Longitudinal Variability Research Articles

Within-subject variation of C-reactive protein and high-sensitivity C-reactive protein: A systematic review and meta-analysis.

C-reactive protein (CRP) and high-sensitivity C-reactive protein (hsCRP) are measures of inflammation used in diagnosis, to guide treatment decisions, and in disease prediction. Variability in measured CRP and hsCRP may affect their clinical utility but estimates of within-subject variability are based on limited data. A systematic review and meta-analysis was performed to estimate longitudinal within-subject variability of CRP and hsCRP over any time period. Follow-up studies of any design in adults or children, with repeated measures of CRP or hsCRP were sought. Multiple databases were searched from inception to November 2022. Titles and abstracts were screened in duplicate. Full text screening and data extraction were performed by one reviewer and verified by a second. Risk of bias was assessed with a modified Consensus-based Standards for the Selection of Health Measurement Instruments (COSMIN) tool. Intraclass correlation coefficient (ICC) results were pooled with a meta-analysis and coefficient of variation (CV) results were described by median and range. Of 2675 studies identified, 60 met the inclusion criteria: 34 reported CRP and 26 reported hsCRP. For CRP, median CV was 0.41 (range 0.11 to 0.89), and the pooled estimate of ICC was 0.55 (95% CI 0.35 to 0.74). For hsCRP, median CV was 0.44 (range 0.27 to 0.76) and the pooled estimate of ICC was 0.62 (95% CI 0.58 to 0.67). Assessment of variability was not the main aim of many of the included papers, and it is possible that some relevant papers have been missed. Many of the papers included had low numbers of participants and/or low numbers of repeated measurements. Estimated within-subject variability is high for both CRP and hsCRP, but estimates are based on small numbers of participants and measurements. There is a need for better estimates of within-subject variability from analysis of larger numbers of repeated measurements in larger numbers of subjects.

The Southern Ocean Hub for Nutrients, Micronutrients, and Their Isotopes in the Global Ocean

The sustenance of marine primary productivity depends on the supply of macro- and micronutrients to photosynthesizers in the ocean’s sunlit surface. Without supply from the deep, sinking particles would deplete the upper ocean of these vital elements within decades. Over the last 20 years, it has been recognized that the Southern Ocean, where nutrient-rich deep waters are brought to the surface and the water masses that fill much of the upper ocean are formed, plays a pivotal role in replenishing upper-ocean nutrients. Photosynthesizers that grow and take up nutrients within the Southern Ocean circulation “hub” thus have an outsize influence on global-​scale distributions of macronutrients and many micronutrients. The GEOTRACES program has contributed observations of the concentration and stable isotope composition of “nutrient-​type” metals like zinc, cadmium, and nickel, within the Southern Ocean and beyond it, that are driving a sea change in our understanding of their marine cycles. Simultaneously, our understanding of Southern Ocean circulation has been refined, with recognition of the importance of longitudinal variability and subtropical overturning. Here, we aim to bring together these two strands of progress, review insights gained into marine micronutrient cycling, and consider the questions that remain to be resolved.

Geodesic shape regression based deep learning segmentation for assessing longitudinal hippocampal atrophy in dementia progression

Longitudinal hippocampal atrophy is commonly used as progressive marker assisting clinical diagnose of dementia. However, precise quantification of the atrophy is limited by longitudinal segmentation errors resulting from MRI artifacts across multiple independent scans. To accurately segment the hippocampal morphology from longitudinal 3T T1-weighted MR images, we propose a diffeomorphic geodesic guided deep learning method called the GeoLongSeg to mitigate the longitudinal variabilities that unrelated to diseases by enhancing intra-individual morphological consistency. Specifically, we integrate geodesic shape regression, an evolutional model that estimates smooth deformation process of anatomical shapes, into a two-stage segmentation network. We adopt a 3D U-Net in the first-stage network with an enhanced attention mechanism for independent segmentation. Then, a hippocampal shape evolutional trajectory is estimated by geodesic shape regression and fed into the second network to refine the independent segmentation. We verify that GeoLongSeg outperforms other four state-of-the-art segmentation pipelines in longitudinal morphological consistency evaluated by test–retest reliability, variance ratio and atrophy trajectories. When assessing hippocampal atrophy in longitudinal data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI), results based on GeoLongSeg exhibit spatial and temporal local atrophy in bilateral hippocampi of dementia patients. These features derived from GeoLongSeg segmentation exhibit the greatest discriminatory capability compared to the outcomes of other methods in distinguishing between patients and normal controls. Overall, GeoLongSeg provides an accurate and efficient segmentation network for extracting hippocampal morphology from longitudinal MR images, which assist precise atrophy measurement of the hippocampus in early stage of dementia.

Longitudinal Evaluation of Biomechanical Indices in Fellow Eyes of Patients With Keratoconus Classified as Having Very Asymmetric Ectasia With Normal Topography.

To evaluate the biomechanical longitudinal variability and progression of tomographically normal fellow eyes of patients with keratoconus. Of 513 patients with keratoconus, 30 patients with tomographically normal fellow eyes were included in this study. Tomographic and biomechanical parameters of the Pentacam and Corvis ST (Oculus Optikgeräte GmbH) were analyzed in multiple follow-up examinations, including the ABCD grading, Belin/Ambrósio Enhanced Ectasia total deviation index (BAD-D), Corvis Biomechanical Index (CBI), Corvis Biomechanical Factor (CBiF), and Tomographic and Biomechanical Index (TBI). A mixed regression model was applied. The results were compared to a healthy control group (n = 17) and a keratoconus group (n = 20). Within a maximum observation period of 3.3 years, no fellow eye (0%) showed a progression to tomographically evident keratoconus. No significant change in tomographic or biomechanical parameters was detected over the study period. The indices BAD-D, CBI, CbiF, and TBI exhibited a certain variability over time, whereas the tomographic ABC parameters and maximum keratometry barely changed. This was also shown in the control group and for all parameters in the keratoconus group, except the TBI. During the observation period none of the normal fellow eyes progressed to tomographically detectable keratoconus. However, biomechanical parameters CBI, CbiF, and TBI showed pathological values in 43.3% of eyes and certain variability. Subsequent studies with a longer observation period are warranted to confirm the biomechanical trends seen in this study and to rate the ability of single measurements to diagnose early keratoconus. [J Refract Surg. 2024;40(1):e48-e56.].

Ionospheric longitudinal variability in the Northern Hemisphere during magnetic storms in March 2012 from ionosonde and GPS/GLONASS data

A comprehensive study of spatio-temporal variations of geomagnetic, ionospheric, and atmospheric parameters in the middle and high latitudes of the Northern Hemisphere during a series of magnetic storms in March 2012 has been expanded by including vertical total electronic content (TEC) data from measurements at the chains of dual-frequency phase receivers GPS/GLONASS in the analysis. The features of longitudinal variations in ionosphere ionization over mid-latitude Eurasia, found earlier from vertical sounding data, are confirmed by vertical TEC data. We emphasize the complex physics of the long magnetically disturbed period in March 2012 with switching between positive and negative effects of an ionospheric storm during the same magnetic storm phases for spaced mid-latitude regions of the Eastern Hemisphere. Such changes in the ionospheric storm effects might have been caused by the superposition of competing processes in the mid-latitude region of the Eastern Hemisphere due to variations in the thermospheric composition, thermospheric winds, and large-scale electric fields affecting ionospheric ionization. We have observed significant differences in the nature of the ionospheric ionization reaction between the Eastern and Western hemispheres to the prolonged geomagnetic disturbance in March 2012. According to TEC data, there was an effect of reduced ionization of the ionosphere at longitudes of the Western Hemisphere, unlike the Eastern one. The effect of a negative ionospheric storm was caused by the formation of vast areas of atmospheric gas with a reduced density ratio [O]/[N2] over the mid-latitude region of the Western Hemisphere in the zone of maximum penetration of geomagnetic disturbances from high latitudes to middle latitudes. According to the INTERMAGNET magnetometer chain data for the analyzed period of magnetic storms on March 7–20, 2012, at midlatitudes of the Northern Hemisphere the maximum geomagnetic field variations were observed in the Western Hemisphere.

Gender differences in midlife to later-life cumulative burden and variability of obesity measures and risk of all-cause and cause-specific mortality.

Previous studies have reported the gender-specific association between general and central obesity measures, using snapshot assessments, and mortality events. This study seeks to further explore this link by examining how the longitudinal cumulative burden and variability of obesity measures from midlife to later-life impact mortality events in the Atherosclerosis Risk in Communities (ARIC) study population, specifically in relation to gender differences. Using data from the ARIC study, a total of 7615 (4360 women) participants free of cardiovascular disease, cancer, and early mortality events were included in the data analysis. Longitudinal cumulative burden (estimated by the area under the curve (AUC) using a quadratic mixed-effects method) and variability (calculated according to average successive variability (ASV)) were considered as exposures, separately and all together. Cox proportional hazard regression models were used to estimate multivariable-adjusted standardized hazard ratios. The mean age was 62.4 and the median follow-up was 16.9 years. In men, AUCs of waist-related obesity measures, and also ASVs of all obesity measures were associated with increased all-cause mortality risk. In women, waist circumference and waist-to-height ratio AUCs were associated with increased all-cause mortality risk. Regarding cardiovascular mortality, all adiposity measures ASVs in both genders and waist-related obesity measures AUCs in men were associated with increased risk. Significant gender differences were found for the associations between cumulative and variability of waist-to-hip ratio for all-cause mortality and all adiposity measures ASVs for cardiovascular mortality risk with higher impact among men. Cumulative burden and variability in general and central obesity measures were associated with higher all-cause and cardiovascular mortalities among men. In women, general obesity measures variability, as well as cumulative and variability of central adiposity measure, increased all-cause mortality risk.

Trajectories of adaptive functioning from early childhood to adolescence in autism: Identifying turning points and key correlates of chronogeneity.

Previous research has demonstrated heterogeneous adaptive outcomes across the autism spectrum; however, the current literature remains limited in elucidating turning points and associated factors for longitudinal variability (chronogeneity). To address these empirical gaps, we aimed to provide a finer-grained characterization of trajectories of adaptive functioning from early childhood to adolescence in autism. Our sample (N=406) was drawn from an inception cohort of children diagnosed Autistic at ages 2-5. Adaptive functioning was assessed with Vineland Adaptive Behavior Scales (VABS, 2nd Edition) across 6 visits from the time of diagnosis by age 18. Parallel-process latent growth curve modeling were used to estimate domain-level VABS trajectories, followed by latent class growth analysis to identify trajectory subgroups. Child characteristics at diagnosis, family demographics, and participation outcomes at adolescence were compared across subgroups. Piecewise latent growth models best described VABS trajectories with two turning points identified at around ages 5-6 and 9-10, respectively reflecting transitions into school age and early adolescence. We parsed four VABS trajectory subgroups that vary by level of functioning and change rate for certain domains and periods. Around 16% of the sample exhibited overall adequate functioning (standard score >85) with notable early growth and social adaptation during adolescence. About 21% showed low adaptive functioning (standard score ≤70), with decreasing slopes by age 6 followed by improvements in communication and daily-living skills by age 10. The other two subgroups (63% in total) were characterized by adaptive functioning between low and adequate levels, with relatively stable trajectories entering school age. These subgroups differed most in their cognitive ability at diagnosis, household income, and social participation in adolescence. We identified key individual and family characteristics and time windows associated with distinct adaptive functioning trajectories, which have important implications for providing timely and tailored supports to Autistic people across developmental stages.

Longitudinal clustering of Life's Essential 8 health metrics: application of a novel unsupervised learning method in the CARDIA study.

Changes in cardiovascular health (CVH) during the life course are associated with future cardiovascular disease (CVD). Longitudinal clustering analysis using subgraph augmented non-negative matrix factorization (SANMF) could create phenotypic risk profiles of clustered CVH metrics. Life's Essential 8 (LE8) variables, demographics, and CVD events were queried over 15ears in 5060 CARDIA participants with 18years of subsequent follow-up. LE8 subgraphs were mined and a SANMF algorithm was applied to cluster frequently occurring subgraphs. K-fold cross-validation and diagnostics were performed to determine cluster assignment. Cox proportional hazard models were fit for future CV event risk and logistic regression was performed for cluster phenotyping. The cohort (54.6% female, 48.7% White) produced 3 clusters of CVH metrics: Healthy & Late Obesity (HLO) (29.0%), Healthy & Intermediate Sleep (HIS) (43.2%), and Unhealthy (27.8%). HLO had 5 ideal LE8 metrics between ages 18 and 39years, until BMI increased at 40. HIS had 7 ideal LE8 metrics, except sleep. Unhealthy had poor levels of sleep, smoking, and diet but ideal glucose. Race and employment were significantly different by cluster (P < .001) but not sex (P = .734). For 301 incident CV events, multivariable hazard ratios (HRs) for HIS and Unhealthy were 0.73 (0.53-1.00, P = .052) and 2.00 (1.50-2.68, P < .001), respectively versus HLO. A 15-year event survival was 97.0% (HIS), 96.3% (HLO), and 90.4% (Unhealthy, P < .001). SANMF of LE8 metrics identified 3 unique clusters of CVH behavior patterns. Clustering of longitudinal LE8 variables via SANMF is a robust tool for phenotypic risk assessment for future adverse cardiovascular events.

Cognitive status modifies the association between elevated cerebrospinal fluid phosphorylated tau and longitudinal cerebral blood flow changes over a 7‐year follow‐up period

AbstractBackgroundAlzheimer’s disease (AD) is associated with widespread reductions in cerebral blood flow (CBF), but the precise link between AD pathology and CBF is unclear. Hypoperfusion is a late‐stage event associated with tau, but not amyloid, pathology. We assessed whether baseline levels of cerebrospinal fluid (CSF)‐measured core AD pathology [β‐amyloid42 (Aβ42), phosphorylated tau (p‐tau)] related to CBF changes over a 7‐year period and if associations differed by baseline cognitive status.MethodsVanderbilt Memory and Aging Project participants free of clinical dementia or stroke at enrollment [n = 153, 72±6 years, 37% mild cognitive impairment (MCI), 33% female] underwent lumbar puncture to obtain CSF at study entry and serial multimodal 3T brain magnetic resonance imaging over a 7‐year follow‐up period (mean = 5.4 years). Enzyme‐linked immunosorbent assays quantified CSF Aβ42 and p‐tau181. Pseudo‐continuous arterial spin‐labeling captured CBF in total and lobar grey matter regions of interest (ROI). Linear mixed‐effects models related biomarker x time to longitudinal CBF variables adjusting for baseline age and cognitive status, sex, race/ethnicity, education, Framingham Stroke Risk Profile (minus age), ROI volume, APOE‐ε4 status, and time. Models were repeated testing a cognitive status x biomarker x time interaction term.ResultsIn main effect models, neither CSF Aβ42 (p‐values&gt;0.31) nor p‐tau (p‐values&gt;0.28) related to CBF outcomes. Aβ42 did not interact with cognitive status on CBF outcomes (p‐values&gt;0.37), but p‐tau interacted with baseline cognitive status on longitudinal frontal, temporal, parietal, occipital, and total grey matter CBF trajectories (p‐values&lt;0.005). Among cognitively unimpaired participants, higher p‐tau was associated with slower reductions in frontal (β = 0.01, p = 0.03), parietal (β = 0.01, p = 0.04), occipital (β = 0.01, p = 0.02), and total grey matter CBF (β = 0.01, p = 0.04). Among individuals with MCI, higher p‐tau was associated with faster reductions in frontal (β = ‐0.03, p = 0.008), temporal (β = ‐0.02, p = 0.01), parietal (β = ‐0.04, p = 0.002), occipital (β = ‐0.03, p = 0.0004), and total grey matter CBF (β = ‐0.03, p = 0.005).ConclusionsResults linking CSF p‐tau to CBF decline in MCI support the theory that grey matter hypoperfusion is a later‐stage event tied to neurofibrillary tangle pathology. Associations between higher p‐tau and slower rates of CBF decline among cognitively unimpaired individuals warrants further investigation. Funding: R01‐AG034962, K24‐AG046373, F31‐AG079640, T32‐AG058524, P20‐AG068082

Higher baseline cerebrospinal fluid platelet‐derived growth factor receptor‐β levels are associated with slower reductions in longitudinal cerebral blood flow among cognitively unimpaired individuals

AbstractBackgroundCerebrospinal fluid (CSF) platelet‐derived growth factor receptor‐β (PDGFR‐β) is an emerging biomarker hypothesized to reflect pericyte damage. CSF PDGFR‐β levels are increased in Alzheimer’s disease (AD), but it is unknown if pericyte injury contributes to cerebrovascular changes common in AD. We examined if PDGFR‐β relates to changes in cerebral blood flow (CBF) and white matter injury in older adults.MethodVanderbilt Memory and Aging Project participants (n = 152, baseline age 73±7 years, 43% mild cognitive impairment [MCI]) underwent baseline lumbar puncture and serial brain MRI over a 7‐year (5.5±2.3) follow‐up period. CSF PDGFR‐β concentrated was measured by immunoassay. MRI included pseudo‐continuous arterial spin‐labeling to quantify CBF, T2‐FLAIR to quantify white matter hyperintensities (WMHs), and diffusion tensor imaging (DTI) to quantify white matter microstructural integrity. Ordinary least squares regression and linear mixed‐effects models related baseline CSF PDGFR‐β to baseline and longitudinal whole brain and regional CBF, WMH, and DTI metrics, adjusting for baseline demographic variables, modified Framingham Stroke Risk Profile, cognitive status, apolipoprotein‐ε4 status, and regional tissue volume. Longitudinal models adjusted for follow‐up time. Follow‐up models tested a CSF PDGFR‐β x cognitive status interaction.ResultIn main models, higher CSF PDGFR‐β levels were cross‐sectionally associated with lower mean diffusivity (indicating better microstructural integrity) in white matter tracts in the frontal, temporal, parietal, and occipital lobes (p‐values&lt;0.03) but were unrelated to WMHs (p‐values&gt;0.18). Models investigating associations with longitudinal white matter variables were null (p‐values&gt;0.35). CSF PDGFR‐β cross‐sectionally interacted with diagnosis on temporal lobe CBF, such that higher CSF PDGFR‐β related to higher CBF in MCI participants (p = 0.01). In longitudinal models, baseline CSF PDGFR‐β interacted with cognitive status on whole brain and occipital lobe CBF (p‐values&lt;0.04). Stratified models showed higher baseline CSF PDGFR‐β related slower reductions in longitudinal CBF in whole brain (p = 0.02), frontal (p = 0.04), temporal (p = 0.02), parietal (p = 0.01), and occipital lobes (p = 0.008) in cognitively unimpaired participants only.ConclusionAmong older adults, higher CSF PDGFR‐β levels are associated with better white matter integrity at study entry and predict slower reductions in longitudinal CBF among cognitively unimpaired individuals. Results suggest that higher PDGFR‐β levels may reflect subclinical vascular remodeling that provides protection to future cerebrovascular compromise.

A Centiloid cut‐off to help predict true amyloid accumulation

AbstractBackgroundLongitudinal amyloid‐PET‐based imaging endpoints are often included in clinical trials, where they provide critical evidence of treatment efficacy. We assessed the longitudinal variability of the Centiloid (CL) scale and its ability to predict early amyloid accumulation.MethodLongitudinal amyloid‐PET with 18F‐flutemetamol and 18F‐florbetaben (N = 711 at follow‐up 1, time interval: 3.0±1.3years; N = 98 at follow‐up 2, 5.1±0.7years) was conducted in predominantly cognitively unimpaired participants of the AMYPAD Prognostic and Natural History Study (PNHS). Quantification of scans was performed using the MR‐based CL pipeline and visually read (VR) by local certified assessors. Longitudinal change in CL was modelled using linear mixed effects models corrected for age, sex, and education. Subjects were classified based on VR status over time (i.e., Stable VR‐, VR‐Converters or Stable VR+) and based on amyloid accumulation over time (i.e., Accumulator or Non‐accumulator; Accumulator if annualised rate of change [ARC]&gt;95th percentile of a subset of 110 subjects for whom no amyloid accumulation was expected, selection criteria in Table‐1). Finally, a baseline CL cut‐off to identify future Accumulators was derived from a receiver operating characteristic (ROC) curve analysis.ResultParticipants had a median age of 65 years, 57% were females, 41% were APOE‐ε4 carriers, and 18% had baseline VR+ scan (Table‐1). The 95th percentile of ARC in the stable 110 subjects was 2.7 CL/year. Baseline CL was higher in Accumulators compared to Non‐accumulators (24.8 versus 8.7 CL, p&lt;.001). Based on the ROC analysis, the optimal baseline CL threshold to predict Accumulators was 12.1 CL (sensitivity = 67% and specificity = 74%, Figure B‐C‐D), in line with the baseline CL threshold that best predicts VR conversion (11.6 CL). Baseline CL values and ARC were also higher in Stable VR+ and VR‐Converters compared to Stable VR‐ (p&lt;.001), but no significant difference in ARC was found between VR‐Converters and Stable VR+ (Table‐1, Figure A). Results were robust across tracers.ConclusionIncreases over 2.7 CL/year can be indicative of reliable amyloid accumulation. A CL cut‐off of ∼12CL can help identify subjects more likely to accumulate amyloid in the short future. Annualised rates of change in amyloid deposition where highly comparable across tracers.

A Centiloid cut‐off to help predict true amyloid accumulation

AbstractBackgroundLongitudinal amyloid‐PET‐based imaging endpoints are often included in clinical trials, where they provide critical evidence of treatment efficacy. We assessed the longitudinal variability of the Centiloid (CL) scale and its ability to predict early amyloid accumulation.MethodLongitudinal amyloid‐PET with 18F‐flutemetamol and 18F‐florbetaben (N = 711 at follow‐up 1, time interval: 3.0±1.3years; N = 98 at follow‐up 2, 5.1±0.7years) was conducted in predominantly cognitively unimpaired participants of the AMYPAD Prognostic and Natural History Study (PNHS). Quantification of scans was performed using the MR‐based CL pipeline and visually read (VR) by local certified assessors. Longitudinal change in CL was modelled using linear mixed effects models corrected for age, sex, and education. Subjects were classified based on VR status over time (i.e., Stable VR‐, VR‐Converters or Stable VR+) and based on amyloid accumulation over time (i.e., Accumulator or Non‐accumulator; Accumulator if annualised rate of change [ARC]&gt;95th percentile of a subset of 110 subjects for whom no amyloid accumulation was expected, selection criteria in Table‐1). Finally, a baseline CL cut‐off to identify future Accumulators was derived from a receiver operating characteristic (ROC) curve analysis.ResultParticipants had a median age of 65 years, 57% were females, 41% were APOE‐e4 carriers, and 18% had baseline VR+ scan (Table‐1). The 95th percentile of ARC in the stable 110 subjects was 2.7 CL/year. Baseline CL was higher in Accumulators compared to Non‐accumulators (24.8 versus 8.7 CL, p&lt;.001). Based on the ROC analysis, the optimal baseline CL threshold to predict Accumulators was 12.1 CL (sensitivity = 67% and specificity = 74%, Figure B‐C‐D), in line with the baseline CL threshold that best predicts VR conversion (11.6 CL). Baseline CL values and ARC were also higher in Stable VR+ and VR‐Converters compared to Stable VR‐ (p&lt;.001), but no significant difference in ARC was found between VR‐Converters and Stable VR+ (Table‐1, Figure A). Results were robust across tracers.ConclusionIncreases over 2.7 CL/year can be indicative of reliable amyloid accumulation. A CL cut‐off of ∼12CL can help identify subjects more likely to accumulate amyloid in the short future. Annualised rates of change in amyloid deposition where highly comparable across tracers.

Patterns of Emergency Department Utilization of Adolescents and Young Adults Living with Sickle Cell in North Carolina

Introduction Despite improvements in survival and life expectancy in high-income countries, adolescent and young adults (AYAs) living with sickle cell disease (SCD) remain at significant risk for morbidity and mortality. Our previous work using group-based trajectory modeling (GTBM) has demonstrated that a subset of AYAs with SCD are at risk for significant morbidity with SCD with increased healthcare utilization, vasoocclusive pain episodes, and death, particularly after transition to adult care (Kayle et al., Pediatric Blood &amp; Cancer, 2018). Yet, there is limited insight into which AYAs with SCD are at high risk for poor outcomes. The objective of this study was to identify patterns of emergency department (ED) utilization of AYAs with SCD in NC as individuals age, with the overarching future goal of better predicting which individuals are at highest risk. Methods The NC Hospital Discharge Data (Inpatient, Ambulatory Surgery/Outpatient and ED Databases), 2013-2020 were used. A patient with SCD was defined as a patient who had three or more visits (inpatient, ambulatory surgery, or ED visit) with International Classification of Diseases 9 or 10 codes for SCD within a 5-year period. Date of birth, sex and zip code were used to link patients across the datasets. Demographic variables including age, sex, race and ethnicity were extracted from visit level data. ED visits were defined as all ambulatory ED visits and ED visits that resulted in inpatient admissions between 2013 and 2019. Descriptive statistics were calculated to describe the sample characteristics such as sex and race. The number of ED visits per year across age 14 to 30 years were used as longitudinal variables in GTBM to identify trajectories of latent classes of ED visits. The number of latent classes was determined by model-fit indices such as BIC and entropy. Results There were 3643 participants included in this study. Participants were 95% Black or African American, 53.8% female and 46.2% male, with age ranging from 14-30 years. Our GTBM models identified three latent classes of participants (Figure 1) with similar ED utilization at age 14 years, but divergent predicted annual ED trajectories beginning around 18 years of age. Class 1, comprising 95% of the participants (n=3463) and 54% female, had minimal ED utilization throughout all the years. Class 2, comprising 4% of the participants (n=150) and 51.3% female, had a steady increase in annual ED use, peaking at 20 visits per year by at age 24 years. Class 3, comprising 0.8% of the participants (n=30) and 47% female, had more extreme ED utilization, up to 80 visits per year at age 24 years. Conclusions Using a large longitudinal dataset, we demonstrate distinct patterns of ED utilization among AYAs with SCD by age. Although ED utilization was similar between the groups at age 14 years, a small proportion (~5%) of individuals accounted for the highest ED utilization at older ages. Additionally, amongst the patients with highest ED utilization, there was a marked increase in ED utilization from age 18-24 years, highlighting the highest risk ages within the AYA population, when there is often a transition from pediatric to adult care. There were slightly more males in the highest ED utilization class, highlighting potential gender differences in ED utilization. Our next steps will be to assess which clinical and social factors predict ED utilization trajectory. Limitations in this study include the use of hospital discharge datasets, which may have limited the capture of patients with infrequent acute care visits and limited our ability to examine the role of SCD genotypes and disease-modifying therapy use on ED utilization patterns. Despite these limitations, the study identifies distinct classes of ED use among AYAs with SCD. Understanding and predicting patterns of ED utilization are critical in targeting early interventions towards individuals with the highest risk of healthcare utilization and disproportionate risk for future morbidity.

Cortical thickness modeling and variability in Alzheimer’s disease and frontotemporal dementia

Background and objectiveAlzheimer’s disease (AD) and frontotemporal dementia (FTD) show different patterns of cortical thickness (CTh) loss compared with healthy controls (HC), even though there is relevant heterogeneity between individuals suffering from each of these diseases. Thus, we developed CTh models to study individual variability in AD, FTD, and HC.MethodsWe used the baseline CTh measures of 379 participants obtained from the structural MRI processed with FreeSurfer. A total of 169 AD patients (63 ± 9 years, 65 men), 88 FTD patients (64 ± 9 years, 43 men), and 122 HC (62 ± 10 years, 47 men) were studied. We fitted region-wise temporal models of CTh using Support Vector Regression. Then, we studied associations of individual deviations from the model with cerebrospinal fluid levels of neurofilament light chain (NfL) and 14–3-3 protein and Mini-Mental State Examination (MMSE). Furthermore, we used real longitudinal data from 144 participants to test model predictivity.ResultsWe defined CTh spatiotemporal models for each group with a reliable fit. Individual deviation correlated with MMSE for AD and with NfL for FTD. AD patients with higher deviations from the trend presented higher MMSE values. In FTD, lower NfL levels were associated with higher deviations from the CTh prediction. For AD and HC, we could predict longitudinal visits with the presented model trained with baseline data. For FTD, the longitudinal visits had more variability.ConclusionWe highlight the value of CTh models for studying AD and FTD longitudinal changes and variability and their relationships with cognitive features and biomarkers.

Accurate prediction of HCC risk after SVR in patients with hepatitis C cirrhosis based on longitudinal data

BackgroundMost existing predictive models of hepatocellular carcinoma (HCC) risk after sustained virologic response (SVR) are built on data collected at baseline and therefore have limited accuracy. The current study aimed to construct an accurate predictive model incorporating longitudinal data using a novel modeling strategy. The predictive performance of the longitudinal model was also compared with a baseline model.MethodsA total of 400 patients with HCV-related cirrhosis who achieved SVR with direct-acting antivirals (DAA) were enrolled in the study. Patients were randomly divided into a training set (70%) and a validation set (30%). Informative features were extracted from the longitudinal variables and then put into the random survival forest (RSF) to develop the longitudinal model. A baseline model including the same variables was built for comparison.ResultsDuring a median follow-up time of approximately 5 years, 25 patients (8.9%) in the training set and 11 patients (9.2%) in the validation set developed HCC. The areas under the receiver-operating characteristics curves (AUROC) for the longitudinal model were 0.9507 (0.8838–0.9997), 0.8767 (0.6972,0.9918), and 0.8307 (0.6941,0.9993) for 1-, 2- and 3-year risk prediction, respectively. The brier scores of the longitudinal model were also relatively low for the 1-, 2- and 3-year risk prediction (0.0283, 0.0561, and 0.0501, respectively). In contrast, the baseline model only achieved mediocre AUROCs of around 0.6 (0.6113, 0.6213, and 0.6480, respectively).ConclusionsOur longitudinal model yielded accurate predictions of HCC risk in patients with HCV-relate cirrhosis, outperforming the baseline model. Our model can provide patients with valuable prognosis information and guide the intensity of surveillance in clinical practice.

112 Definition and Genetic Parameters Estimation for Climatic Resilience Indicators Derived from Longitudinal Vaginal Temperature Records in Lactating Sows Under Heat Stress Conditions

Abstract Climatic resilience (CR) can be defined as the ability of an animal to maintain euthermia under thermally stressful conditions. However, CR based on longitudinal physiological variables has not been included in current swine genetic selection schemes. This is the first study deriving novel CR indicators using automatically recorded vaginal temperatures (TV) and estimating genetic parameters for them. Therefore, the study objectives were to: 1) derive novel indicators of CR based on automatically recorded vaginal temperature (TV); and 2) estimate (co)variance components and genetic parameters for these indicators. TV of 1,381 lactating sows (parities 2 to 7; Landrace × Large White) were measured automatically every 10 minutes from June 5th, 2021 to July 30th, 2021 using a vaginally implanted thermochron data recorder. A total of 1,639 F1 (Landrace x Large White) sows within the studied population were genotyped using the PorcineSNP50K Bead Chip. A total of 18 HS indicators derived from Tv, their definitions, heritability estimates, and mean genomic estimated breeding value (GEBV) accuracies are shown in Table 1. Multi-trait animal models were fitted for estimating the (co)variance components, considering the effects of week of data collection, parity, and location (barn and room within barn) as fixed. Heritability estimates for these resilience traits ranged from 0.0004 ± 7.16E-6 (SlopeDe) to 0.291 ± 0.047 (HSAB). Most traits derived directly from vaginal temperature (e.g., Duration: 0.201 ± 0.033, MaxTv: 0.203 ± 0.032) are moderately heritable with substantial additive genetic variance. Besides, LnVar(Med) that is derived from the deviation between the real and predicted value had moderate heritability (0.196 ± 0.041) and high mean GEBV accuracy (0.616 ± 0.042). The indicators defined based on the slope of Tv and time had low heritability estimates, which are close to zero, and low mean GEBV accuracy (0.106 ± 0.050 – 0.337 ± 0.065). Among all the CR indicators, Duration, MaxTv, Nor_medvar, Nor_avevar, HSAA, and HSAB outperformed the other indicators with moderate heritability estimates and high average GEBV accuracies. Genetic correlations were estimated among the six indicators with moderate heritability and high mean GEBV accuracy. Moderate to high positive genetic correlations (0.508 ± 0.034 to 0.992 ± 0.025) were estimated among Duration, MaxTv, Nor_medvar, HSAB, and HSAA. LnVar(Med) showed low negative genetic correlations (-0.218 ± 0.028 to -0.039 ± 0.018) with the traits Duration, MaxTv, HSAA, and HSAB, except with Nor_medvar, which was positive (0.156 ± 0.022). Thus, Duration, LnVar(Med), MaxTv, Nor_medvar, HSAB, and HSAA might be promising CR indicator and contribute toward improving lactating sows’ resilience to thermal stress through genetic selection. New studies about the relationship between these new indicators and other economic important traits in pigs are required before including them in selection indexes.

A Hybrid Decision Tree and Deep Learning Approach Combining Medical Imaging and Electronic Medical Records to Predict Intubation Among Hospitalized Patients With COVID-19: Algorithm Development and Validation.

Early prediction of the need for invasive mechanical ventilation (IMV) in patients hospitalized with COVID-19 symptoms can help in the allocation of resources appropriately and improve patient outcomes by appropriately monitoring and treating patients at the greatest risk of respiratory failure. To help with the complexity of deciding whether a patient needs IMV, machine learning algorithms may help bring more prognostic value in a timely and systematic manner. Chest radiographs (CXRs) and electronic medical records (EMRs), typically obtained early in patients admitted with COVID-19, are the keys to deciding whether they need IMV. We aimed to evaluate the use of a machine learning model to predict the need for intubation within 24 hours by using a combination of CXR and EMR data in an end-to-end automated pipeline. We included historical data from 2481 hospitalizations at The Mount Sinai Hospital in New York City. CXRs were first resized, rescaled, and normalized. Then lungs were segmented from the CXRs by using a U-Net algorithm. After splitting them into a training and a test set, the training set images were augmented. The augmented images were used to train an image classifier to predict the probability of intubation with a prediction window of 24 hours by retraining a pretrained DenseNet model by using transfer learning, 10-fold cross-validation, and grid search. Then, in the final fusion model, we trained a random forest algorithm via 10-fold cross-validation by combining the probability score from the image classifier with 41 longitudinal variables in the EMR. Variables in the EMR included clinical and laboratory data routinely collected in the inpatient setting. The final fusion model gave a prediction likelihood for the need of intubation within 24 hours as well. At a prediction probability threshold of 0.5, the fusion model provided 78.9% (95% CI 59%-96%) sensitivity, 83% (95% CI 76%-89%) specificity, 0.509 (95% CI 0.34-0.67) F1-score, 0.874 (95% CI 0.80-0.94) area under the receiver operating characteristic curve (AUROC), and 0.497 (95% CI 0.32-0.65) area under the precision recall curve (AUPRC) on the holdout set. Compared to the image classifier alone, which had an AUROC of 0.577 (95% CI 0.44-0.73) and an AUPRC of 0.206 (95% CI 0.08-0.38), the fusion model showed significant improvement (P<.001). The most important predictor variables were respiratory rate, C-reactive protein, oxygen saturation, and lactate dehydrogenase. The imaging probability score ranked 15th in overall feature importance. We show that, when linked with EMR data, an automated deep learning image classifier improved performance in identifying hospitalized patients with severe COVID-19 at risk for intubation. With additional prospective and external validation, such a model may assist risk assessment and optimize clinical decision-making in choosing the best care plan during the critical stages of COVID-19.

Optimal systolic and diastolic blood pressure threshold that associated with lower risk of white matter hyperintensity progression.

The optimal control thresholds for systolic blood pressure (SBP) and diastolic blood pressure (DBP) in patients with white matter hyperintensity (WMH) are still unclear. A longitudinal retrospective study of patients with brain magnetic resonance imaging (MRI) scans with intervals of more than 3 years was conducted. Blood pressure records during hospitalization and from outpatient visits between baseline and the last MRI scan were collected. The outcome was the change in total WMH from baseline to the final visit. Among the 965 patients with MRI scans, 457 patients with detailed longitudinal blood pressure records were ultimately included and classified into the WMH absent group (n = 121), mild WMH group (n = 126), and moderate to severe WMH group (n = 210). Both baseline and longitudinal mean SBP, DBP, and SBP SD were significantly associated with WMH severity (p < 0.05). An average SBP of 130-140 mmHg [vs. <130 mmHg, aOR, 1.80, (95% CI, 1.05-3.07), p = 0.03] was associated with a higher risk of WMH progression. DBP ≥ 90 mmHg [vs. <80 mmHg, OR, 1.81, (95% CI, 0.88-3.74), p = 0.02, aOR, 1.54, (95% CI, 0.66-3.53), p = 0.32] was associated with a higher risk of WMH progression, but was not after adjusted for other covariates. Longitudinal BP variability was not significantly associated with WMH progression. Both SBP and DBP had a stronger relationship with the severity of WMH. A target mean SBP of <130 mmHg and mean DBP of <80 mmHg was associated with a lower risk of WMH progression.

New road surface noise corrections for New Zealand

New Zealand’s transport agency has developed a close-proximity (CPX) noise trailer and primarily operates it as a road surface research tool. Early CPX survey data revealed very high longitudinal variability in the performance of porous asphalts on the highway network (±5 dB). Further study and optimization of porous asphalt mixes has resulted in high-performance low-noise surfaces that are reliably 5 dB quieter than standard NZ porous asphalt. The new surfaces and improved surface data inspired a recalibration of NZ’s implementation of the CRTN road-traffic noise model, for which a novel method based on CPX survey data and individual vehicle pass-by measurements was employed. The relationship between CPX and wayside levels was more complex than anticipated; CPX measurements did not capture all the surface attributes/interactions contributing to the wayside level (such as porosity, directivity). The methodology proved successful and a new table of surface corrections has been published, embodying the recalibration of CRTN for NZ in 2023. CRTN predictions using the new corrections have performed well when evaluated against independent road-traffic noise monitoring data from NZ highways.

Longitudinal variability of thermospheric zonal winds near dawn and dusk

Understanding the morphology and dynamics of the thermosphere is key to understanding the Earth’s upper atmosphere as a whole. Thermospheric winds play an important role in this process by transporting momentum and energy and affecting the composition, dynamics and morphology of not only the thermosphere but also of the ionosphere. The general morphology of the winds has been well established over the past decades, but we are only starting to understand its variability. In this process the lower atmosphere plays an important role due to direct penetration of waves from the lower atmosphere into the ionosphere/thermosphere, secondary waves generated on the way, or internal feedback mechanisms in the coupled ionosphere-thermosphere system. Therefore, knowledge about thermospheric variability and its causes is critical for an improved understanding of the global ionosphere-thermosphere system and its coupling to the lower atmosphere. We have used low-to mid-latitude zonal wind observations obtained by the Gravity Field and Steady-State Ocean Explorer (GOCE) satellite near 260 km altitude during geomagnetically quiet times to investigate the interannual and spatial zonal wind variability near dawn and dusk, during December solstice. The temporal and spatial variability is presented as a variation about the zonal mean values and decomposed into its underlying wavenumbers using a Fourier analysis. The obtained wave features are compared between different years and clear interannual changes are observed in the individual wave components, which appear to align with changes in the solar flux but do not correlate with variations in either El Niño Southern Oscillation or the Quasi Biennial Oscillation. The obtained wave features are compared and contrasted with results from the Climatological Tidal Model of the Thermosphere (CTMT) and revealed a very good agreement between CTMT and the 2009 and 2010 December GOCE zonal wind perturbations at dawn. However, during dusk, the CTMT zonal wind perturbations and in particular the zonal wave-1 component show significant differences with those observed by GOCE.