Latitudinal clines in adaptive traits are well documented, yet how such patterns translate across longitudinal gradients, where climatic variation is reduced, remains poorly understood. We investigated spatial variation in hydrogen cyanide (HCN) - a well-characterized anti-herbivore defense polymorphism governed by two independently segregating loci (Ac and Li), in white clover (Trifolium repens) across nine Midwestern U.S. cities. A total of 269 populations (4,035 individuals) were sampled along urban-rural transects, and cyanotype (cyanogenesis phenotypes) frequencies were quantified alongside regional and site-level environmental variables, including vegetation cover (NDVI), land surface temperature (LST), snow cover, impervious surface intensity, aridity, and potential evapotranspiration. Despite substantial variation in cyanogenic frequency among cities (6.2–19.0
Parkinson's disease (PD) is the fastest-growing neurodegenerative disease affecting 90 thousand new Americans each year. PD includes motor and non-motor symptoms, resulting in progressive disability and difficulty in completing activities of daily living. Freezing of Gait (FoG) is one of the common disabling symptoms of PD, characterized by difficulties in initiating walking, resulting in gait abnormalities and increased risk of falling (RoF) and fear of falling (FoF). Clinical management of FoG is difficult as it is minimally responsive to both pharmacological and surgical interventions. In fact, these interventions can paradoxically worsen of FoG. Additionally, PD patients with FoG have reported worse health-related quality of life (HR-QoL) due to limitations in mobility, activities of daily living (ADL), bodily discomfort, stigma, and social isolation. Despite its increasing treatment and management of FoG is difficult due to its paroxysmal and heterogeneous nature. Therefore, there is a growing need for effective, evidence-based management and intervention approaches for FoG. Some current techniques used to manage FoG are physical therapy, exercise, gait training, and balance training; however, due to a lack of patient adherence, accessibility concerns, and the need for continuous supervision and individualized feedback, the long-term effectiveness of these interventions remains limited and challenging to achieve in real-world settings. A new promising avenue for managing PD is the use of wearable technology, which can provide audiovisual, via augmented and virtual reality (AR/VR), and tactical cueing to offset FoG, thereby enhancing independence in PD patients. In this comprehensive review, we will provide an overview of the symptoms, monitoring, and treatment of PD, with a focus on the neuroanatomy and treatment of FoG. We will review and critique the extant literature on the use of AR/VR technology in the management of FoG. Finally, the challenges and risks associated with wearable technology in FoG management will also be identified.
Although studies suggest that young adults with spina bifida (YASB) often avoid important daily activities (ADL) due to urinary (UI) and fecal (FI) incontinence, no research has prospectively examined how often ADL avoidance occurs, or the context in which it occurs. In this paper, we used ecological momentary assessment (EMA) over 30 days to describe: 1) the frequency of avoidance; 2) the activities YASB most commonly avoid; and 3) preliminary association of daily avoidance with affect, incontinence anxiety and health-related quality of life (HRQoL). As part of a larger 30-day prospective study designed to understand the incontinence in adults with SB (N = 89), participants completed an end-of-day EMA tracking daily ADL avoidance on days when they were worried about possible UI or FI and on days when they had actual UI or FI. Additional day-level measures were: affect and incontinence anxiety. HRQoL was reported in beginning- and end-of-study surveys. We drew a subsample of YASB participants (18-27 years; N = 23 of total 88 participants). ADL avoidance was most frequent with actual FI (20.5%) and least frequent with worry about FI (3.5%, p = 0.02) (UI worry: 8.3; actual UI: 8.9%, p = 0.546). The most common ADL avoided were eating, drinking and spending time with family/friends. Negative mood and incontinence anxiety were significantly higher on days with all types of avoidance. Higher baseline HRQoL was associated with fewer in-study ADLs avoided, which were in turn associated with higher end-of-study HRQoL. Our data suggest that YASB avoid key activities of daily living (e.g., eating, drinking, seeing family/friends) both when they worry about possible incontinence that may but does not occur, and when they experience actual incontinence. Tailored interventions addressing actual incontinence and worry about possible incontinence incorporating daily mood or incontinence anxiety may minimize ADL avoidance and its impact on long-term HRQoL.
Over the past six decades, the labor force participation of working mothers in the United States has grown significantly. This study is motivated by the growing trend of mothers' labor force participation and examines the effect of maternal work intensity on child health. It employs the Panel Study of Income Dynamics' intergenerational survey data and instrumental variable (IV) probit and ordered probit regression techniques. The results show that maternal hours worked has a positive and statistically significant effect on the likelihood of having a healthy childhood, although the estimated magnitude is modest. The fixed effect estimation that accounts for heterogeneity across stages of child development (infancy, early childhood, mid-childhood and late childhood) reveals that maternal hours worked has a negative and significant effect on childhood health for the first two stages (infancy and early childhood) of child development and a positive effect for the late childhood period.
Modern enterprise networks exhibit highly dynamic and relational traffic patterns, where malicious activities often emerge as subtle spatio-temporal behaviors rather than isolated anomalies. While graph-based deep learning models have shown promise for network attack detection, most approaches provide limited insight into why an alert is triggered or which entities are responsible. We propose an explainable spatio-temporal graph learning framework for network attack detection and forensic analysis. Network traffic is modeled as a sequence of temporal graphs and processed using a graph convolutional network combined with a gated recurrent unit to capture spatial dependencies and temporal evolution. To explain detection decisions, we use Integrated Gradients to provide unified attribution across nodes, time steps, and traffic features from a single explanation signal. The framework is validated in a controlled synthetic environment with exact ground truth and evaluated on the Los Alamos National Laboratory enterprise network dataset. Across 704 attack windows, explanations remain sparse and causally meaningful, with ablation of identified attacker nodes yielding an average attack-logit drop of 29.61. These results demonstrate that the proposed approach enables reliable detection while producing actionable explanations suitable for real-world security operations.
