Martin S. Angst

Self-reported pain remains the clinical standard for assessing musculoskeletal pain severity, but its subjectivity limits objective diagnosis, monitoring, and therapeutic development. We present LENS (Latent sEMG Neuromuscular Signature), which decodes exertion-evoked musculoskeletal pain intensity from surface electromyography (sEMG) alone. LENS was pre-trained via a cross-modal self-supervised objective that reconstructs muscle oxygenation from sEMG in 183 adults. LENS was trained exclusively on healthy muscle physiology and generalized to an independent healthy cohort (Spearman's ρ=0.52) and discriminated moderate-to-severe pain (AUROC=0.83). Using a label-free test-time adaptation, LENS generalized to an unseen chronic musculoskeletal pain cohort—myogenous temporomandibular disorder (mTMD)—with comparable performance (ρ=0.61; AUROC=0.86). Reverse transfer, from mTMD to controls, was substantially weaker, an asymmetry attributable to elevated motor variability in chronic pain; excluding highly variable mTMD participants recovered performance, indicating a conserved pain signature masked by pathology-specific motor noise. LENS offers a scalable, physiological correlate of musculoskeletal pain from a single sEMG channel.

High-impact chronic pain (HICP) describes pain that interferes substantially with daily functioning and is associated with diminished quality of life, yet brief tools to identify and track HICP in adolescents are lacking. The 3-item Pain, Enjoyment, and General Activity (PEG) scale derived from the Brief Pain Inventory offers a promising rapid measure, but its psychometric properties and an empirically derived HICP cut-off score based on PEG have not been established in pediatric chronic pain. The present study examined convergent validity, classification accuracy, and a PEG cut-off for HICP in 263 adolescents (10-18 years; 84.4% female) with chronic musculoskeletal pain receiving tertiary pain care. The PEG was strongly associated with primary clinical variables: functional disability ( r = 0.63), quality of life ( r = −0.56), and pain intensity and unpleasantness ( r = 0.58, 0.61), and strongly-to-moderately associated with secondary clinical variables: fatigue ( r = 0.53), depression ( r = 0.49), and pain catastrophizing ( r = 0.49), with modest associations with sleep and anxiety ( r = −0.31, 0.25). Receiver operating characteristic analysis indicated that across all 4 primary clinical variables, the PEG demonstrated the strongest discrimination of upper-tertile functional disability scores (Functional Disability Inventory ≥28, area under the curve = 0.81, sensitivity 0.75, specificity 0.76, positive predictive value 0.64, Youden index 0.51, PEG score cut-off 6) and was cross-validated against pain intensity, pain unpleasantness, and quality-of-life variables. The PEG represents a promising brief monitoring and clinical status tool to support risk-stratified care and clinical research in pediatric chronic musculoskeletal pain.

While medication use is common among pregnant women, medication safety remains insufficiently characterized because studies in pregnant women are challenging due to safety concerns. The recent digitization of healthcare databases and advances in computational methods have created new opportunities for large-scale, retrospective drug safety evaluations. Here, we present PregMedNet, a platform that characterizes multifaceted maternal medication associations on neonatal outcomes during pregnancy, covering more than 27,000 drug-disease pairs across 1,152 medications and 24 outcomes. These results encompass known and additional odds ratios (ORs), adjusted ORs, and drug-drug interactions, systematically analyzed using nationwide claims data and an advanced machine learning pipeline. Notably, one of the associations identified in this study is supported by in vivo experiments, increasing confidence in PregMedNet’s findings and highlighting the utility of claims data and machine learning for perinatal medication safety studies. Additionally, potential biological mechanisms underlying the associations are explored using a graph learning method, providing candidate pathways for future mechanistic investigations. We expect that PregMedNet will contribute to advancing maternal medication safety and improving neonatal outcomes by providing extensive, multifaceted drug safety information on this previously underrepresented population.

Chronic pain (CP) affects approximately 20% of children, significantly impacting their physical, emotional, and social well-being. Hypermobility disorders (HDs) are prevalent in pediatric populations, including those with CP, and often contribute to impairments in psychological and physical functioning. This study examined differences in sensory, functional, and neural circuitry among CP in youth with and without comorbid HD (CP + HD). Contrary to our a priori hypothesis and previous literature suggesting that youth with CP + HD would exhibit greater suprathreshold pain sensitivity and lower pain thresholds, our findings revealed no significant differences in quantitative sensory testing measures between the 2 groups. This suggests that heightened pain sensitivity observed in CP + HD cannot be solely attributable to their hypermobility. However, the CP + HD cohort reported greater pain spread, increased fatigue, and poorer sleep quality, indicating that while they may not be more vulnerable to pain sensitization, they experience greater overall impairment in health and functioning. Furthermore, site-stratified analyses of multisensory task-evoked fMRI data revealed group differences in cerebellar activation and baseline cerebellar-supplementary motor cortex connectivity in youth with CP + HD, indicating altered neural strategies for sensory-motor integration. Notably, these significant fMRI findings were observed at 1 of the 2 study sites. Our results underscore the need for a nuanced understanding of the interplay between hypermobility, CP, and functional impairment to inform targeted interventions that address the multifaceted challenges faced by these patients.

Background Frailty is an important factor in human aging associated with a broad range of adverse outcomes. Frailty metrics are time intensive to collect making them difficult for larger scale application. Methods We apply machine learning to predict these frailty metrics, associated risk factors, and adverse outcomes from activity data. We use activity data collected using Actigraphy wearable accelerometer sensors, which are devices that measure acceleration along three axes of movement. Models were evaluated using Area Under the receiver operator Curve (AUC), Area Under Precision Recall Curve (AUPRC), Spearman rank test, Mann-Whitney U test, or Kruskal-Wallis test on repeated subsampling of train and test sets. All statistical tests are reported using -log₁₀(P-value). Results Machine learning models show strong predictive performance even with small amounts of accelerometry data available. They are also able to better determine adverse outcomes such as hospitalization and mortality than frailty metrics themselves in our geriatric population. Conclusions This approach of wearable activity data-based prediction of frailty offers a surrogate (proxy or estimate) for determining frailty metrics in a scalable manner. It can also be used to determine adverse outcomes such as hospitalizations and mortality, allowing frailty to be used as a metric in other studies or medical practices.

Neonatal prematurity leads to considerable morbidity and mortality, partly because of acquired conditions such as bronchopulmonary dysplasia (BPD), intraventricular hemorrhage (IVH), necrotizing enterocolitis (NEC), and retinopathy of prematurity (ROP). Standard gestational age and birthweight-based classifications of prematurity inadequately capture the variation in newborns’ health outcomes, creating an urgent need to develop risk stratification tools for vulnerable newborn infants to initiate the most appropriate care pathways as early as possible. We hypothesized that the metabolic profiles of newborn infants capture additional risk information beyond current measures. A total of 13,536 newborn screening (NBS) blood spot tests from preterm infants in California with linked clinical outcomes of prematurity were used to develop an NBS-based metabolic health index to stratify preterm infants at risk for BPD, IVH, NEC, and ROP (12,096 cases with one or more conditions and 1440 controls) through a deep learning model that provides a single index score in tandem with subgroup discovery to identify individuals with the strongest metabolite biomarker signals for adverse outcomes of prematurity. This metabolic health index captured risk signals that were distinct from gestational age and birthweight and outperformed other machine learning algorithms and clinical risk variable-based models in stratifying at-risk individuals for adverse outcomes of prematurity. The metabolic health index was externally validated in an independent retrospective cohort of 3299 very premature newborns from Ontario, Canada (2117 cases and 1182 controls), which recapitulated common metabolic risk subgroups. In summary, combining widespread metabolite screening with deep learning established a generalizable biological risk metric of prematurity.

One in ten neonates are admitted to neonatal intensive care units, highlighting the need for precise interventions. However, the application of artificial intelligence (AI) in guiding neonatal care remains underexplored. Total parenteral nutrition (TPN) is a life-saving treatment for preterm neonates; however, implementation of the therapy in its current form is subjective, error-prone and resource-consuming. Here, we developed TPN2.0—a data-driven approach that optimizes and standardizes TPN using information collected routinely in electronic health records. We assembled a decade of TPN compositions (79,790 orders; 5,913 patients) at Stanford to train TPN2.0. In addition to internal validation, we also validated our model in an external cohort (63,273 orders; 3,417 patients) from a second hospital. Our algorithm identified 15 TPN formulas that can enable a precision-medicine approach (Pearson’s R = 0.94 compared to experts), increasing safety and potentially reducing cost. A blinded study (n = 192) revealed that physicians rated TPN2.0 higher than current best practice. In patients with high disagreement between the actual prescriptions and TPN2.0, standard prescriptions were associated with increased morbidities (for example, odds ratio = 3.33; P value = 0.0007 for necrotizing enterocolitis), while TPN2.0 recommendations were linked to reduced risk. Finally, we demonstrated that TPN2.0 employing a transformer architecture enabled guideline-adhering, physician-in-the-loop recommendations that allow collaboration between the care team and AI.

Background: Preclinical evidence suggests that young plasma has beneficial effects on multiple organ systems in aged mice. Whether young plasma exerts beneficial effects in an aging human population remains highly controversial. Despite lacking data, young donor plasma infusions have been promoted for age-related conditions. Given the preclinical evidence that young plasma exerts beneficial effects by attenuating inflammation, this study examined whether administering a young plasma protein fraction to an elderly population would exert anti-inflammatory and immune modulating effects in humans, using surgery as a tissue injury model. Methods: This double-blind, placebo-controlled study enrolled and randomized 38 patients undergoing major joint replacement surgery. Patients received four separate infusions of a plasma protein fraction derived from young donors, or placebo one day before surgery, before and after surgery on the day of surgery, and one day after surgery. Blood specimens for proteomic and immunological analyses were collected before each infusion. Based on the high-content assessment of circulating plasma proteins with single-cell analyses of peripheral immune cells, proteomic signatures and cell-type-specific signaling responses that separated the treatment groups were derived with regression models. Results: Elastic net regression models revealed that administration a young plasma protein fraction significantly altered the proteomic (AUC = 0.796, p = 0.002) and the cellular immune response (AUC 0.904, p < 0.001) to surgical trauma resulting in signaling pathway- and cell type-specific anti-inflammatory immune modulation. Affected proteomic pathways regulating inflammation included JAK-STAT, NF-kappa B, and MAPK (p < 0.001). These findings were confirmed at the cellular level as the MAPK and JAK/STAT signaling responses were diminished and IkB, the negative regulator of NFkB, was elevated in adaptive immune cells. Conclusion: Reported findings provide a first proof of principle in humans that a young plasma protein fraction actively regulates inflammatory and immune responses in an elderly population. They provide a solid rationale for elucidating active principles in young plasma that may be of therapeutic benefits for a range of age-related pathologies. Trial registration: ClinicalTrials.gov, NCT 03981419.