See “Celiac Crisis/Refeeding Syndrome Combination: New Mechanism for an Old Complication” by Catassi on page 442.
Refeeding syndrome is a term that describes a constellation of metabolic complications that arise when initiating the nutritional rehabilitation of malnourished patients. Despite being well described, it is an underrecognized entity. It was first described in 1948 on starved Asian prisoners after World War 2 (1,2). This manifested as hormonal changes leading to hypophosphatemia, hypokalemia, hypomagnesemia, hypocalcemia with or without hyperglycemia, fluid retention, hypoalbuminemia, and thiamine deficiency (3). It is a potentially lethal condition and needs early detection and prompt management. Similarly, celiac crisis has been described as a life-threatening complication of celiac disease that presents with diarrhea, vomiting, and multiple metabolic derangements (4). Because there are no standardized diagnostic criteria for celiac crisis, Jamma et al (5) has given a working definition of celiac crisis, which is “acute onset or rapid progression of gastrointestinal symptoms attributable to celiac disease along with at least 2 objective signs of malnutrition, dehydration, or electrolyte disturbance (hypernatremia/hyponatremia, hypocalcemia, hypokalemia, or hypomagnesemia).” Clinically and biochemically, both refeeding syndrome in celiac disease and celiac crisis look similar, except that the former condition develops after the introduction of a gluten-free diet (GFD) and the latter is the initial presentation of celiac disease (while the patient is receiving gluten) (6).
Malabsorption syndrome, as with celiac disease in children, is often associated with poor nutritional status especially in developing countries such as India, and is a predisposing condition for refeeding syndrome. Various studies from India have shown that 62% to 100% of children with celiac disease are undernourished at the time of diagnosis (7–9). Once GFD is introduced, the appetite of these children improves and a sudden increase in food intake puts them at risk of developing refeeding syndrome; however, when a severely malnourished patient with celiac disease does not improve or worsens while receiving GFD and develops dehydration and electrolyte and metabolic imbalances, it is often labeled as celiac crisis. So far, no one has looked into this aspect of celiac disease. There is only 1 report of refeeding syndrome in an adult with celiac disease, but there are none in children (10). It is important to recognize refeeding syndrome because management is a gradual increase in caloric intake with correction of electrolyte and mineral deficiencies and not steroids, which are often used in celiac crisis. Recently, we had the opportunity to treat 5 children with celiac disease with refeeding syndrome. We present our experience to stress the need for recognition of this potentially fatal condition and to treat it appropriately.
From January to December 2010, a total of 35 children with celiac disease were managed in the Department of Pediatric Gastroenterology at Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow, India. Chart review of all of these cases was done to discover which children had refeeding syndrome. Celiac disease was diagnosed as per the modified European Society for Pediatric Gastroenterology and Nutrition criteria (11). Refeeding syndrome was ascribed in celiac patients who deteriorated clinically after the initiation of GFD with parenteral or enteral feeding and had biochemical parameters suggestive of refeeding syndrome such as hypophosphatemia, hypokalemia, hypocalcemia, and hypoalbuminemia (12). Of 35 children with celiac disease, 5 developed or presented with refeeding syndrome and were reviewed in detail. The workup of such cases includes complete blood count, serum electrolytes, liver function tests, IgA-anti-endomysial antibody or IgA-anti-tissue transglutaminase antibody (recombinant human), and upper gastrointestinal endoscopy (oesophagogastroduodenoscopy) with duodenal biopsy. Special attention was paid to correct electrolyte and mineral deficiencies, prophylactic thiamine therapy, fluid restriction, and gradual buildup of GFD. The target level of required energy was reached in 2 weeks. Informed consent was obtained from parents before intervention (oesophagogastroduodenoscopy). Ethical clearance was obtained for this clinical audit. The clinical features of these children are summarized in Table 1 and the biochemical features are summarized in Table 2.
The median age of 5 patients with refeeding syndrome in celiac disease was 6.5 (range 2.2–10) years and 3 were boys. Details are provided in the following sections.
Patient 1, a boy, presented with chronic diarrhea (stools: 10–12 times/day, large volume, watery, explosive) since 1 year of age. He received multiple courses of antibiotics and even antitubercular therapy with equivocal response. He was evaluated elsewhere 2 months before admission and diagnosed as having celiac disease on the basis of positive tissue transglutaminase antibody (IgA-tTG 298 IU/mL) and total villous atrophy on duodenal biopsy and was put on GFD. He received intravenous fluid for 1 week because of poor oral intake and subsequently developed severe weakness, pedal edema, and generalized exfoliating skin lesions. At admission to our hospital, he was grossly emaciated and listless. His weight was 12 kg (weight z score −8.5) and his height was 106 cm (height z score −3.9). He had pallor, angular cheilitis, pedal edema, and generalized exfoliating hyperpigmented skin lesions. Chest examination revealed basal crepitations, and abdominal examination revealed 3-cm palpable liver with mild ascites.
The investigation at our hospital is shown in Table 1 and Table 2. His retroviral serology was negative. His stool examination did not reveal any pathogens.
Initially, the child was thought to have celiac disease with celiac crisis. Hence, a trial of intravenous methyl prednisolone was started in addition to GFD. Despite doing this, the child did not improve; rather he developed hyperglycemia without ketosis caused by steroids and required subcutaneous insulin injections. Because the diarrhea continued despite GFD and steroid therapy, our attention was diverted to refeeding syndrome because his biochemical parameters were classical (Table 2). Steroids were stopped after 1 week and he was started on intravenous supplementation of potassium, magnesium, and calcium and oral supplementation of phosphate and thiamine along with reduced caloric intake of 400 kcal (requirement 1600 kcal/day), protein 1 g · kg−1 · day−1. Thereafter, a gradual increment of caloric intake (90 kcal/day during 10 days) of GFD was done. Because he had fluid retention, salt and fluids were restricted and he was given multiple courses of albumin infusion (total 5 g/kg in 2 weeks) and 1 packed red blood cell transfusion for anemia.
With continued supplementation of electrolytes and slow increments in oral dietary intake, his dyselectrolytemia was corrected. Subsequently, diarrhea improved with subsidence of pedal edema, ascites. Skin lesions also healed. The patient was discharged after 3 weeks, and by then, he gained 2.5 kg weight and his edema and ascites had subsided. He was discharged receiving GFD with caloric intake of 1500 kcal/day.
Patient 2 presented with chronic diarrhea (5–6/day watery, explosive, and voluminous stools) for 4 months with significant failure to thrive and anorexia. On examination, the child was emaciated and his weight was 10.5 kg (weight z score −4.47) and height was 96 cm (height z score −1.48). He had pallor and distended abdomen but no organomegaly. His investigations (Table 1) confirm the diagnosis of celiac disease and he started receiving GFD. In view of poor oral intake, the patient was given partial parenteral nutrition but diarrhea increased with worsening hypokalemia and further lowering of phosphate levels, reflecting the development of refeeding syndrome (Table 2). At this point, oral supplementation of phosphorus, thiamine, and B complex and intravenous replacement of calcium, magnesium, and potassium along with gradual buildup of caloric intake were done.
The patient improved gradually, diarrhea subsided, and appetite and activity improved. He was discharged after 18 days, receiving GFD with oral supplementation of calcium and phosphorus.
Patient 3 presented with chronic diarrhea that he had since 6 months of age (3–5/day semisolid, explosive stools), associated with abdominal distension and failure to thrive. He received antitubercular therapy in the past (3 courses) with no relief of symptoms. The child was evaluated elsewhere, and barium enema and rectal biopsy were done with a suspicion of Hirschprung disease. On examination, the child weighed 10 kg (weight z score −8.8) and his height was 93 cm (height z score −4.9). He had pallor, clubbing, widening of wrists, and distended abdomen with no organomegaly. In view of his chronic small bowel type of diarrhea with failure to thrive, rickets, and pallor, celiac disease was considered. Investigations (Table 1) confirmed the diagnosis, and he was discharged, receiving GFD and multivitamins and iron supplementations; however, within 2 weeks of discharge, the child was readmitted with complaints of facial puffiness, pedal edema, increased fatiguability, and increased frequency of stools despite receiving strict GFD. On examination, the child had cold, clammy extremities and bilateral pitting pedal edema, and his chest had bilateral basal crepitations. This time, investigations (Table 2) were suggestive of refeeding syndrome and the child was put on restricted diet of 500 kcal/day (50% of his caloric requirement), and a gradual escalation of energy was achieved with salt restriction and supplementation of phosphorus, potassium, calcium, magnesium, thiamine, iron, and multivitamin. The child was discharged after 2 weeks, and by the time of discharge, his diarrhea subsided, edema and electrolyte disturbances were corrected, and he was receiving GFD with a caloric intake of 1000 kcal/day.
Patient 4 presented with complaints of diarrhea for 4 months (4–5/day watery, large-volume stools) and abdominal distension and intermittent abdominal pain (periumbilical) of the same duration. She was admitted with hypovolemic shock, and she required volume resuscitation and inotropic support for hemodynamic stability. Her weight was 10 kg (weight z score −14.28) and height was 110 cm (height z score −4.29). Investigations (Table 1) confirmed the diagnosis of celiac disease and she started receiving GFD. Although diarrhea continued, her appetite became voracious within 5 days of starting GFD; however, she developed pedal edema, tachypnea, and hepatomegaly. Her biochemical profile (Table 2) showed features of refeeding syndrome, so her caloric intake was reduced and correction for potassium, calcium, and phosphorus was started. She improved clinically and laboratory parameters normalized within 1 week. At discharge, she had gained 2 kg in 2 weeks of her hospital stay with no edema and diarrhea.
Patient 5 presented with complaints of diarrhea for 3 months (5–6 watery stools per day), and abdominal distension and vomiting for 15 days. She had low hemoglobin (5.5 g/dL) 1 month ago, and 1 unit of packed red blood cell was transfused elsewhere. Also, there was a history of weight loss of 2 kg in 2 months and loss of appetite and lethargy. On examination, her weight was 7.9 kg (weight z score −4.8) and height was 77 cm (height z score −3.1). She was emaciated with frontal bossing, Harrison sulcus (suggestive of rickets), and distended abdomen. Investigations (Table 1) confirmed the diagnosis of celiac disease and GFD was started. Within 3 days of GFD, her appetite became voracious, but she started developing pedal edema and tachypnea in addition to diarrhea. Her clinical and biochemical profile (Table 2) were suggestive of refeeding syndrome, so her caloric intake was reduced and biochemical correction was started. She improved clinically and laboratory parameters normalized within 1 week. At discharge, she had gained 1 kg weight in 2 weeks with no edema and diarrhea. During a period of 3 to 6 months, all 5 children were asymptomatic while receiving GFD.
In this series, all 5 children were severely malnourished (body mass index <14 kg/m2) and all of them had hypophosphatemia, hypokalemia, hypoalbuminemia, and hypocalcemia, meaning that they had the perfect setting for developing refeeding syndrome. There was no ambiguity in the diagnosis of celiac disease in any of these cases because all of them had Marsh grade III (villous atrophy) changes on duodenal biopsy with positive serology. Because celiac crisis, although rare, is a more-often-discussed condition than refeeding syndrome, we made a diagnosis of celiac crisis in the first case and tried steroids before realizing that refeeding syndrome may mimic celiac crisis. Subsequent cases (cases 2, 3, 4, 5) were diagnosed and treated as refeeding syndrome in the setting of celiac disease. The lack of awareness about refeeding syndrome in celiac disease is substantiated by relatively little information in the literature about this association. Besides this series, there is only 1 case report of refeeding syndrome and that in an adult. Stanga et al (10) described a 65-year-old man diagnosed as having celiac disease who developed refeeding syndrome subsequent to aggressive parenteral nutrition.
Celiac crisis is a term that has been applied to patients with celiac disease of acute onset that is severe enough to be potentially fatal (6). In 1972, Lloyd-Still et al (4) described celiac crisis in 3 children who presented with profound diarrhea, dehydration, metabolic abnormalities, and weight loss. The electrolyte abnormalities in this condition are akin to that described in refeeding syndrome such as hypokalemia, hypomagnesemia, hypocalcemia, and hypoalbuminemia. Jamma et al (5) have described 12 adult cases of celiac crisis, so far the largest series, and have shown that the majority of cases had hypoalbuminemia, hypokalemia, hypocalcemia, and hypomagnesemia. Half of them were treated with steroids and nutritional supplements and the remaining half with nutritional supplements only. Unfortunately, no one has paid any attention to phosphate levels in this condition. Others have shown that nutritional supplements with correction of metabolic and electrolytes deficiencies are enough to treat celiac crisis, and steroids are not always necessary (13–16). It is possible that some of the reported patients with celiac crisis, especially those who had worsening of symptoms after the introduction of GFD, actually had refeeding syndrome. Hence, emphasis should be placed on the sequence of events to differentiate these 2 entities (celiac crisis as a presentation and refeeding syndrome as worsening while receiving GFD) because the management protocol is different (correction of metabolic abnormalities vs steroids).
Refeeding syndrome refers to severe metabolic and biochemical changes that occur in malnourished patients when they are aggressively fed orally, enterally, or parenterally (17). Clinical features are nonspecific initially and often go unrecognized; severe manifestations include hypotension, arrhythmias, seizures, respiratory or cardiac failure, coma, and sudden death. Hypophosphatemia, the hallmark of refeeding syndrome, is caused by a combination of depletion of total body phosphorus stores during catabolic starvation and increased cellular influx of phosphorus during anabolic refeeding. In addition, a high demand for phosphate occurs in the production of phosphorylated intermediates of glycolysis (ATP; 2,3-diphosphoglycerate) in the anabolic refeeding state (3). Hypophosphatemia can impair cardiac, neuromuscular, and respiratory muscle function, and is associated with rhabdomyolysis, hemolysis, thrombocytopenia, and leukocyte dysfunction (12,18).
Unlike celiac crisis, the pathophysiology of refeeding syndrome is well established. During refeeding, a metabolic transition from catabolism to anabolism occurs. Glycemia leads to increased insulin and decreased glucagon secretion. Insulin stimulates glycogen, fat, and protein synthesis; all of these processes require minerals such as phosphate, magnesium, and cofactors such as thiamine. Insulin also promotes the intracellular transport of potassium through the Na-K-ATPase symporter, which also transports glucose into cells; magnesium and phosphorus are transported into the intracellular compartment, and to maintain osmolality, sodium and water are retained in the extracellular fluid. These processes result in a decrease in serum levels of phosphate, potassium, and magnesium, all of which are already depleted (12,17,18).
Because this syndrome is usually asymptomatic unless electrolyte disturbances are severe, it is important for clinicians to be aware of this condition and identify and monitor patients at risk of developing it. In at-risk patients, that is, those who have been chronically undernourished or have poor oral intake for 7 to 10 days, baseline serum concentration of electrolytes, especially phosphate, potassium, magnesium, calcium, sodium, urea, and creatinine, should be ascertained before the initiation of nutritional support and should be measured once daily for 1 week and and 3 times in the following week. The mainstay of treatment for refeeding syndrome is correction of metabolic and electrolyte abnormalities and replenishing thiamine and other water-soluble vitamins before starting enteral or parenteral feeding. Caloric intake should be commenced at 50% to 75% of the estimated requirement to reduce the risk of refeeding syndrome (17). Furthermore, within the next 7 to 10 days, the caloric intake should be increased gradually to attain optimal energy requirement (10). The basic approach should be “start low and go slow.” Fluid and salt intake should be minimized to prevent fluid overload (12).
Patient 1 had low body mass index with a history of aggressive intravenous fluid therapy and initiation of GFD after the diagnosis of celiac disease, and presented to us 1 month later with severe manifestations of refeeding syndrome such as cardiac failure, beriberi-like skin lesions, severe electrolyte disturbances, edema, ascites, hyperglycemia, and hypophosphatemia. Patient 2 received parenteral nutrition initially because of poor oral intake after the diagnosis of celiac disease was made, and that led to manifestations of refeeding syndrome, which resolved once it was correctly identified. Patients 3 and 5 had been evaluated elsewhere as having Hirschprung disease, delaying the correct diagnosis of celiac disease in already malnourished status. None of cases 2 to 5 received steroids, although they all had laboratory parameters to classify them as celiac crisis but improved dramatically once they were managed as refeeding syndrome. Rather, patient 1, who received steroids, developed complications of steroid therapy as hyperglycemia requiring insulin. Identification of the first patient led to the early diagnosis of refeeding syndrome in other 4 patients, and that explains the disparity in severity in electrolyte disturbances between patient 1 and the other 4 patients. Thus, early detection is the key to reducing the morbidity and mortality of this state.
Severely malnourished patients with celiac disease are at risk of developing potentially life-threatening refeeding syndrome. Refeeding syndrome may mimic celiac crisis, and the treatment goal of these patients should be correction of metabolic and electrolyte abnormalities with reduced caloric intake and not steroids.