Wearable activity trackers for young people: usability, acceptability, feasibility, and behaviour change

Background: Physical activity (PA) has numerous immediate and long-term benefits for child and adolescent health and well-being. However, PA levels in children and adolescents are typically low, with few achieving the recommended 60 minutes of moderate-to-vigorous intensity physical activity (MVPA), on average, per day (in 5- to 18-year olds). To date, interventions targeting child and adolescent PA have been limited in their effectiveness. The use of technology (e.g. apps, pedometers) as PA intervention tools have received growing interest, given their increasing availability, affordability and popularity. Previous research has found that wearable activity trackers (wearables) can increase step counts and MVPA in adults, but less is known about their use, acceptability, feasibility, and ability to change behaviour in children and adolescents.

Thesis Aims: This thesis aimed to: use theoretical frameworks to investigate what factors impact the habitual use of wearables in children and adolescents (Chapter 2); systematically review the literature investigating the acceptability, feasibility, and effectiveness of wearables for increasing PA, in children and adolescents (Chapter 3); identify behaviour change techniques (BCTs), targeting PA, sedentary time and sleep, embedded within the wearable used in the original studies of this thesis (Fitbit Alta HR; Chapter 4); explore the acceptability of families using wearables (Chapter 5); use behaviour change and co-design approaches to develop a family-based wearable intervention to increase family PA levels (Chapter 6), and outline a protocol for implementing and evaluating the developed intervention (Chapter 7).

Methods: Chapter 2: Parents/guardians of 5- to 17-year olds completed a cross-sectional survey, with questions aligned with the theoretical domains framework (TDF) and capability, opportunity, motivation and behaviour (COM-B) model; Chapter 3: Six databases were searched. A narrative review and thematic synthesis were conducted to synthesise the findings, and BCTs in ‘effectiveness’ studies were identified using the BCT taxonomy v1(BCTTv1); Chapter 4: Two authors used the Fitbit Alta HR for four consecutive weeks and used the BCTTv1 to identify BCTs embedded in the Fitbit Alta HR, targeting PA, sedentary time and sleep; Chapter 5: Twenty-five families took part in a five week study where all family members received a Fitbit Alta HR for four weeks. Families’ acceptability of using wearables were explored using questionnaires and focus groups. The technology acceptance model (TAM) and TDF were used as theoretical underpinnings, and quantitative and qualitative data were integrated using the pillar integration process (PIP); Chapter 6: The behaviour change wheel (BCW) and co-design workshops with seven families and seven public health professionals were used to develop a family-based wearable intervention, using composite vignettes that represented barriers to families using wearables to be active (based on previous research; Chapters 3 and 5); Chapter 7: A protocol outlining the implementation and evaluation of the intervention, following recommendations by the Medical Research Council (MRC), was produced. 

Results: Chapter 2: Most children and adolescents had never used a wearable (n=429; 51.6%), and 252 (30.3%) and 150 (18.1%) currently or had previously used a wearable, respectively. Reasons for use included interest in technology, and to monitor PA levels, with reasons for disuse or discontinuing use including lack of interest and the cost of wearables. Child and adolescent demographics (age, sex), capability (knowledge of PA), opportunity (parent wearable use), and motivation (enjoyment of PA) were associated with child and adolescent wearable use; Chapter 3: Thirty-three studies were identified. There was some, but mixed, evidence that wearables increased step counts and MVPA, and reduced sedentary time, with no differences based on the number of BCTs incorporated in the intervention. Qualitative accounts suggested wearables increased motivation via self-monitoring, goal setting, feedback, and competition, however, children and adolescents reported technical difficulties and a novelty effect when using wearables; Chapter 4: The Fitbit Alta HR incorporated 25 BCTs, with most BCTs aimed at increasing PA (n=18) and increasing/improving sleep quantity or quality (n=14), with fewer aimed at reducing sedentary time (n=8); Chapter 5: Families’ acceptability of using wearables were synthesised into nine pillars reflecting the six key components of the TAM, and 12 components of the TDF. Families found the Fitbits easy to use and acceptable, but use varied, and perceived impact on PA were mixed, with external variables (e.g. work and school commitments) contributing towards this; Chapter 6: The ‘Move & Connect’ intervention was developed, which is a multi-component flexible intervention targeting family PA and wearable use via education, training, modelling, persuasion, incentivisation and environmental restructuring, and 24 BCTs. Components include Fitbit support resources, PA challenges, reviewing and amending goals and engagement prompts; Chapter 7: The ‘Move & Connect’ intervention is intended to be implemented and evaluated using a pilot randomised controlled trial (RCT) cross-over design, with its acceptability, feasibility, fidelity, and preliminary effectiveness, on family PA levels, co-engagement in PA and physical health, measured. 

Conclusions: The work conducted in this thesis represents some of the few studies exploring the use, acceptability, and feasibility of using wearables in children and adolescents. It is the first known series of studies to systematically develop a family-based wearable intervention, targeting PA (‘Move & Connect’), using theoretical underpinnings, previous findings, and co-design approaches. This intervention now warrants implementation and a detailed evaluation to demonstrate its potential feasibility and effectiveness. If shown to be feasible and effective, this intervention could have a positive impact on combating child, and parent, physical inactivity levels.