measures; associations did not differ by child sex. CONCLUSION: While most associations were weak, exposure to certain phthalates during gestation and childhood may be associated with adolescent body composition, particularly lean mass.

Maternal urinary OPE metabolite concentrations and blood pressure during pregnancy: The HOME study.

BACKGROUND: Few studies have examined the association between maternal exposure to organophosphate esters (OPEs) and systolic/diastolic blood pressure (SBP/DBP) during pregnancy. METHODS: We analyzed data from 346 women with a singleton live birth in the HOME Study, a prospective birth cohort in Cincinnati, Ohio, USA. We quantified four OPE metabolites in maternal spot urine samples collected at 16 and 26 weeks pregnancy, standardized by specific gravity. We calculated intraclass correlation coefficients (ICCs). We extracted the first two recorded BP measurements (<20 weeks), the two highest recorded BP measurements (≥20 weeks), and diagnoses of hypertensive disorders of pregnancy (HDP) via chart review. Women with two BP measurements ≥140/90 mmHg or HDP noted in the chart at ≥20 weeks pregnancy were defined as HDP cases. We used linear mixed models and modified Poisson regression with covariate adjustment to estimate associations between OPE concentrations as continuous variables or in tertiles with maternal BP and HDP. RESULTS: ICCs of OPEs were 0.17-0.45. Diphenyl phosphate (DPHP) had the highest geometric mean concentration among OPE metabolites. Increasing the average bis(2-chloroethyl) phosphate (BCEP) concentrations were positively associated with two highest recorded DBP ≥20 weeks pregnancy. Compared with women in the 1st DPHP tertile, women in the 3rd tertile at 16 weeks pregnancy had 1.72 mmHg (95% CI: -0.01, 3.46) higher DBP <20 weeks pregnancy, and women in the 3rd tertile of the average DPHP concentrations had 2.25 mmHg (95% CI: 0.25, 4.25) higher DBP ≥20 weeks pregnancy. 33 women (9.5%) were identified with HDP. Di-n-butyl phosphate (DNBP) concentrations at 16 weeks were positively associated with HDP, with borderline significance (RR = 2.98, 95% CI 0.97-9.15). Other OPE metabolites were not significantly associated with HDP. CONCLUSION: Maternal urinary BCEP and DPHP concentrations were associated with increased BP during pregnancy. Maternal urinary DNBP concentrations were associated with HDP, with borderline significance.

Changes in Bone Marrow Adipose Tissue in Transgender and Gender Non-Conforming Youth Undergoing Pubertal Suppression: A Pilot Study.

Pubertal suppression with gonadotropin-releasing hormone (GnRH) agonists in transgender and gender non-conforming (TGNC) youth may affect acquisition of peak bone mass. Bone marrow adipose tissue (BMAT) has an inverse relationship with bone mineral density (BMD). To evaluate the effect of pubertal suppression on BMAT, in this pilot study we prospectively studied TGNC youth undergoing pubertal suppression and cisgender control participants with similar pubertal status over a 12-month period. BMD was measured by dual-energy X-ray absorptiometry and peripheral quantitative computed tomography. Magnetic Resonance T1 relaxometry (T1-R) and spectroscopy (MRS) were performed to quantify BMAT at the distal femur. We compared the change in BMD, T1-R values, and MRS lipid indices between the two groups. Six TGNC (two assigned female and four assigned male at birth) and three female control participants (mean age 10.9 and 11.7 years, respectively) were enrolled. The mean lumbar spine BMD Z-score declined by 0.29 in the TGNC group, but increased by 0.48 in controls (between-group difference 0.77, 95% CI: 0.05, 1.45). Similar findings were observed with the change in trabecular volumetric BMD at the 3% tibia site (-4.1% in TGNC, +3.2% in controls, between-group difference 7.3%, 95% CI: 0.5%-14%). Distal femur T1 values declined (indicative of increased BMAT) by 7.9% in the TGNC group, but increased by 2.1% in controls (between-group difference 10%, 95% CI: -12.7%, 32.6%). Marrow lipid fraction by MRS increased by 8.4% in the TGNC group, but declined by 0.1% in controls (between-group difference 8.5%, 95% CI: -50.2%, 33.0%). In conclusion, we observed lower bone mass acquisition and greater increases in BMAT indices by MRI and MRS in TGNC youth after 12 months of GnRH agonists compared with control participants. Early changes in BMAT may underlie an alteration in bone mass acquisition with pubertal suppression, including alterations in mesenchymal stem cells within marrow.

Frequency drift in MR spectroscopy at 3T.

PURPOSE: Heating of gradient coils and passive shim components is a common cause of instability in the B0 field, especially when gradient intensive sequences are used. The aim of the study was to set a benchmark for typical drift encountered during MR spectroscopy (MRS) to assess the need for real-time field-frequency locking on MRI scanners by comparing field drift data from a large number of sites. METHOD: A standardized protocol was developed for 80 participating sites using 99 3T MR scanners from 3 major vendors. Phantom water signals were acquired before and after an EPI sequence. The protocol consisted of: minimal preparatory imaging; a short pre-fMRI PRESS; a ten-minute fMRI acquisition; and a long post-fMRI PRESS acquisition. Both pre- and post-fMRI PRESS were non-water suppressed. Real-time frequency stabilization/adjustment was switched off when appropriate. Sixty scanners repeated the protocol for a second dataset. In addition, a three-hour post-fMRI MRS acquisition was performed at one site to observe change of gradient temperature and drift rate. Spectral analysis was performed using MATLAB. Frequency drift in pre-fMRI PRESS data were compared with the first 5:20 minutes and the full 30:00 minutes of data after fMRI. Median (interquartile range) drifts were measured and showed in violin plot. Paired t-tests were performed to compare frequency drift pre- and post-fMRI. A simulated in vivo spectrum was generated using FID-A to visualize the effect of the observed frequency drifts. The simulated spectrum was convolved with the frequency trace for the most extreme cases. Impacts of frequency drifts on NAA and GABA were also simulated as a function of linear drift. Data from the repeated protocol were compared with the corresponding first dataset using Pearson's and intraclass correlation coefficients (ICC). RESULTS: Of the data collected from 99 scanners, 4 were excluded due to various reasons. Thus, data from 95 scanners were ultimately analyzed. For the first 5:20 min (64 transients), median (interquartile range) drift was 0.44 (1.29) Hz before fMRI and 0.83 (1.29) Hz after. This increased to 3.15 (4.02) Hz for the full 30 min (360 transients) run. Average drift rates were 0.29 Hz/min before fMRI and 0.43 Hz/min after. Paired t-tests indicated that drift increased after fMRI, as expected (p < 0.05). Simulated spectra convolved with the frequency drift showed that the intensity of the NAA singlet was reduced by up to 26%, 44 % and 18% for GE, Philips and Siemens scanners after fMRI, respectively. ICCs indicated good agreement between datasets acquired on separate days. The single site long acquisition showed drift rate was reduced to 0.03 Hz/min approximately three hours after fMRI. DISCUSSION: This study analyzed frequency drift data from 95 3T MRI scanners. Median levels of drift were relatively low (5-min average under 1 Hz), but the most extreme cases suffered from higher levels of drift. The extent of drift varied across scanners which both linear and nonlinear drifts were observed.

Maternal Urinary Organophosphate Esters and Alterations in Maternal and Neonatal Thyroid Hormones.

Production of organophosphate esters (OPEs), which represent a major flame-retardant class present in consumer goods, has increased over the past 2 decades. Experimental studies suggest that OPEs may be associated with thyroid hormone disruption, but few human studies have examined this association. We quantified OPE metabolites in the urine of 298 pregnant women from Cincinnati, Ohio, in the Health Outcomes and Measures of the Environment Study (enrolled 2003-2006) at 3 time points (16 and 26 weeks' gestation, and at delivery), and thyroid hormones in 16-week maternal and newborn cord sera. Urinary bis(1,3-dichloro-2-propyl)-phosphate concentrations were generally associated with decreased triiodothyronine and thyroxine levels and increased thyroid-stimulating hormone levels in maternal and newborn thyroid hormones in quartile dose-response analyses and multiple informant models. There was weaker evidence for thyroid hormone alterations in association with diphenyl-phosphate and di-n-butyl-phosphate. Bis-2-chloroethyl-phosphate was not associated with alterations in thyroid hormones in any analyses. We did not observe any evidence of effect modification by infant sex. These results suggest that gestational exposure to some OPEs may influence maternal and neonatal thyroid function, although replication in other cohorts is needed.