Do previous pediatric inpatient interventions predict better outcomes for psychiatric inpatient treatment of anorexia nervosa?

Introduction

Anorexia nervosa (AN) is a complex psychiatric disorder with extensive physical and psychosocial consequences. The disease has one of the highest morbidity and mortality rates amongst psychiatric disorders, is often characterized by co-morbidity with other physical and mental illnesses [1], and its peak occurrence is during adolescence [2], 3]. AN has a chronic course with repetitions and high relapse rates [1].

The etiology of the disorder is complex and multifactorial; however, there is a distinct genetic component [4]. Risk factors include anxiety, perfectionism, cognitive rigidity, and early feeding problems [5] but other known risk factors for eating disorders, such as sexual abuse or assault, have not been specifically established for AN [6]. It seems that knowledge, experience, and understanding of this disorder are suboptimal in many cases, along with a lack of understanding of the associated risks [7].

Epidemiologically, while the general incidence of the disorder has been relatively stable in recent decades, there is evidence of increasing incidence amongst adolescents (under 15 years) and young women [8], 9]. Incidence also increased during the COVID-19 epidemic [10]. Whether these changes were due to earlier onset or diagnosis, the above findings are of great importance for understanding risk factors and establishing early prevention plans for AN. The growing number of incidents is causing a significant increase in hospitalizations and therefore requires more efficient and precise treatment.

In terms of prognosis, the current estimation is that ∼50 % of individuals with AN will attain full recovery, 30 % improve, and 20 % remain with this chronic disorder [11].

Beyond the mortality caused by the disorder, many patients suffer from physical or psychosocial morbidities and thus require high levels of care for their medical or psychological stabilization, such as inpatient or day patient treatment. A study that followed female patients for an extended period of 9–14 years showed that more than 20 % of female patients were unable to support themselves independently [12]. Another study of a large cohort of cases showed poor psychosocial outcomes even at age 35 years [13]. Severe cases of AN require hospitalization, which is a great burden for patients, their families, and the medical system, from both a therapeutic and psycho-social standpoint.

It is important to acknowledge that patients under the age of 15 years are at higher risk than older age groups of hospitalization during the course of the disorder [14]. In recent years, there has been a significant increase in the hospitalization rate of this pediatric age group (e.g., Germany [15] and England [14], 16]). This escalation may be attributed to the higher incidence of AN in recent years (as previously mentioned), increases in its severity, or greater awareness and diagnosis among primary care physicians [17].

Given that the main causes for inpatient treatment are extremely low weight and rapid weight loss, one of the most common treatments required during hospitalization is refeeding [18]. Refeeding is comprised of rapid feeding and physical stabilization with a full menu adjusted by a specialist dietician that includes caloric calculation adjusted for age, gender, weight, and activity level.

In recent years, there has been a debate on the optimal speed of refeeding. On the one hand there are arguments in favor of slower feeding aimed at preventing refeeding syndrome, while on the other hand previous studies have demonstrated rapid improvement in vital indices with a shortening of hospitalization days under a feeding regime with more calories at the beginning of hospitalization as compared to slow feeding [18], 19], without adverse events. This suggests a shorter hospital stay is a favorable course of action.

Different departments use diverse approaches regarding feeding options for girls who refuse to cooperate with the meal plans. In some cases, it is customary to convert food into high-calorie formulas such as Ensure^® [18], while in others and at our institution tube feeding is frequently used. It is essential to note that some studies suggest that rapid feeding methods, as detailed above, may harm the interpersonal relationship between the therapist and patient [20].

Weight restoration is a central goal in the hospitalization of patients with AN. Therefore, various variables related to weight have been examined in many studies. For example, BMI calculation at admission for hospitalization and at discharge, reaching a target weight set in the plan at admission, and maintaining a target weight for a specified length of time before discharge [21], 22]. In recent years, research has increasingly concentrated on weight gain during hospitalization, with studies examining the impact of weight gain while hospitalized.

A 2014 RCT study found that rapid weight gain was associated with longer remission [23]. Wales et al. (2016)’s study of 102 hospitalized patients with AN showed a distinct prognostic advantage for patients who had gained weight during a shorter period of time [24].

Pemberton et al. (2013) described how hospitalizations of adolescents with eating disorders are complex with many aspects: physiological, psychological, social, and more. While some patients require treatment that is mainly physiological due to heart rhythm disturbances, electrolyte disturbances, and more, others require treatment focusing on psychological or behavioral aspects [25]. Therefore, some female patients at our institution were admitted to a psychiatric inpatient unit, where they receive treatment from a multidisciplinary team. The mainstay of this admission is setting a target weight and establishing an orderly treatment plan to achieve it alongside other therapeutic goals. It should be noted that the psychiatric unit is an open and voluntary inpatient facility which relies on patients’ motivation, and there is no use of feeding tubes, only menus.

However, some patients are first admitted to the pediatric ward, due to urgent physiological needs, for predominantly physiological treatments by a pediatric team. A dietician prepares a menu, after which weight, vital signs, and laboratory tests are closely monitored. The usage of a feeding tube is optional, depending on severity.

Due to significant challenges associated with the pediatric hospitalization of adolescents with eating disorders that are often accompanied by adverse effects and other complications, we aimed to examine the benefits of pediatric hospitalization that preceded hospitalization in a psychiatric inpatient unit that specializes in eating disorders. To the best of our knowledge, this is the first study to examine the benefits of both short- and long-term pediatric hospitalization prior to psychiatric hospitalization.

Our hypothesis holds that the purpose of hospitalization among adolescents with AN in the pediatric department is primarily to address urgent physical needs, such as bradycardia and arrhythmias, hypoglycemia, various electrolyte disorders, kidney failure, and more. We hypothesized that addressing these needs as quickly as possible and shortening the hospitalization time in the pediatric wards, as compared to longer hospitalizations before transfer to the psychiatric inpatient unit for further treatment, would lead to improved prognosis and reduction of peripheral hospitalization damages. We further hypothesized that patients admitted to the psychiatric department following pediatric hospitalization would have a shorter hospitalization period and better prognosis (in terms of achieving their target weight), as compared to patients referred from ambulatory care.

Methods

Results

In this retrospective study, 145 patients met our inclusion criteria and 41 were excluded due to missing data (Figure 1). The 104 patients included in the study, aged 12–18 years (M=15.4, SD=1.5), were divided into two main groups: pediatric group with patients previously admitted to a pediatric ward (n=68) and Ambulatory group with patients referred directly from ambulatory care (n=36).

Figure 1:

Flow of participants into the study.

T-tests were used to assess differences between the two groups regarding epidemiological variables, weight variables, and vital signs on arrival from the pediatric ward. The results are shown in Table 1.

No differences were found between the groups with regards to age of onset and age at hospitalization. However, there was a longer mean duration of morbidity in the Pediatric group as compared to the Ambulatory group (15.00 ± 12.11 vs. 19.89 ± 10.94, respectively, p≤0.05). The Pediatric group also had more mean previous admissions than the Ambulatory group (0.87 ± 0.87 vs. 0.32 ± 0.053, respectively, p≤0.001). Regarding vital signs, no significant differences were found between the groups for all the indices tested.

Since the observed trends in variables such as BMI and heart rate changes did not reach statistical significance, we did not conduct further analysis on their potential clinical implications. However, we acknowledge that non-significant trends could still have clinical relevance, and future research with larger sample sizes may be needed to explore these patterns in more depth.

To test the relationship between type of preliminary hospitalization and success in the criterion of reaching the target weight, a chi squared test was conducted and results presented in Figure 2. No significant relationship was found between the groups and achieving target weight.

Figure 2:

Correlation between arrival from pediatric and reaching the weight target (n=87).

T tests were used to test for differences in the length of hospitalization, as shown in Table 2. No difference was found between the groups with regards to hospitalization time.

To test the relationship between type of preliminary hospitalization and presence of depression, a chi squared test was conducted. The results of the analysis are shown in Table 3. A significant dependence was found between depression and arrival from the pediatric ward, with a medium-low Cramer’s V coefficient (V=0.23), indicating the group with depression was associated with arrival from the pediatric ward.

To test for differences in the number of psychiatric medications, a t-test was conducted and results presented in Table 4 (see appendix). A significant difference was found in the mean number of psychiatric medications with the Pediatric group prescribed fewer than the Ambulatory group (0.76 ± 0.90 vs. 1.29±0.29, respectively, p≤0.05).

Discussion

The purpose of the current study was to investigate the implications of pediatric hospitalization prior to psychiatric hospitalization in adolescents with AN. We examined adolescents admitted to a psychiatric facility that specializes in eating disorders, where some patients had previously sought treatment in a pediatric ward and others in ambulatory care settings. Specifically, we examined whether pediatric hospitalization is beneficial in achieving target weights and overall treatment outcomes and prognosis as compared to patients who were not previously hospitalized in a pediatric ward.

One of the most interesting findings in this study referred to indicators of success during psychiatric hospitalization for eating disorders, which is achieving a target weight. Our findings yielded no significant difference between the groups regarding the achievement of target weight during psychiatric hospitalization [26].

Although our analysis did not reveal statistically significant differences in certain clinical parameters, such as BMI and heart rate changes, it is important to acknowledge that non-significant trends may still hold clinical relevance. For example, even small differences in weight restoration or vital signs could influence long-term treatment trajectories and overall prognosis. Given the relatively small sample size, it is possible that the study was underpowered to detect subtle yet meaningful differences. Future research with larger cohorts may be better equipped to explore whether these trends represent clinically significant effects that impact treatment outcomes.

According to our hypothesis, our findings indeed demonstrate that the major benefit of pediatric hospitalization was the immediate treatment of life-threatening conditions. The significant weight gain was carried out within the period of psychiatric hospitalization and therefore our findings indicate that preliminary pediatric hospitalization did not have any prognostic effect.

We estimate that one of the reasons for the difficulty in treating eating disorders in pediatric departments is due to the conflict that arises around the need for immediate feeding, either PO or by tube feeding. Assuming that the importance of the therapist-patient relationship is very important for this type of illness [27], it seems imperative to separate the purely physiological treatment from the psychiatric treatment and supportive treatment.

We found that patients required to be hospitalized in a pediatric ward prior to psychiatric hospitalization had been diagnosed with AN for a shorter time than those admitted to psychiatric hospitalization from any ambulatory setting. This may reflect cases in which deterioration is rapid, with insufficient time to start an ambulatory process. In addition, it was found that patients who required pediatric hospitalization had a greater number of past hospitalizations. However, despite these findings, we did not find any significant epidemiological differences between the groups. Patients in both groups were hospitalized at a similar age after being diagnosed at a similar age. In terms of vital indicators (heart rate, blood pressure, weight, and BMI) no significant differences were found between the groups. Also, no significant difference was found in the degree of change between these indices at the beginning of the hospitalization in the psychiatric ward and at discharge. In terms of comorbidities, patients requiring pre-hospitalization in a pediatric ward were diagnosed with depression at similar rates to those not hospitalized in a pediatric ward; however, not due to anxiety or previous attempts at self-harm.

We acknowledge the importance of exploring potential mechanisms underlying our findings. One possible explanation for why pediatric hospitalization did not improve outcomes is the increased psychological stress associated with medical hospitalization. Hospitalization in a pediatric ward often involves strict monitoring, medical interventions (e.g., tube feeding), and loss of autonomy, which could contribute to increased anxiety and resistance to treatment. Additionally, the treatment milieu differs significantly between pediatric and psychiatric units; pediatric wards primarily focus on medical stabilization, whereas psychiatric inpatient settings emphasize behavioral, psychological, and multidisciplinary therapeutic interventions tailored to eating disorders. It is possible that prolonged stays in pediatric wards delay engagement with specialized psychiatric care, limiting opportunities for early therapeutic intervention. Further research is needed to examine how different hospitalization settings impact long-term recovery trajectories in adolescents with AN.

What is already known on this subject?

Some adolescents with anorexia nervosa are first admitted to pediatric departments due to severe acute medical complications, such as bradycardia, hypoglycemia, and electrolyte imbalances. However, it remains unclear whether pediatric hospitalization contributes to eating disorder recovery or if it merely stabilizes immediate medical risks. The long-term impact of pediatric hospitalization compared to direct psychiatric admission has not been well studied.

What this study adds?

This study provides evidence that pediatric hospitalization does not improve long-term recovery compared to direct psychiatric admission. While hospitalization is necessary for acute medical stabilization, it does not appear to offer additional benefits for achieving weight restoration or reducing relapse. These findings suggest that prolonged pediatric hospitalization may be unnecessary, emphasizing the need for shorter medical stays and earlier transitions to specialized eating disorder units.

Conclusions

This study focused only on psychiatric hospitalization, which provided useful insights into the importance of short pediatric hospitalization. A significant percentage of patients with AN require urgent pediatric hospitalization due to various physiological disorders. Despite the different therapeutic approaches in the pediatric wards, using an optimized menu or feeding with tube in a short or long period of treatment, no significant prognostic benefit (in terms of achieving target weight) was found for pediatric hospitalization over ambulatory care prior to psychiatric hospitalization.

As we found no actual benefits to long pediatric hospitalization over short hospitalization, we assessed hospitalization duration by defining “short” and “long” based on the median hospitalization time, within our cohort. Given the peripheral side effects of long hospitalizations in pediatric wards, we believe that it might prove beneficial to all parties considered to set the goal for as short as possible pediatric hospitalizations combined with a rapid refeeding approach.

In addition, further research is necessary regarding the negative results accompanying long hospitalizations of adolescents with eating disorders, such as deterioration of psychiatric illnesses, social harm, infections, hospitalism, and even deterioration of the eating disorder.