Is More Aggressive Refeeding Safe for Patients with Anorexia Nervosa?

When a patient with anorexia nervosa (AN) is admitted for treatment, the primary goal is physiologic stabilization with nutritional rehabilitation. However, the need to rapidly correct serious weight loss must be balanced against the risk of the refeeding syndrome. In two separate studies, Canadian researchers have concluded that more aggressive initial refeeding for inpatients is more effective than more traditional approaches.

The refeeding syndrome, marked by electrolyte shifts secondary to insulin surges brought on by rapidly and aggressively refeeding severely malnourished patients, can lead to muscle weakness, delirium, and cardiac failure or arrest.  More recent reviews of the dilemma of refeeding determined that limiting caloric intake does not eliminate the risk of refeeding syndrome, and high-risk patients remain at high risk. In one report,  the authors suggested that the risk of developing refeeding syndrome may not lie solely in the amount of calories administered but also in the composition of those calories; specifically, a lower carbohydrate load may prevent insulin surges that can lead to electrolyte abnormalities  (Int J Eat Disord. 2011;44;182).

A study tests the theory

According to a team at Montreal Children’s Hospital, providing higher initial caloric intake via nasogastric (NG) tube feeding leads to improved early weight gain and shorter hospitalization, without increased risk of the refeeding syndrome (J Adolesc Health. 2013).

Dr. Holly Agostino and colleagues at Montreal Children’s Hospital performed a retrospective review of patients admitted for restrictive eating disorders from December 2003 to December 2011. Beginning in May 2010, a standardized continuous nasogastric tube feeding protocol was begun at the hospital, and all patients admitted for treatment of restrictive eating disorders had a NG tube inserted when they arrived on the medical ward. In the new standard treatment, patients receive an initial daily caloric intake of 1500 kcal/day or 1800 kcal/day, depending on age, administered exclusively through continuous NG feeds with a formula containing 44% carbohydrates (Nutren Junior, with Fiber), a lower carbohydrate percentage than that used in their older feeding programs. Caloric intake is then increased by 200 kcal/day until maximal caloric intake is achieved. All patients are started on prophylactic phosphate with supplementation and electrolytes are monitored twice daily in the first week.

Comparing the new and old protocols

The authors compared the courses of all patients treated with the NG feeding protocol (NG) with all patients admitted prior to the new protocol. In the earlier protocol, patients were treated with standard bolus feeds of daily calories divided into 3 meals and 3 snacks beginning with 1000 to 1200 kcal/day and increased by 150 kcal/day. The mean caloric intake of the bolus-fed cohort was 1069 kg/day (800 to 1500 kcal/day). The initial prescription for the NG group was significantly higher, with a mean of 1617 kcal/day (1200 to 2000 kcal/day).

The mean rate of weight gain in the first week of hospitalization among the bolus-fed group was .08 kg/week (-2.9 kg to +2.6 kg/week). Only 32% of those patients attained a minimally acceptable rate of weight gain in the first week, and in fact 51% lost weight or gained no weight during this time. Mean weight gain improved to 0.69 kg/week. In contrast, the mean weight gain in the NG cohort was significantly improved during the first 2 weeks compared to the bolus-fed group. Only 6% of the NG group lost weight or gained no weight during week 1, and 84% met the minimal goal of gaining at least 0.5 kg/week during the first week of admission.

The study results supported treating undernourished inpatients with restrictive eating disorders with a higher initial caloric intake (60% to 75% of estimated needs). And, despite the significantly higher caloric start and initially rapid weight gain among the NG group, no increased incidence of electrolyte abnormal ties was reported. To further reduce the risk of refeeding syndrome; patients were prescribed a nutritional formula containing less than 50% carbohydrates.

The authors noted that because placement of an NG tube is perceived as intrusive to patients, most centers use NG feeding only to manage patients who refuse food. The authors believe that the answer is to define NG feeding as a standardized practice supported for early discharge. Although the initial weight gain was greater among the NG cohort, no difference was noted in total weight gain by both groups during hospitalization.  However, the hospital stay was reduced by 17 days in the NG group, and no increase in the rate of readmission to the medical ward was reported during the 6 months after discharge.

An all-food nutrition rehabilitation protocol for adolescents

In contrast to the program at Montreal Children’s Hospital, Anick Leclerc RD, and a team of dietitians and an eating disorders specialist at McMaster Children’s Hospital, Hamilton, Canada and the Hospital for Sick Children in Toronto, found that a more aggressive food-based approach to nutritional rehabilitation with macronutrients resulted in a consistent and safe weight increase, with negligible adverse effects. The aggressive structured nutrition rehabilitation protocol (NRP) used with 21 patients with AN and eating disorders not otherwise specified (EDNOS) safely led to weight gain with little risk of the refeeding syndrome (J Adolescent Health 2013[Epub ahead of print]).

The NRP was begun on the first full day of admission with 1500 kcal per day, with 3 meals and 3 snacks, each containing 250 kcal. All meals and snacks were provided as food and taken orally. The prescribed nutrition was increased by 250 kcal each day on days 2 and 3, and then by 250 kcal every other day until day 7, when patients were receiving 2500 kcal/day. Thereafter nutrition was increased as required based on a minimum weight gain of 1.0 kg/week. The inpatients were supported through all meals and snacks by parents with the support of frontline staff or staff alone. When a food item was refused, an oral nutritionally complete liquid supplement was given; any patient needing NG support was removed from the study.  And, during the first 2 weeks of hospitalization, patients were either placed on bed rest or minimal activity, such as walking to and from meals and their therapy group sessions.

All but one patient completed the study; patients were hospitalized for a mean duration of 35.8 days, and medical stability was achieved on day 14 (mean). The mean length of inpatient stay was similar to that in the Montreal study: 33.8 days vs. 35.8 days, respectively, and the patient populations were similar inage—14.7 years in this study, compared to 14.9 years in the Montreal study. None of the patients required nasogastric feedings; the 1 patient who dropped out of the study left the hospital before becoming medically stable. None of the patients had low serum phosphate levels or electrolyte abnormalities.

All patients had significant weight gains (mean: 0.24 kg/day) on the NRP. A cumulative change in body mass index over the 14 days demonstrated sustained daily weight gain among all the inpatients, with a weekly average gain of 1.7 kg.

The authors point out that using an all-food program to refeed inpatients with eating disorders involves not only balancing macronutrient intake but helping acclimatize patients to eating at regular intervals with variations in food quantity and quality.