Meynard John L. Toledo

An intervention to reduce sitting and increase light-intensity physical activity at work: Design and rationale of the ‘Stand & Move at Work’ group randomized trial

BACKGROUND American workers spend 70-80% of their time at work being sedentary. Traditional approaches to increase moderate-vigorous physical activity (MVPA) may be perceived to be harmful to productivity. Approaches that target reductions in sedentary behavior and/or increases in standing or light-intensity physical activity [LPA] may not interfere with productivity and may be more feasible to achieve through small changes accumulated throughout the workday METHODS/DESIGN: This group randomized trial (i.e., cluster randomized trial) will test the relative efficacy of two sedentary behavior focused interventions in 24 worksites across two states (N=720 workers). The MOVE+ intervention is a multilevel individual, social, environmental, and organizational intervention targeting increases in light-intensity physical activity in the workplace. The STAND+ intervention is the MOVE+ intervention with the addition of the installation and use of sit-stand workstations to reduce sedentary behavior and enhance light-intensity physical activity opportunities. Our primary outcome will be objectively-measured changes in sedentary behavior and light-intensity physical activity over 12months, with additional process measures at 3months and longer-term sustainability outcomes at 24months. Our secondary outcomes will be a clustered cardiometabolic risk score (comprised of fasting glucose, insulin, triglycerides, HDL-cholesterol, and blood pressure), workplace productivity, and job satisfaction DISCUSSION: This study will determine the efficacy of a multi-level workplace intervention (including the use of a sit-stand workstation) to reduce sedentary behavior and increase LPA and concomitant impact on cardiometabolic health, workplace productivity, and satisfaction. TRIAL REGISTRATION ClinicalTrials.gov Identifier: NCT02566317 (date of registration: 10/1/2015).

Long-term effects of sit-stand workstations on workplace sitting: A natural experiment

OBJECTIVE Sit-stand workstations may result in significant reductions in workplace sitting. However, few studies have examined long-term maintenance under real-world conditions. The purpose of this study was to evaluate workplace sitting time, cardio-metabolic biomarkers, and work productivity during a workplace re-design which included the installation of sit-stand workstations. DESIGN Natural experiment with appropriately matched comparison. METHODS Office workers from distinct worksites in the same unit were recruited (Intervention, n=24; Comparison, n=12). Intervention arm participants received a sit-stand workstation and 4 months of sitting-specific motivational support. The comparison arm received 4 months of ergonomic focused motivational support. Time spent in sitting, standing, and other physical activity were measured by activPAL3c for a week. Cardio-metabolic biomarkers and work productivity were also measured. Assessments occurred at baseline, 4 months, and 18 months. RESULTS At 4 months, work sitting time was reduced by 56.7±89.1min/8h workday (d=-0.64), relative to comparison. Standing time (37.4±69.2min/8h workday; d=0.54) and sit-to-stand transitions (3.3±0.4min/8h workday, d=0.44) were also improved relative to comparison. At 18 months, work sitting time reductions (52.6±68.3min/8h workday; d=-0.77) and standing time improvements (17.7±54.8min/8h workday, d=0.32) were maintained in the intervention group relative to comparison. Cardio-metabolic and work productivity changes were mixed; however, strongest effects favoring the intervention group were observed at 18 months. CONCLUSIONS Sit-stand workstations, accompanied with behavioral support, were effective in reducing workplace and overall daily sitting and increasing standing time in a real-world setting. The effect appears to have been sustained for 18 months, with mixed results in cardio-metabolic and productivity outcomes.

Enrollment Strategies, Barriers to Participation, and Reach of a Workplace Intervention Targeting Sedentary Behavior

Purpose: To review enrollment strategies, participation barriers, and program reach of a large, 2-year workplace intervention targeting sedentary behavior. Approach: Cross-sectional, retrospective review. Setting: Twenty-four worksites balanced across academic, industry, and government sectors in Minneapolis/Saint Paul (Minnesota) and Phoenix (Arizona) regions. Participants: Full-time (≥30+ h/wk), sedentary office workers. Methods: Reach was calculated as the proportion of eligible employees who enrolled in the intervention ([N enrolled/(proportion of eligible employees × N total employees)] × 100). Mean (1 standard deviation) and median worksite sizes were calculated at each enrollment step. Participation barriers and modifications were recorded by the research team. A survey was sent to a subset of nonparticipants (N = 57), and thematic analyses were conducted to examine reasons for nonparticipation, positive impacts, and negative experiences. Results: Employer reach was 65% (56 worksites invited to participate; 66% eligible of 56 responses; 24 enrolled). Employee reach was 58% (1317 invited to participate, 83% eligible of 906 responses; 632 enrolled). Postrandomization, on average, 59% (15%) of the worksites participated. Eighteen modifications were developed to overcome participant-, context-, and research-related participation barriers. Conclusion: A high proportion of worksites and employees approached to participate in a sedentary behavior reduction intervention engaged in the study. Interventions that provide flexible enrollment, graded participant engagement options, and adopt a participant-centered approach may facilitate workplace intervention success.

Comparing Gaussian Mixture Model and Hidden Markov Model to Classify Unique Physical Activities from Accelerometer Sensor Data

With the recent interest in physical therapy through sufficient physical activity, considerable efforts have been made to monitor and classify daily human activities, especially for people who need physical rehabilitation. In our previous study, we designed a classifier to identify 25 unique physical activities performed by 92 healthy participants between the ages of 20 and 65. In this study, with the use of a GENEActiv accelerometer to monitor a wide range of daily activities, we present a learning approach to identify unique activities performed by a varied group of participants with various health conditions. The dataset is comprised of 99 senior participants and 23 participants who are significantly taller in height than the general population, performing 8 unique activities. We have extracted 130 different features in time and frequency domain and selected the most efficient features with the sequential forward selection algorithm. With two stages of classification, the first is utilized for combining similar classes, and the second determines the final decision. We have tested two classifiers for our learning approach, the Gaussian mixture models (GMMs) and the hidden Markov models (HMMs) and compared their performances. We have improved the GMM classifier from our previous study and it has shown more promising results for this dataset. We achieved an accuracy of 88.92% when classifying the 8 unique activities with GMM and 93.5% with HMM when classifying 7 activities.

A statistical estimation framework for energy expenditure of physical activities from a wrist-worn accelerometer

Energy expenditure (EE) estimation from accelerometer-based wearable sensors is important to generate accurate assessment of physical activity (PA) in individuals. Approaches hitherto have mainly focused on using accelerometer data and features extracted from these data to learn a regression model to predict EE directly. In this paper, we propose a novel framework for EE estimation based on statistical estimation theory. Given a test sequence of accelerometer data, the probability distribution on the PA classes is estimated by a classifier and these predictions are used to estimate EE. Experimental evaluation, performed on a large dataset of 152 subjects and 12 activity classes, demonstrates that EE can be estimated accurately using our framework.Energy expenditure (EE) estimation from accelerometer-based wearable sensors is important to generate accurate assessment of physical activity (PA) in individuals. Approaches hitherto have mainly focused on using accelerometer data and features extracted from these data to learn a regression model to predict EE directly. In this paper, we propose a novel framework for EE estimation based on statistical estimation theory. Given a test sequence of accelerometer data, the probability distribution on the PA classes is estimated by a classifier and these predictions are used to estimate EE. Experimental evaluation, performed on a large dataset of 152 subjects and 12 activity classes, demonstrates that EE can be estimated accurately using our framework.