Malnutrition can be broadly defined as a deficiency, excess or imbalance of energy, protein and other nutrients, which can have an adverse effect on the human body, function and clinical outcomes (British Association for Parenteral and Enteral Nutrition (BAPEN), 2018). The term ‘malnutrition’ also relates to obesity and other diet-related conditions (World Health Organization (WHO), 2016), not just undernutrition. This article will discuss the topic of malnutrition, with a focus on undernutrition. It will use a case study taken from clinical practice in an acute hospital setting (Table 1a; Table 1b). In addition, this article will analyse and synthesise the relevant evidence relating to malnutrition in older adults and discuss the role of the advanced clinical practitioner (ACP) in the assessment and management of malnutrition to reduce the risk of refeeding syndrome.
| Characteristic/field | Details | ||
|---|---|---|---|
| Age (years) | 76 | ||
| Sex | Male | ||
| Presenting complaint | Fall | ||
| History of presenting complaint | Attended emergency department (ED) following a fall at home; transferred to the clinical decision unit for frailty team assessment | ||
| Patient able to recall a fall when standing in the kitchen. Felt well prior to the fall and no injuries | |||
| Unable to get himself off the floor so called 999 and paramedics conveyed to hospital | |||
| Patient presented following a fall, initially assessed by ED doctors who identified no acute medical issue and transferred to the Clinical Decision Unit for frailty team assessment | |||
| Past medical history | Raised cholesterol | ||
| Depression | |||
| Active smoker; pack year history: 35 years | |||
| Alcohol consumption: 10–12 units a day | |||
| Drug history | Regular medications: Simvastatin 40mg OD, Citalopram 20mg OD | ||
| Acute medications: none | |||
| No allergies | |||
| No over the counter medication or herbal remedies taken | |||
| Family history | Nil to note | ||
| Occupational history | Retired teacher | ||
| Social history | Lives alone in a house | ||
| Independent mobile with a stick | |||
| No carer or family support | |||
| Systematic enquiry | No shortness of breath, difficulty in breathing or cough | ||
| No chest pain, palpitations or presyncope symptoms | |||
| No fever, no change in bladder or bowel and no abdominal pain | |||
| No pain or injuries reported | |||
| Reduced oral intake at home, fatigued and low motivation to cook, noticed gradual weight loss | |||
| Investigations | Observations | Self-ventilating on air SaO2 | 97% |
| Respiratory rate | 14 | ||
| Blood pressure | 128/75 | ||
| Heart rate | 72 regular | ||
| Temperature | 36.7C | ||
| Capillary blood glucose | 6.5mmol/L | ||
| Body mass index | 18 | ||
| Malnutrition Universal Screening Tool score | 3 | ||
| Characteristic/field | Details | ||
|---|---|---|---|
| Investigations | Blood results | White cell count | 10.2 |
| C-Reactive protein | <5 | ||
| Sodium | 142 | ||
| Potassium | 3.1 | ||
| Urea | 5.3 | ||
| Creatine | 65 | ||
| Glomerular filtration rate | 84 | ||
| Phosphate | 2.8 | ||
| Magnesium | 1.7 | ||
| Examination | Alert and orientated | ||
| Chest clear, no added sounds, heart sounds normal, no peripheral oedema | |||
| Abdomen not distended, soft with no pain or palpable masses | |||
| Global muscle wastage, no injuries and remained mobile | |||
| On assessment, the patient was physically frail, with low body mass index and hypokalaemic on initial blood tests. Holistic assessment revealed that he was a lifelong smoker with a pack year history of 35 years, and he was a drinker of 10-12 units of alcohol daily. It was identified he was not meeting his nutritional need resulting in sarcopenia and increased frailty and at risk of refeeding syndrome | |||
| Diagnosis | Multifactorial fall: sarcopenia, poor nutritional status, alcohol excess, at risk of refeeding syndrome | ||
| Differential diagnosis | Sinister pathology | ||
| Management | Refeeding blood tests (blood tests were requested to include phosphate and magnesium to assess refeeding syndrome risk prior to possible commencement of nutritional supplements) | ||
| Oral potassium replacement | |||
| Input charts were commenced and referral to dietitian for further advice | |||
| Patient education and lifestyle advice on nutritional status and alcohol excess, including the risks and harms associated with action plan agreed | |||
In alignment with NHS England (2023), this article defines an ‘older adult’ as a person chronologically aged 65 years and over. However, the ACP, should remain astute that due to heterogeneity, people can biologically age at different rates; identifying the presence of frailty in individuals, regardless of age, may improve clinical outcomes (WHO, 2022).
Malnourishment
Cost to the NHS
In the UK, it is estimated that over 1 million people over the age of 65 years are malnourished or at risk of malnutrition (Malnutrition Task Force, 2021). Up to 25–34% of patients admitted to hospital are at risk of malnutrition, which annually costs the NHS over £19 billion (BAPEN, 2018).
In 2017, there were 350 deaths in which malnutrition was recorded as a cause or contributory factor. Although these numbers could be considered as low, it is important to note that numbers have significantly increased over the last 10 years (Mayor, 2018). Malnutrition accompanies other states of ill-health, and can be both a cause of, and an effect on, disease processes (National Institute for Health and Care Excellence (NICE), 2012); it is often seen in those with multi-comorbidities. A sample study conducted by Glickman (2018) reported that out of those that died as a result of malnutrition, 10% also had Alzheimer's and 29% had concurrently been diagnosed with frailty. These results could be attributed to the UK's growing elderly population, and the increased recognition and diagnosis of frailty over the last decade (Kojima et al, 2019). ACPs should be able to respond to the change in the healthcare needs of the population (Health Education England (HEE), 2017) and be placed across services to help support the elderly demographic.
Causes of malnourishment in the older adult
In older adults, malnourishment is multifactorial and can be caused by insufficient dietary intake, increased metabolic demand from disease processes, poor absorption or increased nutrient loss (BAPEN, 2006). Physiological changes during the ageing process—such as poor dentition, altered taste and smell of food, slower gastric emptying and constipation—can reduce appetite (Pilgrim et al, 2015). Reduction in lean body mass and slower metabolic rate can also be seen in older adults (Evans, 2005). Cognitive impairment can reduce orientation to mealtimes and affect recognition of food (Pilgrim et al, 2015), and, towards the end stages of dementia, individuals can develop dysphagia, which further impacts their nutritional status (Hansjee, 2019).
Socioeconomic issues and loss of functional independence to shop and prepare food also contribute to malnutrition in older adults (BAPEN, 2006). Inadequate social support, loneliness and depression can exacerbate malnutrition further, as individuals develop apathy towards nutritional intake (Evans, 2005). Some of these contributory factors to malnutrition are preventable or reversible. ACPs should identify those with these risk factors and work collaboratively with other professionals or agencies to empower the individual to make choices that could improve their health outcomes (HEE, 2017).
Refeeding syndrome
In those that are malnourished, the risk of refeeding syndrome (RFS) should be considered. RFS is ill-defined, but broadly recognised as a severe fluid and electrolyte shift when malnourished patients recommence feeding (Crook, 2014). It is not represented by a singular condition or specific marker, but rather a spectrum of possible effects that can occur in high-risk groups (Khan et al, 2011).
RFS was first recognised in individuals liberated from Japanese prison camps after the Second World War, who had experienced prolonged fasting (Mehanna et al, 2008). They developed peripheral oedema and neuropathy, and approximately 20% of liberated prisoners died unexpectedly following the onset of feeding (Schnitker et al, 1951). RFS as a term was proposed by Weinsier and Krumdieck (1981) following the death of two malnourished patients who were given parental nutrition support. The patients developed severe hypophosphataemia and acute cardiopulmonary decompensation, which led to their death.
While incident rates are variable within population groups, RFS is predicted to occur in 14% of older malnourished patients (Friedli et al, 2018). However, in a recent survey, only 20% of physicians were able to describe the syndrome (Janssen et al, 2019). With no gold standard definition of RFS and no specific clinical marker to measure, gaining accurate prevalence data is difficult (Crook, 2014).
The evidence-base for clinical outcomes from RFS is limited; a small cohort of authors have generated a variety of results. A literature review by Friedli et al (2017) found no relationship between RFS and adverse outcomes. A further randomised control study by Friedli et al (2020) concluded a statistical significance in RFS to be associated with increased length of hospital stay (1.57 extra days, 95% confidence interval (CI) 0.38–2.75, p value=0.01) and increased mortality at 180 days (odds ratio 1.97, 95% CI 1.18–3.29, p value=0.01) (Table 2).
| Term | Definition |
|---|---|
| Confidence intervals (CI) | A range of values in which the researchers are reasonably confident that the population lies within those values, eg 95% CI means that the researchers are 95% confident that the population lies within those limits |
| Odds ratio | The odds of an event happening in that one group, by the odds of it happening in another group |
| P value | This is usually used to test a null hypothesis. The p value gives the probability of any observed differences |
| Null hypothesis | A hypothesis that there is no difference between the groups being tested |
Pathophysiology
The exact pathophysiology of RFS is unknown but is believed to be caused by a change from catabolic to anabolic metabolic activity when nutrition is reintroduced after a period of fasting (Aubry et al, 2018). This is a normal physiological reaction (Martini, 2018); however, in those that are malnourished, the resulting sudden serum electrolyte and fluid shift is detrimental (Friedli et al, 2018). The severity of RFS is proportionally linked to the length of time fasting occurred (Friedli et al, 2018).
During digestion, carbohydrates from food are metabolised primarily into glucose, absorbed through the intestinal mucosa and, via facilitated diffusion, enters the portal circulation which can result in a rise in blood sugar levels. Glucose provides the body with most of its energy requirements and is used in glycolysis for the production of adenosine-triphosphatase (ATP) in cells (Martini, 2018). In response to hyperglycaemia, insulin is released by beta cells in the pancreatic islets, activating GLUT-4 receptors in skeletal and adipose tissue to absorb glucose (Sherwood, 2013).
Additionally, the presence of insulin increases the permeability of cells to potassium, phosphate and magnesium, which are all largely intracellular ions. Insulin activates the sodium-potassium (ATP) pump, transporting two potassium ions into the cell and three sodium ions out of the cell (Sweeney et al, 2001). Magnesium is the cofactor required to catalyst the phosphorylation of glucose, thus maintaining glucose within the cell for glycolysis, as the cell membrane is impermeable to phosphorylated glucose (Crook, 2014). Additionally, the cell has an increased demand for thiamine, a co-factor in carbohydrate metabolism catalysing the conversion of glucose to ATP in the Krebs cycle (Aubry et al, 2018).
Glycogenesis
Furthermore, insulin stimulates glycogenesis (the conversion of glucose into glycogen for storage) and occurs largely in the liver, kidneys and skeletal tissues (Dashty, 2013). Once glycogenesis capacity is reached in the liver, excess glucose is synthesised by hepatocytes into triglycerides and stored as fat in adipose tissues known as lipogenesis (Martini, 2018). Concurrently, insulin inhibits processes that convert stores into glucose, glycogenolysis and gluconeogenesis, until homeostasis of blood sugar levels has been achieved (Dashty, 2013).
If blood glucose levels start to fall and are not corrected by ingestion of carbohydrates, the body enters a catabolic state, metabolising glycogen stores in the liver into glucose (Sherwood, 2013). Stores are usually sufficient to maintain homeostasis of blood sugar levels between feeding, but evidence on duration of liver glycogen stores is varied. Some reports state stores are depleted after as little as 4 hours of fasting (Martini, 2018) and others state stores last as long as 24 hours (Cherkas and Golota, 2014). However, there are many variants that could affect duration of stores, including activity levels and the volume of ingested carbohydrates prior to fasting (Hearris et al, 2018).
Gluconeogenesis
Once liver glycogen stores are exhausted, gluconeogenesis occurs, during which amino acids are converted into glucose (Dashty, 2013). After 24 hours of fasting, breakdown of skeletal muscle occurs, which releases alanine into the circulation that is subsequently converted by the liver into glucose. On initial fasting, gluconeogenesis uses approximately 75g of skeletal protein per day, which can lead to loss of muscle strength and function (Berg et al, 2002). This kind of impact will be the most severe in those with pre-existing reduced muscle mass such as the elderly (Gingrich et al, 2019).
If catabolic metabolism continues, in effort to preserve protein in skeletal tissue, adipose tissue stores of triglyceride are metabolised by lipolysis into fatty acids (Griffiths, 2012). In turn, ketogenesis in the liver is stimulated, which releases high levels of ketones into the circulation that are used by cells in the Kreb cycle to yield energy, with length of survival during fasting proportionate to the size of triglyceride stores (Berg et al, 2002). Ketones are acidic, disassociating in solution and, in prolonged fasting, can lead to acidosis and eventual cell death (Martini, 2018).
Upon recommencing feeding, hyperglycaemia occurs and the body reverts back to anabolic carbohydrate metabolism as its energy source (Crook, 2014). In response, a sudden release of insulin is initiated, which triggers an intake of glucose, potassium, phosphate, magnesium and thiamine, intracellularly (Reber et al, 2019). With mineral stores diminished from a period of fasting, serum levels of the ions can fall rapidly. Consequently, to maintain osmolality, sodium and water are retained (Khan et al, 2011); the resulting electrolyte shift can lead to development of RFS.
Importance of recognising those at risk
The ability to recognise those that are, or at risk of malnutrition and RFS, is fundamental to improving health outcomes for the ageing population, with evidence suggesting improving nutrition can reduce frailty (Roberts et al, 2019). All care providers are responsible for identifying people at risk of malnutrition (NICE, 2012). In acute hospital care, screening for malnutrition should occur on admission and be repeated weekly where there is clinical concern, as mandated by the Department of Health's (DoH) (2010) Essence of Care benchmarking document and NICE's guidance on nutrition support in adults (2012).
Assessment tools
Body mass index
NICE (2017) recommend that screening should include an assessment of body mass index (BMI) (Table 3). BMI is endorsed by the WHO (2020) as a marker of an individual's nutritional status and a risk indicator of disease, referencing normal weight/height ratio between 18.5–24.9kg/m2. In this case study (Table 1), BMI was calculated at 18 and therefore classified as underweight (WHO, 2020).
| Patient has one or more of the following | Body mass index (BMI) less than 16kg/m2 |
| Unintentional weight loss greater than 15% within the last 3–6 months | |
| Little or no nutritional intake for more than 10 days | |
| Low levels of potassium, phosphate or magnesium prior to feeding | |
| Patient has two or more of the following | BMI less than 18.5kg/m2 |
| Unintentional weight loss greater than 10% within the last 3–6 months | |
| Little or no nutritional intake for more than 5 days | |
| A history of alcohol abuse or drugs including insulin, chemotherapy, antacids or diuretics. |
The need for age-adjusted body mass index
The applicability of the BMI as a reference value is the subject of debate; some studies suggest that they need to be adjusted for older patients. BMI measurements can be influenced by a variety of external and internal factors, such as:
Winter et al (2014) reported increased mortality in patients aged over 65 years with a BMI less than 23kg/m2 and recommend that the ‘healthy range’ for older adults be adjusted accordingly. Additionally, malnutrition may be overlooked in patients with higher BMIs if full nutritional evaluations are not complete—the older patient's nutritional need and percentage weight loss may not be considered (Bahat et al, 2012).
Malnutrition Universal Screening Tool
To assess nutritional status, a validated tool that provides reliability of results across all care settings should be used (NICE, 2012). The Malnutrition Universal Screening Tool (MUST) (Figure 1) has been advocated as an appropriate tool in a variety of studies (Beck et al, 2009; DoH, 2010; NICE, 2012; NICE, 2017). The MUST has been shown to have high sensitivity and specificity, at 100% and 98%, respectively (Harris et al, 2008); several studies have established the predictive validity of the test, correlating high MUST scores with mortality. Mountford et al (2016) have reported higher mortality in those with MUST scores of 2 (p value 0.004). Similarly, Stratton et al (2006) reported 24% predictive validity of hospital readmissions and length of stay (Rasheed and Woods, 2013; Bakewell et al, 2020). Following a review of nutritional screening tools in the elderly, Power et al (2018) concluded that the MUST had the greatest validity within the hospital setting.
Compliance in using the MUST tool in clinical practice remains low, despite national recommendations (Wong et al, 2022), with poor education and training, and perceived low clinical importance by ward staff reported as barriers to its effective implementation (Frank et al, 2015). Additionally, the MUST has received criticism due to its reliance on the use of BMI and inability to differentiate between types of mass loss. Patients with higher BMI can mask lean muscle mass depletion by the accumulation of fluid or fat mass (van Vliet et al, 2021). van Vliet et al (2021) identified that the MUST tool incorrectly calculated nutritional status in 90% of obese patients in their study. Therefore, ACPs should ensure that when assessing nutritional status, the MUST score has been completed accurately and interpreted with caution, particularly in those with higher BMIs, and reviewed alongside a nutritional history.
Diagnosis and stratification of RFS
NICE (2017) outline criteria to identify those at risk of RFS on the commencement of nutritional support; however, currently, there are no validated diagnostic criteria. Friedli et al (2018) propose an algorithm to assist with the diagnosis and stratification of RFS. They suggest if serum phosphate levels drop by 30%, or if two or more of the electrolytes—phosphate, potassium, magnesium and thiamine—drop below normal range within 72 hours of feeding, a diagnosis of RFS is probable. If there are no associated symptoms, the patient is stratified as imminent RFS; if clinical symptoms are present, the patient is stratified as manifest RFS. The suggested algorithm is based on a literature review and no evidence is currently available of its validity or reliability, so the approach is currently unable to be translated into clinical practice.
Clinical symptoms
Clinical symptoms of RFS are diverse due to the range of electrolytes potentially affected (Reber et al, 2019). Symptoms can overlap in older people, with the clinical manifestations of multi-comorbidities (Pourhassan et al, 2018), which can be non-specific, such as onset of delirium (Janssen et al, 2019). A common characteristic reported is the rapid clinical deterioration in the patient's condition (Janssen et al, 2019), typically seen within 72 hours of initiating feeding; this characteristic is commonly noted in patients recommenced on oral and artificial feeding methods (Pourhassan et al, 2018). Friedli et al (2018) have reported the most prevalent symptoms documented are tachycardia, tachypnoea and oedema.
Current research suggests that the baseline serum levels of electrolytes should be measured in those at risk of RFS (Khan et al, 2011; Friedli et al, 2017; Aubry et al, 2018). NICE (2017) recommend those on nutritional support should have their glucose, sodium, potassium, magnesium and phosphate levels monitored daily, as well as their baseline levels of calcium, albumin, C-reactive protein, full blood count, iron, folate and vitamin B12 similarly recorded. However, it is important to consider that serum levels of electrolytes prior to feeding may remain normal, as homeostatic mechanisms maintain serum levels at the detriment to intracellular levels (Khan et al, 2011).
ACPs must use professional judgement to manage risk appropriately (HEE, 2017) and consider the possibility that refeeding syndrome may develop in individuals with risk factors at onset of feeding, even in the presence of normal blood results. Adhering to research guidelines (HEE, 2017), ACPs should measure baseline and daily blood values as per NICE (2017) recommendations in those commencing nutrition, so that RFS can be identified and appropriately managed. In those at high risk of RFS, electrocardiogram monitoring is also recommended during the initial refeeding stages to detect any cardiac arrhythmias secondary to hypokalaemia (Skinner, 2005).
Low phosphate
In a literature review completed by Friedli et al (2017) 38 of 45 studies defined hypophosphataemia as a measure of RFS. Hypophosphataemia can be a result of many physiological processes and disease states. Potential differential diagnoses of low serum phosphate include alcohol abuse, malabsorption secondary to bowel disease, increased renal excretion in hyperparathyroidism, vitamin D deficiency, concurrent use of phosphate-binding medication and respiratory alkalosis.
Respiratory alkalosis secondary to hyperventilation occurs due to low serum carbon dioxide, producing alkalaemic conditions that stimulate intracellular glycolysis and an uptake of phosphate, resulting in hypophosphataemia (O'Brien and Coberly, 2003). In a case-controlled study, the researchers found sepsis diagnosis to be threefold more likely in those with hypophosphataemia (Kagansky et al, 2005). ACPs must complete a detailed assessment of the patient including detailed history taking to underpin expert clinical decision-making skills (HEE, 2017). These measures ensure the timely diagnosis and treatment of RFS, while ensuring differential diagnoses are appropriately excluded.
Approaches
Food-first approach
Working in conjunction with the patient, family and carers, a ‘food-first’ approach should initially be used. This includes monitoring intake, frequent small meals high in energy and nutrients, fortifying food to increase nutrient density and supporting the individual at mealtimes if required (Beck et al, 2009). If unable to meet full nutritional needs, oral nutritional supplements (ONS) should be considered (NICE, 2017). Elia (2012) reports a cost saving of £330 per elderly hospitalised patient using ONS. A Cochrane review in 2011 found no significant difference in mortality, morbidity or hospital length of stay 3 months after commencing ONS, but reported improved grip strength compared to dietary advice alone (Baldwin and Weekes, 2011).
The first 72 hours after reintroducing nutrition is considered time critical for identifying and treating RFS (Aubry et al, 2018). Current NICE (2017) guidelines advise that those at high risk of RFS should be commenced on hypocaloric nutritional support, which is slowly increased over 4 to 7 days, with immediate thiamine for the first 10 days of refeeding. Oral or intravenous supplementation of potassium, phosphate and magnesium should also be corrected alongside feeding (NICE, 2017).
A prophylactic prescription of phosphate and thiamine, prior to commencing feeding, has shown positive outcomes in preventing RFS from developing and reducing mortality rates (Friedli et al, 2018). However, Friedli et al (2018), in their systematic review of the literature, found that the positive findings are not transferable to the frail elderly population, as they were completed in younger patients diagnosed with anorexia nervosa. There is a growing research consensus that to prevent further harm from malnutrition, feeding and electrolyte replacement can occur simultaneously (Khan et al, 2011; Janssen et al, 2019). This supports NICE (2017) recommendations that the correction of electrolytes is not required pre-feeding. However, at present, the principal evidence base is only observational or expert opinion, and there have been no published primary research studies on the approach.
Due to the electrolyte shift during the period of refeeding, it is important to monitor sodium and fluid balance to prevent fluid overload (Aubry et al, 2018). This is particularly prudent for the older population who may have coexisting comorbidities such as heart failure or chronic kidney disease. For those at high risk of RFS, sodium restriction should be considered. Attention should also be directed towards the provision of any additional fluid resuscitation for concurrent dehydration due to the high sodium content of Hartman's solution or 0.9% sodium chloride solution (Reber et al, 2019).
A multidisciplinary approach
Successful management of malnutrition and RFS is achieved through a multidisciplinary approach (Khan et al, 2011). Recognising limitations of competence and professional scope (HEE, 2017), ACPs should refer the patient to specialists such as dietitians or nutrition support teams for a detailed nutritional assessment (Beck et al, 2009).
A care plan should be implemented to ensure their complete nutritional requirements are met (NICE, 2012) and reviewed at regular intervals ensuring treatment is appropriate and effective (NICE, 2017). Reber et al (2019) advise that management plans should be individualised and adapted to patient populations, aligning to the ACP framework to promote person-centred care (HEE, 2017). ACPs should support and educate patients and carers in malnutrition, empowering the individual to make decisions about their long-term management of nutritional support (HEE, 2017; NICE, 2017).
Reflecting on this case study, timely identification and treatment of malnutrition was achieved through a patient-centred nutritional plan. Recognising the potential to develop refeeding syndrome by monitoring potassium, phosphate and magnesium levels at onset of feeding additionally improved the patient's outcome.
Conclusion
Malnutrition in the frail elderly population is increasing and can lead to a devastating personal and socioeconomic burden. By undertaking a detailed history, holistic assessments, initiating and evaluating interventions, while working collaboratively with the multidisciplinary team, ACPs are perfectly placed to identify those at risk, initiate treatment and support these patients by explaining what is occurring and the necessity of treatment. This ensures the patient is fully informed and the centre of care. Future studies should focus on the validation of RFS diagnostic criteria and research into the management of RFS in older adults. Such measures could support the development of specific guidelines in this cohort of high-risk patients.
Recommendations for practice
The authors suggest the following recommendations for practice:
Limitations
This clinical review has been influenced by several limiting factors, specifically the: