- Top of page
The discipline of developmental psychopathology emphasizes that disorders must be viewed in terms of age and developmental level. Research on young people with eating disorders (EDs), particularly children (<13 years), is limited.[2, 3] The predominance of eating disorders not otherwise specified (EDNOS) (51–77%)[2, 4, 5] may have rendered patients “unsuitable” in research samples which historically focused on Diagnostic and Statistical Manual (DSM) anorexia nervosa (AN) and bulimia nervosa (BN). The tendency to group children and adolescents together may also have trivialized important developmental differences. Biological, social, and cognitive processes en route to adulthood conceivably influence clinical presentation of EDs.
During puberty, adipose tissue increases and hips widen to support pregnancy and lactation, changes which may move female adolescents away from the thin beauty ideal. Interpretations of pubertal bodily changes, shaped by media, peers, gender roles, and self-schema may catalyze psychopathology. Heterosocial relationships and emerging sexuality in adolescence heighten the saliency of physical attractiveness, and self-objectivity increases. Children may not possess sufficient higher-order cognitive skills to endorse pathological domains, such as shape and weight overvaluation, considered by some as the core psychopathology of EDs or devaluing the seriousness of underweight, a diagnostic criterion of AN, which involve abstract reasoning, consequential thinking, and future time perspective. Etiological contributions from genetics and environment are substantially different between prepubertal and pubertal twins, with genetic risk for the transmission of eating pathology increasing from puberty. Putatively, ovarian hormones modulate gene transcription within neurostransmitter systems implicated in EDs. Thus, developmental changes associated with puberty and adolescence may result in differentiable clinical presentations among children and adolescents.[9, 11] In the only identified study comparing cognitions of children and adolescents with EDs, Arnow et al. compared premenarcheal (mean = 10.9 years) and postmenarcheal (mean = 16.9 years) females with AN. The premenarcheal group self-reported lower eating pathology; however, parent ratings showed equivalent AN behaviors. Arnow et al. interpreted that premenarcheal patients had limited capacity to reflect on complex concepts.
Developmental differences among children and adolescents on behavioral symptoms are apparent clinically. Children with EDs are less likely than adolescents to engage in binge eating and purging. Parental oversight is typically greater for children, precluding privacy to carry out these behaviors, children lack financial resources to access binge foods and weight loss aids and may lack sophistication of thinking to understand the role of these behaviors in weight reduction. Australian and British clinical populations (<14 years) most commonly present with “restrictive” eating behaviors and food avoidance (96–98% cases).[2, 13] In contrast, 43% report excessive exercising, 19% self-induced vomiting, 5% binge eating, and 1% laxative or diuretic misuse. Incidence rates show that younger individuals are less likely to be diagnosed with BN spectrum disorders, because of the absence of binge eating and compensatory behaviors. Animal models have shown that excessive exercise is perpetuated by starvation accomplished through severe dietary restriction and a normative decline in exercise participation among both sexes from puberty, which is inversely related to level of body dissatisfaction. A tentative conjecture is that the intermittent binge eating and social changes around exercise may offer relative protection to adolescents with EDs from compulsive exercise, a form of exercise with risks to physical and psychological health and linked to poor prognosis.
Biological and course of illness factors unique to children may translate to higher physical and medical risk when presenting for care when compared with adolescents. Medical instability linked to malnutrition in EDs has been defined by hypotension, bradycardia, hypothermia, poor peripheral perfusion, and amenorrhea.[16, 17] Children with EDs have been reported to lose weight at a faster rate,[3, 18, 19] present to care at a lower percentage of ideal body weight, have feeding patterns with more consistent self-starvation devoid of intermittent binging, and, biologically, have less adipose tissue than adolescents to buffer negative impacts of starvation. In a surveillance study of children aged 5–13 years presenting to specialist and nonspecialist mental health services with an early-onset ED, hypothermia (33%), bradycardia (40%), and hypotension (20%) were prevalent. A study of females with AN aged 10–20 years seen at a tertiary hospital-based ED service found higher prevalences of hypothermia (41%), bradycardia (94%), and hypotension (70%). These studies are not directly comparable because of diagnostic and service setting differences. A comparison controlling for service setting is needed.
Overall, little research exists regarding differences in clinical presentation between children and adolescents with EDs and in nontreatment trial samples. To the authors' knowledge, the study by Peebles et al. is the only existing study that examined differences in diagnosis, behavioral symptoms, and weight parameters. The study found out that children were more likely to be male and more likely to be diagnosed with EDNOS, lost weight more rapidly, and presented at a lower percentage of ideal body weight. Adolescents had a significantly longer duration of disease, were more likely to report purging, binge eating, diet pill, and laxative misuse, and were more likely to be diagnosed with BN. Research on course of illness suggests that the BN finding is likely to hold with DSM-5; however, differences in EDNOS ought to have diminished with removal of the amenorrhea criterion and emphasis on clinician observation and collateral reporting to enunciate psychopathology (i.e., parents). Although informative, this line of research could benefit from replication and extension. Thus, we examined the differences in psychopathology and aspects of physical and medical presentation besides weight.
This study aims to compare the cognitive, behavioral, and physical presentation of children and adolescents with EDs presenting for care using a naturalistic sequential tertiary sample acquired over two decades. The hypotheses, tentatively formulated on the basis of limited empirical data, are as follows: children will be more likely to be male, less likely to have BN spectrum disorders, less likely to exercise as a method of weight control, endorse less severe eating pathology, exhibit more extreme body mass index (BMI) z-score in the direction of underweight, have lost a greater percentage of body weight, have lost body weight at a faster rate, and will have a higher prevalence of complications associated with malnutrition. Adolescents will more commonly engage in binge eating, purging, self-induced vomiting, and laxative misuse and will more commonly have a bulimic spectrum diagnosis.
- Top of page
Healthcare professionals require knowledge of developmental manifestations of illness to ensure early detection and effective care. To this end, this study examined differences in clinical presentation among children and adolescents with EDs in the HOPE Project, a naturalistic sample referred to a statewide specialist hospital-based pediatric ED service over two decades. Eating pathology, binge eating, and bulimic behaviors were less pronounced among children. Children were more likely to be male and had a faster degree of weight loss between onset of illness and admission, whereas adolescents were more likely to present with bulimic spectrum disorders.
When compared with adolescents, children reported lower scores on measures of ED cognitions, consistent with previous studies comparing premenarcheal and postmenarcheal females with AN and adolescents and adults with AN and mixed EDs. There are several interpretive possibilities. Formal operational thinking brings about changes in abstract reasoning, logical thinking, risk and future-oriented thinking, self-objectivity, and introspection and is not fully developed in children, thus children may not report traditional cognitive symptoms of EDs to the same extent as adolescents.[40, 41] DSM-5 changes, fortunately, increase emphasis on collateral reporting (i.e., parent and teacher) and behavioral observations in assessment. A recent study by our group showed that standardized parent-informant assessment can detect eating problems less readily apparent by the young person's report, particularly in cases of AN. A future research enquiry would be to evaluate the discriminant validity of the EDE among children with EDs, other mental illnesses, and healthy controls with respect to ED cognitions and to characterize its usefulness and limitations in a younger population. Health professionals of children who present with concerning weight loss or failure to make expected growth gains should be aware that limited or absent self-report of eating pathology (i.e., concern about caloric intake and body dissatisfaction) does not rule out the presence of an ED.
Surveillance studies suggest that more than two-thirds of children with EDs report ED cognitions,[2, 13] and studies of nonclinical child populations have similarly found that children experience and are capable of reporting body- and food-focused worries.[43-45] Diagnostic and assessment systems rely on eating pathology being present to extreme degrees, such as an “intense” fear of weight gain or self-evaluation with “undue influence” of body weight and shape.[23, 46] Eating pathology may be absent or relatively lower for children until puberty and adolescence. These pivotal developmental periods may “switch on” genes implicated in EDs, as suggested by findings based on twin research and animal models, or bring about physical changes with social-cognitive contingencies. This study supports recent discussion that highlights symptom severity thresholds for diagnosis as too high for children and links this to delays in diagnosis. Worth considering is the idea that professionals and adults in general have applied a weight- and shape-centric model to the explanation of EDs, whereas presentations such as avoidant/restrictive food intake disorder or those undocumented may genuinely be free of or involve limited shape- and weight-related disturbance. Alternative explanations that may account for lower scores on eating pathology measures among children are stronger need for social approval, children's tendency toward more positive self-appraisals, more egosyntonic presentations among children with EDs, and lower desire for help as younger people are often compelled to treatment by their parents.[12, 39]
Children were significantly less likely to binge eat (10% vs. 20%) or purge (7% vs. 36%), supporting study hypotheses and in line with previous studies.[3, 12] We anticipated that children would be more likely to engage in exercise for weight and shape control, but this was not supported. Binge eating is a major trigger for compensatory behaviors and more common among adolescents with EDs, which could explain this finding. Furthermore, a new exercise routine among adolescents is likely to be viewed positively in the beginning and may be difficult to curtail once established; parents often have reasonable control over children 12 years or younger. Future research could improve upon this study by using a more sensitive measurement of exercise, such as an actigraph, or measures that better represent a theoretically based definition of exercise in EDs. The rates at which children engaged in disordered eating behaviors were similar to previous studies that collectively found intense exercise for weight or shape control to be the most common compensatory behavior and binge eating and laxative and diuretic use to be the least common within this younger population.[2, 13] Given that children are less likely to binge eat and purge, these findings emphasize the importance of targeted health assessment of behaviors more common to younger children, such as food restriction and compulsive exercise. There is no suitable measure of compulsive exercise that has been validated among children and adolescents with EDs.
Children were expected to exhibit more physical compromise, such as more extreme BMI z-scores in the direction of underweight when compared with adolescents, given with previous observations of children and adolescents with varied EDs and AN.[2, 6] However, this finding was not supported: children and adolescents had BMI z-scores in the direction of underweight, but average scores were not significantly different. The rate of complications associated with malnutrition was not significantly different either. In retrospect, hypotheses regarding medical compromise were made on the basis of studies that investigated physical complications in children and adolescents as separate populations and on the basis of an expected higher emaciation in children and faster weight loss, which are established risks for physical compromise. Purging behaviors are also risks for medical instability and may have reduced group differences.
A strength of the study was the use of a gold standard clinical interview, the EDE, to assess ED cognitions and behaviors. Clinical interviews are rarely used in research because of time and resource restraints. A study limitation is the unknown factorial validity of the EDE in a pediatric ED population. Approximately one-fifth of our child sample had reached menarche, a recognized sign of entry into adolescence. Our age-based classification has limitations; future studies which take into account biological indicators of maturity may further the knowledge base. Studies have suggested that loss of control rather than the size of the binge episode is the salient feature of binge eating; however, we were unable to evaluate loss of control through HOPE Project data, but encourage future research to do so.
The findings of this study, in addition to those comparing youth and adults, indicate that the clinical presentation of EDs differs across age groups. Findings underscore the importance of further research into developmentally appropriate assessment and treatment, and multi-informant assessment and flexible diagnoses for clinical purposes.