Advancements in accelerometer analytic and visualization techniques allow researchers to more precisely identify and compare critical periods of physical activity (PA) decline by age across the lifespan, and describe how daily PA patterns may vary across age groups. We used accelerometer data from the 2003–2006 cohorts of the National Health and Nutrition Examination Survey (NHANES) (n = 12,529) to quantify total PA as well as PA by intensity across the lifespan using sex-stratified, age specific percentile curves constructed using generalized additive models. We additionally estimated minute-to-minute diurnal PA using smoothed bivariate surfaces. We found that from childhood to adolescence (ages 6–19) across sex, PA is sharply lower by age partially due to a later initiation of morning PA. Total PA levels, at age 19 are comparable to levels at age 60. Contrary to prior evidence, during young adulthood (ages 20–30) total and light intensity PA increases by age and then stabilizes during midlife (ages 31–59) partially due to an earlier initiation of morning PA. We additionally found that males compared to females have an earlier lowering in PA by age at midlife and lower total PA, higher sedentary behavior, and lower light intensity PA in older adulthood; these trends seem to be driven by lower PA in the afternoon compared to females. Our results suggest a reevaluation of how emerging adulthood may affect PA levels and the importance of considering time of day and sex differences when developing PA interventions.

No abstract

Summary For multivariate spatial Gaussian process models, customary specifications of cross-covariance functions do not exploit relational inter-variable graphs to ensure process-level conditional independence among the variables. This is undesirable, especially for highly multivariate settings, where popular cross-covariance functions such as the multivariate Matérn suffer from a curse of dimensionality as the number of parameters and floating point operations scale up in quadratic and cubic order, respectively, in the number of variables. We propose a class of multivariate Graphical Gaussian Processes using a general construction called stitching that crafts cross-covariance functions from graphs and ensures process-level conditional independence among variables. For the Matérn family of functions, stitching yields a multivariate Gaussian process whose univariate components are Matérn Gaussian processes, and which conforms to process-level conditional independence as specified by the graphical model. For highly multivariate settings and decomposable graphical models, stitching offers massive computational gains and parameter dimension reduction. We demonstrate the utility of the graphical Matérn Gaussian process to jointly model highly multivariate spatial data using simulation examples and an application to air-pollution modelling.

The #MeToo Movement has brought new attention to sexual harassment and assault. While the movement originates with activist Tarana Burke, actor Alyssa Milano used the phrase on Twitter in October 2017 in response to multiple sexual harassment allegations against Hollywood producer Harvey Weinstein. Within 24 hours, 53,000 people tweeted comments and/or shared personal experiences of sexual violence. The study objective was to measure how information seeking via Google searches for sexual harassment and assault changed following Milano’s tweet and whether this change was sustained in spite of celebrity scandals. Weekly Google search inquiries in the United States were downloaded for the terms metoo, sexual assault, sexual harassment, sexual abuse, and rape for January 1, 2017 to July 15, 2018. Seven related news events about perpetrator accusations were considered. Results showed that searches for metoo increased dramatically after the Weinstein accusation and stayed high during subsequent accusations. A small decrease in searches followed, but the number remained very high relative to baseline (the period before the Weinstein accusation). Searches for sexual assault and sexual harassment increased substantially immediately following the Weinstein accusation, stayed high during subsequent accusations, and saw a decline after the accusation of Matt Lauer (talk show host; last event considered). We estimated a 40% to 70% reduction in searches 6 months after the Lauer accusation, though the increase in searches relative to baseline remained statistically significant. For sexual abuse and rape, the number of searches returned close to baseline by 6 months. It appears that the #MeToo movement sparked greater information seeking that was sustained beyond the associated events. Given its recent ubiquitous use in the media and public life, hashtag activism such as #MeToo can be used to draw further attention to the next steps in addressing sexual assault and harassment, moving public web inquiries from information seeking to action.

Predictions during the early stage of an epidemic are essential to inform quarantine protocols, plan medical resources, and implement economic strategies. One of the major obstacles for making accurate predictions is imperfect available data on the current state of the disease dynamics typically summarized via the number of infected and recovered cases, and disease-related deaths. In the case of rapidly evolving COVID-19 pandemic, major challenges include underreporting of infected and recovered cases due to the shortage of available tests, non-uniform performance and not sufficiently high sensitivity/specificity of currently available tests, a long incubation period, and a significant number of asymptomatic/unconfirmed cases, inconsistent accounting practices in death classification. All these make it quite challenging to accurately predict the future trajectory of this pandemic.Putting aside the imperfections in the currently available data, choosing an appropriate model is another essential component of the modelling endeavor. Modifications of the classical SIR (Susceptible-Infectious-Removed) and SEIR (Susceptible-Exposed-Infectious-Removed) models are among the most popular modelling frameworks used during the early stage of this pandemic. Ironically, even the notorious Institute for Health Metrics and Evaluation at the University of Washington that originally proposed an misleadingly inflexible curve-fitting model (The Institute for Health Metrics and Evaluation, 2020) that received vigorous criticism within statistical and epidemiological communities (see (Brookmeyer, 2020;Jewell et al., 2020;Jin