It is common for all people receiving palliative care to have a reduced oral intake, reasons can include physical obstruction, swallowing difficulties, cachexia, weakness, loss of desire to eat or the person can become increasingly sleepy and therefore less able to receive oral nutrition.  While the literature does not specifically focus on older adults, many of the recommendations for nutrition support are applicable to this cohort as much as the adult population.
Three systematic reviews [1-3] are specific to nutrition and palliative care. However, it is important to note that only one of the systematic reviews  addressed the medical and quality of life (QoL) benefits of artificial nutrition and hydration in palliative care. This review actually excluded all studies found in the initial search and concluded that inadequate evidence is available on which to base recommendations for practice.
Two of the palliative care specific systematic reviews [2,3] addressed the decision making and emotional impact of hydration and nutrition provision in palliative care and conclude that while it is essential to consider emotions and health beliefs of the person, family and staff, that currently there is insufficient evidence to understand how such beliefs and emotion may influence decision making.
Three reviews and guidelines related to the nutrition in dementia. [4-6]
In addition, four nutrition guidelines [6-9] have been included in this synthesis. Although the guidelines are not palliative care specific, they include references to the management of cancer and advanced disease process. [7,8] Additional papers have been drawn on for context. [10-14]
Overall, the quality of the evidence is of moderate - low quality, many of the reviews are based on prospective non-controlled trials only, some without a robust methodology and others not adequately considering bias in the discussion of their results. Guidelines are frequently not specific to end-of-life care or relate broadly to multiple disease states. The ethical issue concerning withholding nutrition intervention in randomised controlled trials is cited as the likely reason for the lack of RCT’s in this area. It is therefore prudent to base practice on the next-best evidence available.
Poor appetite, weight loss and oral nutrition support
ESPEN guidelines for the nutritional management of people with cancer suggests that nutrition therapy for people with cancer who are at risk of malnutrition has been shown to improve or maintain nutritional status but not survival.  First line management should be advice to manage symptoms, such as incorporating energy enriched foods and fluids and where necessary inclusion of oral nutrition supplements.  This can also include counselling to manage symptoms of adjunct therapies or the disease itself. 
The use of novel diets, particularly those which would promote undernutrition through removal of whole food groups or calorie restriction should be discouraged with no evidence to support their use.  Ideally, nutrition support would be provided by a dietitian on the basis of the nutrition care process as outlined by the Dietitians Australia in their practice guidelines for the management of malnutrition across the continuum of care.  Unfortunately data to define the best time to introduce nutrition support is still lacking. However, at the point in which palliative patients are experiencing metabolic disturbances related to the disease process or mechanical and side effect issues with eating, nutrition therapy is unlikely to reverse malnutrition.  As such nutrition therapy should preferably be initiated before the person becomes malnourished, particularly if their prognosis indicates known symptoms of anorexia or GI disturbance.
Nutritional screening tools are useful for rapid, early identification of malnutrition. The available screening tools for older adults include Malnutrition Universal Screening Tool (MUST), Mini Nutritional Assessment (MNA), Short Form Mini Nutritional Assessment (MNA-SF), Geriatric Nutritional Risk Index (GNRI) and the Simplified Nutritional Appetite Questionnaire (SNAQ).  When a risk of malnutrition is identified, a full nutritional assessment and an appropriate intervention is recommended.  Further research is needed to validate the nutritional screening tools for older adults in the hospital setting. 
There is no consensus as to the most appropriate screening or assessment tool for people with dementia. The Mini Nutritional Assessment (MNA) or Short Form Mini Nutritional Assessment (MNA-SF) contain specific questions regarding neurophysiological problems, as well as capturing high-level information regarding quantity of food and fluids consumed, and mode of feeding.  While these tools enable a structured approach across settings, they are not replacements for observations of appetite and intake by carers and health professionals. 
For people receiving palliative care who are nutritionally stable or who are in the early stages of palliative care, malnutrition screening tools can be useful in highlighting onset of undernutrition and malnutrition risk  This is important as widely available screening tools, such as the Malnutrition Universal Screening Tool (MUST), use traditional BMI cut-offs of <18kg/m2 for underweight which are considered low in older adults 65yrs and over. 
Unplanned weight loss in people residing in residential aged care facilities receiving end of life palliative care are exempt from reporting as part of the National Aged Care Mandatory Quality Program 7.2.1. 
Dysphagia and effect on nutrition
In the review by Arends,  dysphagia as a consequence of diseases such as head and neck cancer is discussed. Evidence for the maintenance of swallowing function with texture modified oral intake is of low quality but consensus is for the use of enteral nutrition (EN) and swallowing exercises to maintain function where it may be restored. Such recommendations are less applicable to a palliative care cohort where physical obstruction e.g. a tumour, or deterioration in health means improvements in swallow function are less likely and EN may be the only viable option.
Medically assisted nutrition: enteral and parenteral nutrition support
The systematic review conducted by Good et al.  specifically looked at studies that evaluated medically assisted nutrition to palliative care patients via either a tube being placed into the gastrointestinal system (enteral) or the venous system (parenteral). While there were no randomised controlled trials or prospective controlled trials, and all search results were excluded based on these criteria, five prospective non-controlled trials and one updated Cochrane systematic review were retrieved. Of the non-controlled trials, one study included participants with advanced dementia and the other four included only participants with advanced cancer. The participants either received only parenteral nutrition (PN) or only enteral nutrition (EN), with the exception of one study where participants could have received either.
Enteral nutrition (EN): Both studies in the review by Good et al.  looked at the survival rate associated with EN in people receiving palliative care, while one also looked at outcomes for nutritional status and quality of life (QoL). Overall, the authors note the lack of evidence to support the use of artificial nutrition via an enteral feeding tube when oral intake declines despite it being a common intervention for older adults with dementia. In the study of individuals with advanced dementia, median survival was 6 months despite the use of artificial nutrition or parenteral fluids.
Feeding tubes are not recommended for older adults with advanced dementia as there is no evidence to support long-term benefit to the person.  Observational studies discuss complications (e.g. aspiration pneumonia, visits to ED) arising from tube feeding in advanced dementia without a clear association to an observed increase in mortality.  The ESPEN guidelines on nutrition in dementia do not recommend artificial nutrition in severe dementia or for people in the terminal phase. 
Similar studies are cited in a variety of settings including aged care, whereby results also conclude that EN does not extend life. In addition, the systematic review by Good et al.  looked at papers discussing nutrition and QoL outcomes. The evidence for weight gain or other measures of improved nutritional status overall was weak. Similarly, of the two studies that considered QoL, neither found evidence for improvement with a percutaneous endoscopic gastrostomy (PEG). 
Parenteral nutrition (PN): In the prospective study identified in the review by Good et al.  the impact of home parenteral nutrition (HPN) on mortality and QoL is evaluated in cancer patients over three years. Although the study does not refer to the patients as being palliative, they are all noted as being no longer suitable for oncologic therapy and required PN predominantly due to chronic obstruction which would cause death by starvation unless artificial nutrition support was provided. Overall, the study found that only 38% of the patients provided PN at home (HPN) may have benefited from prolonged survival as a result of assisted nutrition.
In regard to QoL, Good et al.  noted HPN appeared to maintain QoL outcomes until approximately 2 months before time of death, which suggests that only patients with a prognosis of more than 3 months might benefit from HPN. Bozzetti also identified a randomised controlled trial (2017) which suggested an improvement in older people with cancer undergoing palliative chemotherapy when administered with supplemental HPN for a period of 6 months. 
Another study in the review by Good et al.  utilised semi structured interviews with patients with advanced cancer and their family members to evaluate the experience of HPN. The main themes to emerge were that of ‘relief’ and ‘security’ that nutrition needs were being met and had a subsequent positive effect on QoL and body weight. However, negative impact was also discussed and reported to be restrictions in family life and social contacts for the person and their family. Despite this, participants reported the benefits outweighed the negatives. 
Medically assisted hydration
The systematic review by Del Rio et al.  presents the traditional arguments for and against medically assisted hydration. The argument presented in favour of hydration is:
- provides a basic human need
- provides comfort and prevents uncomfortable symptoms
- assists in achieving better QoL
- provides minimum standards of care.
Arguments presented against hydration:
- interferes with acceptance of the terminal condition
- IV therapy is painful and intrusive
- unnecessary in the instance of an unconscious patient who will not experience symptoms of thirst or pain
- ketones and metabolic by-products of dehydration act as a natural anaesthetic, causing reduced consciousness and suffering for the patient
- minimises coughing, congestion, oedema, and ascites.
Certainly, the evidence for medically assisted nutrition and hydration support is weak for the improvement of nutritional status and QoL and no improvement in mortality is observed when they are administered.  However, it is important to note the emotional and comforting impact of providing hydration as with nutrition, and how this fits with cultural norms and values for the patient and family. 
Cultural values and norms for nutrition and hydration in palliative care
In addition to supporting health, food and eating also has an impact on social and psychological experiences for people receiving palliative cares and also their families.  The administration or withdrawal of medically assisted nutrition or hydration can be complex and emotional for all involved.  Particularly in the case of withholding or withdrawing support, studies suggest that this can be perceived by the patient and family as an act of ‘giving up’.  In particular, two studies discussed in the systematic review by Del Rio et al. note that reduced interest in food, inability to digest and fatigue all caused serious psychological distress to the person receiving palliative care.  Western cultures are most likely to associate reduced oral intake with acceleration to end of life, but this concept is also prevalent in other cultures such as Taiwan. Taiwanese people believe if a person dies hungry their soul will be restless and hungry after death.  Thus there is a preference for artificial nutrition and hydration at end of life.
However, the review paper by Gent et al.  suggests that when families and patients are prepared for such changes in their health and understand ‘nature taking its course’, their focus becomes more on providing other forms of care such as mouth care, with less negative emotional impact.  There are cultures that are more accepting of decline in oral intake at end of life, such as in the Hindu faith who recognise this is a sign of end of life and not the cause. 
Open discussions with the patient and family from an early stage can be beneficial. These discussions could include disease progression, nutrition support [2,3] and likely benefit of an intervention. While it is important to acknowledge the significance of artificial nutrition/hydration from the cultural perspective, there is a duty of care to fully inform the patient and family of the lack of evidence in support of providing it. 
Essentially, for people receiving palliative care the desired outcome of nutritional management is to maximise the person’s comfort and QoL. Oral nutrition support should be based on recommended daily intake of vitamins and minerals and dietary advice individualised for each patient to maximise QoL while maintaining nutritional status while possible.  Alternative diets based on the concept of ‘starving’ the disease or which promote mega dosing of vitamins and minerals should be discouraged as having no clinical evidence to support positive outcomes. 
Medically assisted nutrition support (EN or PN) may be appropriate due to a medical need for assisted nutrition or maintaining the cultural values and norms for the individual and their family.  Enteral nutrition and PN is not recommended in the case of people with advanced dementia.  Overall, EN does not appear to offer any survival advantage for people receiving palliative care in relation to mortality, and there is only weak support for the maintenance of nutritional status or QoL, although each person should be considered on the basis of their individual needs and prognosis.  In the event of administering EN, survival advantage with PEG placement is weakly positive for people with some degenerative health conditions, although best timing for PEG placement is not known. 
Parenteral nutrition would not be the first choice for medically assisted nutrition. However, in the case of no enteral access and a prognosis of more than 3 months is estimated, PN or HPN may be beneficial in improving QoL outcomes and stabilising nutritional status. It is unlikely to significantly lengthen life expectancy and the risks of hospitalisation due to risk of infection should be discussed prior to commencing. 
Similarly IV hydration presents complications much like PN, with risk of infection and symptoms of fluid collection such as oedema and ascites  with little benefit other than providing comfort to the person and their family which should not be overlooked in palliative care pathways. 
Ultimately, medically assisted nutrition and hydration may only be justified when the patient is at higher risk of dying from starvation than the disease progression  but discussion of withdrawing or withholding nutrition and hydration support should be discussed and normalised with the person and their family early in palliative care planning. 
Currently there is a lack of good quality evidence to assist best practice guidelines for commencing EN or PN support, which patients would be most appropriate, criteria for commencing and ending this type of nutritional support and how this treatment is evaluated.
Page updated 22 June 2021