OBJECTIVES: Regular physical activity (PA) and reduced sedentary behaviour (SB) have been associated with positive health outcomes, but many older adults do not comply with the current recommendations. Sensor-triggered ecological momentary assessment (EMA) studies allow capturing real-time data during or immediately after PA or SB, which can yield important insights into these behaviours. Despite the promising potential of sensor-triggered EMA, this methodology is still in its infancy. Addressing methodological challenges in sensor-triggered EMA studies is essential for improving protocol adherence and enhancing validity. Therefore, this study aimed to examine (1) the patterns in sensor-triggered EMA protocol adherence (eg, compliance rates), (2) the impact of specific settings (eg, event duration) on the number of prompted surveys, and (3) participants' experiences with engaging in a sensor-triggered EMA study. DESIGN: Two longitudinal, sensor-triggered EMA studies-one focused on PA and the other on SB-were conducted using similar methodologies from February to October 2022. Participants' steps were monitored for seven days using a Fitbit activity tracker, which automatically prompted an EMA survey through the HealthReact smartphone application when specified (in)activity thresholds were reached. After the monitoring period, qualitative interviews were conducted. Data from both studies were merged. SETTING: The studies were conducted among community-dwelling Belgian older adults. PARTICIPANTS: The participants had a median age of 72 years, with 54.17% being females. The PA study included 88 participants (four dropped out), while the SB study included 76 participants (seven dropped out). PRIMARY AND SECONDARY OUTCOME MEASURES: Descriptive methods and generalised logistic mixed models were employed to analyse EMA adherence patterns. Simulations were conducted to assess the impact of particular settings on the number of prompted EMA surveys. Additionally, qualitative interview data were transcribed verbatim and thematically analysed using NVivo. RESULTS: Participants responded to 81.22% and 79.10% of the EMA surveys in the PA and SB study, respectively. The confirmation rate, defined as the percentage of EMA surveys in which participants confirmed the detected behaviour, was 94.16% for PA and 72.40% for SB. Logistic mixed models revealed that with each additional day in the study, the odds of responding to the EMA survey increased significantly by 1.59 times (OR=1.59, 95% CI: 1.36 to 1.86, p<0.01) in the SB study. This effect was not observed in the PA study. Furthermore, time in the study did not significantly impact the odds of participants confirming to be sedentary (OR=0.97, 95% CI: 0.92 to 1.02, p=0.28). However, it significantly influenced the odds of confirming PA (OR: 0.81, 95% CI: 0.68 to 0.97, p=0.02), with the likelihood of confirming decreasing by 19% with each additional day in the study. Furthermore, a one-minute increase in latency (ie, time between last syncing and starting the EMA survey) in the PA study decreased the odds of the participant confirming to be physically active by 20% (OR: 0.80, 95% CI: 0.72 to 0.89, p<0.01). Simulations of the specific EMA settings revealed that reducing the event duration and shorter minimum time intervals between prompts increased the number of EMA surveys. Overall, most participants found smartphone usage to be feasible and rated the HealthReact app as user-friendly. However, some reported issues, such as not hearing the notification, receiving prompts at an inappropriate time and encountering technical issues. While the majority reported that their behaviour remained unchanged due to study participation, some noted an increased awareness of their habits and felt more motivated to engage in PA. CONCLUSIONS: This study demonstrates the potential of sensor-triggered EMA to capture real-time data on PA and SB among older adults, showing strong adherence potential with compliance rates of approximately 80%. The SB study had lower confirmation rates than the PA study, due to technical issues and discrepancies between self-perception and device-based measurements. Practical recommendations were provided for future studies, including improvements in survey timing, technical reliability and strategies to reduce latency.

• Keywords
• Aging, Behavior, Feasibility Studies, Observational Study, PUBLIC HEALTH, eHealth,
• MeSH
• Exercise * MeSH
• Fitness Trackers MeSH
• Humans MeSH
• Longitudinal Studies MeSH
• Ecological Momentary Assessment * MeSH
• Independent Living * MeSH
• Sedentary Behavior * MeSH
• Aged, 80 and over MeSH
• Aged MeSH
• Check Tag
• Humans MeSH
• Male MeSH
• Aged, 80 and over MeSH
• Aged MeSH
• Female MeSH
• Publication type
• Journal Article MeSH
• Geographicals
• Belgium MeSH

BACKGROUND: To design effective tailored interventions to promote physical activity (PA) among older adults, insights are needed into the contexts in which older adults engage in PA and their affective and physical experiences. Sensor-triggered event-based ecological momentary assessment (EMA) is an innovative method for capturing real-life contexts, as well as affective and physical states, during or immediately after specific events, such as PA. This study aimed to (1) describe the physical and social contexts, and the affective and physical states during PA among older adults, (2) evaluate how these constructs fluctuate during PA episodes, and (3) describe affective states during PA according to the context. METHODS: An intensive longitudinal sensor-triggered event-based EMA study was conducted with 92 Belgian older adults (65 + years). During seven days, participants were monitored using a Fitbit, which triggered a smartphone-based questionnaire on the event-based EMA platform 'HealthReact' after a five-minute walk. Participants reported on contexts and affective (positive/negative valence) and physical states (pain and fatigue) during the PA event. Descriptive statistics and generalized mixed models were used for data analysis. RESULTS: Older adults predominantly engaged in daily physical activities, such as walking for transport, leisure walking, and gardening, rather than structured exercise. They consistently reported high positive affect, low negative affect, and minimal physical complaints during PA. Furthermore, older adults mainly engage in physical activities alone, particularly in outdoor settings. Variations in contexts, affect, and fatigue were mostly driven by within-subject differences. The model showed significant differences across times of day, with negative affect being highest in the evening and fatigue lowest in the morning. Additionally, the physical and social context influenced negative affect (but not positive affect), with outdoor activities performed alone and indoor activities performed with others being associated with lower negative affect. CONCLUSIONS: While these findings could enhance the effectiveness of tailored PA interventions, it remains unclear whether the observed affective and physical states are causes or effects of PA, and whether the contexts in which the activities were performed align with older adults' preferences. Further research is needed to explore these relationships and to better understand older adults' preferred PA contexts.

• Keywords
• Affect, Ecological momentary assessment (EMA), Older adults, Physical activity, Physical and social context, Sensor-triggered event-based experience sampling,
• MeSH
• Affect * MeSH
• Walking MeSH
• Exercise * psychology   physiology   MeSH
• Humans MeSH
• Longitudinal Studies MeSH
• Ecological Momentary Assessment * MeSH
• Surveys and Questionnaires MeSH
• Aged, 80 and over MeSH
• Aged MeSH
• Fatigue MeSH
• Check Tag
• Humans MeSH
• Male MeSH
• Aged, 80 and over MeSH
• Aged MeSH
• Female MeSH
• Publication type
• Journal Article MeSH
• Geographicals
• Belgium MeSH

Ecological Momentary Assessment (EMA) enables the real-time capture of health-related behaviours, their situational contexts, and associated subjective experiences. This study aimed to evaluate the feasibility of an EMA targeting physical and eating behaviours, optimise its protocol, and provide recommendations for future large-scale EMA data collections. The study involved 52 participants (age 31±9 years, 56% females) from Czechia, France, Germany, and Ireland completing a 9-day free-living EMA protocol using the HealthReact platform connected to a Fitbit tracker. The EMA protocol included time-based (7/day), event-based (up to 10/day), and self-initiated surveys, each containing 8 to 17 items assessing physical and eating behaviours and related contextual factors such as affective states, location, and company. Qualitative insights were gathered from post-EMA feedback interviews. Compliance was low (median 49%), particularly for event-based surveys (median 34%), and declined over time. Many participants were unable or unwilling to complete surveys in certain contexts (e.g., when with family), faced interference with their daily schedules, and encountered occasional technical issues, suggesting the need for thorough initial training, an individualised protocol, and systematic compliance monitoring. The number of event-based surveys was less than desired for the study, with a median of 2.4/day for sedentary events, when 4 were targeted, and 0.9/day for walking events, when 3 were targeted. Conducting simulations using participants' Fitbit data allowed for optimising the triggering rules, achieving the desired median number of sedentary and walking surveys (3.9/day for both) in similar populations. Self-initiated reports of meals and drinks yielded more reports than those prompted in time-based and event-based EMA surveys, suggesting that self-initiated surveys might better reflect actual eating behaviours. This study highlights the importance of assessing feasibility and optimising EMA protocols to enhance subsequent compliance and data quality. Conducting pre-tests to refine protocols and procedures, including simulations using participants' activity data for optimal event-based triggering rules, is crucial for successful large-scale data collection in EMA studies of physical and eating behaviours.

• MeSH
• Exercise * MeSH
• Adult MeSH
• Middle Aged MeSH
• Humans MeSH
• Young Adult MeSH
• Ecological Momentary Assessment * MeSH
• Surveys and Questionnaires MeSH
• Data Collection * methods   MeSH
• Feeding Behavior * MeSH
• Feasibility Studies MeSH
• Check Tag
• Adult MeSH
• Middle Aged MeSH
• Humans MeSH
• Young Adult MeSH
• Male MeSH
• Female MeSH
• Publication type
• Journal Article MeSH