Is More Aggressive Refeeding Safe for Severely Malnourished Inpatients?

Vol. 28 / No. 2 — Eating Disorders Review

More evidence on this controversial question.

Recently there has been increased interest in the rate of refeeding people with AN. A new article from University of California, San Diego researchers suggests that more aggressive refeeding may be appropriate in selected young patients hospitalized with restricting eating disorders, including AN,  avoidant/restrictive food intake disorder (ARFID), or other specified eating disorders (J Eat Disord. 2017; 5:1, published online before print). The study group of 87 patients 8 to 20 years of age included a subgroup of severely malnourished patients who presented at less than 75% of expected body weight (%EBW).

On admission, most patients were started on an oral nutritional rehabilitation diet, and the initial caloric level was based on the patient’s recent dietary history. Typically the patients received from 1500 to 1800 kcal/day, but lower-calorie diets (such as 1200 kcal/day) were used if the patient had been using extreme dietary restriction, for example, fewer than 500 kcal/day over several weeks. Daily caloric intake was titrated to achieve 2 kg weight gain per week. If the daily target weight wasn’t achieved for 2 days in a row, the caloric intake was increased; this was also the case for those with persistent or severe cardiac complications, such as bradycardia overnight, symptomatic postural changes in heart rate or blood pressure, or arrhythmias.

Nasogastric or nasojejunal tubes were used if needed until the patient could consume 100% of daily nutritional and fluid needs by mouth for at least 24 hours. Intravenous fluids were reserved for dehydrated patients unable to tolerate oral fluids.

On admission, all patients were screened with complete blood counts and complete metabolic panels, urinalysis, and electrocardiograms. Refeeding laboratory findings were assessed daily, concentrating on serum magnesium and phosphorus levels, and urinalysis. Continuous cardiac monitoring was used throughout the patient’s entire hospital stay.

According to Dr. R. Tamara Maginot and colleagues, %EBW was determined through evaluation of several clinical factors. Although the conventional approach uses the 50th percentile body mass index (BMI, or kg/m2) to calculate expected body weight for children and adolescents, the San Diego researchers customized %EBW to return each child or teen to the growth trajectory in which he or she was healthy. For example, if the patient’s premorbid BMI was greater than the 85th percentile, and the patient was otherwise healthy, the initial %EBW was selected to restore the patient to a BMI at the 75th percentile level. If the premorbid growth patterns were unknown, the authors used the conventional 50th percentile for sex and age.

Most of the study group had been diagnosed with AN-restrictive subtype (66.7%), while 16.1% had diagnoses of AN-binge-purge subtype, 5.7% were diagnosed with Other Specified Eating Disorders, and 11.5% were diagnosed with ARFID. Fifteen patients required the use of nasogastric/nasojejunal feeding tubes during hospitalization. About 75% of patients were assigned to a higher-calorie diet on admission (66 patients; range: 1500 to 3000 kcal/ day). The lower-calorie group (21 patients) was given a mean of 1185 kcal/day, and had significantly longer hospital stays than other patients (38.3 days versus 31.9 days, respectively).

The authors note that just as in previous studies, the incidence of electrolyte abnormalities in their sample was not linked to the rate of caloric increase or initial calorie level. Instead, low phosphorus levels were more common among patients who had a lower %EBW on admission. This suggested that the degree of body depletion may be a more important predictor of electrolyte abnormalities than caloric levels. In this study, with every 1% decrease in %EBW on admission, the odds of hypophosphatemia increased by 6%. However, starting severely malnourished patients at on a higher-calorie regimen was not associated with a higher risk of hypophosphatemia, hypomagnesemia, or hypokalemia.

An important caveat to this study is that refeeding occurred on a medical unit with continuous cardiac monitoring. Although the regimen was well tolerated in this sample, this degree of monitoring would not be available in most ED treatment settings. Future studies of larger patient populations will be needed to help define the safety of administering higher-calorie diets, especially to severely malnourished patients. It is important to define rates that are safe, but also as effective and promptly delivered as possible.

Tracking Recovery from AN and BN over the Long Term

Vol. 28 / No. 2 — Eating Disorders Review

Encouraging results for most, but not all, patients over more than 2 decades.

Gaining weight is a major worry among people with EDs. The Massachusetts General Hospital Longitudinal Study of Anorexia and Bulimia Nervosa has once more provided valuable information about the long-term effects of AN and BN. Dr. Kamryn T. Eddy and colleagues recently reported the results of their 22-year follow-up of women with AN and BN (J Clin Psychiatry. 2017. 78:184). At follow-up, two-thirds of the women had recovered. The Longitudinal Study was begun in 1987 with 246 women, all of who were seeking treatment for AN or BN. As part of the study, the women were interviewed every 6 months for a median time of 9.5 years to assess weekly symptoms, comorbidities, participation in treatment, and psychosocial functioning. Subsequently, 22-year follow-up was completed.

Dr. Eddy and coworkers reported that one of the indictors of long-term recovery was early recovery. At 22-year follow-up, 62.8% of women with AN and 68.2% of those with BN had recovered. This was vastly improved over the 9-year follow-up evaluation, which showed 31.4% of women with AN and 68.2% of those with BN had recovered.

Notably, about half of the women who were still ill at the 9-year point had recovered from their ED by the 22-year point. Recovery happened earlier for the women with BN, while recovery for AN patients was slower and continued over the long term.

A second study shows mortality trends 

An earlier study by Debra L. Franko, PhD, and her coworkers at Massachusetts General Hospital identified risk factors for mortality in people with AN and BN over time (Am J Psychiatry. 2013. 170:917). These researchers found that a long duration of illness, substance abuse, low weight and/or poor social functioning all raised the risk for death from AN. Among the 246 women who participated, 16 deaths (6.5%) were reported; among the 186 women with a lifetime history of AN, 14 (7.5%) died. Four of the deaths were from suicide Among the 60 women with BN and no history of AN, 2 (3.3%) had died. The authors also reported that the majority of women who had been initially diagnosed with AN experienced crossover from AN to BN, while crossover from those with an initial diagnosis of BN to AN was much less common.

A return to normal weights for most in a third study

H.B Murray and colleagues recently published very valuable long-term weight outcome results from a well-described cohort of people with AN and BN (Int J Eat Disord. 2017. Feb 11. doi: 10.1002/eat.22690).

The authors described the results of their 22-year longitudinal study of an original group of 225 patients with AN and BN. Two hundred and twenty-five were followed for 10 years and 177 were still in the study at the 22-year follow-up point. Over the course of this long study, most weight gain occurred within the first 2 years, and this lessened between years 2 and 5. Thereafter, the subjects’ weight remained relatively stable. Body mass index increased most rapidly during the early years of the study for those with lower weights at baseline (i.e., AN) and plateaued over time, settling in the normal range for most. At the 22-year follow-up point, 69% of participants were in the “normal weight range.” Seventeen percent were overweight or obese, and 14% were underweight.

Clinically, it seems as though people with AN or BN have various fears related to weight; these include the fear of ending up overweight or obese, but also the fear of gaining any weight. These results will be helpful for addressing the former fear.

As the longitudinal study continues, we will undoubtedly learn much more, and there is always the hope that recovery rates will move upward.

Tracing the Path of Disordered Eating Over Time

Vol. 28 / No. 2 — Eating Disorders Review

Four risk factors emerged in a community sample followed over 10 years. 

Little is known about how disordered eating behaviors and symptoms that appear in early adolescence may change over time. Results of a recent study gives us new information about this trajectory over time, including the fact that there is a 75% chance that those who had disordered eating would continue to have symptoms a decade later.

Carolyn M. Pearson, PhD, and collaborators at the University of Minnesota and Columbia University studied a community sample of teens over 10 years to see how risks of ED change throughout young adulthood (Int J Eat Disord. 2017. doi: 10:1002/eat.22692, published online before print). The team studied data from 2,287 participants in Project EAT, a valuable longitudinal study that has assessed eating-related, anthropomorphic, and psychological factors using surveys of male and female middle and high schools students, who are now young adults living in and around Minneapolis/St. Paul, MN. The EAT I survey analyzed teens 11 to 18 years of age and then revisited them as young adults 10 years later (EAT III).

Stability and transitions

The researchers found that those who had no symptoms in EAT I tended to remain in the same group 10 years later. For those identified in the dieting group in EAT I, dieting usually persisted. There was a 75% that chance that those with disordered eating symptoms during adolescence would continue to have disordered eating or to transition to the dieting group over the 10-year follow-up period.

Lower self-esteem, depressive symptoms, and substance use were important predictors of transition to disordered eating, and female respondents had greater odds than males of transitioning to increased symptoms a decade later. Family weight teasing also increased the odds of a teen transitioning to disordered eating; this was nearly statistically significant, and was a more powerful element than peer teasing.

Two factors were protective: higher self-esteem and family communication/caring. Some positive findings also emerged. One was that most teens who were asymptomatic at the first survey had not developed disordered eating behaviors by the 10-year follow-up. And, 1 in 4 of those with ED behaviors at the first survey was asymptomatic at the 10-year follow-up point. This suggested that some young adolescents with disordered eating do get better over time.

Screening for and intercepting risk factors early on

The study underscores the value of early detection and intervention for teens with disordered eating. Screening for psychological and socio-environmental risk factors, particularly low self-esteem, depressive symptoms, substance use, and poor family communication/caring, especially among adolescent girls, may help interrupt transition to EDs in young adulthood.

Transcranial Direct Stimulation for Bulimia Nervosa: Early Encouraging Results

Vol. 28 / No. 2 — Eating Disorders Review

A new treatment shows benefits over electroconvulsive therapy.

Neuromodulation is currently an area of great interest in the field of mental health treatment. One example of a neuromodulation technique is transcranial direct current stimulation (tDCS). Like electroconvulsive therapy (ECT), this method involves application of current to the scalp, but tDCS is quite different from ECT. In tDCS, patients are awake, there are few side effects, a very low level of current is supplied to the scalp, and no seizures follow the application of the currents. In most studies, the targeted site has been the dorsolateral prefrontal cortex (DLPFC). This region is part of the dorsal cognitive frontostriatal circuitry and is the major neural structure involved in executive functions, including self-regulatory control. It is also implicated in reward processing.

tDCS affects behavior in many different domains when applied precisely to the correct area of the scalp. In fact, previous evidence has shown some changes in eating behavior after even a single session of tDCS (Biol Psychiatry. 2010. 67:793). To further evaluate the potential of tDCS in eating disorder treatment, Dr. Maria Kekic and colleagues at Kings College, London recently conducted a study of tDCS in 39 patients (37 females and 2 males) with bulimia nervosa (PLoS One. 2017. 12: e0167606).

In this study, the subjects received either tDCS or sham application of tDCS in a double-blind fashion for 3 sessions. The results showed when a particular scalp electrode configuration was used (with the anode on the right and the cathode on the left), eating disorder cognitions were diminished, and mood was improved.

It seems likely that tDCS could ultimately be used as an eating disorder treatment that augments other treatments, rather than as a single modality. The authors are correct that this work is encouraging, and further study of tDCS is definitely needed.

Fear of Weight Gain and Dieting Affect Many Women throughout Life

Vol. 28 / No. 2 — Eating Disorders Review

A large Dutch study highlights distinctly different attitudes between men and women. 

Dieting and fear of gaining weight are common all across the lifespan for women, while a much smaller but still sizable number of men also have the same fears. These were a few findings from a recent study conducted in the Netherlands (Int J Public Health. 2017; published online February 20 before print). Dr. Margarita C. T. Slof-Op ‘t Landt and colleagues in Leiden and Amsterdam wanted to explore factors that might lead to the fear of weight gain throughout life, including gender, exercise habits, education, and lifestyle. This study is noteworthy in that it describes dieting and fear of weight gain in both men and women from adolescence to older age.

The researchers evaluated data from the Young or Adult Netherlands Twin Register, two national registers that compile information on health, lifestyle, personality and/or behavioral problems in twins and their family members. Each survey covers a 2- to 3-year period. In the youth-oriented register, twins and infants from multiple births (such as triplets) are registered at birth by their parents, and surveys are then repeated at fixed ages. At 14, 16, and 18 years of age, twins, triplets, and additional siblings complete self-report surveys with their parents’ consent. For those who completed several surveys, the most recent assessment was used. The adult cohort is made up of adult twins and multiple-birth adults born between 1909 and 1996, as well as their parents, non-twin siblings, spouses, and offspring. The authors invited 47,122 subjects to participate in their study, and data were available for 31,636 participants.

Women had the greatest weight-related concerns

Across all ages, the distribution of fear of weight gain was significantly different between men and women. Most men (up to 74%) reported not being afraid of gaining weight or of becoming “fat”; in contrast, up to 46% of women had no such concerns. Most women who responded had been on a diet at some point in their lives. A fear of gaining weight or becoming fat peaked among women between 16 and 25 years of age, and these concerns continued throughout their lives. A particularly troubling finding was that many women in the 16- to 25-year age group were very or extremely afraid of weight gain despite the fact that only 12% to 14% were actually overweight.

Most respondents were female (60.2%) and twins (61.9%). More than a third of the participants lived in densely populated areas; men and women 25 to 35 years of age were most likely to live in urban settings. More than half of the study participants reported having a high education level. Underweight or normal weight was most common among 13- to 16-year-olds. Overweight was most common among men between 55 and 65 years of age and in women older than 65. The level of exercise (determined by metabolic equivalents, or METs) was highest in 13- to 16-year old females and 16- to 18-year old males, and the lowest METs scores were found in men and women over 65 years of age

The study did have a few limitations: twin samples may differ from non-twins in various ways, including the fact that 62% of birth weights are usually lower in twins than in single births, and these body mass index differences continue into adulthood (Twin Res. 2001. 4:464). Still, the results stress that weight and diet concerns continue across the lifespan.

Some Good News for Eating Disorders Patients Concerned About Weight Gain

A return to normal weights for most in a long-term study.

Gaining weight is a major worry among people with eating disorders. H.B Murray and colleagues recently published very valuable long-term weight outcome results from a well-described cohort of people with anorexia nervosa (AN) and bulimia nervosa (BN) (Int J Eat Disord. 2017. Feb 11. doi: 10.1002/eat.22690).

The authors described their 22-year longitudinal study of an original group of 225 patients with AN and BN. Two hundred and twenty-five were followed for 10 years and 177 were still in the study at the 22-year follow-up point. Over the course of this long study, most weight gain occurred in the first 2 years, and this lessened between years 2 and 5. Thereafter, the subjects’ weight remained relatively stable. Body mass index increased most rapidly during the early years of the study for those with lower weights at baseline (i.e., AN) and plateaued over time, settling in the normal range for most. At the 22-year follow-up point, 69% of participants were in the “normal weight range.” Seventeen percent were overweight or obese, and 14% were underweight.

These results will certainly be useful in helping to address and help reduce some of the fears about weight gain that people with anorexia and bulimia have. Clinically, it seems as though people with AN or BN have various fears related to weight; these include the fear of ending up overweight or obese, but also the fear of gaining any weight. These results will be helpful for addressing the former fear.

The Positive Effects of Early Treatment Response

Vol. 28 / No. 2 — Eating Disorders Review

For many patients, early response can predict remission.

Early response to treatment for an eating disorder is often linked to a positive outcome, and to remission. Changes in behavior in the first weeks of treatment, such as a reduction in binge- eating and purging behaviors, or early weight gain in anorexic patients, can predict remission after a variety of treatments, regardless of the type or duration (Appetite. 2009. 52:805). This factor also enables clinicians to personalize treatment to improve outcome.

A group from King’s College, the University of Palermo, Italy, and the Federal University of Rio de Janeiro, Brazil, conducted a systematic review and meta-analysis of the literature to see how predictive early response actually is for patients with anorexia nervosa (AN), bulimia nervosa (BN), binge eating disorder (BED) or eating disorders not otherwise specified (Eur Eat Disord Rev. 2017. 25:67).

Early responses were keyed to the disorder

One of the initial findings reported by Dr. Bruno Palazzo Nazar and his colleagues was how the definition of “early response” varied according to the individual eating disorder. For patients with AN, the main outcome was attaining 95% of ideal body weight (IBW). In contrast, for patients with BN, it was a percentage reduction in binge eating and/or purging over a variable time period. For AN patients, early weight gains by outpatient adolescents with AN were predictive of remission in studies that used family-based therapy (Psychothotherapy, Chicago IL. 2016. 53:251) and cognitive behavioral therapy. This pattern was the same for inpatient AN treatment, where early weight gains at different time points during the first 6 weeks of therapy could predict remission (Psychother Res. 2007. 17:218; J Eat Disord Assoc. 2016. 10:1002/erv.2443). In contrast, for BN patients, the predictive pattern was far weaker, and early response was only moderately linked to better outcome. In BED studies, early response had only a fair capacity to predict good outcome.

Clinical implications of the findings

The study results indicated that failure to respond to outpatient psychological treatment during the early phase of treatment is linked to persistent symptoms in patients with AN at one year. The authors also reported there is a 77% chance that a patient who responds early will have a remission of symptoms at the end of treatment.

Among patients with BED, early response to treatment showed a high specificity and low sensitivity in predicting remission—there was an approximately 30% increase in the chances of remission in a patient with BED with an early response, and these individuals were 5 times more likely to achieve remission than were patients with slower responses (Clin Psychol Rev. 2012. 32:292).

In contrast to the positive effects of early response among patients with AN and BED, early response among BN patients did not strongly predict remission. In fact, an early response to treatment by patients with BN increased their chances of remission only by 15%.

These results should spur interest in identifying modifications to existing treatments to increase the likelihood of early response.

An Option for Those with Severe Anorexia Nervosa: Deep Brain Stimulation

Vol. 28 / No. 2 — Eating Disorders Review

After 1 year, AN patients had weight gains and improved mood.

Deep brain stimulation, or DBS, has been helpful for a number of neurological disorders, including Parkinson’s disease, and has been studied for the treatment of refractory psychiatric disorders, including major depression and obsessive-compulsive disease (OCD). Similar to a cardiac pacemaker, DBS uses implanted electrodes to send low-level electrical currents to specific structures deep within the brain. Implantation of the electrodes into the brain requires precise surgery and careful brain mapping for each individual.

Nir Lipsman, MD, and a team at the University of Toronto recently reported that DBS of the subcallosal cingulate, deep in the brain’s white matter, can significantly improve anxiety and depression, along with affective regulation, and can help increase body weight for patients with extreme treatment-resistant AN. This is the largest study thus far of DBS used for patients with long-term treatment-resistant AN (Lancet Psychiatry 2017; Feb 23, published online before print). All the 16 women in the pilot study were markedly underweight at the beginning of the study and had AN for a mean of 18 years. Most (88%) of the women also had comorbid mood and anxiety disorders.

Analyzing the effects one year after surgery

The Canadian researchers reported that DBS can affect the brain circuitry that might maintain anorexia, making the disorder so resistant to treatment. In their pilot trial of 16 women with severe, highly treatment-refractive and chronic AN (mean age: 34 years), one year after the surgery, DBS of the subcallosal cingulate was found to be safe. The procedure also improved mood and, over time, BMI as well. The implants provided continuous stimulation of the subcallosal cingulate for 12 months.

Even though most of the women tolerated the procedure and implants well, nearly half had adverse effects. Two women asked that the devices be removed or deactivated. The most common adverse effect was pain related to the surgical implantation itself or to positioning of the electrodes. Seven, or 44%, of the subjects had serious adverse effects, including electrolyte disturbances, but most appear to have been related to AN itself (all patients had severe and unrelenting symptoms). One woman had an unexplained seizure several months after her electrodes were implanted.  Importantly, the overall tolerability of the procedure seemed similar to what has been previously observed in depression and OCD, despite the precarious physical state of people with severe AN.

Eleven of the 14 women who completed the study had an increase of at least 1 BMI unit after 12 months of DBS, and the average BMI after 12 months was 17.34 kg/m2, vs 13.83 kg/m2 at the beginning of the study. DBS also improved depression and anxiety and emotional dysfunction scores. The mean Hamilton Depression Rating Scale score was 8.79 at the end of the 12 months, vs 19.4 at baseline. The mean Beck Anxiety Inventory score was 38.0, compared to 27.14 at baseline, and mean dysfunction in emotional regulation score was 131.8 vs 104.4 at baseline.

At the end of the 12 months, positron emission tomographic images showed significant changes in cerebral glucose metabolism in key brain structures, suggesting that DBS can directly affect AN-related brain circuitry. The authors noted that a longer-term study will help determine how long the positive effects of DBS last in AN patients with such severe disease.