“It's not as easy as saying, ‘just get them to eat more veggies’”: Exploring healthy eating in residential care in Australia

Young people living in residential out-of-home care (henceforth OoHC) are at increased risk of becoming overweight or obese. Currently, recognition of the everyday mechanisms that might be contributing to excess weight for children and young people in this setting is limited. The aim of this study was to better understand the barriers and complexities involved in the provision of a 'healthy' food environment in residential OoHC. Heightening awareness of these factors and how they might compromise a young person's physical health, will inform the development, refinement and evaluation of more sensitive and tailored weight-related interventions for this population. The paper presents a nuanced picture of the complexity of everyday food routines in residential care, and illustrates the ways in which food is 'done' in care; how food can be both symbolic of care but also used to exercise control; the way in which food can be used to create a 'family-like' environment; and the impact of traumatic experiences in childhood on subsequent behaviours and overall functioning in relation to food. It is argued that a health agenda designed for a mainstream population ignores the very complex relationship that children in residential OoHC may have with food. It is recommended that future intervention approaches account for personal food biographies, trauma and children's social backgrounds and how these are implicated in everyday practices and interactions around food.


Introduction
For decades research has found an association between child maltreatment and a number of adverse outcomes across the lifespan (Ferraro, Schafer, & Wilkinson, 2016;Fratto, 2016;Greenfield, 2010). Broadly, maltreatment can be defined as ' … ill-treatment (that results in) actual or potential harm to the child's health, survival development, or dignity in the context of a relationship of responsibility, trust or power' (World Health Organization, 2006, p. 9). Maltreatment includes a range of behavioural phenomenon but commonly relates to acts of abuse (including sexual, physical, and emotional) and/or neglect (Greenfield, 2010). Despite the current national focus in Australia on early intervention and provision of family support services to minimize the number of children who experience maltreatment, the most recent published statistics indicate that between July 2013 and June 2014, 143,023 Australian children, or 1 in 37 children aged 0e17 years, received child protection services (Australian Institute of Health and Welfare [AIHW], 2015). This includes 54,438 substantiations relating to 40,844 children (i.e., after notification and subsequent investigation, child protection concluded there was reasonable cause to believe that the child had been, was currently being or was at risk of being abused, neglected or harmed) (AIHW, 2015). In England, the number of children on the child protection register is similar, with 49,700 recorded in 2015, however, lower rates were reported for other parts of the UK: 2935 children in Wales, 2751in Scotland and 1969in Northern Ireland (Scottish Government, 2016. For some children, this elevated risk of harm will have resulted in them being removed from the care of their primary caregiver(s) by child protection authorities and placed in OoHC (AIHW, 2015).
Although OoHC provision differs slightly across each Australian state and territory, overall, there are five different placement types: (1) foster care; (2) relative or kinship care; (3) family group homes; (4) residential care (where paid staff provide 24-h care for up to four young people in a residential unit or house); and (5) independent living (Department of Families, Housing, Community Services and Indigenous Affairs, 2011). The children living in such OoHC arrangements experience a wide array of adverse physical and mental health outcomes, most likely as a consequence of maltreatment and potentially compounded by adverse experiences and placement disruption in care (Bromfield & Osborn, 2007). Particular health challenges that these young people may face include, developmental delays, disability, learning difficulties, poor dental health, lower levels of immunisations, higher levels of general health problems (including illnesses and accidents), mental health issues, behavioural disorders, and risky health behaviours (including higher rates of teenage pregnancy and self-harm) (Department of Families, Housing, Community Services and Indigenous Affairs, 2010; Wise & Egger, 2007).
There is emerging evidence that being overweight or obese is also a significant issue for children and young people living in OoHC. Indeed, a recent study in Victoria (Australia) reported the prevalence of obesity in this population to be almost three times higher than young people in the general community . This finding is consistent with international studies, which also report high rates of overweight/obesity within this group (Skouteris et al., 2011). While the contributors to obesity within this vulnerable population are undoubtedly complex and multifaceted, a number of biological and behavioural mechanisms linking maltreatment to obesity have been proposed (Mason et al., 2016). In particular, research has focused on understanding the impact of stress-related changes to neurobiology, physiology, affect and behaviour (Hemmingsson, Johansson, & Reynisdottir, 2014;Mason et al., 2016;Vamosi, Heitmann, & Kyvik, 2010). While understanding the pathways linking child maltreatment to excess weight is important for identifying trauma-informed targets for prevention and treatment of maltreatment-related obesity (Mason et al., 2016), there is a need to understand the everyday mechanisms (i.e., structural, personal and relational barriers) contributing to excess weight for children and young people in OoHC.
The Healthy Eating, Active Living [HEAL] Study, a 12-month randomised trial conducted in Australia, sought to identify and address risk factors contributing to the physical health of young people in residential OoHC, specifically being overweight or obese . The HEAL Study was comprised of three different phases: (1) Phase one established the need for intervention, including examination of the rates of overweight/obesity in a sample of young people living in residential OoHC and their carers ; (2) Phase two involved identifying possible determinants of overweight/obesity in the target population; and (3) Phase three consisted of intervention development and evaluation. This paper draws on data collected in the second phase of the project, focusing on the ways in which young people and carers in residential OoHC experience food. Specifically, we sought to expand the current understanding of food and eating in residential settings through examination informed by the Food in Care Study [FaCS] Emond, McIntosh, & Punch, 2013;.
The FaCS used an ethnographic approach to explore food practices and the meanings food can hold for children and young people in OoHC, as well as their carers [FaCS]. The aim of this research was to provide an in-depth analysis of the micro-level food practices that are carried out day-to-day in residential children's homes in Scotland. Some of the key findings of the FaCS include: (1) food can provide a window into how care is given but also how it is received and experienced; (2) food practices are powerful mechanisms of socialization, which emerge as a medium for expressing feelings and relationships, across many contexts; (3) food has enormous potential to help children be nurtured and recover from the absence of nurture, to feel as if they belong, that they are cared for and to feel connected; and (4) food can be used to exercise control (Punch, Dorrer, Emond, & McIntosh, 2009).
In the current study, our specific focus was on understanding the barriers and complexities involved in the provision of a 'healthy' food environment in residential OoHC in Australia. Arguably, food and weight are imbued with moral and ideological meaning (Evans, Davies, Rich, & DePian, 2013), and in a residential context, the condition of children's bodies is increasingly seen as an indicator of 'good' or 'bad' care. Food has become a nexus for measuring care, yet very little is known about the day-to-day experience of how food is 'done' in this context and the meanings given to the practices that surround it. This paper argues that by deepening understandings of these factors and how they might compromise a young person's physical health, the development, refinement and evaluation of more sensitive and tailored weightrelated interventions for this population will result. The paper critically engages with the current discourse of 'healthy' living and draws on the FaCS Emond et al., 2013;McIntosh et al., 2010;Punch et al., 2010) with a view to expand understandings of food and food practices that contribute to a 'healthy' food environment in residential OoHC.

Methods
Eleven focus groups with residential staff and 18 face-to-face interviews with young people were conducted with representatives from one participating community service organisation, and one therapeutic residential care facility run by the Department of Health and Human Services. Sixty-nine staff were invited to take part in a focus group, and 56 agreed (81.2% response rate; mean age ¼ 38.0 years (SD ¼ 11.9), 78% were female, 62% held certificate or diploma qualification, and the average time spent working in residential care was 28 months (SD ¼ 30.0)). Eighteen of the 32 young people who were approached took part (56.2% response rate; mean age ¼ 13.0 years (SD ¼ 2.0), 55% were male, 27.8% Aboriginal or Torres Strait Islander, and the average time spent living in residential care was 24 months (SD ¼ 26.1)). This study was approved by the Deakin University Human Ethics Research Ethics Committee and the (former) Department of Human Services Research Coordinating Committee. Participation was voluntary and young people and carers were eligible if they could provide informed consent.
For both groups, a semi-structured interview schedule including both open-and closed-ended questions was used to explore the barriers to creating a healthy eating environment in residential care. This included a series of interview/focus group questions developed for the UK FaCS study (Punch et al., 2009) and adapted for the HEAL study context. Staff were asked specific questions relating to food preparation and storage, mealtime routines, the role of food in residential care, and current strategies to support healthy eating. In turn, the interviews with young people explored their likes and dislikes in relation to food and the surrounding practices, how they experienced mealtimes, and the varied regulatory practices around food. All focus groups and interviews were audiotaped.
Techniques drawn from a framework analysis approach were used in the current study (Ritchie, Spencer, & O'Connor, 2003) to gain a contextualised understanding of micro-level food practices in residential OoHC and the meanings given to this by staff and young people. This approach allowed relevant themes to be generated from the interviews during analysis, rather than developing themes in advance and coding them as such. Initial thematic coding in NVivo was undertaken by one researcher whilst a second verified the codes. Any differences were discussed and a decision was made as to the final code. Inter-rater agreement between the themes identified by the two coders was 90%.

Findings
The key themes to emerge from the analysis included: (1) food routines; (2) food in context (3) food, care and control; (4) creating a family-like environment; and (5) the challenges of the past on the present. The remainder of the paper briefly outlines the findings in relation to these themes and discusses them in the context of wider research. In particular, parallels are drawn with the FaCS study (Punch, McIntosh, & Emond, 2012), and the current obesity discourse. Implications for residential practice in Australia and internationally are identified.

Food routines
In line with the findings of the FaCS (Punch et al., 2012), it was apparent that the 'normal' food routines which staff [HEAL] employed in their own homes were seen as a template of good practice but one which was often very difficult to implement within the context of residential care units. For many of the residential staff involved in FaCS, challenges around managing their unit's food routines included: (1) having to lock food storage areas; (2) food being used as a reward or a bribe; (3) food being used to wield control; and (4) many young people finding rituals around food threatening or completely unfamiliar. Similarly, staff in the HEAL study also indicated they exerted control over food provision and access to food storage areas; often food was locked away or kept in an area that young people could not access, e.g., the staff office. Underpinning these practices seemed to be a wide range of factors including the young people's perceived inability to self-regulate their food intake (especially in regards to energy-dense snack foods): The snack boxes have always been in the office. Because we had a 20 bag of chips [crisps] and it would go in two seconds. You only need to look at it and its gone -that's always been kept in the office and we distribute it to them. [RCW1] A number of staff [HEAL] also reported a high incidence of food wastage, both in terms of food not being eaten as well as food being 'contaminated' or spoiled by young people as a means of resisting the offered care or exerting power and control: Urinate into it or spit into it … crushed up laxatives in cake, in chocolate icing. And you don't leave the table unless another staff members watching your plate or you take your plate over. [RCW6] Many of the young people [HEAL] interviewed shared these views. Most talked about 'junk' food being locked away, offering the same explanations as staff for why this was done, including: (1) food is eaten too quickly if left out, either due to boredom or young people eating during the night; (2) food wastage; and (3) food being used as a weapon to 'get back' at others. Many responses were similar to the following: Nutella, Milo, eggs, and technically everything that you can poke someone with or make someone cough or make someone itchy with. So pretty much just not biscuits … because someone might stab someone with it or burn their eyes with chilli or something or pepper spray someone. [YP10] Most young people indicated they didn't mind that food was locked away and could see the rationale for this practice. However, others clearly stated their frustration and powerlessness in relation to this. For example: [Interviewer]: How do you feel about that?
Not happy. Because sometimes I get hungry in the middle of the night and there's no food. [YP4] Paralleling the findings from the FaCS , food could also be used for a range of social and behavioural purposes both by staff and young people. Interestingly, the majority of HEAL staff described using food as a 'treat' or reward for 'good' behaviour. However, officially, three of the participating residential units [HEAL] clearly stated that food was not to be used as a reward as there were already "so many issues around food". It was of note that some staff [HEAL] made a distinction between using food as an incentive, tool or strategy for encouraging a young person to engage with them versus using food as a bribe. As can be seen in the extract below it was unclear whether this distinction was evident or whether particular terms were discouraged because of their moral and ethical underpinning. In this example, a member of staff is corrected by her team leader (TL) for framing this type of food practice as a bribe: RCW3: Incentive sorry. For coming here today, sometimes they'll still be in bed now, so we'll say, "come to the meeting, we'll get you Macca's [McDonald's] on the way".

RCW11:
Well yesterday I got them out of the house only because I said to them, "we'll go and have coffee and cake". And we ended up doing that. When asked whether food was ever used as a form of punishment, the majority of staff [HEAL] stated that it was never used in this way. However, a few did describe taking away 'treats', as a consequence for 'misbehaving': … But if behaviours have not warranted that reward, then they would have to have a home meal that night and not have takeaway. [RCW13]

Food in context
Staff in the HEAL study were prompted to provide further detail on the everyday activities surrounding food preparation and eating, including cooking ability. Vast differences in cooking and meal preparation knowledge and skills were revealed across teams. Many identified poor planning and communication around food and the difficulties of having one staff member do the shopping, if a weekly meal plan was not in place. This was especially difficult in units that rely on a large number of casual or relief staff. For example: … everyone cooks differently. The person who's done the shopping might have steaks pulled out and they might have had an idea for steaks, but when I see steak I think of something different. So there's usually an ingredient missing. [RCW1] Variation in each staff member's cooking abilities and interest in cooking was also discussed. Staff talked about some colleagues having limited ability to cook, having little interest in cooking, or not knowing what to prepare with the ingredients purchased. Differences in skill levels was frequently linked to inconsistencies in the quality of meals prepared: … So you've got some staff that are cooking home cooked meals and trying to do activities, and then you have some staff that just come in and put frozen pies in the oven …. [RCW4] This also impacted on young people's enjoyment of food [HEAL]. When asked if there was anything they disliked about the meals prepared for them, responses included: meals being unhealthy; simply not liking what was cooked; having an allergic reaction to the meal prepared; staff having poor cooking skills (including burning food); and staff not making foods the young person had requested: Some are overcooked or some are undercooked and it tastes yuck. [YP6] There's probably more unhealthy food than there is healthy. [YP09] Interestingly, staff [HEAL] views regarding the importance of being able to cook (to fulfil their job role) were varied, with some identifying it as 'an important skill that somebody needs to have', while others placed less value on this aspect of their role. For example: At the end of the day we're not cooks, we're not chefs, we're workers. [RCW5] Staff [HEAL] were also asked about involving the young people in meal preparation. It was explained that at a bare minimum, the young people were able to contribute by helping decide the upcoming menu during a weekly house meeting. However, despite encouraging their participation, difficulties involving the young people in the shopping and/or food preparation were described. These included: staff experiencing increased stress and/or anxiety related to having young people in the kitchen or taking them to the supermarket, as well as hygiene concerns. Staff from four of the units involved in the HEAL study indicated that young people often refused to eat a meal if they knew that a particular young person had helped prepare it. For example: … [Client 2] will say, "Oh, I'll help, I'll help". And it's so hard because you want her to help but if the boys find out that she's helped, they won't touch it. [RCW14] One young person who stated they would prefer other children were not responsible for meal preparation confirmed this: Well I wouldn't let the kids cook cause I think when the kids cook, they probably did something -cause I felt sick afterwards and I couldn't go to school for about a week. [YP6] Meal preparation could also be interrupted by the context of the residential unit, including responding to crises or incidents. This was seen as placing limits on the time staff had available to prepare 'healthy', home cooked meals.
… if you're allocated to do the preparation of the cooking and it's really chaotic you don't actually have time to do it … so sometimes we have to take short cuts because the meal isn't prepared. [RCW10] 'Challenging behaviours' as identified by staff, also influenced their enjoyment of preparing a meal, as well as their ability to make something from scratch: If you've had a massive critical incident, and you haven't prepared the roast that you planned to, it'll be something quick and easy … if you've had kids throwing hissy fits all day and you're highly stressed, you're not going to have a perfect meal planned. [RCW9] The perceived need for specific training around food practices and eating was varied. Some staff [HEAL] indicated 'it would be nice if we were a little bit educated', and others believed their own knowledge and life experiences are sufficient. This seemed to exemplify the notion that food was an everyday, taken for granted resource and set of practices: You'd like to think that we're not all idiots and didn't need to be trained. [RCW27] When prompted with questions about whether they had ever experienced any tensions with young people over "healthy" versus "junk" foods, responses were mixed across groups. Some staff commented that this was not an issue for them, whilst others described instances where a young person reacted aggressively to a staff member offering a healthy meal or limiting unhealthy foods: "Yesterday I had a client who I'd arranged for her to have a ham sandwich. As I was getting the ham out of the fridge, she wanted an icy pole and I said, "How about you have your sandwich and then I'll get you an icy pole?" And she trashed the unit and smashed a window and continuously threatened me. These examples illustrate how in the interest of health, staff try to regulate the young people's eating. However, this interaction can be a trigger of conflict and aggression. Staff in the FaCS study talked about how these types of outbursts were often not about the food itself (Punch et al., 2009). Instead, food was recognised as a safe outlet for expressing emotions, and was used by young people as a means of communicating how they are feeling (Punch et al., 2009). This highlights how a young person's resistance to staff regulating food -e.g. the staff member not allowing the young person to have chips -may represent much more than just defiance of authority. It is thus important that we start to reflect on the significance of food, especially as a medium for opening communication between young people and staff.

Food, care & control
When you bring a number of young people (who are likely to be unrelated) and carers together in a residential care unit, they will each bring their own food-related meanings, values, rituals and routines. Hence, it is important to be conscious of, and sensitive to, what food may symbolise for different people in the unit and how this may impact on food provision (Emond et al., 2013). Understandably, food is often considered just an ordinary, everyday aspect of care provision -the interactions and meaning around food practices are often not considered in any detail. However, food can be regarded as carrying significant symbolic social and emotional meaning. For young people in care, food can be a way of demonstrating care to others and also a means of receiving it (Emond, Steckley, & Roesch-Marsh, 2016). As with those in FaCS , staff in the HEAL study also talked about how food can be an important way to show they cared for the young people in the unit: I think being European it's, "if you're feeding them, you're loving them". I think I do that a lot with the kids. I'm feeding them so I must be loving them in some way. It doesn't matter what it is e soup, nachos, whatever. Whatever it is, the intention of putting it together and giving it to the client is to look after them in some form. [RCW19] In FaCS a number of staff also described the ways in which food could be used by children as a means of gaining control, often because it was felt that they had very little control over many key aspects of their lives . Staff referred to food being used as a "power tool", a means for "power tripping", a "weapon" and "a way to kick off" when children were upset (p. 98). McIntosh et al. hypothesised that this was in part the consequence of children in residential care living highly regulated lives and therefore seeing food as a vehicle for regaining some personal control and/or exerting control over others. Staff in the HEAL study presented similar examples where food appeared to be used by children as a mechanism for reclaiming control and/or a mechanism to control others. For example: Well again, [Client 1] when he first came in, that was a way he could control his world e he was refusing to eat. But as time has gone on that's not an issue. But that was very much the only thing he could control in his life at that stage. He didn't want to be here, he wanted to be home. [RCW15] The young people [HEAL] were also asked a series of questions to try and better understand what food means to them and whether this might have changed since they moved into OoHC. The majority of young people indicated that the type and/or meaning of food had not really changed from when they were living at home. However, five young people did agree that the experience of food is different in OoHC. Three specifically stated that they eat more food now they are in care, for example: I didn't really eat food at all. Like if I wanted dinner I had to make it myself and I just couldn't be bothered really so I never ate … I actually get people to cook my own meals for me, instead of me doin' it. [YP1]

Creating a family-like environment
Similarly to participants in FaCS , staff in the HEAL study made a considerable effort to make the unit 'homely' and food was frequently referred to as a tool for creating a 'family-like' atmosphere. Staff [HEAL] stated that they used practices such as having mealtimes at the table to create what they considered to be a "home-like" environment. In reality, though, how frequently this occurred was mixed. When asked how often they come together for a meal and to sit at the table, the young people [HEAL] initially responded using a ten-point Likert scale, where 0 ¼ never and 10 ¼ every day. More than half of the young people scored a 5 or less and only five indicated they sit at the table every day. This view was supported by staff, with the majority advising that despite making a concerted effort to get the young people to the table for dinner, such meals happened infrequently: I would love to eat with them. I would love to sit down and have a meal with them but they normally go off into the lounge room, computer room or their room and take their meal and eat it … [RCW27] Staff [HEAL] reasoned that young people preferred not to eat at the table because they might not 'feel safe', because it can be a 'foreign experience', or can be perceived as 'confronting'. Many staff saw this as a result of missing out on these kinds of experiences when living with their biological families. More generally, for adolescents, over time, eating dinner with the family is a less frequent phenomenon (Walton et al., 2016). However, for those staff and young people who regularly sat together for dinner this social practice was well liked and seen as providing some sense of 'family', enjoyable conversation, and they liked being around others. It also appeared to rest on a powerful, if mixed, constructions of 'family life': We need to make it as much as a normal, healthy, family life for them as we possibly can. So it's highly encouraged that staff eat, especially the night time meal, with the young people. And we try to eat lunch with them most days. Interestingly, when children in the FaCS were asked about whether they felt 'at home' in residential care, the majority indicated they were living in an institution as opposed to 'their home' (Punch et al., 2009, p. 24). An important factor in making residential care feel more 'home-like' was building relationships with others: 'It becomes 'home' when you get to know people and when you feel that people care about you like they care about a friend or their own family and they make an effort to understand you as a person' (Punch et al., 2009, p. 24).
The children identified that food can really only make you feel 'at home' when it is shared with family and friends, or if it is prepared by a family member (Punch et al., 2009). However, if the person preparing and sharing food with the young person was able to form a trusting relationship, this was identified as being fundamental to how food is experienced (Punch et al., 2009).
3.5. The challenges of the past on the present Barton, Gonzalez, and Tomlinson (2012) suggest that many young people who have suffered deprivation, trauma and/or abuse can be very anxious in relation to food. When asked to comment on the role of food in residential care, staff from every HEAL focus group identified a range of problematic eating behaviours, either from past or present residents. Each of these highlights the possible impact of the young person's past experiences on their current food behaviours: Whole range of issues. This lot of kids don't have unusual food behaviours. Others have been bulimic, refuse to eat in front of staff, strange food combinations, and other kids will only eat the same type of food and not try anything new. [RCW18] Perhaps most commonly discussed, was children's emotional relationship with food. Many young people [HEAL] were described as having tendencies to comfort eat or eat in response to particular (usually negative) emotions. For example: And then one of our other clients who is constantly wanting love and affection she feels that she can get that through food, that love and affection. [RCW9] I think there's a degree of comfort. Food is a comfort … often when they've been emotionally upset they'll raid the fridge. [RCW21] This was confirmed by many of the young people who, when prompted to describe whether food 'said something about how you feel', agreed that the type of food someone eats could be an indication of feelings. The young people were able to link different foods with moods, emotions and/or comfort eating: It is important to note that these responses to food may not correlate with a history of trauma. Indeed, a number of researchers have suggested that looking for comfort through food and using food to express or repress emotion is very much a feature of day-today life outside of residential care (Tanofsky-Kraff et al., 2007;Troisi & Gabriel, 2011;Wansink, Cheney, & Chan, 2003;Wildermuth, Mesman, & Ward, 2013). However, in the HEAL study, food was often described by staff as an emotional 'trigger', whereby particular foods might initiate disruptive behaviours. This was thought to result from particular foods being associated with negative past experiences or food being a focal point for the expression of feelings: It's a huge focus for them. Especially if they're feeling frustrated about something, food can become the centre of their frustration. [RCW25] Well we had an incident the other night because a child was served a food that he hates, we have no idea why but again it's a trigger. [RCW6] Research on learned food preferences indicates that individuals learn to like or dislike foods, depending on the context in which they're eaten and whether the experience is enjoyable or not (Benton, 2004). For children with a history of maltreatment, some foods might trigger positive memories and others may be associated with trauma and abuse (Barton et al., 2012). Staff also linked early neglect and food insecurity to a range of behaviours, including: only eating very specific foods; having a tendency to favour easy to prepare, pre-packaged foods (often with little nutritional value); hoarding behaviours; and taking food and hiding it in their bedrooms: A lot of them will, not so much the kids we've got at the moment but certainly in some kids we've had in the past, they will steal food from the unit and store it in their bedroom. And this might be because they've had long periods of their life where they haven't had food always in the cupboard, so they sort of stock it up. [RCW2]

Health implications of everyday food practices
Interviews with residential staff and young people [HEAL] provided insight into the complexity of everyday food routines in residential care. The current discourse around obesity provides some insight into possible practices around food that might be contributing to excess weight in this population, including: (1) restricting access to food; (2) using food as a reward; and (3) comfort eating. It is important that these findings are thought about in the context of residential care. For example, it appeared that concerns and regulation in relation to 'health and safety' resulted in food practices being highly regulated in residential care. The focus on providing a protective environment (i.e., maintaining safety) can place limits on the quality and nature of the eating experience and the food practices that are employed in each residential unit. Additionally, increased regulation around food may represent an important barrier to the provision of 'healthy' foods. For example, staff in one group explained that they were unable to leave a fruit bowl out for young people to help themselves: A lot of the fruits put away cause we don't want it to be thrown around. It's not there for them to see. [RCW32] Whilst removing food resulted in the immediate behaviour ceasing, the cause of the behaviour and the impact of not having access to such foods was less frequently described. Given that healthy food availability in the home has been found to be associated positively with daily consumption of fruit (and vegetables) and associated negatively with consumption of sugar-sweetened beverages and energy-dense snack foods (Loth, MacLehose, Larson, Berge, & Neumark-Sztainer, 2016;Stephens, McNaughton, Crawford, & Ball, 2014), this may represent one pathway leading to weight gain for young people in care. This is not to imply that carers who restrict access to fruit are being 'bad parents' but highlights a need for further reflection on the frequency and causes of the behaviour.
A further unintended consequence of the regulatory practices described by participants may be the development of food preferences that promote weight gain. Certainly there is literature which indicates that restricting access to particular foods decreases a child's ability to self-regulate their intake, and increases rather than decreases preference for and consumption of that food (Benton, 2004;Loth et al., 2016;Stephens et al., 2014). Again, this highlights a need for further reflection on how food is currently being 'done' in residential care, especially since restricted access to energy-dense snack foods was common practice in many units. Although this approach is well intentioned, and intuitively, a sensible method for promoting healthy eating behaviours, it may lead to overconsumption of these foods once staff control is removed. Likewise, using food as a reward for approved behaviour can also enhance a person's preference for those foods (Benton, 2004;Loth et al., 2016). This is a key consideration, given carers frequently talked about restricting access to 'junk' foods and gave examples of using food like McDonalds to reward children for compliance. While these observations are not unique to residential care, and commonly occur in 'typical' families (Benton, 2004), it signals a need to develop strategies that will help staff increase availability of nutritious foods and consider alternative incentives for good behaviour, within the limits of an often pressurized environment.
Another possible trajectory for weight gain is using food as a means to cope with tensions and discomfort, as a form of selfsoothing or as a way of trying to keep difficult feelings at bay (Barton et al., 2012;Goossens et al., 2009;Kelly et al., 2016). There is a growing body of evidence which links stress to consumption of highly caloric and palatable foods (Mason et al., 2016). This is likely due to the satisfaction that people experience when eating these types of food (Benton, 2004). Staff in the current study frequently identified that the "comfort" foods chosen by the young people are not necessarily a 'healthy' food choice: … They want instant gratification. Which sometimes fast food gives you with that high sugar and fat content -which they may not get from other foods. [RCW22] Not surprisingly, high consumption of these foods is generally accompanied by weight gain. Given it is not uncommon for children in care to have a history of eating "junk" foods, high in fat, salt and/ or sugar (Barton et al., 2012), high familiarity with these types of foods may represent another contributing factor (Benton, 2004).
If we consider these behaviours through the lens of a 'healthy eating' agenda, reducing a young person's intake of 'junk' foods is likely to have a positive impact on a young person's physical health. However, it must be recognised that making a shift from this type of food to 'healthier' alternatives might not be easy for a young person in care. Besides being addictive (Danese and Tan, 2014), if food is being used as a form of self-soothing, potentially as a substitute for the kind of soothing they would normally receive from a primary caregiver, than one cannot expect the child to simply give this up (Barton et al., 2012). Instead, staff need to explore beyond the behaviour and try to connect with the feelings and needs of the young person. Doing so might present an opportunity to help the young person find other ways of feeling soothed and subsequently become less reliant on using food for comfort (Barton et al., 2012).
The discussion so far has focused on food behaviours and practices that might contribute to weight gain. However, other practices in care, for example, eating meals as a 'family' group, may operate as a protective factor for a healthy diet. Research investigating the frequency of 'family' meals and subsequent dietary patterns has shown that eating dinner together is associated with improved dietary intake in adolescents (Stephens, McNaughton, Crawford, MacFarlane, & Ball, 2011;Leech et al., 2014;Walton et al., 2016). Possible explanations for this association include increased opportunities for caregivers to role model 'healthy' eating and reinforce intake of 'healthy' foods and reduced television screen time (Leech et al., 2014;Stephens et al., 2011). The potential importance of young people and carers eating together is strengthened by research which shows that eating dinner with family members is associated with lower prevalence of disordered eating behaviours, lower levels of substance abuse, and improved academic and wellbeing outcomes among adolescents (Walton et al., 2016). Other social benefits include, providing structure (which may reduce some anxiety around food) and opportunities for learning through modeling, as well as fostering a sociable culture within the unit (Barton et al., 2012;Holden, 2009). Despite these benefits, it was evident from discussions with staff and young people that daily routines of eating together are often difficult to establish and maintain in every residential unit. The examples provided by participants, highlight how work pressures and 'institutional' routines can often compromise staff efforts to develop an ethos of 'homely-ness'. Accordingly, strategies that bring people together around the table must consider the reality of the varied nature, organisation and experience of mealtimes within private homes/family settings/residential care contexts. Importantly, whilst staff should be encouraged to try and come together for meals, this must be done in a way that is sensitive to a child's needs, especially if they are anxious about eating in a group.
Another protective factor that might promote healthy eating is the involvement of the young people in meal preparation and other food-related activities. Evidence suggests this is associated with better dietary patterns in adolescents (Leech et al., 2014). Involvement affords an opportunity to learn new skills, engage in conversations around cooking and health and improved self-efficacy when it comes to selection and preparation of healthy foods (Leech et al., 2014). Despite noted benefits, some staff were hesitant to involve the young people: … I encourage him to be in the kitchen area but I don't encourage him to cook due to hygiene. [RCW12] Without diminishing the carers' concerns, failing to involve the young person suggests the dominance of the 'risk' agenda in OoHC, highlighting its impact on the preventative, reparative, and normative food practices so essential for future wellbeing.
In the context of this paper, food was originally thought about primarily in the domain of health -linked to children's rights to stay healthy and well, and have access to healthy food (Department of Human Services, 2007). However, these findings illustrate that it is also important to consider the experience of food in everyday life, particularly in the context of food providing nurturance and care. The work by Punch et al. (2009) acknowledges that in the interest of health, it is important that we aim to provide a balanced diet but it is also important for carers to be sensitive to the role of food in meeting a child's emotional needs. For young people in care, who might not have internalised these types of experiences, it is integral that staff seize the opportunity to meet such needs by ensuring food is provided in a thoughtful and caring way (Barton et al., 2012). Taking the time to consider a young person's eating habits and attitudes towards food can also deepen our understanding of them and provide important insights into their current emotional states and experiences (past or present) (Barton et al., 2012).

Conclusion: implications for practice, policy and research
By drawing on the model established in the FaCS (Emond et al., 2013;McIntosh, Dorrer, Punch, & Emond, 2011;Punch et al., 2010), the complex and nuanced ways in which food and the surrounding practices can impact on health outcomes for children in residential care in Australia have been illuminated. Whilst originally HEAL sought to establish a more robust understanding of the nutritional content of foods being offered in residential care, the emphasis on the symbolic power and potential of food, as developed by FaCS, allowed for a richer analysis and deeper understanding of how food was being used and experienced. What began as a simple correlation between care and obesity revealed a far more complex picture; one requiring more extensive and value-free investigation. The initial findings suggest that mainstream approaches to addressing weight are unlikely to be appropriate or successful in residential OoHC. A health agenda that is designed for a 'typical' adolescent population has the potential to ignore the very complex relationship that children may have with food e the memories that it holds, the sense of belonging and identity it might engender, or the trauma that it evokes. Instead, approaches must account for personal food biographies, behaviours and social backgrounds, and how these are implicated in the everyday practices and social interactions around food. This includes recognising the unique life experiences of children and the role that food may have played in that.
Moving forward, researchers and practitioners must recognise the central importance of understanding the context, and the individuals involved and how we can best support staff to reflect on and develop their practice in this area. One approach is to reflect on the social systems that can impact food provision -staff and young people highlighted a number of limitations relating to the procedural management of food in residential care. Staff often struggled to balance competing interests across child and staff needs, resources, and regulations e perhaps because they have not been afforded the opportunity to reflect on the very real tensions of balancing the needs, histories and care plans of a wide range of young people living together. Instead, staff need to given the space to reflect on the what, when, who and why of daily routines and functions around food. For staff, this includes thinking about the power that their own histories and ways of doing food can have in communicating care, and how this impacts on their approach to food at work. Where practice is determined by risk avoidance, it is important that it does not become so rigid that children's needs are not being met (Punch et al., 2009). Instead, staff need to be given flexibility and space to consider their own assumptions and practices around food. As Punch et al. (2009) point out, food can be a catalyst for understanding how care is provided and received by young people.
It is also important to consider the physical environment that surrounds the young people and how this might influence the food culture in the residential unit e for example, is the dining room setting arranged in a way that encourages desired interactions around food? Finally, it is also important to acknowledge the strategies that children have developed to cope with life and to keep themselves psychologically safe (Emond et al., 2016). Whilst residential staff need to actively promote a relationship with food that values the 'body' and ensures that it is nurtured and repaired, it must be done at a pace appropriate to the child's current capacity and with an understanding of the context surrounding food behaviours. Given food also presents a lens through which social connectedness can be explored in everyday experiences, the meanings they take on must also be considered in the development of a coherent strategy to address behaviours and overall functioning around food. Broadly, the results of this work indicate, for this population, the emotional aspects of food are just as important as the nutritional quality (Emond et al., 2013). It is essential that in future work both are recognised.
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Conflict of interest
There is no conflict of interest to be declared.