Malnutrition in older people

Malnutrition, defined as ‘a state of nutrition in which a deficiency, excess or imbalance of energy, protein or other nutrients, including minerals and vitamins, causes measurable adverse effects on a person's body function and clinical outcome’ (Royal College of Physicians (RCP) 2002) remains a source of concern for older people in the UK and other parts of the world (Beck and Ovesen Reference Beck and Ovesen2002; Margetts et al. Reference Margetts, Thompson, Eli and Jackson2003). Older people living in care homes are deemed to be particularly vulnerable (British Association of Enteral and Parenteral Nutrition (BAPEN) 2007; Cowan et al. Reference Cowan, Roberts, Fitzpatrick and While2004; Malnutrition Advisory Group 2000). Although malnutrition is not an inevitable consequence of ageing, the physiological changes that occur in older adults increases the risk (Hickson Reference Hickson2006). Malnutrition may lead to physical weakness and poor health (BAPEN 2003), and is related to increased mortality and morbidity rates (Lehmann Reference Lehmann1991); moreover older people are less likely to recover from malnutrition (European Nutrition for Health Alliance 2006).

Wales and UK policy context

Older people are recognised as a priority group in the nutritional strategy for Wales (Welsh Assembly Government (WAG) and Food Standards Agency Wales 2003) and deemed vulnerable to malnutrition. Malnutrition affects over 10 per cent of the UK population aged over 65 and throughout the UK costs the National Health Service more than £7 billion a year (BAPEN 2005). The Strategy for Older People in Wales (WAG 2008) recommended further action to address malnutrition in community settings, including care homes. The document Fundamentals of Care (WAG 2003) recommends that people who are unable to feed themselves adequately are quickly identified and that assistance is provided. The delivery of improved nutrition and food security for users and their carers is part of the government's ten-year social services strategy Fulfilled Lives, Supportive Communities (WAG 2007). The Welsh Consumer Council (2008) also recommends that ensuring people are well nourished should be a key element of the ‘Dignity in Care’ agenda for Wales. In the UK the need to detect, prevent and treat malnutrition in older people has been recognised (BAPEN 2007; Department of Health 2000, 2001; WAG 2003, 2006) and incorporated into national guidelines, for example, the National Institute for Health and Clinical Excellence (2006) guidelines for nutritional support in adults. However, there is little research into the impact of these guidelines on residents' experiences.

In the UK a number of ongoing studies to improve nutrition for older people in hospitals and other care settings are being conducted as part of the ‘New Dynamics of Ageing’ programme (www.newdynamics.group.shef.ac.uk). This eight-year programme is a multidisciplinary research initiative aimed at improving the quality of life of older people and completes in 2013. Studies include: the ‘Multidisciplinary approaches to develop prototype for the prevention of malnutrition in older people’ (MAPPMAL) project, which explores production and delivery of food to older people and the ‘Novel assessment of nutrition and ageing’ (NANA) project to help identify older people at risk of under-nourishment and improve targeting of interventions. Whilst these studies will hopefully yield insightful data with regard to identifying and addressing malnutrition in hospital and community settings, they do not address the issue of the provision of nutritional care and the mealtime experiences of older people living in care homes. Yet food and drink provide more than nutrition and hydration and the social context of nutrition in care homes is now explored.

The social context of nutrition in care homes

Sources from sociological and anthropological disciplines illustrate that ‘food is a cultural affair … a product and reflection of the norms and values of the society in question’ (Murcott Reference Murcott1982: 678), and that the sharing of meals constructs and communicates relationships between people (Crogan et al. Reference Crogan, Evans, Seversten and Shultz2004; Douglas Reference Douglas1975; Evans, Crogan and Schultz Reference Evans, Crogan and Schultz2005; Helman Reference Helman2000; Lupton Reference Lupton1996; Murcott Reference Murcott1982). Mealtimes provide temporal structure to the day for people in residential care homes (Gubrium Reference Gubrium1997; Nijs et al. Reference Nijs, de Graaf, Kok and van Staveren2006; Savishinsky Reference Savishinsky and Stafford2003) and are important ‘indicators of time, place, social interaction and “normality” ’ (Pearson, Fitzgerald and Nay Reference Pearson, Fitzgerald and Nay2003). Mealtimes and coffee breaks bring people together in care homes (Savishinsky Reference Savishinsky and Stafford2003) and are fundamentally social occasions, providing opportunities for conversation and companionship (Caroline Walker Trust 2004; CSCI 2006; Wikby and Fagerskiold Reference Wikby and Fagerskiold2004; Wright, Hickson and Frost Reference Wright, Hickson and Frost2006). In Gubrium's (Reference Gubrium1997) ethnographic study of a North American care home, he noted the centrality of mealtimes to residents' daily lives and that preparing themselves for and talking about mealtimes were major ways in which residents passed their time. Similarly, mealtimes have been described as the ‘highlight of the day’ for people in care homes (CSCI 2006).

The physical environment of a care home is particularly important to all aspects of residents' experiences, including their mealtime experiences (Department of Health 2003; Johnson, Rolph and Smith Reference Johnson, Rolph and Smith2010; Parker et al. Reference Parker, Barnes, McKee, Morgan, Torrington and Tregenza2004; Townsend Reference Townsend1962; Willcocks, Peace and Kellaher Reference Willcocks, Peace and Kellaher1987). The maintenance of homeliness in the care home environment has informed the rhetoric of care policy for some time (CPA 1996; Department of Health 2003; WAG 2004) and much of the literature. There is a potentially fruitful body of literature from the sub-discipline of environmental gerontology, which explores the ways in which spatial settings hold meaning and shape relations between older people (Wiles Reference Wiles2005). Additionally, Willcocks, Peace and Kellaher (Reference Willcocks, Peace and Kellaher1987) highlight the importance and meaning of ‘place’ in older people's lives, whilst also noting that the physical environment is just part of the larger whole of a person's experience in a care home. Spatial arrangements in dining areas and the organisation of meal delivery have also been shown to be influential, although there is a dearth of UK research in this area. A North American study by Remsburg et al. (Reference Remsburg, Luking, Barab, Luking, Barab, Radu, Pineda, Bennett and Tayback2001) reported ‘buffet-style’ dining, where food was served on a steam table allowing residents to choose from a selection of foods and choose second helpings of favourite foods, was effective in increasing sociability and enjoyment. In a Dutch study Nijs et al. (Reference Nijs, de Graaf, Kok and van Staveren2006) found that ‘family-style’ mealtimes, which included table dressing and seating organisation, were effective in maintaining quality of life, physical performance and body weight. However, Kofod and Birkemose's (Reference Kofod and Birkemose2004) small qualitative study conducted in Denmark of the influence of ‘stay-and-living’ environments, where residents eat in small groups, indicates the complex nature of residents' perceptions of mealtimes. In this study of 19 nursing home residents, although half of the group reported positive reactions, the other half found this environment embarrassing or difficult. Similarly, Sidenvall, Fjellstrom and Ek's (Reference Sidenvall, Fjellstrom and Ek1994) qualitative study of 18 patients in a Swedish nursing home indicated that patients with eating difficulties found it a strain to share tables to eat with other people and would have preferred to eat alone. Moreover, people without eating difficulties were sometimes disgusted by ‘inappropriate behaviour’ from other patients. Whilst these studies are useful there may be issues of transferability from the Nordic, Dutch and North American context to the UK situation and all were conducted in nursing homes.

A situated study – location and access

Ethical approval was gained from the University Ethics Committee and approval and permissions were gained from appropriate gatekeepers within the Social Services Department in the study area to conduct the study.

The care homes were owned and managed by the same local authority and registered to provide personal care only. Both care homes also had ‘day centres’ attached, which provided services for older people who need support but are able to continue living at home. In the local authority under study, there were seven care homes for older people requiring personal care, each with between 24 and 48 beds, of which five were based on communal living and two were unit-based, that is the homes comprised six separate eight-bedded units. We purposively selected one home of each type.

Home One (H1) was a communal living care home and had 29 single rooms spread over two floors with communal bathrooms, toilets and lounge facilities. There were communal dining rooms on each floor, with residents usually using the dining room on the same floor as their bedrooms. Meals were provided to each communal dining room from a large central kitchen.

Home Two (H2) was a unit-based care home and had 48 single rooms which were divided into six separate units spread over two floors. Each unit had eight single bedrooms, a lounge and dining area, a kitchenette with cooking facilities, communal bathroom and toilets. A large central kitchen provided all of the units with main meals. There were also communal lounge facilities and other areas for shared use on both floors, including an internet room.

In terms of geography, both homes were similarly situated in traditional working-class urban areas on the outskirts of the same city. Both areas had developed in the mid to late 19th century to accomodate local workers, mainly skilled and semi-skilled manual workers in the coal mining and steel industries, and much of the terraced housing remains. However, H1's location is described by the local authority as a ‘suburb’, two miles from the city centre, whereas H2's situation is described as a ‘village’, five miles from the city centre. According to local authority data, the population of people aged 65 and over living in the localities in which the homes are situated was lower than that for Wales (15.7% for H1, 17.7% for H2 compared with 18.4 and 16.4% for Wales and the UK, respectively) (Office for National Statistics 2012). The proportion of people aged 75 and over in the local population surrounding H1 was higher (8.3% compared with 7.1% for H2 and 7.9% for Wales, respectively). The homes also differed in terms of where their residents came from inasmuch as H1 tended to admit people from further afield, whereas many (although not all) H2's residents' original homes were quite close and part of the community where the home was situated. Consequently, many H2 residents maintained links with people and places (such as the chapel and the public house) in the surrounding locality; they were also more likely to know people who came into the attached day centre. Hence, the community inside H2 residential home appeared to be more anchored into the surrounding community than that of H1.

There were also important differences between the two homes in terms of access to their attached day centres. H1's access was through a separate outside entrance; there was also a walkway from the home to the day centre but visitors from outside the home would use the separate entrance. However, access to H2's day centre was through a secure door from the central foyer of the home, drawing outside visitors first into the main body of the home. Thus H2 had less of a structural boundary between itself and the outside world.

Sampling of participants

Participants were recruited following a presentation given by two of the researchers at each home and information letters to residents, staff and informal carers. Individuals were invited to indicate through a reply slip if they wished to participate. A total of 15 catering (including cooks and kitchen assistants) and care home staff from both care homes (eight from one and seven from the other) were purposively selected (according to criteria set out below) from the reply slips. The two care home managers and two assistant managers (with responsibility for managing catering services) from each care home were also invited and agreed to be interviewed. The criteria for selection of staff (both managers and care home staff) were that they were willing to take part in the study, had worked in the care home for at least six months and were involved with meeting the nutritional needs of residents as part of their role. All the staff who met these criteria participated.

Eight residents from each care home (28.5% from H1 and 16.6% from H2) who had spent at least one month in the care home were purposively selected from the reply slips, giving a sample size of 16 residents. The inclusion criteria for residents were that they were over 65 years of age, taking food orally, had the ability to communicate and able to give consent. Those unable to give consent included residents with moderate to severe cognitive impairment. The residents' sample comprised three men and 13 women (ratio of men to women in H1 was 8:29 and in H2 was 3:48) and their time in the home ranged from 12 months to seven years. The average age of the residents' sample was 80 years in H1 and 93 years in H2 (average age in H1 was 84 years and H2 was 87 years, respectively). A quarter of the residents interviewed had special diets, mainly diabetic diets (27.5% in H1 and 10% in H2 had special diets, mainly diabetic and low-fat diets). The functional ability of residents interviewed ranged from those who were mainly self-caring to those who used wheelchairs or walking aids and had chronic medical conditions including diabetes, arthritis and a neurological disorder and who were dependent for their care needs. All were mentally capable of decision making. The functional ability of residents in the homes varied as H2 had two units which were specifically for residents with moderate to severe dementia and these units were excluded from the study.

The functional ability of residents is presented in Table 1, which indicates that over half of the residents in H1 were physically and mentally active and mainly self-caring compared with 11% in H2. Another important difference was that in H2, 32% of residents used a wheelchair and had no capacity to support themselves and had increasing dementia, whilst a further 8% were totally dependent for their care needs and were in the later stages of dementia. Most but not all of these residents resided in the two units which were excluded from the study.

Table 1. Functional ability of residents in Home One (H1) and Home Two (H2)

Five informal carers of residents from each care home were purposively selected, giving a sample of ten informal carers. The inclusion criteria for informal carers were that they were willing to take part in the study, had a relative who had spent at least one month in the care home and were able to give consent. The informal carers' sample comprised two men and eight women; eight daughters, one son and one husband. No participants who wished to participate were excluded in that all those that responded met the above criteria.

Methods of data generation

Data were generated through a range of methods, which included focus groups with care home and catering staff and individual semi-structured interviews with managers, residents and informal carers. Additionally, observation of food preparation and mealtimes throughout the day and analysis of appropriate documents such as assessment tools and residents' case notes were conducted. The issues explored through these methods were informed by reference to the National Minimum Standards for Care Homes for Older People (Department of Health 2003; WAG 2004) and nutritional guidelines and recommendations for care homes (e.g. BAPEN 2007; National Institute for Health and Clinical Excellence 2006).

Focus groups

A focus group, using a topic guide, was conducted with a sample of catering and care home staff (seven in H1 and eight in H2) in each home by two of the researchers. In total two focus groups (N = 15) were conducted. The focus groups were useful in uncovering shared views and understandings of particular issues within the groups of staff. The topic guide derived from the literature on nutritional guidelines and standards for care homes (BAPEN 2007; National Institute for Health and Clinical Excellence 2006; WAG 2004) included issues such as the use of nutritional assessment tools, assistance provided with eating and drinking, and their knowledge regarding older people's nutritional needs.

Interviews

Semi-structured interviews, using an interview schedule, were conducted with managers in a private area in each care home. Such interviews allowed flexibility in the way questions were asked; whilst the use of an interview schedule ensured the research questions would be addressed (Green and Thorogood Reference Green and Thorogood2009). The interview schedule was also derived from the literature and explored issues such as procurement of food and ensuring food quality, staff education on nutrition, and use of nutritional screening and assessment tools to identify residents at nutritional risk.

Additionally, semi-structured interviews, using an interview schedule, were conducted with residents (N = 16) and residents' informal carers (N = 10) in residents' own rooms or in a private area in the care homes; this helped to ensure anonymity and the familiar surroundings facilitated rapport between the interviewer and participants. Issues explored with both residents and informal carers included choice, quality and quantity of food, and in addition informal carers were asked whether or not they supplemented their relatives' meals. With participants' permission the focus groups and all interviews were audio-taped.

Observation

All observation of food preparation and mealtimes throughout the day was undertaken by two of the five members of the research team; this included informal conversations with staff. This involved observation of food preparation in the kitchens and observation of all mealtimes during the week (breakfast, lunch and dinner, mid-morning and mid-afternoon refreshments) in each care home; where it was possible and appropriate researchers would sit down and share residents' meals with them. Observation was conducted only during weekdays and the researchers adopted the role of ‘observer as participant’ (Gold Reference Gold1958), interacting with the residents and staff but not adopting any specific role or being involved in any specific tasks. One researcher observed food preparation in the kitchens of each home from 8.45 am to 12.00 noon (6.5 hours of observation in total). During this time period, observations included food delivery, storage and preparation. Observation of mealtimes was organised so that one researcher observed the serving of breakfasts in both homes (three hours of observation in total), one researcher observed the serving of lunches in both homes (three hours in total) and these same researchers observed the serving of an evening meal in each home (two hours in total). Observations also included mid-morning and mid-afternoon refreshments and attending residents' meetings in the care home, such as menu-planning meetings (1.5 hours in total). Sixteen hours of observation was conducted in total and recorded as field notes. Finally, a purposive sample of documents pertaining to nutrition in both care homes, such as policies, residents' care plans, assessment forms and menus, were reviewed and analysed with permission from residents where appropriate. These documents were considered in terms of their substantive content and also their context, that is the ways in which they were produced and how much attention was paid to them by those who used them.

The total sample size was 45 participants, which included 19 care home staff including managers, 16 residents and 10 informal carers; this sample size was considered adequate and appropriate to address the research aim and objectives (Sandelowski Reference Sandelowski1995).

Ethical considerations

The researchers provided a presentation about the study's purpose to all care home staff (catering, managers and care staff) and residents in each home. A written information sheet was provided to all those who attended the presentations and information packs were left with the managers to distribute to care home staff who were unable to attend the presentation and to informal carers when they visited their relatives. A poster about the study was also displayed prominently in each care home. All participants were given a written information sheet and the opportunity to ask questions about the study prior to giving written consent. Written permission to conduct observation of food preparation and mealtimes was obtained from the care home managers. All data presented have been anonymised.

Data handling and analysis

Data from the focus groups and individual interviews were transcribed verbatim by those who had conducted the interviews in order to ensure accuracy. Observation of food preparation and mealtimes were recorded in the form of field notes. The data generated were read by four members of the research team and analysed using thematic analysis. The aim of the thematic analysis was to identify key patterns and themes in the data using a process of coding, developing categories from clusters of coded data and then generating themes from these categories (Bowling Reference Bowling2002). On a practical level, the initial coding was done using a highlighter pen accompanied by notes in the margins as the data transcriptions were read and re-read. Initial codes were grouped into categories and finally into themes; then the data were returned to in order to check that all the initial codes were encompassed into these themes. Analytical rigour was strengthened through the use of an iterative process, involving all four researchers moving between the data and the agreed coding framework (Barbour Reference Barbour2008).

Similarly, documents were analysed thematically in terms of their content and the context of their production (Prior Reference Prior2003). Through making explicit the details and rationale of our key decisions throughout the study, we have provided an ‘audit trail’ (Guba and Lincoln Reference Guba and Lincoln1989) which may be followed by readers and enhances the rigour of the study.

In discussing findings, excerpts use the abbreviations ‘I’ for interviewer, ‘P’ for participants (care home staff), ‘M’ for care home managers, ‘R’ for residents and ‘C’ for informal carers. Notes taken during the periods of observation are reported as field notes.

The physical environment

Mealtimes provided in care homes are organised according to a routine which has been viewed as structuring the day for residents (Nijs et al. Reference Nijs, de Graaf, Kok and van Staveren2006) and provide important opportunities for social interaction and companionship (Wright, Hickson and Frost Reference Wright, Hickson and Frost2006). In line with government policy, both homes provided a menu on a four-week rotation, offering a choice of meals, which was available for residents to read or have read to them prior to the meal to allow them to make a choice. Each home provided a summer and a winter menu, and residents were included in menu design and could raise any concerns about the menu at the monthly residents' meeting. Whilst both homes sought to provide personalised care and offered choice in respect of food options, where meals were consumed and to some extent when, especially with respect to breakfast, mealtimes were usually dictated by the home's routine. Mealtimes in each home were served around set times which took into account staff break times. Lunch was served between 12.45 and 1.00 pm and the evening meal between 5 and 5.30 pm in each home. In H1 breakfast included a cooked breakfast and was served from 9 am, whilst in H2 the timing of breakfast was more flexible and some residents were able to serve themselves breakfast of, for example, cereals. Whilst residents in H2 could have a cooked breakfast, cooking of hot food was rarely undertaken on the units due to health and safety concerns, so this was cooked in the main kitchen and brought up separately, so few residents opted for this choice. The last hot drink of the day with a snack, if requested, was served at 8 pm in each home. Snacks were available at any time, if requested by the residents, and fresh fruit was always readily available in both homes.

Regarding the spatial environment for mealtimes, residents in both homes took their meals in the dining rooms where the seating arrangements were broadly similar, as can be seen from the floor plans (Figures 1–3).

Figure 1. Dining area, Home Two.

Figure 2. Downstairs dining room, Home One.

Figure 3. Upstairs lounge/dining room, Home One.

The dining rooms in both locations represented a significant community meeting point for residents, some of whom may not have seen much of each other during the rest of the day. Residents who used wheelchairs were sometimes brought to the dining room first but sometimes arrived in their own time; there were no strict rules about it. They also stopped to socialise on their way to (and through) the dining room.

In both homes the positioning of chairs and tables was described by staff as ‘family style’ inasmuch as residents ate at small, four-seater, circular dining tables, pleasantly laid with linen cloths and napkins, pretty china and flowers. However, as can be seen from Figures 2 and 3, H1 had two larger communal dining areas with three tables upstairs and six downstairs; whereas as Figure 1 indicates, H2, was organised into six separate eight-bedded units, each with its own smaller dining area containing just two tables, which undoubtedly enhanced the ‘homeliness' of its dining areas. In addition, as can be seen from Figure 1, H2's small kitchenettes adjacent to each dining area, with cupboards, a refrigerator, a sink, a toaster, microwave, dishwasher and a small hob, engendered the sense of a ‘normal’ kitchen-dining room. H1 had a similar set-up in its upstairs dining room, although there was no hob or sink; the downstairs dining room though was much larger, in more of a ‘canteen style’, and did not need a kitchenette as the main kitchen led off this room. The main kitchens in both homes housed large stoves, refrigerators, dishwashers and storage cupboards. In H2, whilst there was no large communal dining area, when whole home events were organised, for example at Christmas, these were held either in the attached day centre or in the large foyer area on the ground floor.

The spatial arrangements also contributed to important differences in the manner in which food was served to residents in both homes. In H1's upstairs dining room, due to its distance from the main kitchen, food was served from a hot trolley into tureens from which residents served themselves at the table; whereas in the downstairs dining room, which was adjacent to the main kitchen, food was plated up by staff directly from the hot counter. Residents were served their meals by the staff and did not collect their meals directly from the hot counter as per restaurant-style dining, to avoid any potential safety risk. In H2, where the small unit-based dining rooms were some distance from the kitchen, staff plated up the meals from the hot trolley. Whilst staff did check with residents regarding their food choices, as staff were familiar with residents' food preferences, developed through their long-standing relationships with residents, there was a tendency to make assumptions about what residents liked and disliked. The implications of this in terms of paternalism or beneficence are explored in the section below titled ‘Food Choices and Identity’ and is discussed further elsewhere (Merrell et al. Reference Merrell, Philpin, Warring, Hobby and Gregory2012). With respect to portion sizes, portions were observed to be generous and there was always sufficient food. Indeed as one of the residents commented:

Having the downstairs dining room in H1 adjacent to the main kitchen meant that residents in this dining room were in close proximity to the smells of cooking; this was generally perceived as a good thing, adding to sensuous anticipation, stimulating the appetite and providing links to earlier memories. Whereas the other dining areas were quite distanced from the cooking areas and their sensory stimuli, which according to the code of good practice for residential and nursing care homes, can be viewed by some as a deprivation (CPA 1996).

There are different ideas as to what constitutes homeliness and concepts of community, as suggested by the following comment from a staff member in H1 who argued that the larger dining setting created a greater sense of community in that it provided opportunities to mix with a broader range of residents:

This view supports the findings by Willcocks, Peace and Kellaher (Reference Willcocks, Peace and Kellaher1987: 131) who report that unit or group living can be socially isolating. A further socially constraining factor inherent in the homeliness of the smaller units was that their size also led to congestion, especially if each table had a number of wheelchairs in position. This discouraged other people, such as friends and relatives or staff, from joining them for meals. Sharing meals with friends and relatives is one of the ways in which they may be encouraged to take part in the daily life of the care home (Brown Wilson, Davies and Nolan Reference Brown Wilson, Davies and Nolan2009; CPA 1996), enhancing its sense of community. This rarely happened in either of the care homes apart from special occasions, when relatives could share the buffet, which would be served in a larger communal area.

Choice of seating and social relationships

Residents would almost invariably sit in the same chair and hence with same people for each meal and in most cases had done this for many years. However, the ‘choice’ was constrained by which seats were available upon their admission to the care home. Sitting in the same seat was encouraged by the staff and appeared to be conducive to conversation, support and companionship which others have reported (Pearson, Fitzgerald and Nay Reference Pearson, Fitzgerald and Nay2003; Sidenvall, Fjellstrom and Ek Reference Sidenvall, Fjellstrom and Ek1994), as exemplified in this interview excerpt with a resident:

There were apparent social protocols and conventions for residents to negotiate in terms of joining dining tables where there were spaces; the following excerpt from a female resident suggests a certain amount of choice for both the would be joiner and the incumbent. The reference to the need to create a ‘family’ is also significant and will be returned to later:

Our observations also indicated a sense of sociability, companionship and community:

However, sociability and companionship are greatly influenced by the nature of one's companions (Sidenvall, Fjellstrom and Ek Reference Sidenvall, Fjellstrom and Ek1994) and not all residents perceived their mealtimes as conversational opportunities. Some reported very limited conversation on their table as their dining companions were very quiet, hard of hearing or simply had nothing to talk about, as exemplified by these residents in H2:

The desultory nature of conversation between residents, chiefly revolving around their food and the weather, was also apparent in some of our observations in both homes; however, mealtimes were still essentially ‘social’ occasions and residents made efforts to ensure that those with impaired hearing were included in the conversations. There was evidence of reciprocity and mutual support in the social relationships between residents as they helped each other by passing condiments, pouring drinks for each other and ensuring that everyone had what they needed; thus encouraging a sense of belonging (Godfrey, Townsend and Denby Reference Godfrey, Townsend and Denby2004).

The following excerpt with a resident suggests that sitting on the same table with the same people every day may not necessarily be conducive to meaningful conversation or sociability:

Environmental influences on mealtimes were not entirely structured by the spatial confines of the dining rooms as residents could choose to eat alone in their rooms if they wished. This sometimes happened if residents were unwell or particularly tired, but was not viewed as desirable in the long term by residents, their family or staff. It was apparent that sociability and companionship were viewed as being very important. Some residents evidently thought that it was ‘easier’ for the staff if they went to the dining room, as indicated in this account from one of the oldest female residents; this excerpt also illustrates reciprocity between residents and staff and the ways in which mealtimes also provided mutual support:

A staff member in H2 explained that sometimes residents did prefer to eat alone, suggesting possible reasons for this, which echoes the findings from Sidenvall, Fjellstrom and Ek's (Reference Sidenvall, Fjellstrom and Ek1994) study:

Drawing on the data primarily from the interviews with residents and their informal carers, some residents ate their meals outside the care homes, which is now considered.

Eating and socialising in other environments

Residents were able to move from their dining rooms entirely and take meals in the attached day centre, providing opportunities to meet people (some of whom they knew prior to their admission) from outside their home and to maintain links with the outside community. The design and layout of H2 with its easy access to the day centre directly from the care home facilitated the promotion of links with the outside community which were impeded by the design of H1, which despite having a walkway to the day centre, staff and visitors used the separate entrance. As neither day centre had separate cooking facilities the meals were cooked in the main kitchens of the care homes and transferred to the day centres in hot trolleys. Whilst residents from both homes rarely consumed main meals in the day centre, principally because the day centres were full to capacity, residents from H2, but not from H1, would attend the day centre prior to or after lunch for morning coffee or afternoon tea and cake. This was especially the case for residents who were staying in the home for respite care and for those who were more independent, although H2 care home staff would accompany residents to the day centre on request. The design and physical environment of the homes, as evident from the floor plans (see Figures 1–3), impacted on the way in which food was served and the choices residents had with respect to where and with whom they ate their meals. Our findings reinforce those of Willcocks, Peace and Kellaher (Reference Willcocks, Peace and Kellaher1987: 115) that there is a reciprocal relationship between institutional design and institutional organisation. Moreover, some of the residents from H2 moved even further afield and took meals and refreshments with friends and family outside the care home and some residents attended various clubs outside the home where meals and wider social interaction were provided; some residents from this home also went shopping for food outside the home. The ability to take part in these outside activities depended on residents being physically able and also having family and friends outside the home, which indicates the heterogeneity of the residents and hence of their experiences. Residents in H2 were more likely to have resided within the community in which the home was located which engendered shared identities, whereas the residents in H1 came from a wider and more dispersed geographical area. The geographical location of the home also impacted on the ability to sustain relationships and integrate with the local community so that the community inside H2 appeared to be more ‘of’ the local community as opposed to that of H1 which was primarily just located ‘in’ the community. The lack of organic links and community integration of residential homes has been highlighted as an issue for some time (Willcocks, Peace and Kellaher Reference Willcocks, Peace and Kellaher1987: 2).

The concept of community encapsulates ideas of both locality and social relationships and is notoriously difficult to define; although contested and value-laden, the term is generally viewed positively (Titmuss Reference Titmuss1968; Williams Reference Williams1984). A key element of community relates to shared identity which, C. A. Davies argues, it ‘both promotes and depends on; this aspect of community is constructed around cultural symbols and is closely tied to people's identities and sense of belonging’ (2003: 3). Shared identities for H2 residents was enhanced by the fact that in many cases they knew each other from their previous lives, in some case going back to school days, and there were siblings and spouses in this home. Many of the staff in both homes were also recruited locally. In relation to care homes, S. Davies (Reference Davies, Davies and Jones2003) argues that both the residents and staff benefit from a sense of community due to its potential to enrich residents' relationships and quality of life and to improve staff morale and job satisfaction.

This section has indicated that various aspects of the care homes' physical environments influenced residents' sense of sociability and community and hence their dining experiences. The difference in location and spatial arrangements between the two care homes has also been shown to influence social interactions and mealtime experiences. However, it is also clear that the physical environment is intrinsically bound up with socio-cultural elements; it is to these that we now turn.

Socio-cultural aspects – shared meanings and memories of food

People's understanding of food and its meanings is grounded in their socio-cultural background and we identified a web of shared meanings and memories of food amongst residents, informal carers and staff. It was noticeable that traditional food was viewed positively by all groups, for example, one of the cooks described himself as an ‘old-fashioned cook – tripe and onions, faggots – that's me’ (Field notes, H1). All the women residents and one of the men looked back warmly on their own cooking experiences prior to admission, recalling tastes and smells and preparing meals for their families. There were mouth-watering descriptions of cooking traditional Welsh food, joyfully recounted, rekindling memories of their earlier selves and confirming cultural identity as exemplified below:

Many residents described their enjoyment of making a ‘cooked dinner’; indeed residents, staff and informal carers all held in high regard and had a particular understanding of, what constituted a cooked dinner, often described as a ‘proper meal’. A cooked dinner comprised roast meat, accompanied by potatoes and other vegetables and gravy (although pies could also be substituted for the roast meat) and was served in the middle of the day. The cooked dinner was seen as the appropriate meal for residents and was referred to approvingly by residents, informal carers and staff alike. For example, an informal carer referred to the lunchtime meal as a ‘proper lunch … a cooked dinner and there's a pudding’ (C4, H1).

Staff saw it as important to encourage residents to eat a ‘nice hot, cooked dinner’. In the unusual case of one resident who was unable to eat the proper dinner, the staff liquidised it:

The importance of the ‘cooked dinner’ in this area resonates with Murcott's (Reference Murcott1982) study of women in South Wales who regarded the cooked dinner, which in their case consisted of meat, potatoes, at least one other vegetable and, importantly gravy as the ‘ “proper meal” par excellence’ (Murcott Reference Murcott1982: 677). Murcott refers to the ‘major symbolic significance’ of the cooked dinner and also notes that the women in her study had no concerns for their families' health ‘as long as they were getting their dinners’ (1982: 678). It appears that this time-honoured meal provides significant links for residents, staff and informal carers to their collective memories of family mealtimes, reinforcing their own cultural identity; moreover, the traditional foods and customs are important in maintaining social cohesion (Evans, Crogan and Schultz Reference Evans, Crogan and Schultz2005).

This connection by the women in Murcott's study of the cooked dinner with the maintenance of health suggests the social construction of ideas as to what constitutes a healthy diet. Contradictions between the pleasures of taste and what is perceived as healthy shaped residents' eating experiences. Savishinsky (Reference Savishinsky and Stafford2003) notes that eating may be one of the few sensory experiences left to many people and the tastiness of food and the memories it evoked was frequently mentioned by residents. This contradiction is illustrated in the following resident's account in relation to chips:

The staff had particular understandings, shaped by their socio-cultural backgrounds and political rhetoric, of what constituted a healthy diet, the cooked dinner was frequently referred to, as were fresh meat and vegetables; staff and informal carers used the word ‘fresh’ as a positive indicator of nutritional goodness (Lupton Reference Lupton1996):

Moreover, the (UK) health promotion mantra of five portions of fruit and/or vegetables a day was frequently referred to, as was the avoidance of salt:

The staff's perceptions of the meaning of healthy food were not gleaned from formal training although they had completed mandatory health and safety courses, which were concerned with food hygiene, storage, cleanliness and cooking temperatures for meat. Health and safety policies were strictly adhered to. There was also an expectation that adults would know about food and nutrition from their own personal repertory of skills:

People's own experience, the media and ‘common sense’ were frequently referred to as sources of knowledge about food. Despite this assumption that staff would know what constituted a healthy diet through common sense, instinct or general knowledge, there was generally a lack of awareness of current nutritional advice and recommendations for older people, and evidence of confusion in relation to saturated fats as exemplified in the following comments from staff:

Although there were undoubted gaps in the staff's knowledge of nutrition, which could be rectified with appropriate education, our study highlights the ambiguities inherent in providing nourishment, in the full sense of the word, to older people. Arguably, if food provides pleasure, comfort and links with cultural roots which symbolise identity for residents, it could, in one sense be considered to be ‘healthy’ (cf. Evans, Crogan and Schultz Reference Evans, Crogan and Schultz2005).

The symbolic meanings inherent in taking food and drink together are brought to the fore in their use to celebrate or commemorate significant events; in both homes residents would come together into a large communal area (not their usual dining rooms) to mark these occasions. Buffet meals were also provided for residents and relatives in communal areas to mark the occasion of a resident's funeral. The customary Christmas fare would be served over the Christmas period and traditional Welsh dishes such as ‘cawl’ (a stew/broth of lamb and vegetables) and Welsh cakes would be served to celebrate St David's (the Welsh Patron Saint) Day. These celebrations were enjoyed by residents, staff and informal carers; the latter group commented on how nice the food was on these occasions. Residents were also taken out for celebratory meals at Christmas time and their birthdays were celebrated with specially made cakes and sometimes glasses of wine.

In addition to marking traditional ceremonies by eating foods, residents were encouraged to get involved in their preparation, again as a communal activity. For example, at Christmas residents were encouraged to participate in the preparation of mince pies:

These familiar rituals are symbolic of community, links to earlier memories and important in maintaining people's identity, which it is fruitful to explore further.

Families, food and identity

Lupton (Reference Lupton1996) notes the ways in which the social dimensions of eating and those of emotion are linked through the context of the family. For most of the residents, admission to the care homes entailed separation from their families, although there were instances of spouses and siblings living in the same homes and sharing the same dining tables. Despite this separation, the influence of families permeated residents' and staff's perceptions of food and mealtimes in various ways.

First, the use of ‘family-style’ dining with small tables as described above was viewed very positively by the staff in that it was perceived as engendering a more homely atmosphere, which was regarded as beneficial to residents' dining experiences (cf. Crogan et al. Reference Crogan, Evans, Seversten and Shultz2004; Evans, Crogan and Schultz Reference Evans, Crogan and Schultz2005; Nijs et al. Reference Nijs, de Graaf, Kok and van Staveren2006). Providing homeliness has been part of the rhetoric of care policy in Wales and the UK for some time (CPA 1996; WAG 2004), and it was also valued by residents and informal carers in both homes. Dining was embedded in familiar home and family rituals, such as coming to the small tables, helping each other with drinks and condiments, recapitulating their personal memories of home and family, and reconnecting residents with their identities as individuals (Evans, Crogan and Schultz Reference Evans, Crogan and Schultz2005)

Second, for residents and staff, the meaning of food and mealtimes was grounded in their own family experiences. The residents' accounts of their enjoyment of cooking and sharing meals with their families prior to admission indicate its importance to their sense of identity and continuity; whilst care staff, who tended to come from a similar socio-cultural background as the residents, also drew on their own family experiences for their understanding of cooking:

Family and homeliness are also linked to ideas of normality and again staff set great store by achieving a semblance of normality for residents. The social interactions surrounding food and mealtimes were also demonstrative of ‘normal’ behaviour and residents took pleasure in sitting together for meals. Similarly, their reminiscences of earlier cooking for their families were also accounts of ordinary, everyday activities, connecting people back to their normal lives and selves. The staff in H2 recognised cooking as a normal activity which residents would have done in their own homes and, when possible encouraged this activity, which was also facilitated by the spatial arrangements – the unit-based kitchens – in this home. It was evident that the provision of the kitchenettes within the units and group living encouraged residents to participate more in mealtimes, enabled the storage of residents' own ingredients and provided opportunities to participate in cooking. Maintaining residents' involvement in food preparation is also recommended by the National Minimum Standards for Care Homes (Department of Health 2003). Experiences of normality, reciprocity and pleasure bound up with cooking are evident in the following quotation:

Sometimes staff members in H2 encouraged residents to join them in cooking, usually cakes, which would then be shared at tea-time. Our observations captured happy scenes of busy, chattering people, doing what was very familiar and normal to them, with tempting smells, evoking family memories and links to earlier roles, emanating from the ovens.

Although staff sometimes cooked with residents in H2, they did not take their own meal breaks with residents in either home, which arguably would enhance ‘normal’ social interactions between staff and residents, creating feelings of reciprocity and community (Davies Reference Davies2001). Sharing of meals with residents appears to raise a number of contradictory issues; the key prohibiting factor raised by the staff was that sharing mealtimes would entail staff losing their entitled break (from their work) and they would consequently need additional time off as well as their mealtime. Clearly, despite the rhetoric about mealtimes being framed into categories of ‘homeliness, normality and family’, they are categorised by the staff as ‘work’. Separation of residents and staff at mealtimes is a way of demarcating residents from staff and reinforcing status differentials, as Peace and Holland (Reference Peace and Holland2001) identified in their study of UK residential care homes registered to accommodate four or fewer residents. They found that even within these small care homes which were mainly adapted large domestic houses, activities and the division of domestic labour resembled that of a larger institutionalised residential care home setting rather than that of a ‘normal’ domestic home.

Informal carers also did not share residents' meals, which contradicts the above rhetoric; although they were often offered tea or coffee when afternoon snacks were being brought around for residents and were able to share in this refreshment activity. The importance of encouraging residents to invite guests in for either a meal or for just tea/coffee and biscuits to maintain relationships is highlighted by the Caroline Walker Trust (2004). Some informal carers commented that they felt it was inappropriate for them to ‘intrude’ on communal mealtimes, which suggests a different understanding again of the meaning of mealtimes. Informal carers' discomfort in sharing communal meals with their relatives may stem from the fact that this was not the norm in these homes, perhaps if this practice was established, informal carers would have less qualms about it. Best practice as evidenced by a review of international research evidence (NCHR&DF 2007) emphasises the need to promote a relationship-centred approach to creating community and developing a positive culture in care homes. Maintaining residents' identity through adopting a person-centred approach is encompassed within the eight best practice themes of the My Home Life programme developed by the UK Charity Help the Aged for improving the quality of care for all people (residents, relatives and staff) in care homes in the UK (www.myhomelifemovement.org). Opportunities for enhancing the involvement of residents' family and friends in the care home, through actively encouraging the sharing of meals, promotes a sense of community and contributes towards personalising residents' lives.

Food choices and identity

The ability to make choices relates to both psychological and physical wellbeing; the greater the perceived level of control, the less vulnerable the person feels (Crogan et al. Reference Crogan, Evans, Seversten and Shultz2004; Reker Reference Reker1997; Westerhof et al. Reference Westerhof, Bohlmeijer, van Beljouw and Pot2010), and is central to individual identity (Caplan Reference Caplan1997; Department of Health 2003; WAG 2004). As previously stated, both homes offered a choice of meals and a menu was available for residents to read or have read to them prior to the meal to allow them to make a choice. The following excerpt indicates flexibility and understanding in the event of memory problems:

However, the notion of choice is nuanced through various elements of institutional care. First, the very nature of communal living necessarily restricts choice for the individual, as exemplified by staff in the following excerpt:

Resident choice was further complicated by staff's understandings of what was appropriate for them and ‘knowing what they liked’; understandings that were grounded in their own socio-cultural background, as illustrated in the following excerpt concerning menu-planning, which significantly does not include residents:

Such paternalism indicates restricted choice for residents and suggests an imbalance of power (Phillips Reference Phillips2007). However, some residents expressed pleasure and approval that the staff knew them and their likes and dislikes; it represented ‘knowing’ them as people and enhanced their sense of belonging. As one resident commented:

Residents' choice is also linked to the previously discussed tension between what staff saw as their responsibility for providing what they understood as ‘healthy’ food and attempts to ensure resident autonomy and choice:

For the staff, the complexity of providing a healthy yet pleasurable diet for the residents in their care appears to be summed up in the following excerpt:

This indicates the staff's identification with the residents' position but also a pragmatic approach to issues of health and choice.