Barriers and facilitators to young children's physical activity and sedentary behaviour: a systematic review and synthesis of qualitative literature

Summary Positive activity behaviours (i.e. higher physical activity [PA]/lower sedentary behaviour [SB]) are beneficial from infancy, yet evidence suggests that young children (0‐ to 6‐year‐olds) are relatively inactive. To better understand the perceived influences on these behaviours and to aid intervention development, this paper systematically synthesizes the extensive qualitative literature regarding perceived barriers and facilitators to PA and SB in young children (0–6 years old). A search of eight electronic databases (July 2016) identified 43 papers for inclusion. Data extraction and evidence synthesis were conducted using thematic content analysis, underpinned by the socio‐ecological model (i.e. individual, interpersonal, community, organizational and policy levels). Parents, childcare providers and children perceived seven broad themes to be important for PA and SB, including the child; the home; out‐of‐home childcare; parent–childcare provider interactions; environmental factors; safety; and weather. Each theme mapped onto between one and five levels of the socio‐ecological model; barriers and facilitators at the interpersonal level (e.g. parents, care providers and family) were most frequently cited, reflecting the important (perceived) role adults/peers play in shaping young children's behaviours. We provide an overarching framework to explain PA and SB in early childhood. We also highlight where gaps in the current literature exist (e.g. from male carers; in developing countries; and barriers and facilitators in the environmental and policy domains) and where future quantitative work may focus to provide novel insights about children's activity behaviours (e.g. safety and weather).

development, improved metabolic health and decreased adiposity (3). In contrast, sedentary behaviour (defined as any waking activity characterized by an energy expenditure ≤1.5 metabolic equivalents and a sitting or reclining posture (4)) is associated with higher adiposity and poorer psychosocial health and cognitive development in children 0-4 years old (5).
In recent years, interest in both physical activity and sedentary behaviour during the early years has increased. Prevalence estimates suggest that preschool-aged children engage in low levels of physical activity and are sedentary for a large proportion of their day (6,7). This sedentariness appears to manifest early in childhood, with children spending large amounts of time watching television (TV) before the age of 2 (8). A wide range of interventions have therefore been developed in an attempt to boost activity levels and decrease sedentary time in young children prior to their entry into formal schooling (9)(10)(11). Such interventions often target a range of health behaviours (i.e. diet and physical in/activity) (10)(11)(12) or aim to increase physical activity/decrease sedentary time in isolation (9). Yet regardless of emphasis or setting, few studies report evidence of a positive effect on physical activity; those that do see small gains (<10%) in overall physical activity, which are not sustained over the longer term (9)(10)(11).
The reasons for lack of intervention success are likely to be varied, and in no small part due to the difficulties associated with changing behaviour (13). However, it has been suggested that for a greater chance of intervention success, it is important to establish the relevant influences on the target behaviour (14). To this end, several quantitative systematic reviews have been conducted to determine correlates (i.e. cross-sectional factors) associated with young children's physical activity (15)(16)(17)(18) and sedentary behaviour (19). For both behaviours, these investigate a broad range of potential correlates including demographic, biological, environmental, social and psychological influences. Contrasting conclusions regarding what factors are related to physical activity behaviour tend to be drawn across reviews (15,16,18), but family factors (15)(16)(17), time spent outdoors, and the built or physical environment (15,16) appear to be consistently associated with increased physical activity in preschool-aged children. Evidence of the correlates of sedentary behaviour is less well elucidated, with many (early) studies tending to report TV viewing as a proxy for sedentary behaviour rather than using objective measures (19). In a review of correlates, child sex was assessed in four or more studies (19); it showed no association with TV viewing and an indeterminate association with accelerometer-measured sedentary behaviour. Child's age, body mass index, parental education and race also showed an indeterminate association with TV viewing, whilst outdoor playtime was found to be not associated with TV viewing (19).
It is however difficult to draw firm conclusions about causality from cross-sectional studies. As part of series of reviews, of which this qualitative synthesis is one, we assessed the quantitative determinants (i.e. longitudinal predictors) of physical activity in both prospective and intervention studies (11). Although 14 determinants were assessed in four or more studies, only parental monitoring and provider training were associated with positive change in children's total physical activity and moderate-tovigorous physical activity (MVPA), respectively (11). A further 12 factors, some of which have also been identified as correlates of physical activity (e.g. sex and portable equipment), showed inconsistent or no association with change in preschoolers' physical activity over time. These quantitative reviews do however highlight that studies have tended to focus on the influence of specific domains (namely, child, family and environmental) on children's physical activity and sedentary behaviour. A large number of as-yet unexplored factors (particularly in the community and policy domains) therefore remain, which when targeted may have the potential to effect positive change in young children's activity behaviour at a population level.
Drawing on qualitative research may be one way to explore these hitherto unmapped influences on young children's physical activity and sedentary behaviours. Much work has been conducted to explore what factors those caring for children (e.g. parents and caregivers), and the children themselves, perceive to be important for activity behaviour, but no systematic synthesis of this growing body of evidence has been conducted to date. Exploring and integrating this qualitative literature, particularly in combination with previous quantitative reviews, will enhance our understanding of the influences on young children's activity behaviours. Crucially, it may also lead to identification of potential avenues for intervention that those who are instrumental to children's activity behaviour believe to be important but that have yet to be explored by researchers and policy makers as important components of interventions.
This systematic review therefore aims to synthesize the qualitative literature exploring barriers and facilitators to activity behaviours (i.e. physical activity and sedentary behaviour) in young children (0-6 years), in order to (i) establish perceived influences on activity behaviours and (ii) consider where discrepancies and gaps in the wider (qualitative and quantitative) evidence exist that may inform future research.

Methods
This review was conducted as part of a suite of reviews aiming to establish the determinants of obesogenic behaviours in children 0-6 years (including fruit and vegetable intake; sugar sweetened beverages and unhealthy diet intake; and sedentary behaviour). The protocol for this (20) and a quantitative physical activity review (11) have been described previously; the International Prospective Register for Systematic Reviews registration number is CRD42012002881. This review was carried out in three stages (21,22), according to criteria for the rigorous conduct and reporting of systematic reviews (23). Studies were identified in tandem across all reviews, with smaller teams leading on data extraction for specific health behaviours of interest.

Study identification
We conducted a systematic search (common to all reviews) in August 2012, with four sets of search terms relating to the study population, study design (including qualitative studies), outcome of interest and exclusion of clinical populations (Table S1). An extensive scoping phase was conducted prior to the full search to maximize sensitivity and specificity. We searched eight electronic databases ( . We used ENDNOTE citation management software (Thomson Reuters, Philadelphia, PA, USA) to download relevant references, and the references of included papers and relevant reviews were subsequently searched. We did not limit included papers by language, but they were limited to published full texts. An updated search, identifying qualitative studies describing barriers and facilitators of physical activity and sedentary behaviour only (i.e. not including dietary behaviour search terms), was conducted in July 2016.

Study selection
After an initial fidelity check between reviewers (11), three reviewers leading reviews on the included behaviours (K. H.: physical activity and sedentary behaviour) screened approximately 12,500 titles each. A small amount of basic information (e.g. study type and behaviour) about each study deemed to meet the inclusion criteria was extracted. Two random 5% samples of papers (total n = 3600) were also double screened by two additional reviewers (R. L. and E. V. S.) as a quality check. Relevant full texts were obtained and distributed for the behaviour-specific reviews, with data extraction then occurring in parallel.

Methods for activity behaviour qualitative review
Inclusion/exclusion criteria Qualitative studies were eligible for inclusion if they provided an in-depth discussion of barriers and/or facilitators to physical activity or sedentary behaviour in children between 0 and 6 years of age. Studies conducted with parents, caregivers (i.e. grandparents and childcare providers/educators) and children themselves were considered. Exclusion criteria included (i) studies using quantitative methods, including those conducting analysis of free text in questionnaires and (ii) studies conducted in clinical populations (e.g. children who were malnourished or had asthma, cerebral palsy, cystic fibrosis, autism, etc.). For mixed-methods studies (e.g. qualitative and quantitative/q-sort methods), the qualitative element of the study was considered if presented separately. As studies focused on the general population, parents and care providers of children with certain conditions (e.g. asthma) may have been included. Although the influence of clinical conditions on activity behaviours was not the focus of individual studies, if perceived to be important by parents and care providers, it would be drawn out in the extracted barriers and facilitators.

Quality assessment
A standard tool, specific to qualitative study designs, was used to assess the quality of included studies (24). Each was judged according to 12 pre-defined quality criteria: (i) research question clearly stated; (ii) approach appropriate for the research question; (iii) qualitative approach clearly justified; (iv) context/(v) role of the researcher/(vi) sampling methods clearly described; (vii) sampling strategy appropriate for the research question; (viii) method of data collection clearly described/(ix) appropriate; (x) method of analysis clearly described/(xi) appropriate; and (xii) a sufficiency of evidence to support study conclusions (Table  S2). Studies were scored out of 12, where high quality is ≥10; medium, 9-6; and low, ≤5. This was used as an indication of study quality but was not taken into consideration for the synthesis, as directly reported quotes/results were used for the thematic analysis here.

Data extraction and synthesis
All full texts identified for inclusion were read by K. H. and double screened for inclusion by E. V. S. Standardized forms were used to extract relevant information from each paper including the following: first author; publication year; country; study design, setting and population; and descriptive characteristics of the sample. Study-specific data extraction and analysis were then conducted using a thematic framework approach, underpinned by the socioecological model (i.e. individual, interpersonal, community, organizational and policy levels). Data extraction and synthesis proceeded in five stages. First, all papers citing potential barriers and facilitators to physical activity were read twice by K. H. Second, within each paper, key themes identified by original authors and/or this review team, and direct quotes relevant to these, were extracted. The extracted quotes were subsequently used to illustrate each theme; in some instances, quotes have been shortened here for brevity. Third, an iterative process was used to develop the emergent (or sub-) themes across all papers. Fourth, these emergent sub-themes were summarized and consolidated into summary themes, which were mapped onto levels of the socio-ecological model. Finally, using the summary themes, K. H. and E. V. S. derived the overarching theoretical framework, again mapping this onto the socio-ecological model. Any disagreement was resolved by consensus.

Results
A total of 37,868 and 2,824 references were retrieved in 2012 and 2016, respectively, of which 220 were read in full and 43 qualitative papers (describing 35 study samples) relating to children's physical activity and sedentary behaviour were identified for inclusion (Fig. 1). A brief descriptive summary of studies is included in Table 1 (with  a more detailed summary of each included study provided  in Table 2). Studies were predominantly conducted in the USA, Australasia and Europe, and all bar one paper were published in or after 2004.
Across papers, 77 barriers and facilitators to physical activity and sedentary behaviour were reported (Table 4). Most studies used focus groups and semi-structured oneto-one interviews conducted with childcare providers and parents of preschool-aged children (the majority of whom were female) to elicit responses. A few studies also explored the views of preschool-aged children through drawing and 990 Barriers and facilitators to preschoolers' activity K. R. Hesketh et al. obesity reviews discussion. Thematic/content and inductive analysis were most commonly used to analyse the data. In general, studies tended to include predominantly lower socioeconomic or racial minority groups. Many of the included papers (n = 31 [70%]) were deemed to be of high quality: most adequately explained the rationale for their approach, justified the methods used and explained the type of analysis conducted, although studies were often likely to draw conclusions that were not always fully supported by sufficient evidence.

Barriers and facilitators to physical activity
The derived theoretical framework is shown in Fig. 2. Barriers and facilitators to children's physical activity were classified into seven broad thematic areas: the child, the home, out-of-home childcare, parent-childcare provider interactions, environmental factors, safety and weather. These themes spanned one (i.e. child within the individual level), several (i.e. the home/out-of-home childcare across interpersonal, community and organizational levels) or all (i.e. safety) levels of the socio-ecological model. A synthesis of the findings within each thematic area is provided later (also see Table 3 and 4).

The child
Pre-disposal and preferences At the individual level, parents described activity as being innate, with girls perceived to be naturally less active than boys (25,52,53). Enjoyment of activity was also perceived to facilitate physical activity by care providers (52,54), and children themselves (55). I think they have got this natural desire to run around and play, I don't think you have to do too much to encourage them. (Parent [P]) (53) It's absolutely important. And with my 7-year-old it's not a problem. It's hard to keep him indoors, but my 5-yearold, she's more of an indoor gal. (P) (68) A child's own preferences for more sedentary behaviours were cited as largely negative influences by parents (26,56,57) and care providers (27,54). Parents and carers also stated that children generally had a preference for use of electronic media over more active pursuits (28)(29)(30)(31)52,56,(58)(59)(60). Other adults did however state that children could be active or indeed be encouraged to be more active when in front of the TV (29,32,33).
Children can spend all day watching television nowadays. My daughter is completely spellbound in front of the television. It is not possible to talk to her. This is often a struggle, sometimes I just switch off the TV. Unfortunately, not always …. (P) (26) Necessary respite/downtime This theme encompasses parents' (28,29,31,32) (and care providers' (54)) widely expressed belief that children require a certain amount of daily downtime. Importantly, not all

Development
Parents mentioned the importance of positive early habit formation (37,53,57), and a higher skill level (i.e. both motor skills and skills relating to physical activity more generally) in facilitating higher activity levels (53,54).
If you start them young, then they'll build a lifestyle to carry them right through their entire life. (P) (57) We have to go out of our way to teach them sports because we have found if you don't … you get to a certain age where they just won't do it. … and they won't play because they are the only ones who don't know how to … (P) (53) Parents also thought that regular (61), structured activity (37,52,53,56,57) was (on the whole) related to higher activity levels in their children. Age-appropriate activity and allowing children to grow up and undertake independent exploration were further ways that carers suggested activity may be facilitated (52,53,56).
You know like if they've got swimming lessons, I'm happy to take them, but [father name] is like really kind of on it that they need to get the lessons. (P) (62) [Children's physical activity] is a necessary element as far as what we were saying earlier, their muscle development and their balance and I think there is a lot of social development that goes along with it too. (Childcare provider [CP]) (38) However, there was also recognition of the need for children to be 'school-ready', with academic development often deemed to be (more) important for young children (39,52,63).
Because even though I feel that the gross motor is something that's important for the children to experience I think one important barrier to physical activity in obese children is that it is more difficult to participate in activities because of the obesity itself … I think that exercising and sports are very important to a child's health status. (P) (61) The home The role of parents By far, the most commonly cited influences, both positive and negative, were at the interpersonal level. Both parents and childcare providers frequently stated that active parents, who could act as role models for children, facilitated young children's activity (26,31,33,37,(40)(41)(42)(52)(53)(54)57,62). Parents participating in activities with their young child were also deemed to facilitate activity (29,32,52,53,56,60,62), with other positive parental influences including positive encouragement and support, and parents instilling the value of physical activity in their children (53,56).
Parents are responsible for young children's lifestyle; they can't make their own choices, we have to make their choices …. (P) (26) I feel that adults play a crucial role in the physical activity of children. If adults after dinner sit in front of the TV, kids would sit in front of the TV as well. If you go out for exercises, they would follow you. (P) (42) Figure 2 Overarching theoretical framework.  (27,31,34,36,38,40,49,52,54,57,59,60,64,67) 1004 Barriers and facilitators to preschoolers' activity K. R. Hesketh et al. obesity reviews

obesity reviews
If you're going to the convenience store, walk to the store … instead of taking the elevator, taking the stairs. So with my kids I try to do that in the mall, go up the stairs even though it takes a lot longer and walking to the store instead of taking the car. (P) (58) They did however acknowledge that they themselves may also be a barrier to children's activity (25,38,40,42,52,53,57,64). For example, parents stated that their job (29,34,60,62) resulted in a lack of time (26,29,31,32,34,40,43,52,53,56,57,60) and energy (29,32,40,52,53,60). They also had to make the effort to juggle multiple schedules (within the family) (40,53,58), balancing this with the cost of providing active opportunities (37,40,44,52,56,57), which at times limited their young children's physical activity. These barriers also meant that parents tended to use less active modes of transport (i.e. stroller (30)(31)(32)(33)56) and car (52,56)) as they were more convenient and often quicker, which further impacted their child's opportunities to be active.
My mom never does anything with me, she just takes me places and leaves me there. (Child (C)) (55) … I've just started working full-time and it's a long commute and I'm away from the home for such a long period of the day that [I] do more with the kids on the weekend. Yeah, okay then between the laundry and the cleaning and everything else, especially at this time of year [December] when there's so much extra stuff to do that I can't even begin to think of doing some organized physical activity with the Kids. (P) (57) I think it gets to a point of how much can you actually do? You know I've got three kids, how many places am I supposed to be after school on one night? (P) (53) (30) Related to this, a physically active child was often also deemed to have the potential to cause trouble (27,29,33,34,60): limiting physical activity was reported as one way to punish naughty behaviour (54,56), with children themselves mentioning that bad behaviour meant they were prevented from being active (45).
Yeah, they [the children who are too active] are not controlled by the parents. So, it's not … it becomes a level whereby you start also [to] complain: no, this is now too much. (P) (33) For example, when asked how (bad) behaviour influenced physical activity: One 4-year-old stated that you 'get dead, time out'. A 3year-old explained that another child 'pushed me down. I cried. They got time out'. (C) (45) Siblings and peers Children's wider social networks were also noted to be important: siblings were perceived to both facilitate and inhibit children's physical activity levels (52,57,65), with younger children often wanting to mimic or play with their older siblings. In general, interacting with friends was cited by children (46), and parents as facilitating physical activity (33, 34,41,42,52,56,66). Parents believed that coming together as a family, either at home or on holiday, also positively influenced their young children's activity (29,53,56,62).
They [siblings] go out in the garden and she will go and play with them, I suppose so. She doesn't tend to want to play on her own. In fact, I had to always pull her back because she looks up to her brother who is a very active boy and he has been going to all these things, and she always wanted to be just like him. (P) (66) If you live in a neighborhood with adults only, they would not have play-mates and they would have less exercise. That is my case. If you live in a neighborhood with children who like to play with you, that is different. (P) (42) I have a lot of fun with my friends and I feel better. (C) (46) Well actually all of us together, me my husband and [Participating child] … my husband we always liked kickboxing, never had the chance to do it, so we thought oh we'll ask would you like to go and try. (P) (62) The influence of friends and peers was however limited to developing social interactions with children of the same age, with mixed age groups of children thought to hamper younger children's activity levels (27,54,57).
I want to do more group sports but I've got two different ages [of children] to worry about … Just more sports available for the younger 2 to 3-year olds. (P) (58) The home environment More broadly, living close to a child's childcare setting and therefore using active transport to get there (e.g. walking and cycling) was a perceived positive influence (53,56,61). Parents stated that a sedentary home environment, including a TV being persistently on (52,56,58), and small outside spaces inhibited activity (34,52,53,59,60). This said, both childcare providers and parents suggested that homes with adequate space and resources (i.e. active toys, etc.) facilitated activity in small children (52,53,56,58,64).
One of the priorities for me in terms of where we live will be related to school. I really will like to be close to a school so that we don't have to rely on cars. And we always walk to the local shops and to the hairdressers and things. (P) (53) We have a few parents that walk to school and home and I think that obviously encourages physical activity. (CP) (65) My son walks and jumps inside the house. But if you have a small house, the space, the room for him to exercise is limited. What I am saying is that it depends on the condition of your family. (P) (42) [We] live [in an] upstairs [apartment], so it's hard to have indoor playtime that won't bother the neighbors downstairs … it is very easy for things to get broken inside the house. (P) (60) Out-of-home childcare The role of care providers Childcare providers felt that they were able to facilitate and encourage young children to be active (25,27,38,44,52,64), which was corroborated by children themselves (46). Childcare providers (25,38,52,64) did however acknowledge that they also may hinder children's activity, with barriers to providing active opportunities including the demands of work, the need to prioritize academic outcomes, wanting to prevent noise and their own reluctance to go outside (38,39).
You know, It's all the teacher's decision of how much time they're gonna get and how much they're gonna do … Some teachers just aren't into that and some teachers are into that. So it depends. (CP) (38) Sometimes the children are easier to motivate when I participate myself. When I run around and jump on one leg they have a lot more fun and rather take part than when I just stand there and play with my drum.
(CP) (52) Our teacher says that, if we run too much, we sweat, and thus we get a cold and we can no longer attend the kindergarten … our teacher says that we can not play outside football or basket, because we are too lively and we are at risk of hurting any other. (C) (46) I was born in the Dominican Republic. I am not used to the kind of cold weather we get here in [Massachusetts]. I do not think I will ever get used to it … I just try to get through the winter. It can be difficult. (CP) (27) The childcare environment Space and resources within the childcare environment were often discussed as helping and hindering children's physical activity. This may be particularly important given childcare was perceived by providers to be the only place children could be active (38,44). In general, larger spaces, curriculum materials and play equipment (27,47,52) were perceived to benefit children's activity levels. A lack thereof (i.e. small spaces (36,39,54,55,64) and poor resources (39,44,64)) within a preschool environment was perceived to largely inhibit children's active opportunities.
So we provide opportunities for the children to get a bit active … I think it is important that we do that especially as the children who are here five days a week. If they don't get stuff like that at the weekend, at least when they come here, they are on the go all the time. (CP) (44) We've got more stuff [indoor play equipment] now, so whether it's too hot or too cold or too rainy outside, we take it out, then they can do it [activities] in the classroom. So, yeah, being that we have that stuff now, it does make it easier and they get to do a lot more [physical activity]. (CP) (47) My preschooler actually is in a [private preschool] program and they don't have the room to have the physical activity part of it every day or even once a week, so they get it once a month. (P) (

Parent and childcare provider interactions
Of particular interest was the persistent push-pull between parents and childcare providers regarding children's physical activity. As mentioned previously, care providers believed parents to be important role models for their children's physical activity, and responsible for inhibiting their child's active opportunities by, e.g. supplying inappropriate clothing (48) or being worried about children playing outside in inappropriate (i.e. usually cold and wet) conditions (36,39,46,48). Conversely, parents saw childcare centres as a crucial initiator of physical activity for their in children (26,33,41,43,49,52); the active opportunities provided by childcare centres were mentioned as being central to children's daily needs (56,61).
There is no real point in teaching them here and they go home and they do something different so it is a two-way thing! (CP) (43) I've had problems with parents telling the kids, 'Don't get dirty!' because of what they have on. So that's I guess a pet peeve of mine … They're going outside. They are going to play. They are going to be on the floor. They are going to be, you know, children. (CP) (48) I have a hard time when parents don't respect the rules that I have around children bringing and using electronics such as DS. I don't really like to have to keep reminding them that those devices are not allowed … Communication with parents is … very important because it gives us a chance to learn about the child's home environment, the family's routines and rules. (CP) (36) Nevertheless there was also acknowledgement of this tension, and both parents (52,65) and childcare providers (27,36,47) stated that collaboration with the other party was vital for preschool-aged children's active opportunities. Working together, with improved communication between parents and care providers, was often cited (by childcare providers) as how this might be achieved (27,36,47,52).

obesity reviews
There is no real point in teaching them here and they go home and they do something different so it is a two-way thing! (CP) (43) [j]ust like what you do with us as educators, like professional development for us … it's kind of like for the families; if you could help to promote it or inform them [parents] then maybe they might go [to neighbourhood activities and programs] ….
[i]f we don't educate the parents, how are we going to educate the kids. We need to work together with them. It's very important …. (CP) (65) Information that will help would be like to have more communication [with parents about the importance of physical activity] now that winter's coming. (CP) (27) Communication with parents is also very important because it gives us a chance to learn about the child's home environment, the family's routines and rules, which is really important information to have to understand and care for the child in our FCCHs. (CP) (36) Environment Resources within the community, including parks and playgrounds, presented both barriers (34,37,41,60) and opportunities for activity (44,52). Children spending increased time spent outside (27,32,49) and having community support for activity (41) available were perceived to be universally positive influences.
Because there is a lot of fear around with drug users and a lot of crime and that, people don't let their kids out. … I don't let my son out in this area by himself … so he is stuck inside a flat with me, and unless I take him outside, he doesn't get to go outside. (P) (37) Our town is a pretty secluded area and that helps because parents feel comfortable sending their kids outside, and you know they can do whatever they [parents] want in the house and they know that their kid will be ok. … The environment, it is a great place; there is so much to do outside. (P) (41) In order to prevent obesity in childhood, it is necessary to build inspiring playgrounds and green areas where children can play …. (P) (26) Interestingly, although the majority of parents living in rural locations thought this benefited their children's physical activity (41,49,52), rural locations may also prove to be barrier owing to their isolation and a lack of resources (34).
We used to go to the mall in [urban town] a lot because they had this little brand-new indoor play area. There's nothing really like that indoors around here. Like, a play place for kids, you know? I feel like there are a lot of open places in [rural town] that you could fit something like that in. (P) (49) I haven't seen a park around here. Well, there is one in Wilmington, but it's too far to take them in the afternoon. It takes about an hour to go and come back. (P) (34) Safety A common theme mentioned by parents, care providers and even the children themselves was safety. This influence, which was predominantly mentioned as a barrier, spans all levels of the socio-ecological model. Many parents worried that a physically active child could and would hurt themselves and may therefore be likely to limit their child's activity (29,37,41,49,53,56,58,61). Children also mentioned that adults' fears in relation to their safety and their health limited their activity levels (46).
Mom says that I can play with the ball with my sister, but not with all my friends at the kindergarten, because she is afraid I can get hurt … once a friend of mine pushed me and I was hurt and my mom told my teacher that I had not to play running. (C) (46) I think that the most important barrier to physical activity in children is safety concerns re: letting kids outside unsupervised. (P) (61) The way I see it, I prefer that they watch television rather than being endangered outside. (P) (56) This preoccupation with safety was also widespread in childcare providers (27,39,44,54,64), particularly in relation to restrictive preschool policies preventing children from being active (39,52,64). I don't think they really get their heart rate up much from climbing because with all the new licensing regulations [in childcare centres], our climbing equipment isn't that hard anymore …. (CP) (39) We can go on walks except it's got to be a field trip so it's a little more difficult in the sense that we can't go for a neighborhood walk. It has to have a specific purpose, and then we have to get permission. (CP) (64) In order to offer moving opportunities, more staff is necessary. For instance, a climbing landscape requires more supervisory staff. The children are not allowed to do it by themselves. (CP) (52) Importantly, perceptions about changing societal norms and children no longer being allowed to or wanting to go out to play were cited by parents to be major barriers to their children's activity (26,34,41,49,57,60).
I grew up in a country town where you could just go off and play. There were gangs of kids running around the street playing cricket and stuff. You're not living in the city and your kids are just free. (P) (53) This appeared to relate to both their local environment and parents' and children's circumstances: often parents expressed a fear that something might happen to their child if they did not watch them, they were concerned about danger in their surrounding area and, in some instances, that others would perceive them to be bad parents.
Because, sometimes, they fall or they can run away and get run over by a car. So, it's better inside because I can keep an eye on him, here in the room or in the hall: If they go out, they have to go out with us …. They told us that if they saw them playing by themselves there, they were going to be taken by social services. (P) (34) Weather Finally, and often mentioned in combination with safety, the weather was cited as both a facilitator (parents (34,49,57,60) and childcare providers (38,52)) and more often as a barrier by children (67), by parents (31,40,49,57,59) and childcare providers (27,36,38,54,64) to young children's physical activity. At the extreme, this posed a perceived health risk for children, particularly when playing out in the cold (38), with 'extreme heat' and 'mosquitos' (60) also influencing physical activity at the other end of the weather spectrum.
During the summer their activity behaviors are great. I love how active they are and how they want to be outside doing things. (P) (57) Sometimes, I see on the news that children in the sun can pass out. I worry when the sun is very strong because if it's very hot, her heart can beat faster and she could pass out or something. (P) (34) I think the cool air. In the wintertime, (imitates a parent saying) 'Oh no, don't go outside because it's cold outside'. I think a lot of people have the misconception that you are going to get sick if you go outside in cold weather. Really, it's better for you. (CP) (38) No, my mom does not want me to play outside at the kindergarten, because otherwise I get sick. (C) (46) Children urged the dolls to stay inside saying, 'No, that would make them wet'. (C) (67) When it is raining outside or it is very cold, then the children are inside the school building because we [the preschool] do not have a covered playground. And then we [teachers] put on a movie. (CP) (54)

Discussion
To our knowledge, this is the first review to synthesize the qualitative literature relating to barriers and facilitators to activity behaviours in children 0-6 years old. Barriers and facilitators were classified into seven broad themes: the child, the home, out-of-home childcare, parent-childcare provider interactions, environment, safety and weather. These themes spanned between one and all levels of the socio-ecological model (Fig. 2); although themes were categorized according to a primary focus, these themes are not mutually exclusive within the socio-ecological model (SEM), highlighting the complex interplay between barriers and facilitators of young children's activity behaviours. Parents, care providers and the children themselves most commonly cited influences at the interpersonal and organizational levels as barriers and facilitators. A large number of factors remain unexplored in the qualitative literature in the community and policy domains. Evidence from this synthesis does however provide new as-yet unexplored avenues for intervention (e.g. parental time and resources, available space, weather and safety). Moreover, targeting factors that those caring for young children, and the children themselves, believe to be important may enhance intervention tailoring, ultimately effecting both greater increases in young children's physical activity and decreases in sedentary behaviours.
This review combines qualitative evidence relating to barriers and facilitators to both physical activity and sedentary behaviour (together termed 'activity behaviours'). Whilst researchers often differentiate between these behaviours for research purposes, many of the participants in studies included here failed to make any distinction: e.g. a barrier to physical activity was often perceived to facilitate sedentary behaviour (e.g. TV viewing). This was true of both parents and care providers, suggesting that those looking after young children perceive these behaviours to be equal and opposite. Moreover, guidelines for young children recommend specific amounts of physical activity, whilst limiting sedentary time (e.g. (2)). Therefore activity behaviour can be thought of as occurring on a spectrum, where sedentary behaviour and very vigorous physical activity lie at opposite ends. This evidence synthesis is therefore beneficial to researchers working in the fields of both sedentary and physical activity behaviour, providing novel information about what those who are most influential to young children's activity behaviours perceive to be important.

Barriers and facilitators to young children's activity behaviours
As noted earlier, a wide range of facilitators and barriers to activity behaviour were identified in the qualitative literature that have yet to be explored quantitatively (11). Interestingly, parents (and care providers) frequently expressed the belief that children require a certain amount of daily downtime and should not be constantly active. Moreover, not all sedentary time was perceived to be equal, with more academic or developmental activities (e.g. reading and crafts) deemed to be essential in contrast to TV viewing. This fits with the oft-mentioned perception that children are naturally active, which may result in parents seeking to balance this with more sedentary (calming) activities. In addition, many parents mentioned that allowing their children to engage in more sedentary behaviours for part of the day has the added benefit of allowing them to go about their chores during an already hectic day; parents may therefore also benefit from their child's downtime, providing them with some necessary respite to 'get things done'.
Physical activity guidelines for toddlers and preschoolaged children (i.e. ambulatory infants to age 5 years) in several western countries state that children should accrue 180 min of any intensity activity above sedentary (2, 70,71). Studies suggest that children may meet these guidelines (72,73), in part owing to a large proportion of time spent in light intensity activity (73). Allowing a certain amount of 'down time', which in many cases was stated to be of educational value in the qualitative literature here, may therefore be reasonable. Nevertheless, it is important to distinguish between developmentally beneficial (i.e. reading, drawing and crafts) and non-beneficial or entertainment-based sedentary behaviours (e.g. TV viewing), as recently acknowledged by the American Association of Pediatrics. Their guidelines for screen time published in 2016 state that for children age 2 to 5 years, 'entertainment-based' screen time should be limited to 1 h per day, and that infants aged 18 months or younger should not be exposed at all (74); screen time for educational purposes is not included in this time, but some media use for social interaction (e.g. Skype and FaceTime) is permitted. Clearly, younger children will likely benefit from screen time promoting pro-social behaviour. However, parents frequently acknowledged that media use was a major barrier to their children's physical activity (28)(29)(30)(31)52,56,(58)(59)(60).
Providing strategies to reduce entertainment-based screen time in favour of other developmentally appropriate activities may help parents reduce their reliance on sedentary behaviours involving entertainment-based media.
More generally within the home theme, many parentlevel barriers, including parents lacking time (26,40,43,52,53,56,57) and resources (52,53,56,58,64), were perceived barriers to physical activity in preschoolaged children. Although it is not possible to, e.g. increase space within the home or provide parents with more hours in the day, it may be feasible to provide parents with ideas about how to, e.g. use their available time or space more creatively. Use of wider environmental resources, such as parks and community space, which were deemed to positively influence children's physical activity behaviours, may also help parents and children to actively interact. This may be particularly beneficial to boost activity within families (75). In addition, practical advice would be helpful given the concerns parents expressed about child safety when being active and young children's reliance on parents to take them to these places.
The qualitative literature here suggests that within the childcare centre, care providers perceive themselves to be important for children's physical activity. Quantitative literature suggests that providing training for childcare providers may influence change in children's MVPA, but the precise mechanism for this is not clear given the wide variation in training across interventions and countries (11). As no quantitative studies to date have specifically focused on care provider (physical activity) behaviour as a potential determinant of children's activity behaviour (11), it is difficult to determine the direct role childcare providers play in influencing preschoolers' physical activity. Moreover, qualitatively, providers did acknowledge that their own behaviour may on occasion inhibit preschoolaged children's physical activity (38,52,64). This was evidenced in one cross-sectional study conducted in Belgium preschool-age children, where fewer supervising teachers during recess was associated with higher step counts in girls, but not boys (76). Consequently, childcare providers' own behaviour, e.g. preferring to stand or sit during children's outdoor playtime (76), may inadvertently reduce children's and girls' activity in particular. Future work should therefore focus on providers' own behaviour in relation to the children in their care. In addition, given the widespread preoccupation identified with young children being school-ready, work exploring how physical activity may be integrated into academic tasks across the day, as a learning tool rather than separate need, may prove beneficial to children's activity. Such work would be timely in light of provider perceptions about their ability to encourage children to be active, and given the current growing interest in this sphere of research for children later in childhood (77).
Several cross-sectional quantitative reviews conducted previously suggest that elements in the preschool environment may be positively associated with children's activity (78,79). Childcare providers frequently stated that available resources and space within the childcare environment were both positive and negative influences on preschoolers' activity in the qualitative literature (39,44,52,64). However as yet, no clear association between the childcare environment and change in physical activity has been found in interventions (11). In order to show exactly what influences children's physical activity in this setting, it may therefore be beneficial to assess the way children interact with the staff themselves and the environment, rather than focusing on the role of specific elements within the environment.

Emerging themes for use in intervention development
The perceived influence of parent-childcare provider interactions on children's activity behaviours was a novel and potentially important finding to emerge here. Whilst parents clearly have a significant role in shaping young children's health behaviours, young children now spend increasing amounts of time in out-of-home care (80). It is perhaps unsurprising then that parents, care providers and the childcare environment were deemed to be important initiators of young children's physical activity, or that interventions to increase physical activity have also often simultaneously targeted both home-based and childcarebased elements. However, few to date have incorporated or assess interactions between these elements. Moreover, given the opinions expressed here, that each of the other party (i.e. parents and care providers) may be thought to be responsible for children's physical activity, exploring the exact nature of possible conflicts between parents and care providers could provide vital information about how best to encourage physical activity in home and childcare settings. Interventions focusing on interactions between parents and childcare providers (and their environments) could therefore feasibly have a significant positive influence on children's activity behaviours, with resolution of between-group tensions further augmenting intervention success.
Across the quantitative literature, factors such child and parental knowledge have frequently been targeted as potential determinants of physical activity but are rarely associated with positive change in young children (11). Interestingly, these often-targeted determinants were not mentioned across qualitative studies as barriers and facilitators to physical activity, suggesting that adults (and also children) are already have implicit knowledge that physical activity is beneficial. In contrast, the two final themes identified in this review, safety and weather, were largely cited as barriers to children's physical activity. Yet despite both being plausible and potentially influential determinants of young children's activity behaviours, they have rarely been explored in the quantitative literature (11,16,18).
Extreme weather was frequently mentioned in the context of child safety, but a general preoccupation with ensuring children were protected, usually at the expense of being physically active, was pervasive. In childcare centres, this risk aversion was commonly linked to policies around play and requirements for establishing 'safe' environments before children could be allowed to be active (39,52,64). Indeed, children themselves perceived their play was restricted because physical activity could result in them hurting themselves (45). Notably, the terms 'play' and 'activity' were often used interchangeably across studies here. Play is described as 'some social, locomotor, fantasy, or object-directed activity that is not directly functional' (81), which a child usually engages in voluntarily (81). It can occur in each of four domains (i.e. locomotor, object, pretend and social) and can be, with the exception of social play, enacted alone or in a group (81). Children accumulate their physical activity in each of these domains, and largely in an informal manner during the preschool years (82). Changing societal norms noted here, tending towards risk-averse behaviours that limit children's activity and opportunities to take exploratory risks, may consequently impact not only children's physical literacy/activity (83) but also their social and motor development through play (82). Modification of adult risk perceptions may therefore be one way to positively influence young children's activity behaviours, risk awareness and wider development. Doing so, either by intervention or in natural experiments, e.g. changing health and safety practices, would be an interesting and hitherto unexplored avenue of research.
Moderate weather, either hot or cold, was perceived to facilitate children's activity behaviours (34,38,49,52,57,60); hot and cold extremes were perceived here to result in less physical activity and more sedentary time in preschoolers. This fits with quantitative research assessing weather as a correlate of physical activity. For example, in locations with more temperate weather year round (84,85), samples of US preschoolers tended not to show variation in physical activity. In contrast, in studies conducted in regions where clear seasonal variation in weather exists, there are comparable fluctuations in physical activity (e.g. in samples of UK (73,86) and New Zealand (87) preschool-aged children). This was also shown in a longitudinal study of 3-year-old children conducted in Cincinnati, USA, where hot and cold weather conditions were negatively associated with total activity and MVPA, and positively associated with sedentary time (88). Where possible, learning to adapt to weather conditions that may provide fun opportunities for physical activity (i.e. wet weather), or providing parents and care providers with viable contingency options in locations were weather extremes are common, may be advantageous. This could prevent such marked seasonal variation in young children's physical activity and lead to consistently higher levels of physical activity in this age group year round.

Social context and future research directions
The studies included in this review were from culturally diverse sub-populations, and frequently recruited from lower socioeconomic strata. In general these groups are less likely to participate in quantitative (both prospective and intervention) studies. It is therefore beneficial to ascertain the perceived barriers and facilitators to activity behaviours across strata qualitatively. This work suggests that the concerns and activity promotion ideas of those caring from preschool-aged children (i.e. safety, the influence of the home and childcare centres) are relatively stable across infancy and the preschool-period despite differing socio-demographic profiles. Moreover, parenting during the early years is time intensive regardless of resources. Therefore, that consensus as to the barriers and facilitators to young children's activity behaviours emerged across several themes here hints to a number of determinants of children's activity behaviour being viable future intervention targets. Importantly, based on this work, such suggestions as to how to promote activity behaviours could potentially be externally valid across preschool-aged populations internationally.
Finally, this review highlights were gaps in the current (qualitative) literature exist. Interestingly as in the quantitative literature (11), there was a paucity of research with fathers and male care providers, and in developing countries; greater engagement with these populations is required as perceptions and influences on young children's physical activity may differ by carer sex and cultural practices. In addition, all included studies used focus groups or (semi-)structured interviews to elicit participant responses. Other qualitative data techniques (such as ethnography/observation) may also provide valuable insights into how parents/carers and young children operationalize perceptions around activity behaviours in daily life, creating an important bridge between qualitative and quantitative literature. Further, in an era after the epidemiological transition (89), it is more challenging to determine why some children are more or less active than others as we often lack the heterogeneity in exposures (and outcomes) to explore associations quantitatively. Developing countries are currently undergoing similar transitions towards lower physical activity and higher sedentary time as those seen previously in higher income countries, perhaps at a fast rate. By conducting both qualitative and quantitative research in developing countries, where greater heterogeneity may still exist, we can ascertain potential influences on young children's activity behaviours. From this, we can learn how best to reverse negative trends towards inactivity in the future. Lastly factors in the environmental (community resources, rural locations and changing society) and policy domains of the SEM were rarely studied in the quantitative literature, but were suggested by parents and care providers in the qualitative literature as having a significant impact on young children's activity behaviour. These factors therefore represent key future avenues for quantitative activity promotion research in the preschool years.

Strengths and limitations
This review is the first to specifically explore and systematically synthesize the qualitative barriers and facilitators to activity behaviours in children aged 6 and under. We applied rigorous review methods and did not exclude papers based on time of publication or language (although no foreign language papers were identified). In addition to a systematic literature search, we used a parallel reference examination to yield an additional two studies for inclusion. As this review was restricted to published studies, publication bias cannot be discounted. However, that we included a range of international studies, which used differing methods of qualitative data collection and analysis, is another strength of this review.
Qualitative studies were predominantly published from 2005 onwards, with a large increase in studies after 2012. These studies, conducted across differing continents, provide only a snapshot in time of parental, childcare provider and, to a lesser extent, children's activity behaviour perceptions. A large number of qualitative studies (72%) were classified as high quality, adequately explaining the rationale for their approach, justifying the methods used and explaining the type of analysis conducted. This said, included studies tended to draw conclusions that were not always supported by sufficient evidence (i.e. did not explicitly state methods such as triangulation were used to confirm findings), or made generalizations to populations not included in the study sample.
Qualitative research enables in-depth insight into specified topics of interest, with authors interpreting evidence and subsequently using methods such as triangulation and member checking to confirm the credibility (or 'truth') of their findings (90). This review draws upon author findings and also uses primary evidence from each paper (i.e. synthesizing authorextracted themes, quotes, etc.) to construct an interpretation of the wider qualitative evidence base. Whilst it is not possible to determine the extent of confirmability (90) for each individual study, every effort was made to ensure our findings were not influenced by researcher bias, motivation or interest. Moreover, similar themes, barriers and facilitators were identified across the wide range of included studies. This implies that the perceptions around activity behaviours of those (women) caring for young children are relatively dependable (i.e. consistent) and transferable (i.e. stable across contexts such as location and time) (90).

Conclusions
This synthesis of qualitative evidence of barriers and facilitators to activity behaviours in preschool-aged children identified seven broad themes, highlighting the perceived influences of the child, the home, out-of-home childcare, parent-childcare provider interactions, environmental factors, safety and weather. Similarities were apparent between previously explored determinants from quantitative studies, and the barriers and facilitators to activity behaviours identified here, particularly at the interpersonal and organizational levels. Thoughts and beliefs of men caring for young children, those from developing countries and about barriers and facilitators in the environmental and policy domains are currently lacking. Nevertheless, this qualitative synthesis compliments previous research and provides new as-yet unexplored targets for intervention (e.g. the parent-childcare provider interaction, safety and weather). By focusing on factors that those caring for children 0-6 years of age believe to be important, it may be possible to enhance intervention tailoring, and ultimately effect greater positive change on young children's activity behaviours.