ICAMPAM 2024 Symposia
Tuesday, June 18 at 14h00-15h30
Symposium 1
METHODOLOGICAL CONSIDERATIONS WHEN EXAMINGING PHYSICAL ACTIVITY AND SEDENTARY BEHAVIORS IN OLDER ADULT POPULATIONS
Nancy Glynn, Reagan Garcia, University of Pittsburgh; Jennifer Schrack, Lacey Etzkorn, Johns Hopkins Bloomberg School of Public Health
Our symposium will capitalize on data from several large cohort studies of older adults, the Study of Muscle, Mobility and Aging (SOMMA, N=879), the Baltimore Longitudinal Study of Aging (N=1,300) the Study to Understand Falls Reduction and Vitamin D in You (N=688) and the National Health and Aging Trends Study (N=747), to discuss, debate, and propose solutions to pressing challenges in aging PA research including a lack of uniformity across studies in the data interpretation (e.g., counts, steps, calories, raw data). Ms. Garcia will provide critical insight into PA intensity cutpoints and their implications for future cross-cohort comparison of older adult cohorts; Dr. Schrack will provide novel insights into activity metrics for use in older populations; and Dr. Etzkorn will share the importance of considering age, accelerometer placement, and sampling rate when implementing activity measurement in studies of older adults. Dr. Annemarie Koster will moderate the discussion.
Symposium 2
ACTIGRAPH OPEN-SOURCE ACTIVITY COUNTS: (HOW MUCH) DOES IT MATTER TO THE PHYSICAL BEHAVIOUR MEASUREMENT FIELD?
Alexander Montoye, Alma College; Kimberly Clevenger, Utah State University; Jan Brønd, University of Southern Denmark; Samuel Lamunion, National Institutes of Health
Accelerometer-based activity monitors have long been important tools for understanding the link between physical behaviours and health. The ActiGraph company has traditionally captured raw data and translated it into “activity counts” representing movement magnitude. Traditionally, ActiGraph’s proprietary method restricted the use of activity counts, hindering cross-study comparability and data harmonization. In 2022, ActiGraph made their count calculation method open-source, enabling its use beyond their software and across accelerometer brands. A recent Journal for the Measurement of Physical Behaviour issue explores the implications of counts being made open-source. This symposium, led by guest editors from the special issue, will showcase key papers and discuss the broader implications of open-source activity counts for the field of physical behavior measurement.
Symposium 3
RECENT APPLICATIONS OF COMPOSITIONAL DATA ANALYSIS (CODA) FOR ANALYZING THE HEALTH ASSOCIATIONS OF 24-HOUR MOVEMENT BEHAVIORS
Christian Brakenridge, Swinburne University of Technology; Marjo Seppänen, Maisa Niemelä, University of Oulu; Kristin Suorsa, University of Turku; Antti Löppönen, Lotta Palmberg, University of Jyväskylä
Chair Cristian Brakenridge will provide an overview of utilizing CoDA in examining associations between 24-h movement behaviors and health. Marjo Seppänen will start by presenting associations between the composition of 24-h movement behavior and depressive symptoms in urban and rural environments among middle-aged adults. Maisa Niemelä will continue by sharing her findings about associations between worktime and leisure time movement behavior compositions and cardiometabolic health. Kristin Suorsa will present how 24-h movement behavior compositions on workdays and days off are associated with the incidence of cardiovascular risk factors among retiring adults. Lotta Palmberg will then share her findings on how the 24-h movement behavior composition predicts another aspect of health, quality of life in old age. Chair Cristian Brakenridge will summarize the presentations and highlight directions for future studies.
Tuesday, June 18 at 16h00-17h30
Symposium 4
PREDICTING FALL RISK USING DIGITAL TECHNOLOGIES: CURRENT STATE OF THE ART AND FUTURE RESEARCH DIRECTIONS
Lorenzo Chiari, Luca Palmerini, Alessandro Silvani, University of Bologna; Martina Mancini, Oregon Health & Science University
Falls are a significant health concern, leading to severe injuries and loss of independence. Identifying individuals at risk is critical. Traditional clinical tools have limited validity in fall risk assessment. We will present novel methodologies to estimate fall risk using digital technologies (i.e., wearables and machine learning) and discuss current limitations and the most promising research directions. Lorenzo Chiari will be the chair of the symposium, which will feature presentations from Martina Mancini, presenting recent work on fall risk assessment in Parkinson’s disease, Luca Palmerini, presenting fall risk assessment in community-dwelling older subjects using wearables and machine learning, and Alessandro Silvani, discussing evidence supporting the potential for sleep health and heart rate metrics to improve fall risk estimation.
Symposium 5
ADVANCING PHYSICAL ACTIVITY RESEARCH: NOVEL ACCELEROMETER METRICS, COMPARATIVE ANALYSIS, AND FUTURE DIRECTIONS IN COHORT STUDIES
Joanne McVeigh, Bingyan Pang, Curtin University; Janne Kulmala, Jamk University of Applied Sciences; Lisa Micklesfield, University of Witwatersrand
This symposium shares findings from international cohort studies in Australia, Finland, and South Africa, exploring accelerometer-derived metrics of physical activity and their impact on health outcomes. Recent advances in accelerometer processing enable data harmonization and transparency. The analysis aims to provide evidence-based insights for interventions, public health policy, and personalised health strategies globally. The goal is to understand the complex relationship between accelerometer-derived metrics and health outcomes on a global scale. The audience will be encouraged to participate in a discussion on approaches to synchronise accelerometry data within cohort studies, and how best to ensure inclusivity across diverse populations.
Symposium 6
BEYOND THE AVERAGE: THE POWER OF TEMPORAL PATTERNS OF PHYSICAL ACTIVITY ACCUMULATION
Melvyn Hillsdon, Brad Metcalf, Joshua Culverhouse, University of Exeter; Joss Langford, Activinsights and University of Exeter
Knowledge on the benefits of physical activity (PA) is mainly derived from aggregate statistics such as time spent in moderate to vigorous PA. This maybe useful for gaining a snapshot of levels of PA, but masks information about patterns in PA accumulation that may provide a better understanding of how PA relates to health. Time-series data on contiguous active and inactive events helps researchers know not just how much physical activity individuals engage in, but also when and how frequently, providing a richer narrative about PA-disease associations. Harnessing the power of time series data marks a paradigm shift away from averages of daily volume toward an approach that better reflects the diverse ways people incorporate movement into their lives. The symposium will feature studies where different temporal patterns of PA accumulation have provided insights that would not be possible simply by examining average daily volumes.
Wednesday, June 19 at 15h30-17h00
Symposium 7
SEDENTARY BEHAVIOR – FROM METHODOLOGICAL ANALYSIS TO POPULATION SAMPLE AND EFFECTIVE INTERVENTIONS
Tommi Vasankari, Henri Vähä-Ypyä, Pauliina Husu, The UKK Institute for Health Promotion Research; Ilkka Heinonen, Sini Vasankari, Turku University Hospital
This symposium will focus on sedentary behavior (SB). First, it will present angle for posture estimation (APE) method, which can classify SB measured with hip-worn accelerometer. Presentation will show how and with which angle ranges of APE can identify different components of SB when analyzing raw accelerometer data. Second, it will demonstrate components of SB (lying, reclining, and sitting during waking hours) and their associations with indicators of metabolic health among two population-based samples of working-aged adults (n= 4298). Third, symposium will represent effective interventions how reduced SB is associated with improve whole-body insulin-stimulated glucose uptake in patients with metabolic syndrome and how reduced SB and increased PA are associated with health benefits in patients recovering from cardiac procedures. Both interventions use interactive accelerometer + smart phone application + cloud system.
Symposium 8
PHYSICAL BEHAVIOUR DETECTION FROM TRI-AXIAL ACCELEROMETRY USING WEARABLES AND SMARTWATCHES: DEVELOPMENT OF CLASSIFIERS AND EXTERNAL VALIDATION BASED ON ECOLOGICAL MOMENTARY ASSESSMENT IN THE WEALTH STUDY
Christoph Buck, Annika Swenne, Leibniz Institute for Prevention Research and Epidemiology – BIPS; Roar Fenne, Norwegian University of Science and Technology; Luis Sigcha, University of Limerick; Alan Godfrey, Northumbria University
The development of new machine learning algorithms for behaviour classification creates the need for ongoing collection of high-quality data. In this symposium, we will present recent developments from the WEALTH project where we collected data from 600 participants in four European countries, i.e., Ireland, Germany, Czechia, and France. The study design included a 75-minute scripted study, followed by a nine-day monitoring phase supported by ecological momentary assessment prompts for data labeling in a free-living environment. Classifiers of physical behaviours were modeled based on raw from inertial sensors particularly research grade devices and smartwatches.Recent wearable research and developments in closely related fields will be presented. We will discuss and exchange experiences about machine learning approaches and consider important steps and challenges for the establishment of a harmonized guidance in future research.
Symposium 9
BIDIRECTIONAL ASSOCIATIONS OF 24-HOUR MOVEMENT BEHAVIOURS AND COGNITIVE FUNCTION AMONG OLDER ADULTS: IS IT REVERSE CAUSATION?
Paul Gardiner, The University of Queensland; Dori Rosenberg, Mikael Anne Greenwood-Hickman, Kaiser Permanente Washington Health Research Institute; Mikaela Bloomberg, Laura Brocklebank, University College London
Studies demonstrate that physical activity, sedentary time, and sleep are associated with cognitive function. However, the direction of the causal relationships between these 24-hour movement behaviours and cognitive health is currently unclear. Prior research has demonstrated that declining cognition predicts lower levels of physical activity, perhaps due to the cognitive effort required for planning and sustaining physical activity. Sleep disturbances also become more frequent with age, are a common symptom of prodromal dementia, and may feedback to exacerbate ageing-related or pathological cognitive decline. In this symposium, we explore bidirectional associations of 24-hour movement behaviours with cognitive function from three cohorts of older adults and pose the question of whether reverse causality is at play.
Symposium 10
HOW ARE CHILDREN REALLY USING THEIR DIGITAL MEDIA? HARNESSING THE POWER OF TECHNOLOGY TO OBJECTIVELY MEASURE TRUE SCREEN USE AND DETERMINE RELATIONSHIPS WITH SLEEP HEALTH
Kim Meredith-Jones, Rachael Taylor, Shay-Ruby Wickham, Rosie Jackson, University of Otago
Background: How digital media impact health and wellbeing is currently limited by questionnaire measures of screen time. This symposium will describe how photo and video cameras accurately quantify screen time in children and impact sleep.
Methods: Wearable and stationary cameras assessed screen time over multiple days in youth (8-14y) in a repeated measures observational study (BED) which determined the impact of evening screen time on sleep, and in a randomised crossover trial which induced mild sleep deprivation (DREAM).
Results: BED shows how youth really use their screens before bed, in bed, and even after trying to go to sleep. Not all screen time is bad for sleep, with interactive and multitasking being more detrimental. DREAM shows the minimum interval required between photos to accurately estimate screen use and how sleep deprivation changes screen time.
Conclusion: Cameras offer huge potential to increase understanding of how youth actually use devices.
Thursday, June 20 at 9h45-11h15
Symposium 11
PHYSIOLOGY-DRIVEN BEHAVIOUR ASSESSMENT
Timo Rantalainen, Laura Karavirta, Antti Löppönen, University of Jyväskylä; Nicky Ridgers, University of South Australia; Henri Vähä-Ypyä, The UKK Institute for Health Promotion Research
The symposium will comprise four talks on the topic of identifying aspects of free-living behaviour that would be expected to evoke physiological responses. Laura Karavirta will talk about considering the physiological function of an individual when assessing daily physical activity. Timo Rantalainen will discuss identifying bone generating free-living activities from accelerometry records. Henri Vähä-Ypyä will present on how relating the intensity of activity to cardiorespiratory fitness may affect our view of physical activity in a population representative sample. Antti Löppönen will postulate that sit-to-stand (STS) transitions could be used as a proxy of strength demanding daily activities among older individuals. The symposium will conclude with inviting the audience to discuss the claim that device-based physical activity metrics ought to be aligned with the well-recognised physiological underpinnings regarding organ-system adaptations.
Symposium 12
NOVEL INSIGHTS INTO DAILY PHYSICAL ACTIVITY AND FUNCTION IN FITTEST-TO-FRAILEST OLDER ADULTS: FROM VALIDATION TO REAL WORLD PERFORMANCE
Nina Skjæret-Maroni, Astrid Ustad, Karoline Blix Grønvik, Norwegian University of Science and Technology; Karen Sverdrup, Norwegian National Centre for Ageing and Health; Kjerstin Melsæter, Trondheim municipality and Norwegian University of Science and Technology
Physical activity (PA) and function are key determinants of health and well-being in older adults. To treat and support the older population, and preserve good health and independence, granular insight across levels of functioning is critical. Large population studies such as the Trøndelag Health Study (HUNT) include the healthiest community-dwelling to the frailest older adults in nursing homes and are critical to our understanding of PA and function. This symposium address challenges encountered when trying to accurately identify physical behavior and mobility across the spectrum of older adults. Four early career speakers will present unique but related perspective on older adults’ activity and functioning in real life, spanning from physical performance reference values to validation of activity type recognition models. Following, the symposium is devoted to a discussion focusing on challenges and needs, as well as future directions.
Symposium 13
ADVANCES AND FUTURES OF 24 HOUR ANALYTICS
Sebastien Chastin, Glasgow Caledonian University, Ghent University; Dot Dumuid, University of South Australia; Flora Le, Monash University; Vahid Farrahi, TU Dortmund Univeristy; Paulína Jašková, Palacký University Olomouc
In recent years, compositional data analysis has been applied to develop strong evidence for the relationships between reallocations of time and a large range of health measures. In this symposium we present some of the most important advances in compositional data analytics applied to 24 hour movement behaviour that have been developed in the last year. We will show advances: in combining compositional data analysis with evolutionary algorithms to find optimal composition of time use for specific and combined health outcomes, in multilevel analysis, in functional time series analysis and in data-driven compositional data analysis. The symposium assembles distinguished researchers who pioneered the field together with early career researchers who are moving it into new areas and the future. The symposium will finish with an open debate engaging the audience about the limitations, opportunities, gaps and future developments.
Friday, June 21 at 9h45-11h15
Symposium 14
INTEGRATING INTERSECTIONALITY IN BEHAVIORAL DATA ANALYSIS: LABDA ADVANCING HEALTH INSIGHTS THROUGH 24-HOUR ACTIVITY DATA
Mai Chin A Paw, Child and Adolescent Public Health Research and Innovatiion; Sebastien Chastin, Glasgow Caledonian University, Ghent University; Gaia Segantin, Amsterdam UMC; Marian Paiva Marchiori, University of Southern Denmark
The Learning network for Advanced Behavioral Data Analysis (LABDA) Symposium on “Integrating Intersectionality in Behavioral Data Analysis” focuses on the innovative intersection of data science and human movement epidemiology, emphasizing the integration of intersectionality throughout the value chain of data analysis. Traditional approaches in public health have often overlooked the complex interplay of factors such as gender, age, ethnicity, and socio-economic position. However, LABDA’s novel approach recognizes that these elements map onto social hierarchies, interacting and intertwining to produce unique behavioral patterns and health outcomes. This symposium will explore how LABDA advances this field by tailoring algorithms to specific intersections, moving beyond generalized movement guidelines and delving into the diversity of characteristics like gender, ethnicity, and socio-economic position using multiple datasets from diverse samples and contexts.
Symposium 15
EXPLORING THE IMPACT OF USING WEARABLE MOVEMENT SENSORS IN COLLEGIATE SPORT: FROM RESEARCH TO PRACTICE AND REHABILITATION TO COMPETITION
Alexander Montoye, Alma College; Karin Pfeiffer, Matt Harkey, William Burghardt, Michigan State University
Wearable movement sensors have a rich history in monitoring human behaviors and health impacts. Due to technological and analytic advances, wearable sensors are becoming increasingly prevalent in sports and rehabilitation. This symposium will explore movement sensor use in a Division I collegiate sports program in the United States, covering elements such as 1) logistics of data capture, processing, and analysis; 2) facilitating buy-in from stakeholders (e.g., coaches, administrators); 3) reporting findings to stakeholders; 4) using data for both athletics decision-making and research purposes; 5) practical examples and lessons learned; and 6) preliminary research findings and anticipated future directions. The symposium will conclude with discussion on wearables-based sports data capture, applications across diverse athlete populations, and challenges in various sports, hopefully stimulating dialog on research and practice in sports data analysis and athlete care.
Symposium 16
ACCURACY: THE HOLY GRAIL OF PHYSICAL BEHAVIOR ASSESSMENT—HOW BIG IS THE PROBLEM, ARE THERE SOLUTIONS, AND HOW HAVE OTHER DISCIPLINES SOLVED THIS ISSUE?
Charles Matthews, National Cancer Institute, National Institutes of Health; Aiden Doherty, University of Oxford; Sarah Keadle, California Polytechnic State University San Luis Obispo; Kong Chen, National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institutes of Health
This Symposium has four goals. First, we will describe variation in population level estimates for MVPA (moderate-vigorous physical activity) and step counts in recent epidemiologic studies and the impact of these differences on our ability to translate these findings in clinical and public health practice. Second, we will discuss a Framework for the calibration and validation of prediction algorithms that could minimize differences in accuracy between methods, including the need for unified evidence-based definitions for core physical behaviors. Third, we will describe how similar accuracy problems have been solved for wearables now used in the clinical setting, such as the continuous glucose monitors. Fourth, we hope to initiate a lively and productive discussion that will help improve the accuracy of our methods in the future.