A Higher-Calorie Refeeding Protocol Does Not Increase Adverse Outcomes in Adult Patients with Eating Disorders

Abstract

Background: Patients with eating disorders (EDs) are often considered a high-risk population to refeed. Current research advises using “start low, go slow” refeeding methods (∼1,000 kcal/day, advancing ∼500 kcal/day every 3 to 4 days) in adult patients with severe EDs to prevent the development of refeeding syndrome (RFS), typically characterized by decreases in serum electrolyte levels and fluid shifts.

Objective: To compare the incidence of RFS and related outcomes using a low-calorie protocol (LC) (1,000 kcal) or a higher-calorie protocol (HC) (1,500 kcal) in medically compromised adult patients with EDs.

Design: This was a retrospective pre-test–post-test study.

Participants/setting: One hundred and nineteen participants with EDs, medically admitted to a tertiary hospital in Brisbane, Australia, between December 2010 and January 2017, were included (LC: n=26, HC: n=93). The HC refeeding protocol was implemented in September 2013.

Main outcome measures: Differences in prevalence of electrolyte disturbances, hypoglycemia, edema, and RFS diagnoses were examined.

Statistical analysis performed: χ2 tests, Kruskal-Wallis H test, analysis of variance, and independent t tests were used to compare data between the two protocols.

Results: Descriptors were similar between groups (LC: 28±9 years, 96% female, 85% with anorexia nervosa, 31% admitted primarily because of clinical symptoms of exacerbated ED vs HC: 27±9 years, 97% female, 84% with anorexia nervosa, 44% admitted primarily because of clinical symptoms of exacerbated ED, P>0.05). Participants refed using the LC protocol had higher incidence rates of hypoglycemia (LC: 31% vs HC: 10%, P=0.012), with no statistical or clinical differences in electrolyte disturbances (LC: 65% vs HC: 45%, P=0.079), edema (LC: 8% vs HC: 6%, P=0.722) or diagnosed RFS (LC: 4% vs HC: 1%, P=0.391).

Conclusions: A higher-calorie refeeding protocol appears to be safe, with no differences in rates of electrolyte disturbances or clinically diagnosed RFS and a lower incidence of hypoglycemia. Future research examining higher-calorie intakes, similar to those studied in adolescent patients, may be beneficial.