An online family-based self-monitoring and goal-setting intervention to improve children's physical activity: the FRESH feasibility trial and three-arm pilot RCT

Article history

The research reported in this issue of the journal was funded by the PHR programme as project number 15/01/19. The contractual start date was in September 2016. The final report began editorial review in May 2020 and was accepted for publication in December 2020. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The PHR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Copyright statement

© Queen’s Printer and Controller of HMSO 2021. This work was produced by van Sluijs et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2021 Queen’s Printer and Controller of HMSO

Benefits of physical activity

Regular physical activity in children is positively associated with a wide range of health benefits. 1,2 This includes favourable associations with cardiovascular and metabolic,3–6 skeletal7 and mental8,9 health. Improved cognitive and academic performance has also been shown to be associated with regular physical activity engagement. 10 Furthermore, harmful effects have been reported of excessive or uninterrupted sedentary behaviour, especially screen time. 11,12 Given that children and adolescents have been reported to engage in sedentary behaviours for between 6–913,14 and 5–8 hours per day,15,16 respectively, this is a particularly concerning issue. Inactivity in childhood tracks into adulthood,17 increasing the risk of diabetes, cancer and mortality. 4 The development of interventions to promote and maintain children’s physical activity levels is, therefore, a public health priority.

Children’s levels of physical activity and interventions

The UK’s Chief Medical Officers recommend that children and adolescents engage in an average of 60 minutes of moderate to vigorous physical activity (MVPA) per day. 18 The number of children meeting this guideline dropped between 2008 and 2012,19 and the most recent reports suggest that that around one-fifth of English children and youth aged 5–17 years met the recommended physical activity guidelines. 20 Using device-measured physical activity, Steene-Johansen et al. 21 reported that, across Europe, only 29% of children and adolescents were sufficiently active. Observational data show that children are less active after school and at weekends than during school time,22–24 and that as children enter adolescence their levels of MVPA decline steeply,25 particularly at weekends. 24 Last, physical activity levels vary by children’s home location, with indications not only that rural 9- to 10-year-old children are less active than their suburban counterparts,26 but also that their 4-year decline in minutes per week spent in MVPA is higher than that among children living in suburban or urban environments. 24

The issue of declining levels of physical activity is even more concerning when young people’s physical activity levels are examined separately by sex. Girls are less active than boys throughout childhood21 and their participation in physical activity declines more precipitously than that of boys during the transition to adolescence. 27 Therefore, it is critical for young people to develop positive physical activity habits, as an active childhood can track into adulthood. 17,28

An effort to, at a minimum, maintain sufficient physical activity levels should be considered a public health priority. Therefore, intervening in children’s physical activity before they reach adolescence may be an important strategy. 29,30 To date, the majority of research on young people’s physical activity promotion has focused in and around school time. For instance, a considerable amount of attention has been given to general school-based interventions,31–34 active transport to and from school,35 activity at recess,36 physical activity during physical education lessons37,38 and activity generated through after-school programs. 39 Focusing in and around the school setting is understandable because of near-universal attendance rates and the large portion of young people’s waking hours that are spent at school, which makes school an ideal place to target physical activity interventions. However, the effectiveness of school-based physical activity promotion has been limited,31–34 and out-of-school approaches should be explored.

Parents, the family environment and children’s physical activity

The socioecological model (SEM) of health40 posits that individual behaviour is influenced by factors operating at different levels of influence, including individual, intrapersonal and institutional. Beyond individual-level variables, these include those related to the school, neighbourhood and family environment. For example, children’s activity is influenced by the encouragement they receive from their parents, and modelled on their parents’ own behaviour, which is in turn affected by, for example, the time that parents have available for such pursuits, and access to recreational facilities. 41 Indeed, family factors consistently exhibit positive associations with children’s physical activity, particularly parental support and parental modelling. 42,43

The importance of positive parental role-modelling and direct parental involvement in/support (e.g. transport, co-participation and encouragement) of young people’s physical activity is well known. 29,43–47 A recent cohort study by Abbott et al. 48 reinforced the importance of parental role modelling for both physical activity and sedentary behaviour, demonstrating significant associations between preschool children’s behaviours and their parents’ behaviours. In addition, the authors observed a potentially important role of same- and mixed-sex parent–child relationships. 48 Furthermore, family support has been shown to be associated with physical activity at weekends,23,46 when young people are known to be less physically active than on weekdays. 49,50

Parents may also influence their children’s health behaviours through a variety of other mechanisms, including their general parenting style, their parenting practices (e.g. rule-setting, behavioural consequences, establishing behavioural expectations) and their control of the home environment. 51,52 Interventions that target both the child and the family are particularly effective,29,53,54 and without the involvement of family members it is unlikely that a change in children’s physical activity levels will be maintained long term. 44,55,56 Thus, targeting whole families may create a more supportive, synergistic environment for the promotion of healthy behaviours,29,57 from which wider family members may also be able to benefit. 42

Together, this evidence highlights the need for the promotion of young people’s physical activity to target the family, where wider family members may also be able to benefit. 42 That said, little is known about how best to engage families. 29,44,52 This is highlighted by Tremblay et al. ,58 who state that ‘the role of peers and parents in creating supportive environments for physical activity is unequivocal’ but conceded that they could not draw any firm conclusions from their 38-country comparison.

Previous evidence on family-based physical activity promotion

Family-based physical activity promotion has received less attention than the promotion of young people’s physical activity in other settings. In 2016, investigators on the current project published a systematic review, including a meta-analysis and a realist synthesis, in which we included 40 family-based physical activity studies. 29 The meta-analysis showed moderate efficacy in changing children’s activity levels, but only one high-quality trial was identified. Using a realist synthesis approach, it showed the value of using combined goal-setting with reinforcement in the context of family constraints; focusing on changing the family psychosocial environment, for example through the child as agent of change; and drawing attention to additional (non-health) benefits of spending time, such as family time. In addition, this review highlighted the generally low quality of the evidence base (including self-reported physical activity, small sample sizes and limited blinding), lack of post-intervention follow-up, issues with selection bias, recruitment and retention, and the lack of knowledge on how and why interventions may or may not work.

The review also highlighted that most studies focus only on promoting child physical activity, rather than considering the family as a unit that may work as a team to change behaviour. 59 Intergenerational, family-based programmes targeting, for example, early literacy or prosocial development have shown positive effects, and highlight the potential benefit of including multiple family members in an intervention to improve child health outcomes. 60

Theoretical background

In conceptualising an intervention to improve physical activity in children and families, the investigator team used a socioecological approach. 40 Specifically, the SEM provided a framework for the intervention components. Within this framework, behaviour change strategies were guided by self-determination theory (SDT). 61 Brief descriptions of the theories guiding intervention development and evaluation are provided in the following sections.

Self-determination theory

Self-determination theory is a motivational theory that has received significant empirical support in the context of health behaviour change61,65 and in the physical activity context specifically. 66–68 SDT makes a distinction between intrinsic and extrinsic forms of motivation. Intrinsically motivated individuals engage in a behaviour for its own sake (i.e. for the challenge or enjoyment). On the other hand, those motivated by extrinsic regulations engage in an activity to satisfy external demands that can be experienced as controlling or autonomous to varying degrees. 69 SDT posits that individuals move along a continuum as their extrinsic motives or reasons become more internalised they become more autonomous (or self-determined) to engage in behaviours over time (Figure 1).

FIGURE 1.

The self-determination continuum, showing the motivational, self-regulatory and perceived locus of causality from Deci and Ryan. 70 From ‘The “what” and “why” of goal pursuits: human needs and the self-determination of behavior’, Deci EL, Ryan RM, Psychological Inquiry, 1 October 2000, Taylor & Francis, reprinted by permission of the publisher (Taylor & Francis Ltd, http://www.tandfonline.com).

From ‘The “what” and “why” of goal pursuits: human needs and the self-determination of behavior’, Deci EL, Ryan RM, Psychological Inquiry, 1 October 2000, Taylor & Francis, reprinted by permission of the publisher (Taylor & Francis Ltd, http://www.tandfonline.com).

According to SDT, social environments that support individuals’ basic psychological needs (i.e. autonomy, relatedness and competence) are assumed to foster more autonomous motivational patterns. 71 When individuals are more autonomously motivated or self-determined, ‘they experience volition, or a self-endorsement of their actions’. 69 The highest level of self-determination is intrinsic motivation, whereby behaviours, such as physical activity, are performed for their own inherent rewards, such as enjoyment or challenge. 70

Specifically, SDT argues that there are basic psychological needs for autonomy, competence and relatedness, all of which are critical and universal among individuals for psychological health and to move towards autonomous motivation. 70 Satisfaction of these basic needs results in increased feelings of vitality and well-being. 72 Thus, Deci and Ryan’s concept of need support is what is thought to explain individual differences in the development of motivation across the lifespan. 70 Consequently, behaviour change interventions, including those in the area of physical activity, that enhance the satisfaction of participants’ basic needs may be particularly effective. 73,74

In summary, the broad purpose of SDT interventions is to assist individuals’ progress on the continuum towards more autonomous forms of motivation. Overall, when the complete SDT causal sequence (Figure 2) is used, it creates an intervention outline that has the potential to be quite powerful. 71

FIGURE 2.

The SDT process model for health behaviour change in intervention research from Fortier et al. 71 Reproduced with permission. This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reproduced with permission. This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Summary and rationale for the FRESH project

The above literature highlights the importance of physical activity promotion in young people. This was echoed by an international expert panel, who concluded that developing effective and sustainable interventions to increase physical activity among young people is a key research priority in children’s physical activity. 75 In addition, the National Institute for Health and Care Excellence76 in the UK has identified ‘The effect of community and family interventions on young people’s physical activity levels’ (p. 28) as an evidence uncertainty requiring further primary research.

Much of youth physical activity promotion has been predominantly targeted in and around the school setting; however, this project focuses on the important intrapersonal domain of the SEM. Moreover, most studies focus only on promoting child physical activity, instead of considering the family as a unit that may work together to change behaviour,59 despite the known potential benefit of including multiple family members in an intervention to improve child health outcomes. 60 The Families Reporting Every Step to Health (FRESH) intervention, based on extensive prior work, including input from families themselves, will target the whole family and will be able to investigate whether or not this approach is more effective than solely targeting the child. The project was proposed to show whether this approach is feasible and acceptable and potentially effective in changing whole-day physical activity levels of the child and their family members, informing a potential definitive evaluation.

The FRESH project received funding from the National Institute for Health Research (NIHR) Public Health Research programme in 2015 and commenced in late 2016. FRESH consisted of two phases: (1) intervention optimisation and feasibility testing, and (2) pilot testing. Both phases are described in this report, as is an additional project aimed at optimising family recruitment. As per NIHR definitions,77 ‘a feasibility study asks whether something can be done, should we proceed with it, and if so, how’. Feasibility studies are used to estimate important parameters that are needed to design the main study, but do not evaluate the outcome of interest. ‘A pilot study asks the same questions but also has a specific design feature: in a pilot study a future study, or part of a future study, is conducted on a smaller scale’. 77

Study aims and objectives of FRESH feasibility and pilot project

The investigator team identified several strategic and practical uncertainties that needed to be dealt with before a definitive evaluation of the FRESH intervention could commence. The project reported on here consisted of the feasibility and pilot phases of the FRESH trial to reduce these uncertainties. The results of this project were to inform the decision whether or not to proceed with a definitive trial. As stated in the original funding application, the overall aim of a future definitive trial would be ‘to establish the long-term effectiveness and cost-effectiveness of the family-based FRESH interventions to promote MVPA in 8-10 year-old children and their families living in rural Norfolk’.

The aim of the FRESH feasibility and pilot project was to assess the feasibility of delivering the FRESH intervention and its accompanying evaluation. The specific objectives are listed in Box 1.

The four main research questions addressed were:

1. In what ways does the FRESH intervention(s) need to be optimised prior to a definitive trial?

2. What is the feasibility and acceptability of the FRESH family-based physical activity promotion intervention and accompanying evaluation?

3. Which methods are valid and acceptable for measuring family physical activity?

4. What are the most effective and resource-efficient methods for recruiting families into obesity prevention programmes?

Progression criteria

The FRESH progression criteria were pre-defined at the grant application stage. The following parameters were to be used to inform progression to a definitive trial, taking into account qualitative findings on the acceptability of trial procedures:

• intervention adherence (> 75% of families uploading steps at least six times in the first 3 months of the pilot study)

• demonstrable feasibility of recruiting 20 families per month (based on pilot and accounting for increased staffing in a future definitive trial) and retaining 75% of index children at 1 year

• intervention optimisation feasible (identified adaptations are practical, affordable and acceptable)

• evidence to suggest that an adequately powered trial would require a feasible number of participants (n = 250 is considered logistically feasible and to provide sufficient power)

• discontinuation of trial arm based on evidence of harm or limited acceptability/feasibility

• positive expected net gain of sampling from definitive trial.

FRESH project study management

The overall FRESH project was managed by the FRESH project group, which was chaired by the principal investigator and consisted of all applicants, research associates working on the project, the study co-ordinator and a local stakeholder. Depending on the project phase, the project group met once every 1–3 months. Operational management was led by the FRESH operational group, consisting of the principal investigator, the study co-ordinator and the main research associate appointed on the grant.

At the start of the project, the FRESH Study Steering Committee (SSC) was established, consisting of seven independent members and the principal investigator. The independent members represented various scientific disciplines (young people’s physical activity promotion, public health, family-based interventions, health economics, physical activity measurement, feasibility and pilot trials) and included stakeholders (public health) and members of the public (including with expertise in web-design). The FRESH SSC met once or twice each year. Its stated role was to:

• oversee the development and co-ordination of research activities

• act as a sounding board and provide advice on research matters to ensure the long-term health, development and scientific value of the project

• advise on the continuation of the project after the completion of a pilot study.

At its first meeting, the SSC agreed that, in addition to the pre-established progression criteria, it would consider ‘changes in MVPA’ as ‘evidence of promise’ to inform progression to a full trial.

Introduction

Previous evidence indicates that home-based physical activity interventions are potentially more effective than those that require the family to travel to community or other intervention locations. 63,79 Furthermore, it is unlikely that any change in children’s physical activity levels will be sustained long term without the active involvement of wider family members. 44,55,56 Many previous studies, however, focus only on promoting children’s physical activity instead of considering the family as a unit that may work together to change behaviour. 59 Calls for physical activity research in young people and families highlight the dearth of research in this area76 and the need to develop and evaluate innovative interventions targeting children and families.

Responding to this challenge, we sought to identify and develop a family-based physical activity intervention and evaluation. In this chapter, we describe the development of the FRESH intervention and recruitment strategy, and the protocol of the FRESH feasibility study. The aims of this study were to (1) assess the feasibility and acceptability of the FRESH recruitment strategy, intervention (including intervention fidelity) and accompanying outcome evaluation; and (2) explore how FRESH could be optimised through a mixed-methods process evaluation.

Methods

Study setting

Families were recruited from rural Norfolk, a county in East Anglia, UK (Figure 3). Norfolk has an area of 2074 square miles and an estimated population of 898,400. 82 About half of the population live rurally;83 rural–urban disparities in physical activity have been reported. 24,26 In accordance with the Office for National Statistics84 classification, ‘rural’ was defined as having a postcode falling in a small town, village, hamlet or dispersed settlement.

FIGURE 3.

Map of England showing study location of FRESH feasibility and pilot studies (note that the feasibility study was conducted in Norfolk only).

Description of the FRESH feasibility study intervention

In brief, FRESH was primarily a goal-setting and self-monitoring intervention aimed at increasing physical activity in whole families. The SEM (individual and interpersonal levels)40 and family systems theory85 provided a framework for the intervention components. Within this framework, behaviour change strategies were guided by SDT. 61 A detailed description of the FRESH intervention components and associated behaviour change techniques,86 targeted SDT constructs and hypothesised mediators is provided in Table 1. In addition, the FRESH feasibility study logic model can be found in Figure 4.

TABLE 1 - Summary of the FRESH feasibility study intervention components, their proposed frequency and their theoretical basis

PA, physical activity.

The index child is the child aged 8–10 years in the family.

Intervention components	Dose	Description	Behaviour change techniques	Targeted SDT constructs	Hypothesised mediators
1. ‘Family time’	Minimum once per week, 10–20 minutes	‘Family time’ provided an opportunity for index childrena and family members to plan PA, monitor their week’s steps, discuss any potential PA barriers and strategies to overcome them by logging in their family action planners.27 Regular ‘family time’ was hypothesised to provide index children with: A consistent (re)structured environment, where family members supported index children in choosing an optimally challenging and realistic goal (reflected as an easy, moderate, or difficult challenge on the FRESH website), breaking down goals into proximal (daily) and distal (weekly) step count targets, and providing praise and positive feedback on progress towards those goals. These strategies provide direct support for participants’ perceived competence32 Consistent parental involvement which directly facilitates relatedness. 32 Parental involvement (via co-participation in PA) may also positively affect family connectedness33 An opportunity for consistent autonomy support. Autonomy support has been shown to directly support participants’ autonomy and indirectly support their basic needs for competence and relatedness34 Additionally, index children were named their family’s team captain (i.e. change agent) where they were in charge of initiating ‘family time’. Evidence suggests that children may elicit changes to the psychosocial environment;28 therefore, promoting the index children to the role of family ‘team captain’ may strengthen child buy-in, perceived autonomy, and improve intervention fidelity	Goal-setting Self-monitoring Positive feedback on progress Social support Praise Positive reinforcement	Perceived competence Perceived relatedness Perceived autonomy	Family social norms for PA PA awareness Basic needs satisfaction PA motivation
2. FRESH website	Minimum once per week, 5–20 minutes	The FRESH website facilitated self-monitoring of step counts, and goal-setting through selecting challenges. Specifically, the FRESH website allowed families to choose one of three target cities to ‘walk to’ weekly, with the aim to eventually ‘walk’ around the world. Each week, families chose an easy, moderate or difficult challenge, which represented a 0%, 5% or 10% increase, respectively, relative to the average steps they had taken in preceding weeks. Increases were adjusted to 0%, 2.5% and 5% once adults and children accumulated an average of 10,000 and 12,000 steps per day, respectively. Families also had access to a general resources area that provided suggestions of activities that families could do together and a map to give a visual representation of the locations families had travelled to	Goal-setting Self-monitoring Positive feedback on progress Rewards	Perceived competence Perceived relatedness Perceived autonomy	Social support Family social norms for PA PA awareness Basic needs satisfaction PA motivation
3. Pedometry	Throughout intervention (6 weeks)	Participants were provided with pedometers for self-monitoring and immediate feedback. Pedometers are simple to use and convenient and are associated with effective interventions for increasing parent–child physical activity.35 Index children logged their steps (and their family members’ steps) into the FRESH website and/or onto the family action planners, which allowed participants to view their progress towards their proximal and distal step goals	Self-monitoring Immediate feedback	Perceived competence Perceived autonomy	Social support Family social norms for PA PA awareness Basic needs satisfaction PA motivation
4. Virtual rewards/competence reinforcement	Approximately once per week (6 weeks)	To praise effort (i.e. competence reinforcement), participants received supportive messages, virtual passport stamps (i.e. virtual rewards) and access reinforcement materials (i.e. interactive multimedia information about the cities they have visited) on the FRESH website as they completed challenges to various cities around the world. Participants received 2–4 passport stamps for completed challenges (i.e. as difficulty increased, more stamps were awarded) and one passport stamp for an incomplete challenge	Feedback on progress Rewards	Perceived competence	Basic needs satisfaction PA awareness

FIGURE 4.

The FRESH feasibility study logic model.

To initiate intervention participation, a facilitator visited all families a week after baseline assessments for a ‘kick-off’ meeting to introduce the families to the intervention components and accompanying materials (e.g. family action planner). The main purpose of this meeting was to familiarise families with the website and prompt them to schedule regular ‘family time’ meetings (a suggested minimum of one per week) during which they would review and update their family action planner. All meetings occurred in participating families’ homes and lasted approximately 1 hour. Participant-initiated distant support was available for the duration of the intervention.

A detailed description of the FRESH intervention components can be found in Table 1. At the start of each new weekly challenge, families had ‘family time’, during which they selected a challenge on the FRESH website and filled in their action planners. The FRESH website allowed families to choose one of three target cities to ‘walk to’ each week, with the aim of eventually ‘walking’ around the world. The FRESH website primarily facilitated the self-monitoring of step counts and goal-setting through selecting challenges of varying difficulty. In both study conditions, children were allocated the role of ‘team captain’, leading on destination selection and uploading steps online. Families were to wear their pedometers for as long as possible daily to capture their steps and were asked to upload their step counts at least once weekly. After completing a challenge, families received effort-praising messages and virtual rewards (i.e. virtual passport stamps) and were able to track their progress around the world and access reinforcement materials on the FRESH website (i.e. interactive information about the cities they had walked past and reached during their challenge). If a family did not complete a challenge, to praise their effort, they progressed to a hidden city along their challenge route and still received a supportive message, a virtual passport stamp and access to reinforcement materials. Completing a challenge (or the week coming to an end) initiated the next ‘family time’ meeting, when the above cycle was repeated (see the cycle in Figure 4).

Recruitment and retention

Owing to intervention development delays, we were only able to deploy school recruitment strategies. Of 11 schools approached, three declined (too busy, n = 2; doing enough physical activity promotion already, n = 1), and three did not respond. Five schools with an estimated 437 eligible students in Years 3–5 agreed to disseminate the FRESH recruitment material (reach).

Figure 5 shows the flow of participants from the number of families assessed for eligibility through to the number analysed. Of those families reached, 6.4% (i.e. 28 families) expressed interest; initial interest came from 23 mothers and 5 fathers. Expressions of interest occurred at a rate of three or four families per week or five or six families per school assembly conducted. Fewer than half of the families expressing interest in participation subsequently signed up to participate in FRESH (n = 12 families), and these were enrolled at a rate of one or two families per week. All families were retained at the 6-week follow-up.

FIGURE 5.

Flow of participants in the FRESH feasibility study.

Of the 12 families enrolled, four were whole families and six were dyads (i.e. one parent and one index child); 32 family members participated overall. About two or three family members took part per family (range 2–4 family members); four families had an additional eligible adult, three families had an additional eligible child and one family had both. Table 4 describes the participant characteristics.

In focus groups, families were asked about the perceived challenges to recruitment, which revealed four challenges to be considered for optimising future recruitment. A brief description of the challenges is provided below, with supporting quotations in the first part of Table 5.

Intervention feasibility, acceptability, fidelity and optimisation

Intervention acceptability and fidelity

Findings related to feasibility of outcome evaluation

Data collection took a mean of 91.1 (SD 27.7) minutes per family at baseline and 77.1 (SD 24.5) minutes per family at follow-up. Overall, adults disagreed that there were too many measures and that data collection took too long. All children self-reported that they ‘liked’ being measured. With the exception of accelerometer/GPS and step test assessment (one refusal each), all participants completed all measures at baseline. At follow-up, 91% of participants accepted an accelerometer/GPS and completed the step test; 94% of participants completed all other measures.

At baseline, mean valid accelerometer wear time was 851.5 (SD 54.1) minutes for adults and 755.7 (SD 29.7) minutes for children. At follow-up, mean wear time was 843.1 (SD 78.6) for adults and 742.3 (SD 56.4) for children. The GPS provided a location for a mean of 750.6 (SD 191.4) (adults) and 646.2 (SD 189.0) (children) minutes at baseline and for a mean of 720.0 (SD 237.6) (adults) and 586.8 (SD 262.8) (children) at follow-up. Valid data (600 minutes) on ≥ 4 days (including one weekend day) was available for 83% of adults at both baseline and follow-up; this was slightly lower for children, at 75% and 67%, respectively. A visual inspection of wear-time data revealed a tendency for children to remove their devices at around dinner time, for parents to remove their devices after their child had gone to bed, and for families to put on their devices much later in the day at the weekend than on weekdays.

An initial assessment of family functioning via the video-recorded Fictional Family Holiday activity showed poor to moderate data quality, as discussions were limited and cursory. Three factors may have affected data quality: (1) most families enrolled were dyads, limiting opportunities for whole-family discussion; (2) providing families with a planner to write out their itinerary may have shifted the emphasis away from open-ended discussion; and (3) the activity was completed at the end of the visit, when participants may have been fatigued from data collection.

The physical activity-related expenditure questionnaire developed for this study appeared to have appropriate face validity and was capable of providing rich data related to membership fees and subscriptions (e.g. for sports clubs, fitness centres, after-school clubs) and sports equipment (e.g. sportswear, gadgets).

Optimising recruitment

Previous literature has identified family-based recruitment as being particularly difficult. 44,112 Our formative work30 and other studies (see a review by Morgan et al. 52) recommend a multifaceted recruitment strategy in family-based research. Owing to unforeseen delays, we were unable to employ our planned multifaceted recruitment strategy, which likely contributed to our under-recruitment of families (60% of targeted 20). Of the families enrolled, only one-third included all family members. There was some suggestion that this may have been because of either a lack of confidence in physical activity or a reluctance to be measured. Improved messaging is, therefore, required early in the recruitment process to reassure low-active families, and individual family members, that FRESH is tailored to their activity levels and to highlight that they have the option of opting out of (parts of) the measurements. Allowing family members to be involved in the intervention, regardless of their participation in the evaluation, as was done in FRESH, may improve effectiveness and long-term behaviour change. 44,55,56,59

Interestingly, our findings showed that fathers appeared to be interested in participating in FRESH, but only 5 out of 28 expressions of interest were initiated by fathers. This may be because, among heterosexual parents, tasks such as making telephone calls (e.g. to express interest) or family event preparation (e.g. study participation) are more likely to be performed by mothers than fathers. 113 Therefore, recruiting whole families, whereby any parent can initiate an expression of interest, may be an important catalyst for the inclusion of more fathers in family-based research.

Other key areas of improvement to recruitment include optimising the conversion from children reached to families expressing an interest (e.g. extending the age range of index children to cover the whole of UK Key Stage 2 (Years 3–6, covering ages 7–11 years); reducing the burden on children to explain FRESH to their parents (e.g. by directing parents to a video); targeting adults directly via community- and employer-based recruitment or social media; and obtaining recruitment support from local organisations.

Optimising the FRESH intervention

FRESH was designed as a goal-setting and self-monitoring intervention, aimed at increasing family physical activity. Encouragingly, these behaviour change techniques resonated with most families and align with recommendations to increase family physical activity. 29 Participants reported being aware of what their daily step goals needed to be in order to complete their weekly challenges. Interestingly, the challenge context did not seem to be important to participating families (i.e. choosing challenge cities to walk to virtually). Instead, focus group interviews revealed that meeting daily step goals, completing weekly challenges and intrafamily competition appeared to be the key drivers motivating families throughout the intervention period.

We found that families did not implement all of the intervention components as intended, and strategies may be needed to improve intervention delivery and families’ fidelity to the intervention protocol. For instance, most families did not select new challenges on the FRESH website together during ‘family time’, and families used their action planners only to log their steps and not to also identify family activities or upcoming challenges for the week ahead. This may be resolved by the facilitator placing a greater emphasis on ‘family time’ and helping the family to schedule this at the ‘kick-off’. Facilitators are critical to the delivery of interventions, and a recent review found that facilitators have an important moderating influence on the effectiveness of any programme. 114 Ongoing follow-up with the facilitator would also support this. Other strategies to improve intervention fidelity include sending e-mail reminders to log steps, adding competition elements to the website (e.g. a leaderboard), providing more regular feedback/praise from the research team, and including small, tangible rewards.

Although the FRESH intervention was well received overall, in our focus groups it was evident that families, and in particular fathers in the child-only arm, would have preferred to have their whole family involved in FRESH. Consideration, therefore, should be given to discontinuing the child-only arm. Furthermore, the finding that fathers were particularly interested in having their whole family participate in FRESH is noteworthy. Fathers have an independent influence on their children’s health and development115 and are an important influence on children’s physical activity,116–118 but they are very under-represented in family-based interventions. 119 Fathers’ engagement with FRESH is consistent with recent evidence that fathers are more willing to participate in family-based interventions when the focus is on their children,120 and as a result reported newfound enjoyment for family-based physical activity and a desire to be a positive role model. 121 The online delivery of FRESH may have also appealed to some fathers. 120

Optimising measurement

The duration of data collection at both time points was in line with our estimates, and acceptability of the duration and number of measures was high among both adults and children. Nevertheless, the duration of the data collection process may have acted as a barrier to participation. Minor adjustments are needed to improve the quality of data from the Fictional Family Holiday, and monitor wear time. For example, for the Fictional Family Holiday paradigm, a greater emphasis on recruiting whole families, removal of the written aspect of the activity and shifting the order of measures, so that the assessment occurs earlier during data collection, might improve the quality of the family functioning data. To improve monitor wear time, emphasis should be placed on participants wearing the monitors for as long as possible from the time that they wake up until the time that they go to sleep, as opposed to going by their children’s bedtime. In addition, reminders (e.g. e-mail, telephone) could improve wear time,122 particularly at the end of the week to improve weekend wear.

Study strengths and limitations

This study is among the first physical activity interventions that aim to target and measure whole families, providing novel evidence in an area where more primary research is needed. 76 The phased approach of assessing feasibility and acceptability to inform refinement for pilot study is in accordance with established guidelines. 86 Public involvement was used extensively to inform development and refinement of FRESH, as suggested previously. 123,124 Furthermore, our use of a mixed-methods design provides unique insight and context for our quantitative findings, assisting in identifying strategies to further optimise FRESH. Limitations include that we were unable to fully employ our recruitment strategy and did not have the opportunity to test the efficacy of recruiting families through community-based recruitment. In addition, only one-third of families enrolled in FRESH included all family members.

Conclusions FRESH feasibility study

This study demonstrates feasibility and acceptability of the family-targeted FRESH intervention and provides valuable suggestions for further optimisation. This work informs a future pilot trial testing the impact of these adaptations and the preliminary effectiveness of FRESH on family physical activity. The findings of our upcoming pilot trial will inform sample size/power calculations for a future definitive trial, should the pilot study findings suggest that a definitive trial is warranted.

Introduction

The feasibility study described in Chapters 3 and 4 was among the first physical activity interventions to specifically target whole family engagement. The findings showed that it was feasible and acceptable to deliver and evaluate a family-targeted physical activity promotion intervention with high acceptability from participating families. However, as described in Chapter 4, multiple areas for further improvement of the intervention and its evaluation procedures were identified. This chapter describes the protocol of revised FRESH interventions and evaluation, evaluated in a pilot randomised controlled trial (RCT).

The primary aim of the FRESH pilot trial was to assess the feasibility and acceptability of the revised recruitment strategy, intervention and outcome evaluation. The secondary aims were (1) to explore the preliminary effectiveness of the intervention on potential outcomes of interest and (2) to assess whether or not prespecified criteria were sufficiently met to warrant progression to a full-scale definitive trial.

Trial design

The Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT)80 was used to guide the reporting of this study. We also used TIDieR81 to guide our description of the intervention. All protocol amendments from the FRESH feasibility study described below were made following consultation with the FRESH SSC and the funder (the NIHR Public Health Research programme).

The FRESH pilot study was a three-arm, parallel-group, randomised controlled pilot trial using a 1 : 1 : 1 allocation ratio with follow-up assessments at 8 and 52 weeks post baseline (Figure 6). Following baseline assessment, families were randomly allocated to one of three arms: (1) FRESH arm, (2) pedometer-only arm or (3) no-treatment control arm. All family members in the pedometer and FRESH arms received pedometers and generic family physical activity promotion information. The FRESH arm additionally received access to the intervention website. Those in the control arm were asked to carry on as normal.

FIGURE 6.

Flow chart of process of the FRESH pilot study.

Participants

Families were deemed eligible to participate in this study if there was consent from at least one child in school years 3–6 (aged 7–11 years, hereafter referred to as the index child) and at least one adult responsible for the index child. The adult had to live with the index child in the main household (i.e. the index child’s primary residence as indicated by the parent). No restrictions were placed on family type (e.g. single parent, shared parenting, inclusion of extended family living in the main household) and there was no maximum number of participants per family. If requested, we also enabled the inclusion of parents or siblings who lived outside the main household, or extended family members (e.g. grandparents) living in or outside the index child’s main household. All participants needed to be able to perform light-intensity physical activity (e.g. walking), have access to the internet and understand the English language sufficiently well to provide informed consent. Family members were able to take part in the intervention irrespective of their participation in the accompanying evaluation and vice versa. Exclusion criteria related to participation in specific assessments are outlined below.

The eligible age range of the index child was expanded to reflect the age range of children attending Key Stage 2 in UK primary schools, which meant that all children attending assemblies about the FRESH project were eligible to take part.

Setting

Families were initially recruited from rural Norfolk and Suffolk, counties in East Anglia, UK (see Figure 3). Norfolk is 2074 square miles in size and had a total estimated population of 898,400 in 2017 (most recent estimate)82 and Suffolk is 1466 square miles and had a total estimated population of 756,978 in 2017 (most recent estimate). 127 According to Norfolk and Suffolk County Councils,128,129 approximately 53% and 42% of the Norfolk and Suffolk populations, respectively, are classified as living in a rural area. Based on the Office for National Statistics84 classification, ‘rural’ will be defined as having a postcode falling in a small town, village, hamlet or dispersed settlement. In both counties, existing inequalities have been identified, including physical activity, obesity and other indicators of child ill health, school readiness and attainment. 24,26,130 Following recruitment challenges, the requirement for participating families to live in a rural location was dropped after consultation with, and approval from, the FRESH SSC and the funder.

Recruitment

To overcome the challenges encountered in the FRESH feasibility study, in the pilot study we used a multifaceted recruitment strategy, including targeting adults (parents) directly, and messaging that focused on the wider benefits of research participation (e.g. spending more time together as a family) as opposed to solely focusing on increasing physical activity or obesity prevention. 30 Recruitment was scheduled to be undertaken over an estimated 3 months (with a recruitment rate of ≈ 20 families per month), using two main strategies that target adults and children, as summarised in Figure 7. The first strategy involved recruitment in the school setting and the second in employer-based and community-based settings (e.g. Brownies/Cubs, community centres, GP surgeries). Alternative recruitment settings were also explored (e.g. online/traditional media) as needed, following the same procedure as in the second strategy. For logistical purposes, we aimed to find recruitment settings located roughly within an hour’s drive of Cambridge, UK.

FIGURE 7.

Overview of recruitment methods in the FRESH pilot RCT.

Randomisation

Randomisation occurred after the baseline assessments were completed; the unit of randomisation was at the family level (i.e. the index child and all participating family members). Families were randomised in blocks of six by an independent statistician using a computer-generated algorithm produced with Stata version 14 and stratified by county (i.e. Norfolk or Suffolk). This strategy ensured that families were randomised to the three different groups at an equal rate, and that similar numbers of families were allocated to each arm within the two counties. The random allocation sequence was implemented via a database created in-house on Microsoft Access by independent data management staff. A study co-ordinator used the database after baseline to determine which study arm a family was allocated to. No one directly involved in the evaluation had access to the allocation code or complete sequence.

Interventions

FRESH arm

As in the FRESH feasibility study, the intervention the FRESH arm received is primarily a goal-setting and self-monitoring intervention, delivered online, aimed at increasing physical activity in whole families. The intervention used concepts from the SEM40 and family systems theory85 and operationalised constructs from SDT61 to inform behaviour change strategies. The full FRESH logic model can be found in Figure 8.

FIGURE 8.

FRESH logic model (expanded version of logic model shown in Figure 4).

As in the feasibility study, 1 week after baseline measures, each family allocated to the FRESH arm had an hour-long kick-off meeting, which took place in the family home with a facilitator (who was a member of the study team). The facilitator introduced families to the intervention components and the accompanying materials (e.g. family action planner) and distributed pedometers. In the pilot study, families also received their first of four pieces of generic walking information, similar to the pedometer-only arm (described in Protocol for pedometer and control families). The main goals of this meeting were to familiarise families with the intervention website and their pedometers and to prompt weekly ‘FRESH time’ meetings (described in Table 1; rebranded from ‘family’ time, following feasibility evaluation) during which they were to complete their action planners and select a new challenge city to ‘walk to’ on the FRESH intervention website. All families received a follow-up telephone call 1 week after their kick-off meeting so that they could discuss any issues and ask any clarifying questions. Participant-initiated distant support continued to be available, whereby participants could contact the research team with questions or to report issues (e.g. website bugs, pedometer issues).

Table 7 provides a detailed description of the FRESH pilot intervention components and highlights where changes were made from our feasibility study. In each family, the index child (or children, if multiple) were designated ‘team captain(s)’, which involved taking the lead in selecting challenges and uploading steps online. Evidence suggests that children may act as change agents to elicit changes to the psychosocial environment;29 therefore, promoting the index children to the role of family ‘team captain(s)’ may strengthen child buy-in and perceived autonomy, and improve intervention fidelity. All family members were given pedometers with instructions to wear them for as long as possible daily so that their steps would be captured during challenges. The team captain(s) led weekly ‘family time’ meetings. During these meetings, families completed family action planners and accessed the FRESH website, which enabled them to choose weekly step challenges. Family action planners prompted families to plan weekly family physical activities to help them meet their step challenge for a given week. It was intended that families would plan activities they would do together; however, participants had the flexibility to also set individual goals. The action planners also prompted families to monitor weekly step counts, and discuss any potential upcoming barriers for physical activity and strategies to overcome them. Pedometers are simple to use, convenient and associated with effective interventions for increasing parent–child physical activity. 131

TABLE 7 - Summary of FRESH pilot trial intervention components

PA, physical activity.

The index child refers to the child aged 7–11 years in the family.

Text in bold indicates new elements added for the FRESH pilot trial, i.e. post feasibility study.

Adapted with permission from Guagliano et al. 78 This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (https://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. The table includes minor additions and formatting changes to the original table.

Intervention components	Dose	Description	Behaviour change techniques	Targeted SDT constructs	Hypothesised mediators
1. ‘FRESH time’	Minimum once per week, 10–21 minutes	‘FRESH time’ was expected to provide a weekly (at minimum) opportunity for index childrena and family members to review, revise, and update their family action planners. Family action planners prompt families to plan PA, monitor weekly steps, discuss any potential upcoming PA barriers and strategies to overcome them. Index children will be allocated as their family’s ‘team captain’ leading in challenge selection and uploading steps on the FRESH website	Goal-setting Self-monitoring Positive feedback on progress Social support Praise Positive reinforcement	Perceived competence Perceived relatedness Perceived autonomy	Social support Family social norms for PA PA awareness Basic needs satisfaction PA motivation
2. FRESH website	Minimum once per week, 5–20 minutes	The FRESH website provided a place for families to self-monitor their step counts and set goals by selecting challenges of varying difficulty. Specifically, the website allowed families to ‘walk’ around the world by choosing one of three target cities to ‘walk to’ weekly. The challenges were framed as easy, moderate or hard, which represented a 0%, 5% or 10% increase, respectively, relative to the average steps they took in preceding weeks. Once adults and children accumulated an average of 10,000 and 12,000 steps per day, the step challenge increases were reduced to 0%, 2.5% and 5%, respectively Families also had access to:Families with an ongoing challenge received e-mail reminders to log steps 3 days and 1 day before an impending challenge ends	Goal-setting Self-monitoring Positive feedback on progress Rewards	Perceived competence Perceived relatedness Perceived autonomy	Social support Family social norms for PA PA awareness Basic needs satisfaction PA motivation
3. Pedometry	Throughout intervention (8 weeks)	All family members received pedometers to enable self-monitoring and provide immediate feedback. To allow families to view their progress towards their proximal and distal step goals, they were encouraged to log their steps on the FRESH website and/or in the family action planners	Self-monitoring Immediate feedback	Perceived competence Perceived autonomy	Social support Family social norms for PA PA awareness Basic needs satisfaction PA motivation
4. Competence reinforcement/rewards	Approximately once per week (8 weeks)	After completing a challenge or if the challenge week ended, to praise effort (i.e. competence reinforcement) children received personalised supportive letters in the post and messages on the FRESH website They also received small online and tangible rewardsChildren received two to four passport stamps/cards for completed challenges (i.e. as difficulty increased, more stamps were awarded) and one passport stamp/card for an incomplete challenge	Feedback on progress Rewards	Perceived competence	Basic needs satisfaction PA awareness

After a challenge week was over, whether or not a family had completed their challenge, they received personalised competence reinforcement messages praising their effort online. In response to the feedback from the feasibility study, reinforcement messages were also sent in hard-copy letters (addressed and mailed to all participating children in the family). In addition, both online and tangible rewards were given to participating children after a challenge week. The tangible rewards were sent to each participating child in the family and consisted of playing cards representing the cities the families had visited, which could also be used to play rock, paper, scissors. If a family did not complete a challenge, they still progressed to a hidden city along their challenge route, as opposed to the city they had chosen, and received the accompanying reinforcement message and reward, as described above. In the pilot study, we also implemented reminders as requested by some families in the feasibility study: families with an ongoing challenge would receive e-mail reminders to log steps 3 days and 1 day before a challenge ended. After every challenge week, the above cycle would be repeated, starting with the next ‘FRESH time’ meeting. Following the assessments at 8 weeks post baseline, families retained access to the website and their pedometers, and were able to continue using these for as long as they liked. There was also continued support in terms of website updates (e.g. leaderboard and parental resource updates) and participants continued to receive competence reinforcement letters and rewards.

Outcome evaluation measures

Table 8 outlines a summary of the measures taken as part of the FRESH pilot study, including the order of assessments and the estimated duration of each prior to data collection. Data collection was carried out by two trained research staff and occurred in participating families’ homes. Outcomes were assessed at baseline (prior to randomisation; April–August 2018), at 8 weeks post baseline (June–November 2018) and at follow-up (52 weeks post baseline; May–October 2019) for all consenting family members.

Process evaluation

A mixed-methods process evaluation was conducted after assessments 8 weeks post baseline. Using open-ended and four-point Likert-scale questions (with 1 = strongly disagree and 4 = strongly agree), adults self-reported their overall opinion of FRESH, the intervention components (FRESH arm only; Pedometer arm adults responded to pedometer-related questions only) and the measurements and suggestions for improvement. Children also self-reported on the above topics, responding to dichotomous ‘yes/no’ questions. In addition, semistructured focus groups were planned to be conducted online with 10 out of 20 FRESH arm families and 5 out of 20 families from each of the other two arms (i.e. pedometer and control). Focus groups focused on families’ experiences of taking part in the trial, perceived acceptability of individual intervention components, intervention fidelity, challenges/barriers encountered and suggested improvements, as appropriate based on study arm allocation. All focus groups were transcribed verbatim. We also explored families’ engagement with the intervention website (e.g. page views, challenges accepted/completed) and aspects of the recruitment process (e.g. recruitment duration, resources used, comparisons of recruitment strategies).

Patient and public involvement

FRESH was developed with substantial input from children and families from the public, as detailed in Chapter 2. Since the completion of the FRESH feasibility study, families from the public have been further involved with the optimisation of FRESH in a number of ways. As mentioned earlier, we sought the involvement of families from the public to develop a recruitment video; these families helped develop the script and acted in the video, which can be viewed at www.youtube.com/watch?v=UxUHN1JsjUM. We also asked two families to engage with the FRESH website and provide feedback to inform modifications that could be made.

Sample size considerations

As this was a pilot study, a sample size calculation was not performed. We planned to recruit a sample of 60 families, with a sample size of ≈ 180–240 participants, based on three or four members per family. Our estimated sample size is based on previous study experience30 and on sample sizes of previous pilot studies. 140,141

Progression criteria assessment

The prespecified criteria used to inform progression to a definitive trial have been outlined in Chapter 1, Progression criteria. Where applicable, quantitative and qualitative findings were taken into account to assess whether or not a criterion had been met. As detailed in Chapter 1, FRESH project study management, on the recommendation of the SSC, ‘changes in MVPA’ as ‘evidence of promise’ was to be considered in addition to the pre-established progression criteria to inform progression to a full trial.

Recruitment and retention

Table 10 provides a summary of the recruitment sources used in this study and Figure 9 shows the recruitment flow. Owing to under-recruitment, the recruitment period was extended from 12 to 24 weeks. Expressions of interest occurred at a rate of four or five families per week over the 24-week recruitment period. Approximately 77% of families expressing interest were eligible for participation and 48% of eligible families were enrolled with an enrolment rate of one or two families per week.

FIGURE 9.

Flow chart of recruitment and retention in the FRESH pilot study.

Of the 41 families enrolled, 73% were whole families (n = 30 families) and ≈10% (n = 4 families) were dyads of one parent and one index child. The remaining families either only included index parent–child dyads (n = 4 families) or the index parent, index child, and an additional parent and/or children (n = 7 families); however, in all cases there were additional adults and/or children living with the index child that chose not to participate in the study. We did not recruit any extended family members or any family members that lived outside the index child’s main household. Consent was obtained for 149 participants, averaging four members per family (range 2–6 family members) and included 39 mothers (95%), 31 fathers (76%) and 41 siblings (from 32 families with an eligible sibling; 78%). Eleven siblings were younger than the index child (i.e. aged < 7 years), 15 were in the same age category as the index child (aged 7–11 years) and 15 were older than the index child (aged > 11 years). Table 11 describes the participant characteristics at baseline. Notably, children in the family arm were older than children in the other two arms (FRESH, mean 10.1 years, SD 2.8 years; pedometer only, mean 8.6 years, SD 1.9 years; control, mean 8.9 years, SD 2.7 years) and fewer girls were allocated to the pedometer arm (FRESH, 50.0%; pedometer only, 17.4%; control, 48.3). Approximately 92% of adults reported being married or living as married, 94% of adults reported their ethnicity as white, and the mean age at which adults had finished full-time education was 20.5 (SD 3.5) years.

At the 8- and 52-week assessments, 98% and 88% of families were retained (family dropout: FRESH, n = 2; pedometer only, n = 2; control, n = 1). Participant loss to follow-up at 52 weeks comprised nine adults (FRESH, n = 4; pedometer only, n = 3; control, n = 2) and 11 children (FRESH, n = 4; pedometer only, n = 3; control, n = 4).

Intervention feasibility, acceptability, fidelity and optimisation

Most children reported that they liked taking part in the study (> 90%) and thought that it was fun (> 80%). Compared with the pedometer (45%) and control (39%) arms, a higher percentage of children in the FRESH arm (81%) reported doing more activities with their family at the 8-week follow-up. Table 12 shows adults’ overall perceptions of FRESH. Scores were generally positive and favoured the FRESH arm over the pedometer arm. In particular, adults agreed that FRESH encouraged their family to do more physical activity and made their family more aware of the amount of physical activity that they did. Themes related to family physical activity, physical activity awareness and parental modelling also emerged in our focus groups:

It was really fun, it pushed us to get our steps in and do more activities and sports together, you know. I never really thought about how many steps or exercise I’ve done to be honest, so since these [step] counters, I just look and go ‘3000 [steps] only? I have to do something’. So sometimes they [her three sons] come home and they see me . . . dancing, doing something, or skipping, they say, ‘what are you doing, Mum?’ [laughs] and I say, ‘I’m just putting effort in to get my steps’ and then they join me, you know. It just made your more aware . . . I even started walking for small shopping instead of driving just to get my steps up [laughs] . . . small things, you know, it just made you aware.

Mother 127, FRESH arm

The overall acceptability of the pedometers was fairly high among adults in both the FRESH and the pedometer arms (see Table 12). Families in both groups reported that wearing the pedometers had become habitual. One parent stated:

I think it’s become quite habitual now, we pick them up first thing in the morning and take them off last thing at night and they [her children] were quite happy to do that, so that was good from a parent point of view.

Mother 125, FRESH arm

A greater percentage of pedometer-only children than FRESH children self-reported that they liked wearing their pedometer (86% vs. 62%, respectively). In addition, compared with our previous feasibility study, we replaced a greater number of malfunctioning pedometers (despite using the same make and model of pedometers in both studies) and families’ preference for wearing wrist-worn pedometers was emphasised more strongly during focus group discussions.

Adults in the FRESH arm found the kick-off meeting useful (3.6 ± 1.0) to help them get started, felt that they had enough technical support if needed (3.9 ± 0.6) and found it feasible to schedule ‘family time’, albeit not consistently (see Table 12). The focus groups revealed that families rarely used their action planners. One parent said, ‘we probably didn’t fill that [action planner] in as much as we should’ve . . . we use that [action planner] more to actually record our steps’ (father 146, FRESH arm).

The majority of children in the FRESH arm found the website easy to use (93%), wanted to keep using the website (81%), enjoyed being their family’s team captain (70%) and did not find it too difficult to reach their step goals (65%). Overall, adults’ mean scores in relation to the intervention website were generally positive (see Table 12). In particular, adults strongly agreed that the website was easy to use and found various website elements useful (e.g. the step converter). Parents agreed that their child enjoyed receiving rewards and competence reinforcement after each challenge week (3.5 ± 1.2). When asked in focus groups about suggestions for improvement, families in the pedometer-only arm suggested elements that were delivered to the FRESH arm. For example:

I think if you can walk so many steps and it gets you to a place, like a country or something like that. So maybe there could be mini challenges like you walk to London or walk to Paris, you know, or something. Yeah, something like that would be probably quite good for you guys [referring to her children]. [. . .] We haven’t been around the world, but we’d like to go around the world. [. . .] I think that’s something you can add to this [study].

Mother 114, pedometer-only arm

Google Analytics data indicated that 59 users accessed the website (≈ 4 users/family), with a median of 2 [interquartile range (IQR) 1–5] sessions per user viewing about 5 (IQR 2–11) pages per session for approximately 7 (IQR 3–12) minutes per session. The most common behaviour flow was to log on, access the challenge page (to select a new challenge) and then access the steps page (to add steps to complete the challenge). Families selected an average of 11 challenges and completed nine of those. During a challenge, families uploaded steps at least once 86% of the time.

Findings related to the feasibility and acceptability of the outcome evaluation

Data collection took an average of 119.5 (SD 26.4) minutes per family at baseline and 95.0 (SD 16.7) and 82.3 (SD 35.8) minutes per family at the 8- and 52-week follow-ups, respectively. Overall, adults disagreed that there were too many measures [mean values of answers on a 4-point Likert scale, 1.5 (SD 0.7)] and that data collection took too long [mean values of answers on a 4-point Likert scale, 1.7 (SD 0.8)]. In the focus groups, families mentioned the convenience of home-based data collection and, in some cases, said that this had been essential to their participation. One parent said:

. . . it was a lot more convenient you coming to us and you guys being quite flexible in offering us multiple dates and times you could come . . . if you hadn’t come to us, we probably wouldn’t have participated.

Father 125, FRESH arm

In addition, > 80% of children reported that they ‘liked’ the measurement sessions. At each time point, > 90% of eligible adults and children completed all measures, except for the submaximal step test (86%) and the video-recorded activity assessing family functioning (89%).

Mean valid accelerometer wear time was 835.6 (SD 76.5) and 734.9.4 (SD 62.7) minutes for adults and children across time points, respectively. Valid accelerometer data on ≥ 3 days (including 1 weekend day) were available for 82% of adults and 77% of children at all three assessment points. On average, across time points, the GPS provided a location for 757.0 (SD 126.3) minutes and 541.6 (SD 200.3) minutes for adults and children, respectively.

Preliminary effectiveness

Children’s and adults’ levels of MVPA and sedentary behaviour at all time points are presented in Table 13. Subgroup analyses are given in Tables 14 and 15 and the data on family co-participation in physical activity are available in Table 16 and Figure 10. Children and adults either met or were close to meeting the recommended levels of MVPA at baseline, with the exception of children in the FRESH arm, who accumulated notably less MVPA than children in the pedometer and control arms. Within-family intraclass correlation coefficients ranged from 0.42 to 0.52 and were largely similar at the two cut-off points applied.

FIGURE 10.

Mean minutes of family co-participation in LMVPA (T1, baseline; T2, 8 weeks post baseline; T3, 52 weeks post baseline). (a) Child with sibling; (b) child with parent; (c) parent with other parent; and (d) parent with child.

Among children, no notable between-group differences were found in minutes of MVPA, time spent sedentary or co-participation in physical activity with family members at either of the follow-up time points. However, at 12 weeks a sizeable difference in change of 9.4 (95% CI 0.4 to 18.4) and 15.3 (95% CI 6.0 to 24.5) minutes in MVPA was found for adults in the FRESH arm compared with those in the pedometer or control arm, respectively. Adults in the FRESH arm also increased their activity with other family members (co-participation) more than those in the control and pedometer arms (10.0, 95% CI 2.9 to 17.1 minutes, and 10.6, 95% CI –3.6 to 17.6 minutes, respectively). No between-group group differences were found for time spent sedentary among adults.

Exploratory subgroup analyses (see Tables 13 and 14) showed a greater decline in MVPA among FRESH girls and FRESH children from less deprived areas than among their counterparts. The latter group also showed a greater increase in sedentary behaviour than those in the other groups. By contrast, FRESH adults, in particular men, showed a greater increase in MVPA at 8 weeks than the other groups.

Tables 17–19 display the findings for children and adults for all other outcomes. No notable or consistent between-group or subgroup differences were found for any other outcome measured at 8 or 52 weeks for children or for adults.

Evaluation of costs

The proportion of families who bought any sports items was materially unaltered throughout the study. Table 20 reports the costs incurred by families and the intervention cost (incurred by local authority budgets). There was no difference across groups in the cost observed at baseline. The summation of the costs from randomisation until 1 year showed that for families enrolled in the FRESH arm expenditure was on average £157.92 (95% CI –£154.76 to £484.79) more than the no treatment group. The majority of this cost difference is accounted for by cost of the intervention. Conversely, the cost incurred by the families receiving only the pedometers tended to be smaller than that incurred by the families randomised to the control arm (B: –£90.50, 95% CI –£301.30 to £104.45). Adjusting these differences by the cost at baseline did not qualitatively change the findings.

Assessment against progression criteria

Table 21 outlines the prespecified progression criteria, whether or not the study team considers that the specific criterion was met, and the evidence to support that assessment.

Discussion of the FRESH pilot study results

Our findings show that it was feasible to deliver and evaluate a family-targeted physical activity promotion intervention with generally high acceptability from participating families. In addition, each of the prespecified progression criteria was met at least partially (see Table 21). However, we found a favourable indication of effectiveness only for adults and not for children, that is, a sizeable positive change in MVPA for adults in the FRESH arm compared with the other arms. The between-group difference found for adults’ time spent in MVPA was not maintained at the 52-week follow-up, and we also found no notable between-group differences for any other outcome measured at either time point.

Family recruitment posed a substantial challenge, and this progression criterion was not entirely met (i.e. recruiting 20 families/month). Our average recruitment rate was ≈ 7 families/month (range 2–15 families/month), despite using a multifaceted recruitment strategy that targeted adults and children, included a wide range of settings, and used direct and indirect recruitment strategies. A review of 73 publicly funded trials in the UK (through the NIHR) found that only 55% recruited 100% of their target sample size within their pre-agreed timescale and nearly 45% received an extension of some kind. 145 There is little evidence of major improvement in recruitment into experimental research over time. 146 Several studies have reported that the recruitment of families is particularly challenging,44,52 and we have described specific recruitment challenges that we have encountered previously. 78 Little methodological research is currently available to support the optimisation of family recruitment methods. Therefore, further research into how to recruit families to family-based research is needed.

In terms of recruitment, 94% of adults reported their ethnicity as white. Although this figure is reflective of the population of the counties where recruitment occurred,82 the potential effectiveness of this intervention on minority families is unclear. Several studies have acknowledged the under-representation of minority groups in trials. 147,148 Therefore, further research to better establish how to recruit families in family-based research is needed, and in particular, greater consideration should be given to recruiting families from ethnic minority groups. Targeting specific recruitment settings and tailoring messaging on recruitment materials are strategies that could be used. 148,149

An extensive measurement protocol was applied in the FRESH studies, and it is not possible to disentangle whether the challenges of recruiting families were because families lacked interest in increasing their physical activity or in FRESH in particular, or that the commitment to three rounds of home-based assessment of all family members was a barrier. Families in both the FRESH feasibility and the pilot studies indicated that the level of measurement was acceptable to them, but this is likely to have been a biased perception of a group of families that had made the commitment to take part in the FRESH study. Further research is needed to identify whether families may not be interested in physical activity promotion per se, or whether the research commitment required poses a barrier. With this in mind, researchers and funders should carefully balance the scientific need for detailed data collection (driven, for example, by questions around how interventions work, and impacts on important physical health outcomes beyond the target behaviour) with the burden on participants and the impact that it has on the recruitment of a representative sample of participants.

Encouragingly, we found evidence of preliminary short-term effectiveness for adults and, in particular, for fathers in the FRESH arm. Similar interventions with mothers have resulted in positive physical activity promotion. 150 However, the effect on fathers may be noteworthy as evidence indicates that fathers have an independent influence on their children’s health and development115 and an important influence on children’s physical activity. 116–118

We did not find evidence of preliminary effectiveness for children or for co-participation in physical activity between parents and their children in this study. This may be for a number of reasons. First, our process evaluation and focus groups revealed that family planning time was not being implemented as intended. In a family-based physical activity intervention that included a similar planning component, the authors found that children’s MVPA significantly increased in the short term compared with a condition that received education only. 151 Therefore, without implementing the planning component in our study, the step challenges alone may have not been enough to change children’s MVPA. There were also group differences in children’s sex and age, with fewer girls in the pedometer-only arm and more older children in the FRESH arm. Observational data reveal that children’s physical activity declines with age,22–24 girls accumulate less physical activity than boys throughout childhood,134,152 and girls’ physical activity declines more precipitously than boys’ with age. 49,153,154 These differences may have affected preliminary intervention effectiveness on MVPA, but this issue would likely be resolved through randomisation in an adequately powered trial. Last, there may have been a healthy volunteer bias, as participants across groups were generally already meeting the physical activity recommendations at baseline. In future, excluding families that are sufficiently active could be considered.

Delivery of the FRESH intervention was estimated to cost £90 per family (≈ £15 per participant), including pedometers for all family members, the face-to-face kick-off meeting and personalised follow-up support. The last of these accounted for ≈ 55% of the costs. These costs could be reduced in future, as this part of the intervention delivery had not been automated but was processed manually by research staff. The automation of these processes will help reduce delivery costs, and make it more attractive to funding agencies to consider delivering FRESH as part of their portfolio of physical activity interventions, if it is proven effective. Previous work has estimated the cost of delivering a multicomponent school-based physical activity intervention at ≈ £190 per participant,155 and an after-school intervention at £51 per participant,156 suggesting that cost of delivering the FRESH intervention is low in comparison. However, the definitions, scope and perspectives of the costs considered across studies were quite different, making comparison difficult. Moreover, little is known about how much local authorities or other delivery agents are willing to pay, and future research should explore this.

Conclusion

In conclusion, this study demonstrates feasibility and acceptability of the family-targeted FRESH intervention, as well as satisfying all progression criteria, at least partially. However, we failed to recruit the target sample size and did not find a signal of effectiveness on MVPA particularly long term or in children. Therefore, further refinements around intervention delivery and recruitment may be required prior to progressing to a full-scale trial.

Introduction

As described before, physical activity promotion interventions targeting families may be a valuable way to increase physical activity among children. 158,159 However, the study team’s experience in the FRESH feasibility and pilot studies demonstrated that effective strategies are needed to overcome barriers to recruitment.

As noted earlier in this report, recruitment of participants into intervention research is notoriously difficult. 160,161 Two reviews of publicly funded trials in the UK (through the NIHR) found that only about half of the included trials recruited 100% of their target sample size within their pre-agreed timescale. 146,162 The overall start to recruitment was delayed in 41% of trials, early recruitment problems occurred in 63% of trials,146 and just over one-third of trials received an extension of some kind. 146,162 There is little evidence that recruitment into experimental research is improving over time. 146,160 The recruitment of multiple family members to research projects is particularly challenging. 29,30,44,52 We have described specific recruitment challenges that we have encountered previously,78,125,126 but there has not been a comprehensive assessment of how to recruit families to family-based health promotion research.

The aim of this study was, therefore, to systematically identify effective and resource-efficient strategies for recruiting families into intervention research aimed at improving physical activity or nutrition or reducing levels of sedentary behaviour (including screen time) and overweight/obesity. Our objectives were to (1) describe procedures used and outcomes related to recruitment (e.g. recruitment duration, strategies used, recruitment settings, reach, expressions of interest, enrolment rates); and (2) determine the most optimal family-based recruitment strategies.

Methods

This study was conducted in two phases:

1. a systematic review of family recruitment methods

2. a Delphi consensus study.

Both phases examined the settings (i.e. where recruitment occurred) and the strategies (i.e. how recruitment was implemented) used by researchers conducting family-based intervention research with outcomes related to physical activity, sedentary behaviour (including screen time), nutrition, and obesity prevention. Details of the protocol for this study were registered on PROSPERO (CRD42019140042) and can be accessed at www.crd.york.ac.uk/prospero/display_record.php?RecordID=140042.

Results

Findings of phase 1: systematic review

Phase 1: systematic review findings

Figure 11 shows the study selection process. Fifty-five relevant reviews met the inclusion criteria; from these, 360 references to potentially relevant intervention studies were extracted and 50 were included. An additional 14 intervention studies were identified through forward searching, and, therefore, a total of 64 articles, describing 49 intervention studies, met the inclusion criteria. The study characteristics are detailed in Table 22. Of the 49 separate studies, the majority were undertaken in the USA (57%), were pilot or feasibility studies (43%), aimed to improve physical activity only (37%) and recruited parent–child dyads (53%). Publication dates ranged from 1983 to 2019, with 27% of the included articles published in the last 5 years (i.e. since 2014; 17 of 64 articles). After attempting to contact the authors of the 17 studies published in the last 5 years, we received responses for seven of the 17 studies. Modifications were made or additional information was provided for five out of these seven studies.

FIGURE 11.

Flow chart of identification of family-based intervention studies for systematic review.

TABLE 22 - Study characteristics of studies included in systematic review

CRCT, cluster randomised controlled trial; N/A, not applicable.

Intervention name, study (first author, year of publication, country)	Study design (assessments), study arms, aims/objectives	Families/participants [recruitment target; target and actual sample size; mean (SD) years of age at baseline; % female]	Recruitment (duration; settings; strategies used)	Reach, expressions of interest and enrolment
No intervention name, Alhassan, 2018, USA166	Pilot RCT (pre and 2 post measures) Study arms: child–mother, child alone or control Aims/objective: to examine the feasibility and efficacy of a mother–daughter intervention on African American girls’ physical activity	Recruitment target: African American mother–daughter dyads Target sample size: 60 dyads (20 dyads/group) Actual sample size: 76 dyads (child–mother, n = 28; child alone, n = 25; or control, n = 23) Family characteristics: children: 8.3 (1.3) years (100%); adults: 37.4 (7.7) years (100%)	Duration: not reported Setting: not reported Strategies: not reported	Reach: not reported Expressions of interest: 125 dyads Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
SHARE-AP ACTION, Anand, 2007, Canada167	RCT (pre and 2 post measures) Study arms: experimental or usual care control Aims/objective: to determine if a household-based lifestyle intervention was effective at reducing energy intake and increasing energy expenditure	Recruitment target: families on a Six Nations Reserve (minimum parent–child dyad required) Target sample size: not reported Actual sample size: 57 families (intervention, n = 29; control, n = 28) Family characteristics: children: experimental – 10.9 (2.9) years (62.5%), control – 9.9 (3.2) years (60.5%); adults: experimental – 41.3 (9.0) years (not reported), control – 37.2 (8.8) years (not reported) 57 families (participants: intervention, n = 88; control, n = 86); average three participants/family	Duration: 48 weeks Setting: not reported Strategies: not reported	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
No intervention name, Arredondo, 2014, USA168	Pilot trial (pre and post measures) Study arms: experimental arm only Aims/objective: to examine the acceptability, feasibility and preliminary efficacy of an intervention on physical activity and correlates of physical activity of Latina preadolescents and their mothers	Recruitment target: Latina mother–daughter dyads Target sample size: 11 dyads Actual sample size: 11 dyads Family characteristics: children: 9.6 (1.1) years (100%); adults: 36.7 (6.2) years (100%)	Duration: 8 weeks Setting: church (approached, n = 1; agreed, n = 1) Strategies: Announcements in Spanish from the pulpit; flyers distributed by study staff and church leaders	Reach: ≈ 864 parishioners (the church had 1800 enrolled parishioners and 48% were Latino) Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
No intervention name, Baranowski, 1990, USA;169 Baranowski, 1990, USA170	Randomised controlled feasibility study (pre and post measures) Study arms: experimental or no treatment control Aims/objective: to reduce sodium, saturated fat and total fat, and to increase aerobic activity	Recruitment target: families who self-identified as Black American (minimum parent–child dyad required) Target sample size: not reported Actual sample size: 96 families (intervention, n = 50; control, n = 46) Family characteristics: children: experimental – 10.6 years (51.6%), control – 10.0 years (66.1%); adults: experimental – 31.8 years (79.4%), control – 32.9 years (88.2%) 96 families (participants: n = 63 adults and 64 children intervention; n = 51 adults and 56 children intervention)	Duration: not reported Setting: schools only (number not reported) Strategies: mail, telephone calls and home visits (up to five visits) of all Black-American students identified in listings in the public or private school systems	Reach: 728 Black-American families identified Expressions of interest: N/A. This was not a sample of self-presenting volunteers Initiated expression of interest: N/A Expressions of interest rate: N/A Enrolment rate: not reported
Mothers and Daughters Exercising for Life (MADE4LIFE), Barnes, 2015, Australia171	Pilot RCT (pre and 2 post measures) Study arms: experimental or 6-month wait-list control Aims/objective: to evaluate the feasibility and preliminary efficacy of a mother–daughter programme to improve physical activity	Recruitment target: mother–daughter dyads Target sample size: 40 dyads Actual sample size: 40 dyads (mothers, n = 40; daughters, n = 48) Family characteristics: children: 8.5 (1.7) years (100%); adults: 39.1 (4.8) years (100%)	Duration: ≈ 3 weeks Setting: schools (n = not reported) Strategies: media releases, school newsletter advertisements, school presentations to students and parents, local newspapers and local television news	Reach: not reported Expressions of interest: 122 families Initiated expression of interest: not reported Expressions of interest rate: ≈ 40–41 families/week Enrolment rate: ≈ 13 families/week
Family Affair, Barr-Anderson, 2014, USA172	Pilot trial (pre and post measures) Study arms: experimental arm only Aims/objective: to test the feasibility and acceptability of an intervention designed to impact obesity-related behaviours (physical activity, healthy eating and sedentary behaviour) among African American adolescent girls and their mothers	Recruitment target: African American mother-daughter dyads Target sample size: not reported Actual sample size: 18 dyads Family characteristics: children: 12.4 ± 1.3 years (100%); adults: 36.9 ± 5.7 years (100%)	Duration: not reported Setting: not reported Strategies: radio advertisements, flyers and recruitment letters sent to or posted at youth- and family-serving organisations, health-related businesses, churches, social and professional organisations; e-mail distribution lists; Facebook posts; word-of-mouth	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Healthy Kids & Families study, Borg, 2019, USA173	Quasi-experimental protocol (pre and 4 post measures) Study arms: experimental or attention-control Aims/objective: to test the effectiveness of an intervention to promote a healthier lifestyle and to prevent childhood obesity among low-income and minority families	Recruitment target: parent–child dyads Target sample size: 240 dyads Actual sample size: 247 dyads (intervention, n = 121; attention-control, n = 126) Family characteristics: children: 7.8 (2.1) years (49%); adults: 36.2 (7.4) years (92%)	Duration: not reported Setting: schools only (n = 9 schools) Strategies: letter from the school principal placed in child’s backpack by school staff; automated telephone messages from principals; research staff presented study at school events (e.g. parent nights, family events, parent–teacher organisation meetings); interactions with parents at school drop-off/pick-up and after-school programmes	Reach: not reported Expressions of interest: 605 parents Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Back-to-Basics Healthy Lifestyle program, Burrows, 2013, Australia174	Pilot trial (pre and post measure) Study arm: experimental arm only Aims/objective: to assess the feasibility and acceptability of an after-school obesity prevention strategy for families	Recruitment target: parent–child dyads Target sample size: 10 dyads Actual sample size: 10 dyads Family characteristics: children: 7.3 (3.8) years (80%); adults: 31.0 (7.2) years (100%)	Duration: 2 weeks Setting: schools only (n = 1) Strategies: study flyers; word-of-mouth by school staff	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: 5 dyads/week
No intervention name, De Bourdeaudhuij, 2002, Belgium175	Quasi-experimental (pre and post measure) Study arms: family arm, individual arm (adolescents) or individual arm (parents) Aims/objective: to explore the differences between a family- and an individual-based tailored nutrition education programme on fat reduction	Recruitment target: parent–child dyads Target sample size: not reported Actual sample size: family condition: n = 55 dyads (n = 110 participants); individual condition (adolescents): n = 71 adolescents; individual condition (parents): n = 47 parents Family characteristics: children: range 15–18 years (not reported); adults: not reported	Duration: not reported Setting: schools only (n = 52 classes from two secondary schools) Strategies: not reported	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
MOVE/me Muevo Project, Elder, 2014, USA176	RCT (pre and 2 post measures) Study arms: experimental or control Hypotheses: (1) children in the experimental arm would have lower BMI z-scores than control children after 2 years; (2) children in the experimental arm spend more time in MVPA and less time sedentary, eat fewer high-fat foods and sugary beverages, and have more fruits, vegetables and water vs. control children	Recruitment target: families Target sample size: not reported Actual sample size: 541 families Family characteristics: children: 6.6 (0.7) years (55%); adults: not reported	Duration: not reported Setting: schools, libraries, street fairs, recreation centres (n = not reported) Strategies: targeted telephone calls using telephone numbers obtained from a research marketing company (n = 8600); families contacted via school- and community-based recruitment efforts (n = 1000)	Reach: 9607 families Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
No intervention name, Epstein, 2001, USA177	Randomised trial (pre and post measures) Study arms: increase fruit and vegetable (FV) intake treatment condition or decrease high-fat/high-sugar intake (FS) treatment condition Aims/objective: to evaluate the effect of a parent-focused intervention on parent and child eating changes and on the percentage of overweight changes in families	Recruitment target: families (minimum parent–child dyad required) Target sample size: not reported Actual sample size: 30 families (FV: n = 13 parents and n = 13 children; FS: n = 12 parents and n = 13 children) Family characteristics: children: FV – 8.8 (1.8) years (54%), FS – 8.6 (1.9) years (77%); adults: FV – 39.1 (4.1) years (92%), FS – 42.2 (4.8) years (92%)	Duration: not reported Setting: physician practices (n = not reported) Strategies: physician referrals, posters, newspapers, and television advertisements	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
No intervention name, Fitzgibbon, 1995, USA178	Pilot trial (pre and post measures) Study arms: experimental or control Aims/objective: to examine the effects of an obesity prevention programme on eating-related knowledge and behaviour of low income, Black American girls and their mothers	Recruitment target: Black American mother–daughter dyads Target sample size: not reported Actual sample size: 20 dyads (10 dyads/group) Family characteristics: children: experimental – 11.0 (1.0) years (100%), control – 11.0 (1.0) years (100%); adults: experimental – 31.0 (10.0) years (100%), control – 33.0 (5.0) years (100%)	Duration: not reported Setting: tutoring programme (n = 1) Strategies: advertisements in tutoring newsletter	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Children First Study, Fornari, 2013, Brazil179	RCT (pre and post measures) Study arms: experimental or control Aims/objective: to evaluate whether or not an educational programme for children could improve cardiovascular risk in parents	Recruitment target: children and their parents Target sample size: 150 parents per group Actual sample size: 197 children and 323 parents (intervention: n = 105 children, n = 162 parents; control: n = 92 children, n = 161 parents) Family characteristics: children: experimental – 8.2 (1.5) years (50%), control – 9.0 (1.5) years (51%); adults: experimental – 38.3 (6.0) years (55%), control – 39.3 (6.7) years (53%)	Duration: not reported Setting: schools only (n = 1) Strategies: not reported	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Active Families in the Great Outdoors, Flynn, 2017, USA180	Feasibility trial (pre and post measures) Study arms: experimental arm only Aims/objective: to determine whether changes could be observed in duration, frequency and type of outdoor physical activities performed by families; parent social cognitive outcomes and physical activity support behaviours	Recruitment target: families (minimum parent–child dyad required) Target sample size: not reported Actual sample size: 16 families (N = 52 participants: parents, n = 25; children, n = 27) Family characteristics: children: 10.7 (3.3) years (52%); adults: 41.5 (7.9) years (60%)	Duration: not reported Setting: not reported Strategies: flyers, e-mail, word-of-mouth	Reach: not reported Expressions of interest: 38 families Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Take Action, French, 2011, USA181	CRCT (pre and post measures) Study arms: experimental or control Hypothesis: the experimental group would gain less weight and increase healthful behaviours related to energy balance over 1 year compared with the control group	Recruitment target: families Target sample size: not reported Actual sample size: 90 households (n = 45 households/group) Family characteristics: children, not reported; adults, 41.0 years (93%) ≈ 4 members per family (≈ 2 adults and ≈ 2 children/family)	Duration: 32 weeks Setting: libraries, worksites, schools, day-care centres, health clinics, religious institutions, park and recreation centres, grocery stores, and food co-ops (number not reported) Strategies: not reported	Reach: not reported Expressions of interest: 723 households Initiated expression of interest: not reported Expressions of interest rate: 22 or 23 households per week Enrolment rate: 2 or 3 households per week
Families Reporting Every Step to Health (FRESH), Guagliano, 2019, UK	Feasibility trial (pre and post measures) Study arms: ‘child-only’ or ‘family’ Aims/objectives: to describe intervention and recruitment strategy; assess the feasibility and acceptability of the FRESH recruitment strategy, intervention and outcome evaluation; explore options for optimisation	Recruitment target: families (minimum parent–child dyad required) Target sample size: 20 families Actual sample size: 12 families (children, n = 14; adults, n = 18) Family characteristics: children: 8.3 (1.7) years (50%); adults: 39.8 (8.2) years (61%) Whole families 4, parent–child dyads 6, families with an additional adult or child 2; 2 or 3 members per family (range 2–4 members per family)	Duration: 8 weeks Setting: schools only. N = 11 schools approached, n = 5 agreed, n = 3 declined, n = 3 no response. Recruitment from community-based organisations planned, but not implemented Strategies: assembly delivered to students; study leaflets given to students to bring home and emailed to parents from schools; reminder e-mail sent from schools to parents 2 weeks after assembly	Reach: ≈ 437 students Expressions of interest: 28 families Initiated expression of interest: 23 mothers, 5 fathers Expressions of interest rate: 3 or 4 families per week, 5 or 6 families per school assembly Enrolment rate: 1 or 2 families per week
Scouting Nutrition and Activity Program+ (SNAP+), Guagliano, 2012, USA182	Quasi-experimental (pre and post measures) Study arms: experimental arm only Aims/objectives: to evaluate a physical activity promotion intervention with a channel of communication to parents	Recruitment target: Girl Scouts troops and their parents Target sample size: not reported Actual sample size: three troops (children, n = 32; adults, n = 26) Family characteristics: children: 9.5 (1.4) years (100%); adults: 37.1 (5.4) years (92%)	Duration: not reported Setting: Girl Scouts troops (n = 3 troops invited and agreed) Strategies: not reported	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Active 1 + FUN, Ha, 2019, Hong Kong183	RCT protocol (pre and 2 post measures) Study arms: experimental or control Aims/objective: to evaluate the effectiveness of a family-based intervention on parents and their children’s physical activity	Recruitment target: students and their parents (minimum parent–child dyad required) Target sample size: 204 children Actual sample size: 187 children Family characteristics: children: 9.8 (1.2) years (41%); adults: unknown (78%)	Duration: ≈ 4–6 weeks Setting: Schools only (n = 100 invited; n = 9 responded and agreed; n = 1 dropped out) Strategies: written information was circulated to parents; face-to-face parent–researcher sessions	Reach: unknown Expressions of interest: ≈ 229 Initiated expression of interest: unknown (not collected) Expressions of interest rate: unknown (researchers only received a confirmed list from schools) Enrolment rate: unknown (researchers only received a confirmed list from schools)
Abriendo Caminos, Hammons, 2013, USA184	Pilot trial (pre and post measures) Study arms: experimental arm only Aims/objective: to test the effectiveness of a family-based healthy eating programme aimed to reduce obesogenic behaviours among Latino parents and children	Recruitment target: Latino families; only 1 target child (5–13 years) and 1 parent measured Target sample size: not reported Actual sample size: 73 families Family characteristics: children: 8.5 years (49%); adults: 34.4 years (100%) ≈ 4 family members per family (range 2–9)	Duration: 104 weeks Setting: trailer park (n = 1) and elementary school (n = 1) with known Latino population Strategies: flyers, announcements, and word-of-mouth. Project co-ordinators were Latino and fluent Spanish speakers	Reach: unknown Expressions of interest: unknown Initiated expression of interest: unknown Expressions of interest rate: unknown Enrolment rate: < 1 family per week
Fit ‘n’ Fun Dudes Programme, Hardman, 2009, UK185	CRCT (pre and 2 post measures) Study arms: experimental or control Aims/objective: to increase daily step counts of girls with the support of their parents to maintain increases over time	Recruitment target: parent–daughter dyads Target sample size: not reported Actual sample size: n = 32 children (intervention, n = 14 children; control, n = 18 children) Family characteristics: children: 10.6 (0.7) years (100%); adults: 41.0 (4.7) years (83%)	Duration: not reported Setting: not reported Strategies: not reported	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
No intervention name, Hopper, 1992, USA186	CRCT (pre and 2 post measures) Study arms: school-and-home treatment condition, school-only treatment condition, and standard treatment control condition Aims/objective: to compare the effect of including versus not including a family participation component in a school-based programme to develop children’s heart-healthy exercise and nutrition habits	Recruitment target: parents and children or children only Target sample size: not reported Actual sample size: school-and-home condition, n = 45 children and 42 parents; school-only condition, n = 43 children; control condition, n = 44 children Family characteristics: children: 11.6 (0.7) years (not reported); adults: 37.8 (6.8) years (74%)	Duration: not reported Setting: not reported Strategies: not reported	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Teamplay, Jago, 2013, UK187	Randomised controlled feasibility trial (pre and 2 post measures) Study arms: experimental or no treatment control Aims/objectives: six specific aims related to feasibility of recruitment, retention and data collection; intervention development and optimisation; estimating effect sizes of outcomes of interest (e.g. physical activity, screen viewing) and sample size for definitive trial	Recruitment target: parents of children aged 6–8 years Target sample size: between 80 and 340 participants Actual sample size: 48 participants (intervention, n = 25; control, n = 23) Family characteristics: children: experimental – 6–8 years (62%), control – 6–8 years (69%); adults: experimental – age not reported (100%), control – age not reported (96%)	Duration: not reported Setting: schools, coffee shops, children’s centres, play groups, school playgrounds (number not reported) Strategies: leaflets, advertisements, face-to-face recruitment	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Motivating Families with Interactive Technology (mFIT), Jake-Schoffman, 2018, USA188	Pilot trial (pre and post measures) Study arms: tech or tech+ Aims/objective: to test the feasibility, acceptability and preliminary effectiveness of two family-based programs targeting improvements in parent–child dyad’s physical activity and healthy eating and delivered remotely	Recruitment target: parent–child dyads Target sample size: not reported Actual sample size: 33 dyads (tech+, n = 17; tech, n = 16) Family characteristics: children: 11.0 (0.9) years (64%); adults: 43.0 (5.8) years (88%)	Duration: not reported Setting: not reported Strategies: e-mail announcements, flyers posted in community settings, paid newspaper ads, direct mail postcards	Reach: not reported Expressions of interest: 98 Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Childhood and Adolescence Surveillance and Prevention of Adult Non communicable disease (CASPIAN) Study, Kargarfard, 2012, Iran;189 Kelishadi, 2010, Iran190	Non-RCT (pre and 2 post measures) Study arms: mother/daughter arm or student-only arm Aims/objective: to examine the effect of a physical activity programme for high-school girls and their mothers	Recruitment target: mother–daughter dyads or students only Target sample size: not reported Actual sample size: mother/daughter group (n = 206 girls and n = 204 mothers); student-only group (n = 60 girls) Family characteristics: children: 15.8 (1.0) years (100%) in mother/daughter group; 15.9 (1.3) years (100%) in student-only group; adults: age not reported (100%) in either group	Duration: not reported Setting: schools (number not reported) Strategies: not reported	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
4-Health, Lynch, 2012, USA191	Pilot RCT protocol (pre and 2 post measures) Study arms: experimental or ‘best practices’ control Aims/objective: to develop, implement and evaluate a parent-centred obesity prevention programme for rural families	Recruitment target: children and their parents Target sample size: 75 participants per group Actual sample size: unknown Family characteristics: unknown	Duration: not reported Setting: 4-H (n = 25 4-H extension agents) Strategies: announcements and information at county fairs, announcements in 4-H newsletters, electronic and/or printed announcements to 4-H clubs, e-mails to 4-H listservs, and telephone calls to 4-H leaders	Reach: unknown Expressions of interest: unknown Initiated expression of interest: unknown Expressions of interest rate: unknown Enrolment rate: unknown
No intervention name, Mark, 2013, Canada192	Pilot RCT (pre and post measure) Study arms: GameBike (experimental) or traditional stationary bike (control) Aims/objective: primarily, to compare usage of a GameBike with that of a traditional stationary bike placed in front of the television among parents and children	Recruitment target: families Target sample size: not reported Actual sample size: 30 families (adults, n = 59; children, n = 38) Family characteristics: children: experimental – 6.0 (2.1) years (42%); control – 5.4 (1.7) years (42%); adults: experimental – 37.1 (6.6) years (52%), control – 36.6 (6.1) years (50%)	Duration: not reported Setting: health-care centres, recreation centres, day cares, preschools and shopping malls (number not reported) Strategies: not reported	Reach: not reported Expressions of interest: 58 families Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Kick Start Your Day, Mohammad, 2012, USA193	Pilot trial (pre and post measure) Study arms: experimental or control Aims/objective: to evaluate a family-based nutrition and physical activity programme targeting low-income Latino families	Recruitment target: Latino families Target sample size: not reported Actual sample size: 56 parents (intervention, n = 25; control, n = 31) and their children (n = not reported) Family characteristics: children: range 6–12 years (not reported); adults: 37.0 (7.0) years (not reported)	Duration: not reported Setting: community centre (n = 1) and clinic (n = 1) Strategies: flyers and brochures written in English and Spanish, presentation delivered at a parent–teacher association meeting and community leader forum	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Children, parents and pets exercising together (CPET), Morrison, 2013, UK;194 Yam, 2012, UK195	Randomised controlled feasibility trial (pre and post measure) Study arms: experimental or no treatment control Aims/objectives: to assess the feasibility and acceptability of the CPET intervention and trial, preliminary evidence of its potential efficacy, planning and powering a future intervention, and to improve understanding of the frequency, intensity and duration of dog walking among dog-owning families in Scotland	Recruitment target: families with dogs Target sample size: 40 families Actual sample size: 28 families (experimental, n = 16 families; control, n = 12 families) Family characteristics: children: 10.9 years (76%), adults: 44.8 years (82%)	Duration: not reported Setting: primary schools (approached, n = 37; agreed, n = 35) Strategies: invitation letters sent to dog-owning parents with children attending primary schools in one local authority area	Reach: 350 letters sent Expressions of interest: 127 families Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Dads and Daughters Exercising and Empowered (DADEE), Morgan, 2019, Australia131	RCT (pre and 2 post measures) Study arms: experimental or wait-list control Aims/objective: to evaluate a programme designed to improve father–daughter physical activity and daughters’ fundamental movement skill competency; fathers’, daughters’ screen-time; fathers’ physical activity parenting practices	Recruitment target: fathers and their daughters Target sample size: 86 fathers and 134 daughters Actual sample size: 115 fathers and 153 daughters (DADEE: fathers, n = 57; daughters, n = 74; wait-list control: fathers, n = 58; daughters, n = 79) Family characteristics: children: 7.7 (1.8) years (100%); adults: 41.0 (4.6) years (0%)	Duration: 11 weeks Setting: not reported Strategies: university media release picked up by local television, radio, newspaper news outlets	Reach: not reported Expressions of interest: 160 Initiated expression of interest: not reported Expressions of interest rate: 14 or 15 families per week Enrolment rate: ≈ 10 families per week
Healthy Dads, Healthy Kids, Morgan, 2014, Australia;196 Morgan, 2011, Australia197,198 Williams, 2018, Australia199	Community RCT (pre and post measures) Study arms: experimental or a wait-list control Aims/objective: to evaluate the Healthy Dads, Healthy Kids intervention when delivered by trained local facilitators in the community	Recruitment target: fathers and their children Target sample size: 50 fathers and their children Actual sample size: 93 fathers and 132 children Family characteristics: children: 8.1 (2.1) years (45%); adults: 40.3 (5.3) years (0%)	Duration: ≈ 8 weeks Setting: schools (n not reported) Strategies: school newsletters, school presentations, interactions with parents at school pick-up, local media, and flyers distributed through local communities	Reach: not reported Expressions of interest: 116 Initiated expression of interest: not reported Expressions of interest rate: ≈ 14–15 families per week Enrolment rate: ≈ 11–12 families per week
Healthy Dads, Healthy Kids, Morgan, 2011, Australia197,198 Lubans, 2012, Australia;117 Burrows, 2012, Australia200	RCT (pre and 2 post measures) Study arms: experimental or a wait-list control Aims/objective: to evaluate the feasibility and efficacy of Healthy Dads, Healthy Kids to help fathers lose weight and model positive health behaviours to their children	Recruitment target: fathers and their children Target sample size: 44 fathers and their children Actual sample size: 53 fathers and 71 children Family characteristics: children: 8.1 (2.1) years (45%); adults: 40.3 (5.3) years (0%)	Duration: ≈ 8 weeks Setting: schools (n not reported) Strategies: school newsletters, local media	Reach: not reported Expressions of interest: 107 Initiated expression of interest: not reported Expressions of interest rate: ≈ 13 families per week Enrolment rate: ≈ 6 or 7 families per week
The San Diego Family Health Project, Nader, 1989, USA;201 Nader, 1992, USA;202 Nader, 1983, USA;203 Patterson, 1988, USA204	CRCT (pre and 3 post measures) Study arms: Mexican American experimental, Anglo-American experimental, Mexican American control, or Anglo-American control Aims/objective: to decrease consumption of high-salt, high-fat foods; and increase frequency and intensity of physical activity	Recruitment target: families (only up to two children and two adults measured) Target sample size: not reported Actual sample size: 206 families Family characteristics:  Mexican American experimental: children: 12.1 (1.7) years (55%); adults: 37.1 (6.8) years (88%)  Anglo-American experimental: children: 12.1 (1.9) years (38%); adults: 39.4 (7.1) years (62%)  Mexican American control: children: 12.0 (1.7) years (49%); adults: 35.6 (6.9) years (75%)  Anglo-American control: children: 11.8 (1.4) years (48%); adults: 36.9 (5.1) years (58%)	Duration: not reported Setting: primary schools (n not reported) Strategies: newspaper articles, Parent–Teacher Association meetings, community groups and a family fun night (covered by a local TV station)	Reach: ≈ 6000 children Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Behavior Opportunities Uniting Nutrition, Counseling, and Exercise (BOUNCE), Olvera, 2010, USA;205 Olvera, 2008, USA206	CRCT (pre and post measures) Study arms: experimental or control Aims/objective: primarily, to assess the efficacy of the BOUNCE intervention for improving physical fitness and activity in Latino mother–daughter pairs	Recruitment target: Latino mother–daughter dyads Target sample size: 50 dyads Actual sample size: 46 dyads (experimental, n = 26; control, n = 20) Family characteristics: children: experimental – 9.9 (1.1) years (100%), control – 10.4 (1.1) years (100%); adults: experimental – 33.3 (4.6) years (100%), control – 38.2 (10.6) years (100%)	Duration: not reported Setting: not reported Strategies: flyers mailed to homes of Latino families	Reach: not reported Expressions of interest: 57 parents Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
No intervention name, Owens, 2011, USA207	Quasi-experimental (pre and post measures) Study arms: experimental or control Aims/objective: to examine changes in physical activity and fitness in families after 3 months of home use of the Wii Fit	Recruitment target: families Target sample size: not reported Actual sample size: 8 families (n = 21 participants) Family characteristics: children: 10.0 (1.6) years (50%); adults: 37.8 (4.9) years (78%)	Duration: not reported Setting: not reported Strategies: local newspaper advertisement	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Etude Longitudinale Prospective Alimentation et Santé (ELPAS) study, Paineau, 2008, France208	RCT (pre and post measures) Study arms: group A (experimental), group B (experimental) and group C (no-treatment control) Hypothesis: family dietary coaching would improve nutritional intakes and weight control in free-living children and parents	Recruitment target: families (parent–child dyad minimum) Target sample size: 295 families per experimental group and 420 families in the control group Actual sample size: 1013 families (Group A, n = 297 families; Group B, n = 298 families; Group C, n = 418 families) Family characteristics: children, 7.7. years (52%); adults, 40.5 (82%)	Duration: 16 weeks Setting: schools only (n = 54) Strategies: mailed study information	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Y Living, Parra-Medina, 2015, USA209	Quasi-experimental (pre and post measures) Study arms: experimental arm only Aims/objective: to examine the impact of the Y Living Program on the weight status of adult and child participants	Recruitment target: families Target sample size: not reported Actual sample size: 242 adults, 106 children Family characteristics: children, 12 (interquartile range 10–14) years (49%); adults, 41 (interquartile range 33–53) (81%)	Duration: not reported Setting: churches, schools (number not reported) Strategies: organisational newsletters, neighbourhood newspapers, word-of-mouth	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Choosing 5 Fruits and Veg Every Day, Pearson, 2010, UK210	Pilot trial (pre and 2 post measures) Study arms: experimental or no treatment control Aims/objective: to evaluate the feasibility and effectiveness of a family-based newsletter intervention to increase fruit and vegetable consumption among adolescents	Recruitment target: parent–adolescent dyads Target sample size: not reported Actual sample size: 49 dyads Family characteristics: children: experimental – 12.6 (1.0) years (44%), control – 12.3 (0.7) years (42%); adults: experimental – 44.4 (5.3) years (71%), control – 43.9 (3.6) years (75%)	Duration: 16 weeks Setting: schools, universities, factories, warehouses, clubs/societies (number not reported) Strategies: newspaper and website advertisements, posters in workplaces (universities, factories, warehouses), and letters through schools and activity clubs/societies	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Daughters and Mothers Exercising Together (DAMET), Ransdell, 2004, USA;211 Ransdell, 2003, USA;212 Ransdell, 2001, USA213	Pilot trial (pre and post measures) Study arms: community-based or home-based experimental arms Aims/objective: to assess the effectiveness of home- and community-based physical activity interventions that target mothers and daughters to increase physical activity and improve health-related fitness	Recruitment target: mother–daughter dyads Target sample size: not reported Actual sample size: 20 dyads Family characteristics: children: community-based – 15.2 (1.2) years (100%), home-based – 15.7 (1.5) years (100%); adults: community-based – 46.0 (8.5) years (100%), home-based – 44.0 (6.1) years (100%)	Duration: not reported Setting: not reported Strategies: newspaper articles, local Girl Scout troop announcements, referral	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Generations Exercising Together to Improve Fitness (GET FIT), Ransdell, 2004, USA;214 Ornes, 2005, USA215	Pilot trial (pre and post measure) Study arms: experimental or no treatment control Aims/objective: to compare a 6-month home-based physical activity intervention with a control condition for physical activity and health-related fitness in three generations of women	Recruitment target: grandmother–mother–daughter triads Target sample size: not reported Actual sample size: 17 triads Family characteristics: children: experimental – 10.8 (1.4) years (100%), control – 9.4 (1.5) years (100%); mothers: experimental – 37.8 (4.2) years (100%), control – 36.6 (4.2) years (100%); grandmothers: experimental – 60.7 (4.3) years (100%), control – 62.9 (4.5) years (100%)	Duration: not reported Setting: not reported Strategies: newspaper, e-mail and flyer advertisements, word-of-mouth	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
No intervention name, Rhodes, 2019, Canada;151 Quinlan, 2015, Canada216	RCT (pre and 3 post measures) Study arms: physical activity education + planning (experimental) or physical activity education (control) Aims/objective: to evaluate whether or not a planning condition improves regular physical activity compared with an education-only control condition among families	Recruitment target: families (minimum parent–child dyad required) Target sample size: 160 families Actual sample size: 102 families Family characteristics: children: intervention – 8.8 (2.3) years (50%), control – 9.1 (1.9) years (54%); adults: intervention – 42.2 (5.7) years (76%) intervention, control – 43.0 (5.7) years (83%) Dual-parent families, 52%; single families, 44%; families with siblings, 29%	Duration: not reported Setting: schools, recreation centres, health-care centres, children’s recreation classes, shopping malls and outdoor markets (number not reported) Strategies: newspaper advertisements. Snowball recruitment was also used, whereby families received a CA$25 grocery store gift card if they referred another family Recruitment was conducted by stratifying the city into regions to ensure diversity of families	Reach: not reported Expressions of interest: 188 parents Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
No intervention name, Rhodes, 2010, Canada141	Pilot RCT (pre and post measures) Study arms: standard intervention or standard intervention + planning Aims/objective: to examine the effect of a planning intervention compared with a standard condition on intergenerational physical activity in families	Recruitment target: families Target sample size: not reported Actual sample size: 85 families Family characteristics: children: standard – range 4–10 years (not reported) standard+ – range 4–10 years (not reported); adults: standard – 38.6 (5.30 years (79%), standard+ – 39.0 (5.2) years (90%)	Duration: 52 weeks Setting: day cares, recreation centres, preschools, primary schools (number not reported) Strategies: flyers, poster advertisements	Reach: not reported Expressions of interest: 107 families Initiated expression of interest: not reported Expressions of interest rate: ≈ 2 families per week Enrolment rate: ≈ 1 or 2 families per week
Scouting Nutrition & Activity Program, Rosenkranz, 2010, USA;217 Rosenkranz, 2009, USA218	CRCT (pre and post measure) Study arms: experimental or standard-care control Aims/objective: to evaluate an intervention designed to prevent obesity by modifying Girl Scout troop meeting environments, and by empowering girls to improve the quantity and/or quality of family meals in their home environments	Recruitment target: Girl Scout troops and their parents Target sample size: 8 troops with 20 girls per troop Actual sample size: 7 troops (mean 11 girls per troop) Family characteristics: children: experimental – 10.6 (1.1) years (100%), control – 10.5 (1.3) years (100%); adults: experimental – age and % female not reported, adults: control – age and % female not reported	Duration: not reported Setting: Girl Scouts troops (n = 7 troops) Strategies: not reported	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
No intervention name, Salimzadeh, 2010, Iran219	Quasi-experimental (pre and post measures) Study arms: experimental arm only Aims/objective: to evaluate the effectiveness of an exercise programme on the body composition and physical fitness of mothers and daughters	Recruitment target: mother–daughter dyads Target sample size: not reported Actual sample size: 35 dyads Family characteristics: children: 15.0 (1.6) years (100%); adults: 40.0 (3.8) years (100%)	Duration: not reported Setting: schools only (n = 5) Strategies: not reported	Reach: 300 students Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
No intervention name, Schwinn, 2014, USA220	Pilot trial (pre and 2 post measures) Study arms: experimental or control Aims/objective: to improve the well-being of girls living in public housing by improving dietary intake, increasing physical activity and reducing drug use risks	Recruitment target: mother–daughter dyads Target sample size: not reported Actual sample size: 67 dyads (intervention, n = 36; control, n = 31) Family characteristics: children: 11.9 (0.9) years (100%); adults: 36.2 (6.2) years (100%)	Duration: 4 weeks Setting: public housing development (n = 1) Strategies: Google AdWords, public housing development newspapers, Facebook and Craigslist advertisements	Reach: not reported Expressions of interest: 86 Initiated expression of interest: not reported Expressions of interest rate: 21 or 22 families per week Enrolment rate: 16 or 17 families per week
Brighter Bites, Sharma, 2016, USA221	Quasi-experimental (pre and post measures) Study arms: Brighter Bites (experimental) or school health programme (control) Aims/objective: to evaluate the effectiveness of a school-based food co-op programme to increase fruit and vegetable intake, and home nutrition environment among low-income children and their parents	Recruitment target: parent–child dyads Target sample size: not reported Actual sample size: 717 dyads (intervention, n = 407; control, n = 310) Family characteristics: children: 6.2 (0.4) years (52%); adults: 34.3 (7.4) years (90%)	Duration: 2 school years Setting: schools only (n = 12) Strategies: not reported	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: 358 or 359 dyads per school year
No intervention name, Stolley, 1997, USA222	Pilot trial (pre and post measures) Study arms: experimental or control Aims/objective: to assess the effectiveness of an obesity prevention programme on pre-adolescent girls and their mothers	Recruitment target: mother–daughter dyads Target sample size: not reported Actual sample size: 65 dyads Family characteristics: children: intervention – 9.9 (1.3) years (100%), control – 10.0 (1.5) years (100%); adults: intervention – 31.5 (3.4) years (100%), control – 33.7 (6.8) years (100%)	Duration: not reported Setting: tutoring programme (n = 1) Strategies: advertisement in tutoring newsletter, letters sent to mothers of children registered in tutoring programme, presentation delivered to parents at tutoring programme orientation	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
One Body, One Life, Towey, 2011, UK223	Quasi-experimental (pre and post measures) Study arms: experimental arm only Aims/objective: to evaluate a family-based programme designed to prevent obesity	Recruitment target: families Target sample size: not reported Actual sample size: 272 children and 182 parents. Family characteristics: children: 8.0 years (50%); adults: age not reported (87%)	Duration: not reported Setting: neighbourhood groups, local fetes, community groups, general practitioner surgeries, libraries, children’s centres, print media, schools (number not reported) Strategies: flyers, posters, newsletters, word-of-mouth, referrals from health-care professionals and local newspapers, and making team members visible in the community (e.g. attending events, delivering ‘taster sessions’)	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
Family Eats, Cullen, 2017, USA224	RCT (pre and 2 post measures) Study arms: experimental or control Aims/objective: to improve parent and child fruit and vegetable intake	Recruitment target: families Target sample size: not reported Actual sample size: 126 families (n = 92 intervention; n = 34 control) Family characteristics: children: age not reported (55%); adults: 59% aged < 40 years (98%)	Duration: not reported Setting: schools, churches, health fairs, community centres (number not reported) Strategies: flyers, radio advertisements	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported
No intervention name, Ziebarth, 2012, USA225	Quasi-experimental (pre and post measures) Study arms: experimental arm only Aims/objective: to evaluate a family intervention programme designed to decrease overweight and obesity in Hispanic families	Recruitment target: Hispanic families Target sample size: not reported Actual sample size: 47 families (adults, n = 57; children, n = 54) Family characteristics: children: age and % female not reported, adults: age 32 years (not reported)	Duration: not reported Setting: local churches, medical clinics, schools, self-service laundries and community programmes (number not reported) Strategies: posters, announcements, word-of-mouth	Reach: not reported Expressions of interest: not reported Initiated expression of interest: not reported Expressions of interest rate: not reported Enrolment rate: not reported

Table 23 provides the details of all relevant recruitment data. Overall, a target sample size was presented a priori in 33% of studies, with a median target sample size of 120 (IQR 65–182) participants. Actual sample size was reported in 98% of studies and included a median of 100 (IQR 53–304) participants. Of the 16 studies in which target and actual sample sizes were provided, 56% recruited a sufficient number of participants. The duration of recruitment was reported in 33% of studies, and this was a median of 10 (IQR 8–36) weeks. Few studies reported figures on reach (18%), expressions of interest (33%), expressions of interest rate (16%), who initiated an expression of interest (< 1%) and enrolment rate (22%). Where reported, the median estimated reach was 437 (IQR 350–864) families, of whom 122 (IQR 92–174) expressed interest. The single study describing who expressed initial interest showed that in 82% (23/28) of the cases these were mothers. The median weekly expression of interest rate was 14 (IQR 11–21) families per week, with median enrolment rate at about 5 (IQR 2–11) families/dyads per week.

TABLE 23 - Summary of recruitment figures from intervention studies included in the systematic review

N/A, not applicable.

Median (IQR) values are presented unless indicated otherwise.

	Overall	Number of studies with relevant data (n = 49 studies)
Target sample size (participants)	120 (65–182)	16
Actual sample size (participants)	100 (53–304)	48
Recruitment duration (weeks)	10 (8–36)	16
Reach	437 (350–864)	9
Expressions of interest	119 (95–167)	16
Initiated expression of interest	82% mothers	1
Expressions of interest rate (per week)	14 (11–21)	8
Enrolment rate (families per week)	5 (2–11)	11
Percentage of studies with under-recruitment	44%	N/A

Details on family recruitment settings and strategies were reported in 84% and 73% of studies, respectively. On average, researchers recruited from 2.2 ± 1.9 different settings and used 2.7 ± 1.2 recruitment strategies per study; there was no difference between full-scale trials, pilot/feasibility or quasi-experimental trials in the number of recruitment settings or strategies used. School-based recruitment was the most common recruitment setting, with community-based recruitment second. Community-based recruitment settings included churches, recreation centres, play groups, libraries, fairs/fetes, sports clubs, 4-H, day cares, preschools, tutoring programs, malls, grocery stores, farmers’ markets, cafes, trailer parks and laundromats. Recruitment also occurred through employers, primary care (e.g. general practitioners, health centres, other health-related businesses), and print/electronic media.

Across settings, the most commonly used recruitment strategies included disseminating study information through leaflets, posters or newsletters. School-based recruitment had the most recruitment strategies specific to the setting and included leaflets, posters, newsletters, letters from the head teacher (principal), research teams presenting study information to students and parents at assemblies, research teams presenting study information at other school events (e.g. parent–teacher association meetings), and research teams approaching parents during pick-up/drop-off times. Local newspapers and referral-based recruitment (e.g. word-of-mouth) were also popular recruitment strategies. Less commonly reported recruitment strategies included using: electronic/digital media (e.g. television, radio, social media, Google AdWords, Craigslist), face-to-face recruitment (e.g. home visits, community demonstrations), mail, telephone calls and distribution lists (e.g. via marketing companies).

Discussion

The aim of this study was to systematically identify effective and resource-efficient strategies for recruiting families into physical activity, sedentary behaviour/screen time, diet and obesity prevention intervention research. Our systematic review showed that, despite being checklist items on the Consolidated Standards of Reporting Trials,80,226 data related to recruitment strategies and their effectiveness were scarcely reported among the included studies. Moreover, most studies applied similar recruitment strategies, predominantly through schools, despite the known challenges of recruiting families through school settings. Overall, a multisetting and multistrategy approach may be most effective, and the top 10 identified strategies may help researchers allocate limited resources effectively.

The data shown here indicate that researchers conducting family-based intervention studies were unable to attract sufficient expressions of interest, let alone recruit target sample sizes. We extracted very little information from the included studies related to expressions of interest and enrolment, which was similar to another review. 193 In particular, only one study reported which parent (i.e. mother or father) initiated an expression of interest. 78 That study found that 82% of the parents who initiated an expression of interest were mothers; however, once enrolled, fathers were enthusiastic about and benefited from their family’s participation in the study. The fact that most parents who expressed interest were mothers may not come as a surprise as, historically, mothers have been more likely than fathers to be their family’s social agent and lead on tasks such as family event preparation. 113,227 It may be prudent to consider this in the recruitment of families in two ways. First, recruitment materials that target mothers and their family may be the most efficient method of attracting expressions of interest; and it may also be an important catalyst for the inclusion of more fathers in family-based research. Second, separate recruitment materials that explicitly target fathers may also be useful and should be considered; Morgan and colleagues52,119,228 have written extensively on recruiting and engaging with fathers in family-based research.

Related to target sample size, we found that only 38% and 56% of the studies included in our Delphi study and review, respectively, recruited their target sample size. Similarly, other reviews of publicly funded trials have found that only 33–50% of included trials recruited 80–100% of their target sample size within their pre-agreed timescale. 145,146,162,229 Few studies reported on reach and representativeness, but, generally, healthy and affluent families were recruited. Only one study, as far as we are aware, described formative work in which families were consulted to inform the development of the recruitment strategy. 78 Although the target sample size was not achieved in that study, public involvement should be encouraged and has been highlighted as a good method for helping with participant recruitment, engagement and retention. 230 It is also possible that the chance of being randomly allocated into a study arm that was not a family’s preferred study arm may have negatively affected recruitment;231 however, no study reported that the randomisation procedure hindered their recruitment.

Analogous to our Delphi study findings, one-third of trials received an extension of some kind because of recruitment-related issues. 146,162 Our findings showed the planned median recruitment duration to be about 10–11 weeks and when recruitment extensions were implemented recruitment was extended for an additional 20 weeks, which would have a substantial impact on a study’s timeline. A recent survey on research priorities related to the methodology of trials among directors of the Clinical Trials Units registered with the NIHR Clinical Research Network in the UK identified the recruitment of participants in trials as among the top three priorities for improvement;232 overall, our findings reinforce these concerns.

The majority of included studies reported the settings from which they recruited participants, and our findings indicate that researchers recruited in about two settings per study, on average. In both our review and our Delphi study, we found that the two most common recruitment settings were schools and the community. Other recruitment settings included primary care settings, employers and social media. Although it was positive to find that recruitment occurred in multiple settings, as recommended by others,29,30,233 usually it was not possible to discern what proportion of a study’s sample had been recruited by setting.

On average, researchers used about three recruitment strategies per study. As in other studies,233,234 the most commonly used recruitment strategies for family-based recruitment included disseminating study information through leaflets, posters or newsletters. Placing advertisements in local newspapers, using electronic media (e.g. social media platforms, radio, television) and referral-based recruitment (e.g. word-of-mouth) were also popular. Considering that school-based recruitment was the most used recruitment setting, it was unsurprising that this setting had the greatest number of recruitment strategies. Generally, recruitment strategies were only listed in study manuscripts and not described in any great detail, particularly in terms of how and by whom these strategies were actually implemented. However, among the few studies that recruited a sufficient sample size, many included strategies that targeted adults and children and often while they were together. 204,205,207,208 For example, interacting with parents and children at school drop-off/pick-up, giving study presentations at school events (e.g. parent nights) or making announcements from the pulpit are some of the strategies used in the studies that recruited their target sample size. Furthermore, based on the top 10 recommended recruitment strategies identified here, it appears that leveraging familiar, and perhaps trusted, relationships would be beneficial. For example, disseminating study information in correspondence from head teachers, general practitioners and human resources staff and by word of mouth (e.g. through family friends) are strategies that generally assume that potential participants would have at least some familiarity with the person disseminating the information. Even those Delphi participants who were most enthusiastic about recruiting through social media platforms appeared to be trying to build rapport with their followers (i.e. potential participants). Delphi participants recommended that creating regular content on social media platforms or newsletters (e.g. update posts, quarterly newsletter, blogs) would be more beneficial than one-off posts, advertisements or newsletters. The top 10 strategies also include several face-to-face recruitment strategies (e.g. delivering assemblies, attending parent–teacher meetings). Although face-to-face recruitment can be extremely time (and resource) consuming, studies that included this type of recruitment were more likely to achieve close to their target sample size. 229 In addition, face-to-face recruitment allows for quick rapport-building and gives potential participants the ability to receive responses to their questions about the study in real time, while they are still interested. 235

Acknowledgement of contributions and funding

We thank Active Norfolk and Suffolk County Council for providing funding for the delivery of the FRESH programme. We are grateful to participating families their enthusiastic involvement in the study and to the schools, workplaces and community settings that enabled us to recruit families from their locations. We acknowledge FRESH, Centre for Family Research, Cambridge Research Methods Hub, and MRC Epidemiology Unit staff and students past and present for their involvement in the project. We also thank the FRESH Study Steering Committee and patient and public involvement groups who have provided valuable advice through the project. This work was supported by the National Institute for Health Research Public Health Research programme (project number 15/01/19). This work was additionally supported by the MRC [Unit Programme number MC_UU_12015/7], and undertaken under the auspices of the Centre for Diet and Activity Research (CEDAR), a UKCRC Public Health Research Centre of Excellence. Funding from the British Heart Foundation, Cancer Research UK, Economic and Social Research Council, MRC, the National Institute for Health Research, and the Wellcome Trust, under the auspices of the UK Clinical Research Collaboration, is gratefully acknowledged (087636/Z/08/Z; ES/G007462/1; MR/K023187/1).

Contributions of authors

Esther MF van Sluijs (https://orcid.org/0000-0001-9141-9082) (MRC Programme Leader, Behavioural Epidemiology) gained funding, conceptualised the study, acted as chief investigator of the studies included in this report and drafted this report.

Helen E Brown (https://orcid.org/0000-0001-7632-1714) (Research Associate, Behavioural Epidemiology) gained funding, initially conceptualised the intervention, and advised on study conduct and the interpretation of the results.

Emma Coombes (https://orcid.org/0000-0002-7980-5850) (Research Fellow, Public Health) led on the methodological development of the family co-participation measure.

Claire Hughes (https://orcid.org/0000-0003-2545-3025) (Professor, Psychology) gained funding, conceptualised the study, led on the assessment of family functioning and provided a family perspective on the interpretation of the results.

Andrew P Jones (https://orcid.org/0000-0002-3130-9313) (Professor, Public Health) gained funding, conceptualised the study and provided a public health perspective on the interpretation of the results.

Katie L Morton (https://orcid.org/0000-0002-9961-6491) (Research Associate, Behavioural Epidemiology) gained funding, advised on conduct and analyses of qualitative data and provided a behavioural science perspective on the interpretation of the results.

Justin M Guagliano (https://orcid.org/0000-0002-4450-5700) (Research Associate, Behavioural Epidemiology) operationalised the intervention, led on recruitment and data collection, conducted statistical analyses and prepared the results for publication.

Publications

Guagliano JM, Brown HE, Coombes E, Haines ES, Hughes C, Jones AP, et al. Whole family-based physical activity promotion intervention: the Families Reporting Every Step to Health pilot randomised controlled trial protocol. BMJ Open 2019;9:e030902.

Guagliano JM, Brown HE, Coombes E, Hughes C, Jones AP, Morton KL, et al. The development and feasibility of a randomised family-based physical activity promotion intervention: the Families Reporting Every Step to Health (FRESH) study. Pilot Feasibility Stud 2019;5:21.

Guagliano JM, Armitage SM, Brown HE, Coombes E, Hughes C, Jones AP, et al. A whole family-based physical activity promotion intervention: Findings from the Families Reporting Every Step to Health (FRESH) pilot randomised controlled trial. Int J Behav Nutr Phys Act 2020;17:120.

Guagliano JM, Morton KL, Hughes C, van Sluijs EMF. Effective and resource efficient strategies for recruiting families in physical activity, sedentary behaviour, nutrition, and obesity prevention research: a systematic review with expert opinion. Obes Rev 2021;22:e13161.

Data-sharing statement

All data requests should be submitted to the corresponding author for consideration. Access to anonymised data may be granted following favourable review of a research proposal.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the PHR programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the PHR programme or the Department of Health and Social Care.