StandUPTV: Preparation and Optimization Phases of a mHealth Intervention to Reduce Sedentary Screen Time in Adults

Abstract

Recreational sedentary screen time (rSST) is the most prevalent sedentary behavior for adults outside of work, school, and sleep, and is strongly linked to poor health. StandUPTV is a mHealth trial that uses the Multiphase Optimization Strategy (MOST) framework to develop and evaluate the efficacy of three theory-based strategies for reducing rSST among adults. This paper describes the preparation and optimization phases of StandUPTV within the MOST framework. We identified three candidate components based on previous literature: (a) rSST electronic lockout (LOCKOUT), which restricts rSST through electronic means; (b) adaptive prompts (TEXT), which provides adaptive prompts based on rSST behaviors; and (c) earning rSST through increased moderate-vigorous physical activity (MVPA) participation (EARN). We also describe the mHealth iterative design process and the selection of an optimization objective. Finally, we describe the protocol of the optimization randomized controlled trial using a 23 factorial experimental design. We will enroll 240 individuals aged 23–64 y who engage in >3 hrs/day of rSST. All participants will receive a target to reduce rSST by 50% and be randomized to one of 8 combinations representing all components and component levels: LOCKOUT (yes vs. no), TEXT (yes vs. no), and EARN (yes vs. no). Results will support the selection of the components for the intervention package that meet the optimization objective and are acceptable to participants. The optimized intervention will be tested in a future evaluation randomized trial to examine reductions in rSST on health outcomes among adults.

1. Background

Excessive sedentary time (i.e., time spent sitting or lying with low energy expenditure while awake) has major adverse health consequences, including risk for diabetes, cardiovascular disease (CVD), CVD mortality, and all-cause mortality.^1–3 Television (TV) viewing is the most prevalent type of sedentary behavior and accounts for 55% of all available discretionary time, with 78% of U.S. adults watching TV >3.5 h/day.^1,4,5 TV viewing is associated with eight of the leading causes of death in the U.S., and is more strongly linked to poor health outcomes than other sedentary behaviors.^6–8 Importantly, the negative effects of TV viewing are not eliminated through engaging in recommended amounts of physical activity.⁸ Moderate to vigorous physical activity (MVPA) at levels 2–3 times greater than the current recommendations (~60–75 min/day of walking) have been shown to lower, but not fully attenuate, the excess mortality risk from ≥5 h/day of TV watching.⁸ Given the high prevalence of TV watching and its strong link with mortality outcomes reducing, reducing screen time (including TV) is a public health target warranting further attention.

Although efficacious strategies to reduce sedentary time in other contexts (i.e., workplace) have been developed,^9,10 a paucity of studies have successfully reduced screen time in adults. Two randomized controlled trials (RCT) we have identified have focused exclusively on TV among adults, each demonstrating reduced TV and increased physical activity. However, both were short duration (≤8 weeks) and neither study accounted for contemporary trends in media consumption (i.e., streaming services, multiple viewing platforms). Among children (< 13 years), reduced screen time is associated with improvements in physical activity, diet, and weight management behaviors, but the impact of reduced screen time on behavioral and/or health outcomes is understudies among adults.¹¹ We will use the term recreational sedentary screen time (rSST) to refer to discretionary time (i.e., not for work or educational purposes) spent with screens, such as TV, computers, smartphones, tablets, video games, or social media.¹¹ Adapting to modern media consumption necessitates a mobile health (mHealth) intervention that can monitor and intervene on both traditional television viewing as well as streaming and other screen time consumed through tablets and smartphones.

To address the lack of longer-term rSST interventions among adults, we will use the multiphase optimization (MOST) framework to develop and evaluate a mHealth behavioral intervention (Figure 1). MOST, conceptually rooted in engineering principles, is used to efficiently and systematically develop multicomponent behavioral interventions and has three essential phases: preparation, optimization, and evaluation.¹² This paper describes the StandUPTV study, including its completed preparation phase and the clinical trial protocol for the forthcoming optimization phase, including a 16-week mHealth behavioral intervention to reduce rSST in adults.

Fig. 1.

StandUPTV multiphase optimization strategy for sedentary screen time. Adapted from Collins et al., 2018.

2. METHODS

2.1. Preparation Phase.

The purpose of the preparations phase was to derive a theory and empirically derived conceptual model, identify and adapt candidate intervention components for the mHealth approach and select the optimization objective.

Component selection and conceptual model.

From the literature and our own work, we developed a conceptual model (Figure 2) derived from social cognitive theory (SCT)^13,14 underpinnings of behavior change and identified three candidate intervention components. Each component has distinct behavioral pathways that have been outlined in our conceptual model. We also noted several key potential moderators that would impact the function of the pathway effectively across different ages, sex, and race/ethnicity categories.

Fig. 2.

Conceptual model for the StandUPTV intervention.

The first component identified was rSST electronic lockout, which refers to restricting TV through electronic means. This strategy has been tested previously and has shown to reduce TV viewing by >50% (128 min/day) among middle-aged adults.¹⁵ Electronic lockout—alone and in combination with SCT-based behavioral strategies (i.e., self-monitoring, stimulus control)—has been shown to be feasible and acceptable to participants.^15,16 Therefore, we hypothesize that the lockout component in StandUPTV (LOCKOUT) will encourage users to plan their rSST and increase their barriers-related self-efficacy for reducing rSST.

The second component identified was adaptive prompts, which refer to adaptive messaging based on the most recent rSST bout. Our preliminary data suggest, on average, prompts increased the likelihood of transitioning from sedentary to more active behaviors by 42% (unpublished). Given emerging evidence on adapting prompts to user-specific contexts, we adapted prompts based on time of day, message type, and acute rSST behaviors.^17,18 Therefore, adaptive prompts in StandUPTV (TEXT) will provide education and encouragement to users based on the anticipated benefits of reducing rSST, thus targeting task-related self-efficacy and outcome expectations for reducing rSST.

The third component identified was earning rSST through MVPA participation. The conceptualization of this component leveraged prior studies in youth that used a behavioral systems approach to make rSST contingent on physical activity in an “open-loop” behavioral feedback system,^19–21 successfully modifying both behaviors. Compared to baseline, rSST in youth was reduced by 68 min/day over the 16-week intervention, which was sustained at 12-month follow up (31 min/day). This approach leverages the strong reinforcing value of rSST and provides autonomy over available rSST.^19,22 The StandUPTV EARN component is designed to provide additional rSST as reinforcement for successfully completing bouts of MVPA, increasing behavioral autonomy and intrinsic motivation.

Preliminary User Experience and Design Study.

We operationalized the components into an mHealth application using a community-embedded iterative design framework adapted for use in mHealth applications.^23–26 We engaged target users (N=10), in conjunction with our mobile application programming team in a “think aloud” data solicitation protocol²⁷ with a functional prototype of the core plus three candidate components. Details of the user design process are found in Appendix 1.

The user experience study consisted of three rounds of design and usability testing. In Round One, the goal was to ensure participants understood the definition of rSST, basic app features and presented with three options for displaying real-time rSST for self-monitoring purposes. Users consistently preferred a “gauge” view for self-monitoring that decreased as rSST was consumed. In Round Two, users reviewed StandUPTV functions, including a graphical display of their accrued rSST and intervention components (EARN, TEXT and LOCKOUT). Each intervention component was presented to the users by the design team using a combination of visual sketches and app prototypes. Users were consistent in their preference for the 1:3 ratio (e.g., earning 10 minutes of rSST for an increase of 30 minutes of MVPA) as an easier means of tracking and connecting meaningful bouts of exercise and typical rSST bout lengths. Finally, users were presented with a modified gauge for tracking rSST that incorporated earned rSST for those receiving the EARN component (Figure 3). In Round Three, users were provided the StandUPTV application to use in their homes for a week.

Fig. 3.

Screenshots of key app components for StandUPTV. Panel A shows sample educational content available in CORE, Panel B shows the gauge showing participants how much of their weekly target they have used, available in core. Panel C shows the self-monitoring gauge with additional visualizations around earned screen time and MVPA. Panel D shows daily social media time feedback for a complete week.

Identification of an Optimization Objective.

An optimization objective is an a priori defined goal of the intervention based on specific resource limitations or other constraints. Our optimization objective is to carry forward the smallest number of components that achieve a meaningful reduction in rSST with consideration of user burden and acceptability of the intervention package (i.e., all active components objective). We identified a reduction in rSST of 60 minutes per day as a minimal, but meaningful reduction of rSST based upon previous work demonstrating clinically meaningful improvement in chronic disease biomarkers.¹⁰

2.2. Optimization Phase: The Factorial Experiment.

We describe the clinical trial protocol of the factorial randomized experiment that we will conduct in the optimization phase. The primary aim is to examine the effects of the three individual candidate components on rSST in a 2³ factorial experiment yielding 8 experimental conditions (N=30 per condition), for a total sample of 240 individuals (Table 1). Participants will complete data collection at baseline, mid-intervention (8 weeks), and at post-intervention (16-weeks). This trial length was selected as sufficient to detect whether the candidate components produced initial and stable changes in rSST. The secondary aims are to examine how reductions in rSST may impact related health behaviors and outcomes: (a) time displacements into more active behaviors (e.g., housework, walking), activity intensities (i.e., light physical activity, MVPA), and sleep using device-based measures (activPAL3c; GENEactiv); (b) dietary intake, snacking, and mood using self-report and ecological momentary assessment; and (c) body weight and composition. We will also examine potential intervention moderators: age strata [23–39y v. 40–64y], sex, race/ethnicity, which may differ in response to the intervention due to differences in screen time consumption patterns. We will assess intervention mediators in line with our conceptual model (Figure 2). This trial was registered with ClinicalTrials.gov on July 9, 2020 (NCT04464993).

Table 1.

Factorial design for StandUPTV optimization trial (N = 240).

Experimental Condition	CORE -Educational Materials and Self-monitoring	TEXT	LOCKOUT	EARN	n per condition
1	YES	YES	YES	YES	30
2	YES	YES	YES	NO	30
3	YES	YES	NO	YES	30
4	YES	YES	NO	NO	30
5	YES	NO	YES	YES	30
6	YES	NO	YES	NO	30
7	YES	NO	NO	YES	30
8	YES	NO	NO	NO	30

Recruitment.

Recruitment will take place at two study sites (ASU and CalPoly). Our recruitment and retention plan will be standardized across both sites. Recruitment methods include newspaper advertisements, flyers on campus and throughout the local community, ResearchMatch.com, social media (i.e., Facebook, Twitter) and other social networking sites, and sending electronic newsletters and fliers to local businesses. ASU will target recruitment across the greater metropolitan Phoenix area, which is a highly urban/suburban environment, while CalPoly will target the San Luis Obispo region which has a small-city/rural designation. As the study was initiated during the COVID-19 pandemic, all study visits will be conducted virtually.

Target population and eligibility.

Our target population is female and males between 23- and 64-years old due to their high levels and heterogeneous media consumption (i.e., a mix of traditional television, streaming, and smartphone/tablet usage), as well as their elevated risk for developing chronic diseases. Enrollment and randomization will be stratified by age group (23.0 to 44.9y vs. 45.0 to 64.9y), with a target to recruit equal numbers in each age strata. The full list of eligibility criteria is shown in Table 2. Data will be collected and managed using REDCap (Research Electronic Data Capture) hosted at ASU.^28,29

Table 2.

Eligibility criteria for preparation and optimization phases.

Inclusion Criteria:

23–64.9 years of age

BMI≥25 kg/m2

3+ h/day of SST

Insufficiently active

Own an Apple or Android smartphone/tablet

Access to internet at home or unlimited data plan

Willing to download StandUPTV

Able to read/understand English

Willing to be randomized

Exclusion Criteria:

Current smoker (more than once per week)

Major life event (e.g., move out of study area, pregnant)

MVPA contraindicated

Occupation requires social media use or high amounts of physical activity

Drug abuse, major illness and/or psychological disorder

Participant enrollment and baseline procedures

Screening forms will be completed over REDCap. Eligible individuals will receive an email with the online version of the informed consent document that was approved by the Arizona State University Institutional Review Board (IRB #00012109) and complete a video conferencing visit with staff who will explain study procedures and expectations of study participants. Interested participants will sign and submit the informed consent online via REDCap. After consent, basic information (e.g., address and shipping preferences) are confirmed and participants are scheduled for a Zoom technology setup visit. Participants are mailed a technology kit that includes several devices for tracking rSST. Participants are provided with a Wi-Fi plug per television in the home (WeMo Insight Smart Plug), one WiFi router, one Raspberry Pi (a small computer to record and transmit data), one Samsung Galaxy tablet with pre-loaded apps for social media and streaming and a Fitbit Charge 4. Details on the technology kit, data synchronization and safety are found in Appendix 2. Participants will be encouraged to use the tablet for all screen time that is not otherwise consumed on their home television(s).

During the Zoom technology setup visit, participants will be given instructions on (a) how to navigate the StandUPTV mobile application on the tablet; (b) outfitting their home with necessary streaming services; and c) wearable devices (activPAL, GENEactiv) to be worn for seven consecutive days. We also confirm all devices are collecting and transmitting data successfully.

Baseline assessments will take place over 7 days. All screen time will be monitored and participants will be able to use the study tablet but will not yet have access to intervention content or feedback on rSST. They will complete study measures within this 7-day window. Participants will be required to have ≥4 days of measured screen time during the baseline period and will be asked to repeat baseline if required.

Randomization

A randomization process will be developed by the study statistician and implemented in REDCap using the MOST Randomization module.³⁰ In the randomization visit with the staff via Zoom, participants will be told to which experimental condition they were assigned, review their motivations for joining the study, and reasons for wanting to reduce their rSST. The staff will review the purpose of the study and participants will be informed of their target screen time for the intervention period calculated as a 50% reduction in rSST compared to the baseline week. The StandUPTV app will then be enabled; allowing them to see which components are “on” in the condition to which they were randomized. The staff will provide a brief tour of the app that will include reviewing one static content lesson of their choosing and demonstrating how to navigate to the gauge, tools, MyResults and explain features specific to the components in their assigned group (see Appendix 3 for experimental condition descriptions given to participants).

Intervention Components

A full description of the intervention components is in Appendix 4. The Core TEXT, LOCKOUT, and EARN components are briefly summarized below. With a total of 3 components, this

Core Intervention

A core intervention will be delivered to all participants that includes (a) a target to reduce screen time by 50% compared to their baseline values, (b) rSST self-monitoring tools (i.e., the gauge, graphs and charts) (Figure 3); and (c) 16 lessons including education and behavioral change content.

TEXT

Participants assigned to the “Yes” level of this component will receive regular app prompts (between 1 to 3 prompts per day). These prompts will be sent as notifications within the app and include a combination of static and adaptive messages based upon their current rSST. The overall conversational style of these prompts will be encouraging, but include a range of target messages including (a) encouraging use of the intervention strategies; (b) 7-day summarizes of rSST behaviors; (c) summarizes of specific rSST behaviors (e.g., “You’ve watched %X hours of screen time over the last 7 days in bouts of 1 hour or longer”); (d) general encouraging prompts regarding rSST (e.g., “A 1-hour show includes 16-min of commercials! Take advantage of this by using commercials as reminder to get up and MOVE!”), and (e) contextual messages based on time of day (“The sun is about to set, get outside to enjoy your last bit of light and reflect on the day!”). Participants in the “No” level of this condition will only receive prompts to complete assessments.

LOCKOUT

If rSST reaches the prescribed threshold (50% of baseline), we will use the screen limiting feature for both the traditional TV (via the WiFi Plug) and for those apps identified as rSST on the tablet to restrict any further rSST through the end of the week (Monday-Sunday). The 50% baseline allotment will be reinstated for the next week. Participants are granted a single “mulligan” per week that enables them to delay a lockout by 60 minutes (e.g., to finish a show). In addition to the physical lockout, participants receive a planning tool within the app. This tool provides an interactive interface for scheduling screen time up to one week in advance.

EARN

Participants can earn additional screen time by engaging in Fitbit-assessed MVPA in 10-minute bouts at a ratio of 1:3 (i.e.,10-min exercise earns 30-min rSST). The gauge display is modified for this component to emphasize the overall weekly rSST goal, MVPA, and their earning balance (i.e., the difference between total screen time and amount earned, see Figure 3). Within the app, they receive six educational lessons regarding MVPA (e.g., safely exercising, barriers to exercise). Participants can also plan bouts of MVPA up to one week in advance through an interactive interface.

2.3. Measures

All outcome measures will be assessed at baseline (prior to randomization), 8 weeks (interim), and 16 weeks (end of intervention period). A schedule of assessment is described in Table 3

Table 3.

Optimization trial outcomes and data collection schedule.

		Week
Outcome	Measure	0	8	16
	Health history, demographics	○
Primary outcome	SST (min/day)	●	●	●
Displacement effects/ causal factors	Activity intensity	●	●	●
	Activity type (ACT24)	○	○	○
	Sleep	●	●	●
	Dietary intake (ASA24)	○		○
	Snacking, mood	●●●	●●●	●●●
Mediators	e.g., Barriers, outcome expectations, and self efficacy, see Appendix 7.	○	○	○
Clinical Measures	Weight, waist circumference	●		●
User-experience	Application usage	●●●	●●●	●●●
	User burden and compliance		○	○
	Exit interviews			○

○Survey- or interviewer-administered; ●Device-based measure; ●●●● Ecological momentary assessed; SST = Sedentary screen time. Participants are asked to wear devices for 7-days during assessment weeks and devices are mailed back and forth at each time-point. Week 0 is baseline, prior to randomization.

Primary Outcome Measure:

The primary outcome measure is rSST. Sedentary time will be measured using the activPAL3c micro accelerometer (PAL Technologies Ltd, Glasgow, Scotland). This device is the most valid monitor for detecting sitting time with R² = .94 and mean bias of −2.8% against direct observation.³¹ The device will be waterproofed using a medical grade adhesive covering and attached to the midline of the thigh with breathable, hypoallergenic tape, enabling continuous wear for consecutive days without removing for bathing or other water-based activities.^32–34 Collected data will be processed into events of time in bed (primary lying bout), sitting/lying, standing, or stepping using the activPAL software version 8.10 and converted to minute-by-minute data. Each waking minute classified by the activPAL as lying/seated will be categorized as sedentary time.

We will assess screen time using a combination of direct measurement from WiFi Plugs to monitor television power state (Wemo Insight, Belkin, El Segundo, CA) and tablet app usage (Samsung Galaxy, Samsung). All television was considered recreational. We generated a list of 730 apps and study staff categorized them as video game, television/video, social media, or non-recreational based on our rSST definition (See Appendix 5 for details). The three streams of data (TV, tablet, activPAL) are merged at the minute-level for the primary determination of rSST. A minute must be labeled as both sedentary by the activPAL and recreational screen time by either the tablet or television to be considered labeled as rSST. Users can add, modify, or confirm screen time bouts through the app when screen time occurred outside of the home or on alternative devices, or reject bouts when another household member was viewing the screen or when screen time was not the primary activity.

Secondary outcome measures:

Posture and physical activity intensity (min/day) will be assessed using the activPAL at all assessment windows. Data will be classified as non-screen sedentary (i.e., sitting/lying without screen time), standing or stepping events. Stepping time will be split into periods of light-intensity physical activity ([LPA]; <100 steps/minute) and MVPA (≥100 steps/minute).³⁵ Participants will be asked to wear the device 24-hrs a day for 7-days and we will exclude periods of: (a) continuous sitting or standing behavior >6 hr (considered non-wear); (b) days with ≤10 hours of valid wear time during the wake period; and (c) participants with only one valid day of activPAL wear.

Daily behavioral composition will be assessed using Activities Completed Over Time in 24 Hours (ACT24), an internet-based previous-day recall that provides estimates of time (h/day) spent sleeping (in bed), specific active and sedentary behaviors and energy expenditure associated with these behaviors. ACT24 has been validated compared to direct observation and the activPAL.^36,37 and has been mapped to time-use categories,³² enabling examination of shifts in time-use in response to the StandUPTV intervention.

Body Composition. Body weight and fat percentage (%) will be measured using a Tanita BF-679 scale and waist circumference will be measured using a measuring tape (protocol in Appendix 6). The study staff will schedule a brief virtual assessment with participants to oversee and guide participants through completing these measures on their own and uploading screen shots of the measures.

Dietary intake will be assessed using 24-hour recalls on two random days during the baseline and 16 week assessment periods and complete using the Nutrition Data System Software (NDS)^38–40 by trained staff.^41,42 The primary variables of interest will be: calories, protein, carbohydrates, and fat. Underreporting will be estimated by comparing reported energy intakes with estimated total energy expenditure (TEE) for each participant based on estimated basal metabolic rate (adjusted for age, gender)⁴³ and level of physical activity.⁴⁴ If indicated, analyses will exclude extreme under-reporters.

Sleep duration and quality will be assessed using the wrist-worn GENEactiv accelerometer (Activinsights, Kimbolton, UK), which has been well validated for sleep.^45,46 Participants will be asked to wear the device 24-hrs a day for 7-days and complete a validated sleep diary each night to separate sleep from wake periods in the actigraphy data.⁴⁷ We will use an open source and validated algorithm to process device-based sleep measures, including sleep duration, sleep efficiency, sleep onset latency, and wakefulness after sleep onset (GGIR package, R-software).^46,48

Snacking and mood will be assessed with ecological momentary assessment (EMA), a methodology that uses brief surveys to assess real-time participant experiences.⁴⁹ EMA prompts will be randomly delivered (by time and day) during the measurement period, three times per day for 7-days. For snacking, participants respond to the prompt, “Have you had anything to eat or drink in the past 15 minutes?”, followed by additional items to assess food and drink type.⁵⁰ For mood, brief (<1 min) items modified from Dutta et al ^33,51 will assess stress, fatigue, happiness, and calmness on a 4-point Likert-type scale.

Mediators. Several SCT constructs targeted by StandUPTV components will be assessed (Appendix 7) using modified scales from validated measures of self-efficacy and outcome expectation for physical activity.^52–54 These modifications were informed by a previous sedentary behavior intervention⁵⁵ where we demonstrated high internal consistency (α>0.9).⁵⁶

Moderators. Age (23–44y vs 45–64y), sex, and race/ethnicity will be assessed as potential moderators on the baseline questionnaire.

User experience outcomes will be assessed with detailed StandUPTV app usage statistics across a range of interactions (e.g., number of times application is launched, use of self-monitoring feature). Tracking features built into the app will passively monitor interactions and identify what components of the app are used most frequently (and when). User burden, a key component of sustained intervention use, will be assessed on the User Burden Scale.⁵⁷ This scale measures burden across six domains (difficulty of use, physical, time/social, mental/emotional, privacy, financial) and can be completed in <5 min. Protocol compliance (i.e., whether participants circumvented our screen time limiting protocol) will be assessed with a questionnaire. Participants who report non-compliance will be encouraged to exclusively engage in screen time using our system. Post-hoc sensitivity analyses will determine whether non-compliant participants modified any results from the study. Finally, as is standard for mHealth app evaluation, and in line with our group’s previous behavioral interventions,^58,59 qualitative exit interviews will be conducted among all participants following the 16-week assessment period. All participants will be instructed to report any adverse events to study staff within 24 hours of the event occurring.

2.4. Statistical approach

We aim to enroll 240 participants to achieve a final sample of 200 (20% dropout; n=25 per condition) for the 16-week intervention. For models under the Primary Aim, N=200 will afford power=.80 to detect main effects for each of the intervention components (LOCKOUT, TEXT, and EARN) and resultant 2- and 3-way interactions, assuming a balanced experimental design and a modest baseline to post-intervention correlation (r=.3) and α=.05.⁶⁰ We have chosen to power this study on a Cohen’s d effect size=0.4 as this represents a 60 min/day reduction in sedentary behavior. This magnitude of change produced clinically meaningful improvements in chronic disease biomarkers in a recent review.⁶¹ This effect size is conservative, as previous rSST trials have observed effects d > 1.0,^15,16,62 and will provide adequate statistical power to detect our clinically important outcome of reducing rSST by 60 min/day.

Following an intent-to-treat model, multiple regression models will be used to test hypotheses regarding the main effects of three candidate intervention components on the primary outcome, rSST. Following recommendations for factorial designs, effect coding (will be used for experimental conditions to assess for interaction. We will use path analysis methods to explore mediational effects consistent with our conceptual model (Figure 2). We will examine all data for missing information and loss to follow-up. We plan to consult with a statistician regarding any missing data and will use multiple imputation as appropriate. User experience and user burden outcomes will be summarized using descriptive statistics (e.g., means, standard deviations, frequency). Exit interviews will be analyzed using an applied rapid qualitative analysis approach.⁶³

3. Discussion

rSST consumes most of the average American’s leisure time and observational studies have consistently found associations between rSST and poor health. Our long-term goal is to develop and optimize an efficient and efficacious intervention package using MOST for use in future studies. The lack of randomized trials reducing screen time has led to controversy about whether screen time (e.g., television) associations with disease outcomes are causal and whether public health recommendations for adults should specifically target screen time.⁶⁴ The purpose of this trial is to identify efficient and efficacious intervention strategies (i.e., useful intervention components) which can, in the future, be applied within well-designed experimental trials to differentiate which (or what combination) of causal and non-causal factors may explain the strong observed relationship between rSST and numerous health outcomes.

This study measures rSST in real-time using multiple technologies, wearable sensors, and self-monitoring, making it (to our knowledge) the most comprehensive examination of screen time to-date. We take a novel and rigorous approach by pairing real-time multi-stream 24-hour behavioral data (i.e., minute-level rSST, Fitbit-assessed MVPA, other device-collected data). Previous work^65–67 has heavily relied on self-report measures to collect rSST, with few studies also using mobile applications to capture smartphone use^68–71 or television sounds⁷² as proxies to assess screen time. Additionally, these studies only assess one specific screen-time context (e.g., smartphone use, TV time) and do not clearly distinguish rSST from other non-screen sedentary time. Research using device-based measures of specific behavioral context (i.e., rSST) is needed for a more comprehensive understanding of this behavior.

There are important limitations to consider. The technology set-up includes a rigorous on-boarding process to ensure all systems are operational, which may deter some participants from initiating or completing the study. Additionally, at the time of study initiation, the iOS platform was locked to external API access, so we are unable to obtain detailed screen time information from iPhones. As a work-around, we will provide participants with a Samsung tablet with all apps pre-loaded; however, transitioning participants to a new device increases burden and it is possible that some instances of rSST on the phone will be missed.

The US Health and Human Services 2018 Physical Activity Guidelines for Americans added a recommendation that adults “move more and sit less,” but clear knowledge gaps remain regarding types and duration of sitting and their impact on health.^73,74 The proposed optimization trial (and the preliminary preparation phase user design work) directly addresses these gaps by developing and testing innovative and contemporary strategies for reducing rSST, the most prevalent sedentary behavior. The results of the proposed study will ensure the resultant intervention package is efficacious and efficient before moving into a definitive RCT to evaluate the impact of changing rSST on health outcomes.