Beliefs about inevitable decline among home-living older adults at risk of malnutrition:

Background 5 About 14% of free-living adults aged 65 and over are at risk of malnutrition. Malnutrition 6 screen and treat interventions in primary care are few, show mixed results and advice given is 7 not always accepted and followed. We need to better understand the experiences and contexts 8 of older adults in order to develop interventions that are engaging, optimally persuasive and 9 relevant. Using the Person-based Approach, we carried out 23 semi-structured interviews with 12 purposively selected adults aged 65 and over with chronic health or social conditions 13 associated with malnutrition risk. Thematic analysis informed the development of key 14 principles to guide planned intervention development.


Introduction
Malnutrition in older adulthood is a global issue, though contextual differences between 34 countries will impact on how malnutrition can best be addressed in each country. In the UK, 35 1.3 million (11%) of adults over 65 are believed to be malnourished, rising to 18% of those 36 receiving home or day care [1][2][3] . Global Leadership Initiative on Malnutrition (GLIM) 37 diagnostic criteria for malnutrition include: non-volitional weight-loss, low body mass or 38 muscle strength; plus reduced food intake or assimilation, disease burden or inflammation 4 . 39 Malnutrition risk, measured by e.g. MUST 5 or MNA 6 , is associated with frailty, sarcopenia, 40 falls 7-9 , GP consultations, hospitalisation 10 and reduced quality of life 11 . Malnutrition among 41 older adults in the UK was associated with excess costs of £10 billion in 2011-12, mostly for 42 institutional care or hospitalisation, so early identification of risk, and treatment for free-43 living adults might produce significant savings 3 . Screening and treating malnutrition risk in 44 primary care may also improve patients' health and quality of life 12 13 , but it is unclear how 45 best to do this, or how to engage older adults who may not consider themselves to be 'at risk'. 46 Additionally, consensus is lacking about which malnutrition risk factors can be usefully 47 targeted. More than 120 potential causes of malnutrition have been identified, which 48 individually may be unrelated to malnutrition risk 14 , but which interact to increase risk, 49 though the mechanisms are little understood 15 . Nevertheless, deteriorating health, widowhood 50 and retirement can influence changes in food choices and ways of acquiring and preparing 51 food [16][17][18][19] . Changes in such habits can lead to a deterioration in diet quality and quantity 52 accompanied by reduced personal control, exclusion at social events and changed roles and 53 responsibilities [16][17][18][19] . A range of physical and psychosocial factors can undermine motivation 54 to improve eating habits 20 by promoting unhelpful beliefs and fears. A mixed-methods review 55 identified that patients had reservations about screening and discussing diet 21 ; and difficulty 56 chewing, swallowing, shopping or preparing food are barriers to nutritional self-care 21 . 57 Psychosocial barriers included not considering nutrition important, not recognising personal 58 risk, avoiding 'unhealthy' energy-dense food and loneliness 21 , being told to gain weight and 59 not believing recommendations will work 22 . 60 Beliefs among older adults at risk of malnutrition In previous intervention studies, barriers were addressed through eating pattern advice, such 61 as recommending small portions, energy-rich food, daily snacks; and care pathways, for 62 example dental referral for chewing problems 21 , but participants did not always follow the 63 advice given. Psychosocial barriers or beliefs about personal risk were rarely addressed 21 , for 64 example patients can be surprised, offended or unconcerned when told they are 'at risk' 23  perhaps through apathy or unmet support needs 28 . Luncheon club participants ate more with 73 friends than strangers or at home, highlighting the importance of social eating 29 . These 74 studies capture possible explanations for a lack of adoption of eating advice to address 75 malnutrition risk, such as apathy toward cooking and eating alone, but they do not explain 76 how this apathy is developed or maintained through specific beliefs around eating in older 77 adulthood. In order to design sufficiently engaging and optimally effective behavioural 78 interventions to address malnutrition risk, we need to better understand the role of such 79 psychosocial factors in the eating behaviour of older adults, how they vary between 80 individuals, and how best to address psychosocial barriers. More qualitative work is therefore 81 needed to explore older adults' beliefs and experiences of eating and low appetite, to help 82 understand how support for overcoming barriers can be provided in a way that is relevant for 83 older adults and addresses their diverse specific needs and circumstances.

84
In summary, free-living older adults need support to address malnutrition risk. Barriers to 85 engagement include: reservations about screening and discussing diet, physical barriers to 86 nutritional self-care and psychosocial barriers including considering nutrition unimportant, 87 not recognising risk, avoiding energy-dense food, loneliness, aversion to being told to gain 88 weight, and not believing recommendations will work. Psychosocial barriers are not  Clarifying these issues will inform engaging and persuasive interventions to supplement 92 evidence-based screening and care pathways.

93
Beliefs among older adults at risk of malnutrition In the present study, we used the Person-based Approach (PBA) to clarify issues around 94 eating and appetite in a varied sample of older adults with with a range of health or social 95 conditions associated with malnutrition risk. The PBA systematically applies qualitative 96 research, integrating user perspectives when developing behaviour change interventions in 97 healthcare 24 30 , ensuring they are appropriate, engaging, likely to be useful and used. Study 98 findings will inform the development of an intervention to identify and treat malnutrition or 99 malnutrition risk, specifically a self-management package that is delivered in primary care 100 and supported by healthcare professionals. We propose that the intervention is guided by four 101 principles, from current evidence: a) raise awareness of older adults' nutrition needs; b) 102 motivate engagement in diet and lifestyle change; c) promote self-efficacy for lifestyle 103 change; and d) support and promote autonomy, empowering healthy choices. We will refine 104 the guiding principles, based on the findings of this study.

105
Research aim 106 We aimed to explore how older adults, with health or social conditions associated with risk of 107 malnutrition, experience psychosocial factors relevant to appetite and eating behaviour. The 108 purpose of the study was to inform an intervention comprising a screen and treat policy, 109 incorporating a self-management package, delivered in primary care.

111
This qualitative study is part of a larger project using the Person-based Approach, which 112 involves using qualitative interviews to capture participants' experiences and beliefs 24   The study is reported following COREQ criteria 32 .    Most participants rated their health in the past week as good to excellent but qualified this as 144 'for my age' or 'considering' their health conditions.

145
Topic guide 146 The topic guide was based on evidence and evidence gaps, including findings from a mixed 147 methods review 21 and previous qualitative research, as discussed in the introduction. There   Beliefs among older adults at risk of malnutrition chewing, swallowing or digesting certain foods. They described how any of these physical 185 difficulties could present physical and psychological challenges to shopping or preparing 186 food or making what they considered to be 'good' food choices. For example, pain was 187 described as making it difficult to stand in the kitchen to prepare food as well as reducing 188 motivation to eat. Some participants described their appetite as 'good', 'normally good', 189 'fine', 'healthy' or 'ok', while many described it as 'not that good' or reported noticing their hospitalisation, and some of these were successful. Preparing food, cooking and eating were 203 described as a chore by several participants, who stated they sometimes or often could not be 204 bothered to cook or eat. Although others did not specify that they 'could not be bothered', 205 they reported prioritising other activities above eating, missing meals to look after 206 grandchildren or continuing with activities such as gardening, and stated that hunger soon 207 passed. A few participants described losing a spouse as the point at which they struggled to 208 eat, and reported not being bothered to cook, not fancying food, or feeling too lonely to eat. Participants frequently expressed an understanding that eating is important in order to stay fit 217 and healthy. However, participants described often skipping meals, eating two or fewer meals 218 a day or eating small amounts, which was then perceived as confirmation of the belief that 219 they needed less food. Many stated that appetite and quantity of food consumed is expected 220 to decline with age, and this perceived inevitable decline was attributed to reduced activity 221 and mobility after retirement.  if they found it difficult or didn't understand or accept the rationale for recommendations. 258 One participant stated that the personable approach of a new doctor made them confident to 259 ask for advice, but others were deterred from seeking help based on prior, unsuccessful  Participants often expressed negative feelings about appetite or weight loss or loss of 295 enjoyment around eating, and many considered these changes to be inevitable as they got 296 older. Some expressed a desire to change their eating habits, however difficult it was to eat 297 more, more frequently or regularly, but others accepted decreasing desire to eat and described 298 avoiding social activities that involved eating. A few participants stated they tried to eat well  liked some ONS flavours, or mentioned strategies to make ONS more palatable, but overall 311 ONS were disliked and avoided, due to their texture, a sensation of being too full or difficulty 312 in digesting the milk used to mix them, which was perceived as appetite-reducing.

313
Participants also alluded to ONS reminding them of their spouse's terminal decline. Participants offered multiple reasons, and shared their perceptions and beliefs, when 319 explaining why they did not eat as much as they used to, and many described reduced 320 enjoyment or desire around eating. They outlined how shopping, cooking and eating habits 321 changed in the face of physical challenges, for example relying on others for shopping, 322 making simple food, or eating less when experiencing pain. Participants believed that certain 323 foods were needed for health and fitness, but most expected appetite to decline with age. participants expressed fear about their reduced appetite and weight loss after caring for 335 someone who became frail and died, perhaps worrying that they are also in decline 34 .

336
Behaviour change interventions need to increase understanding of risk, but strategies to 337 address risk and provide reassurance that one can stay well are also needed.

338
Beliefs among older adults at risk of malnutrition We found a widely expressed belief that reduced appetite and food intake are normal in 339 ageing, as noted previously 35 . This is important because ageing-related stereotype beliefs may 340 reduce individuals' confidence to carry out health-promoting behaviours 36 . Novel to our 341 study, participants with long-term eating difficulties, pain, inactivity or reliant on others for 342 everyday needs expressed resignation to reduced appetite and eating alongside physical 343 decline and deteriorating quality of life. Resignation was frequently expressed as no longer 344 being 'bothered' to cook or eat as effortfully as they had. Those with recent weight loss, such 345 as during bereavement or hospitalisation, seemed motivated through fear or hope to find 346 solutions, but also seemed to have a sense of resignation while experiencing the pain of loss.

347
There appeared to be a trajectory towards resignation which started with age-related beliefs, Beliefs among older adults at risk of malnutrition seen as a sign that food was not needed. Novel in research with older adults, some 371 participants distinguished between hunger and desire for food items and were more likely to 372 eat due to desire than hunger. Interventions could therefore encourage eating desired foods.

373
Some available ONS flavours were liked, contradicting previous research 44  make ONS more appealing and easier to drink, including suggestions given by participants.

380
In future, enriched food products may provide a more acceptable alternative to ONS 45 , 381 though the way they are presented to users will also be important.

382
Difficulties maintaining independence 383 In the present study, many participants ate less than they used to, concurring with previous independence, but that strict diets could compromise nutrition, undermining independence. 389 We also concur with Maitre, et al 50 who found that malnutrition risk is associated with food 390 'pickiness', both of which increase alongside growing dependence on others for food-related  Included individuals were currently struggling to shop, prepare food and/or eat, or anticipated 427 such challenges in the near future. Some appeared undernourished, though we used no 428 objective measure of malnutrition risk. We also included individuals who were currently 429 eating regularly, some of whom had experience of unintended weight loss from which they 430 had recovered, giving useful insights. This study could be improved by including more men, 431 or those with a wider range of conditions known to increase malnutrition risk.

433
Beliefs among older adults at risk of malnutrition The key findings are that: 1) sense of resignation; 2) diverse experiences and common 434 significant barriers; and 3) difficulties in maintaining independence underpin the experience 435 of eating and appetite among older adults at risk of malnutrition. There seems to be a 436 trajectory of increasing resignation in the face of common beliefs, values and barriers to 437 eating among older adults with health and/or social conditions known to increase malnutrition 438 risk. Diverse multiple barriers to eating were found, which may be underpinned by common study team has carried out such a study and aims to publish the results shortly. The mooted 449 mechanisms identified in the present study e.g. raising risk awareness, promoting self-450 efficacy, also need to be tested, and the study team is carrying out a RCT in which these will 451 be investigated. It will be important to assess whether behavioural techniques included in 452 interventions address patients' psychological needs and issues (resignation, independence), 453 and influence behavioural and clinical outcomes. It would also be useful to identify which 454 food-related strategies work best to enable continued independence for older adults. The lead author affirms that this manuscript is an honest, accurate, and transparent account 459 of the study being reported. The reporting of this work is compliant with COREQ guidelines. 460 The lead author affirms that no important aspects of the study have been omitted and that any 461 discrepancies from the study as planned have been explained. 462