The Refeeding Syndrome: a neglected but potentially serious condition for inpatients. A narrative review

Malnutrition is a frequent and often unrecognized condition among inpatients [1, 2]. Indeed, 20–50% of individuals are at risk of malnutrition or already malnourished at hospital admission, but malnutrition is diagnosed in 7% only [3]. Older age, low socioeconomic status, lack of organizational support, chronic systemic or psychiatric diseases, polytherapy, poor diet, reduced absorption capacity, excessive nutrient losses are the most frequent conditions underlying malnutrition [4]. The management of malnourished inpatients can be difficult due to the risk of metabolic impairment after the start of nutrition [5]. The adverse outcomes of refeeding were firstly reported during the World War II in rapidly re-fed prisoners who had starved for five to six months [6]. People who have fasted for a long time, developed heart, and/or respiratory failure, peripheral edema, neurological symptoms, and death after the introduction of excessive or even appropriate calorie amount [6‐8]. In the 80 s, the term ‘refeeding syndrome’ (RFS) was introduced to describe severe hypophosphatemia and other electrolyte/metabolic abnormalities and the related cardiovascular and pulmonary manifestations leading to death occurring in two chronically malnourished patients who received aggressive dextrose‐based parenteral nutrition (PN) [9]. Since then, many cases of RFS have been described as a rare, but severe and potentially fatal complication related to re-feeding (either orally, enterally or parenterally) of individuals who have fasted or consumed very few calories over a long period of time [10, 11]. Among the diseases or conditions predisposing to malnutrition and consequently to RFS after re-feeding, anorexia nervosa [12‐14], cancer [15, 16], critical illnesses [13, 17‐20], and frailty in the elderly [21‐27] are the most frequently implicated.

The switch from a catabolic to an anabolic state may be the cause of the clinical manifestations of the RFS, even though the pathophysiological mechanisms are still not fully understood [28]. Furthermore, the lack of a clear definition accounts for the difficulty of diagnosis and uncertainties in treatment [2, 29]. Therefore, the RFS is a potentially serious condition, often overlooked by many physicians [30]. This is of particular concern because of the high prevalence of hospital malnutrition often underestimated even in the internal medicine wards [31, 32].

The objectives of this narrative review are to summarize the knowledge on the RFS and to focus on the most useful topics for the clinical practice.

The following databases were queried: PubMed (National Library of Medicine), the Cochrane Library, Excerpta Medica dataBASE (EMBASE), and the Cumulative Index to Nursing and Allied Health Literature (CINAHL). The search strategy was performed using the following keywords: refeeding syndrome OR phosphate, potassium, magnesium AND anorexia nervosa, cancer, critically ill patients, elderly. The filters ‘humans’ and ‘adults’ were used. Hand searching the references of the identified studies and reviews was carried out too.

The lack of a universally recognized RFS definition makes it difficult to obtain precise estimates of its incidence [33]. Indeed, either hypophosphatemia only or multiple electrolyte abnormalities (with different cut offs) with or without clinical manifestations have been considered in its definition [34, 35]. The reported incidence rates ranged between 0 and 80%, depending on the definition and the patient population studied [34]. RFS has been described in 48% of severely malnourished patients, in 34% of intensive care unit (ICU) patients, in 33% of patients with anorexia nervosa (AN), in 25% of cancer inpatients, and in 9.5% of patients hospitalized for malnutrition from gastrointestinal fistulae [10, 12, 33, 36]. Many factors may lead to underestimation of RFS incidence rate, such as insufficient monitoring of the patients’ electrolytes after nutrition starting, lack of consultation by experts in clinical nutrition, the nonspecificity of the clinical manifestations of the syndrome in patients with multiple co-morbidity and the physician unawareness [11].

To identify patients at risk for RFS is necessary evaluating the risk of malnutrition by validated screening tools first, and then assessing the diagnosis and grading the severity of malnutrition [5, 33, 37, 38]. Distinguishing malnutrition from the other related conditions, such as starvation, cachexia, cancer cachexia, and sarcopenia, is important from a clinical point of view (Table 1) [39‐44]. The screening for the risk of malnutrition should be performed in inpatients within the first 24–48 h through validated screening tools, such as the Nutritional Risk Screening 2002 (NRS-2002), the Mini Nutritional Assessment-Short Form (MNA-SF), the Malnutrition Universal Screening Tool (MUST), the Short Nutritional Assessment Questionnaire (SNAQ) [5, 37, 39]. If an individual is identified to be at risk of malnutrition, an extensive nutritional assessment for diagnosis and evaluation of the severity of malnutrition should be carried out by an expert in nutrition [39, 40].

A great number of diseases or conditions predisposes to malnutrition [21, 28, 33, 34, 37, 39, 45‐47]. These predisposing conditions can be divided into the following categories: predisposing to disease-related malnutrition with inflammation (chronic diseases leading to catabolic inflammatory responses); predisposing to disease-related malnutrition without inflammation (acute disease and injury-related malnutrition); and predisposing to malnutrition in the absence of diseases (hunger, socioeconomic, or psychologic-related conditions, drugs) [39], as summarized in Supplementary Table 1.

In the presence of severe underweight or weight loss, prolonged fasting period, and/or low electrolyte concentrations, the risk of RFS is particularly high [30]. In 2006, the National Institute for Health and Clinical Excellence (NICE) guidelines [48] reported the risk factors to identify people at low or high risk for RSF. In 2018 Friedli et Coll added the very high-risk category [21]. Recently, the American Society for Parenteral and Enteral Nutrition (ASPEN) published updated consensus criteria for identifying adult patients at risk for RFS [33]. These criteria are presented in Table 2.

The difficulty in RFS diagnosing is due to the discrepancy between the onset of the symptoms and the occurring of metabolic shift (see below), and the nonspecific nature of its clinical manifestations [46]. There is a great heterogeneity among the published definitions of RFS, ranging from hypophosphatemia alone [18, 19, 22, 24, 27, 49‐54] to the presence of severe low-serum electrolyte levels along with fluid balance abnormalities and/or organ dysfunction [16, 21, 34, 55]. Only hypophosphatemia has been universally recognized as a feature of the syndrome [38]. Friedli et Coll. proposed diagnostic criteria for imminent or manifest RFS, based on the electrolyte blood concentrations and clinical symptoms to standardize its prevention and treatment [21]. According to this definition, “imminent” RFS is present when a shift in electrolytes occurs within 72 h after the start of nutritional treatment (i.e., > 30% decrease in blood phosphate from baseline or phosphate values < 0.6 mmol/L or any two other electrolyte shifts below normal range) [21]. “Manifest” RFS is considered if any electrolyte shift occurs in conjunction with typical clinical symptoms (see below) [21].

More recently, the ASPEN proposed diagnostic criteria for distinguishing mild, moderate or severe RFS [33] (Table 3). The extent of the decrease in the serum levels of one or more electrolytes (among phosphate, potassium, or magnesium) defines RFS severity: 10–20% (mild RFS), 20–30% (moderate RFS), > 30% and/or organ dysfunction and/or thiamine deficiency (severe RFS) [33]. Thus, either hypophosphatemia and/or hypokalemia and/or hypomagnesemia qualify the presence of the RFS. The timing of onset is determinant for the diagnosis, since the RFS develops shortly (from hours up to 5 days) after having substantially increased the energy provision to individuals who have been undernourished [33].

The pathophysiology of the RFS is probably related to the shift from the catabolic to the anabolic metabolic pathways occurring after the re-start of feeding in undernourished subjects. During early starvation, blood glucose and insulin levels decline while glucagon concentrations increase by stimulating glycogenolysis in the liver. When glycogen reserves become depleted, gluconeogenesis is stimulated in the liver, using amino acids derived from muscle breakdown [56]. During prolonged fasting, the body switches to use fats as the main sources of energy with a decrease in basal metabolic rate of 20–25% [57]. Increased lipolysis in fat reserves leads to the production of ketones that are used by the brain as preferred fuel during starvation [29, 56]. During prolonged fasting, several intracellular minerals become severely depleted, particularly phosphate, potassium, and magnesium. However, the concentrations of these minerals may remain within the normal range in the serum because there is a reduction in their renal excretion and because of the phosphate outflow from the cells into the blood, leading to normal blood phosphate levels despite depleted storages [21].

Symptoms generally appear within 2–5 days of re-feeding and may range from absent/mild to a severe and life-threating clinical syndrome, depending on the pre-existing degree of malnutrition and comorbidity [10, 11, 45]. All the body organs may be involved, leading to cardiac, respiratory, hematologic, gastrointestinal, neurologic, and musculoskeletal manifestations, until death [10, 21, 58].

The identification of patients at risk for RFS is the first step to prevent the onset of the syndrome, and to avoid an excessive nutritional replenishment in those individuals [21, 66]. Risk factors should be carefully investigated before starting either oral, enteral, or parenteral nutrition, because every route of calorie administration is implicated in the occurrence of the RFS [33, 58]. Well-trained medical staff and specialized nutritional support teams, consisting of physicians, dieticians, nurses, and pharmacists, positively impact on the patient outcomes [48]. However, a multidisciplinary team is not available in all hospital settings, and often the evaluation of the risk for RFS is left to the clinician's critical sense at the time of starting nutritional support [11, 33, 36, 38, 67]. After defining the degree of RFS risk, the rate of fluid and nutrition administration, the correction of electrolyte imbalances, and the supplementation of vitamins and micronutrients (zinc, iron, selenium) can be determined [36] (Table 5). If a prolonged nutritional support is required, adjustments over time in accordance with the patient clinical conditions might be necessary [58].

Several therapeutic approaches have been proposed to prevent or treat the RSF [10, 21, 28, 36, 45, 46, 48, 67, 68] (Fig. 1). Since hypophosphatemia occurs after refeeding, according to the grade of RSF risk, phosphate may be administered preventively before the initiation of nutritional therapy, even if blood levels are in the low-normal range [21]. Similarly, thiamine is essential in carbohydrates metabolism and should be supplemented before restart feeding even in the case of normal blood levels [21]. An excessive administration of glucose by stimulating insulin production leads to the consumption of electrolytes (mainly phosphate) through the anabolic pathways. Starting re-feeding very gradually, independently of the route of administration, is therefore mandatory [58]. Owing to the risk of fluid overload, sodium and hydration should be provided cautiously, until the patient is metabolically stable [38]. In case of overt symptoms, energy and fluid intakes should be reduced and adapted to the clinical conditions [30].

Specific conditions might require special attention.

This narrative review provides the latest information on the management of RFS in light of the current evidence. Although RFS is a frequent condition that can have serious consequences above all in specific categories of inpatients, it is often undiagnosed and overlooked by physicians. Its knowledge is essential to avoid rapid and excessive nourishing of at-risk patients; thus, preventing serious complications, long hospital stays, and the increase in health costs.