Eating disorders are life threatening, biopsychosocial diseases that affect children, adolescents, and young adults. Pediatricians, parents, and others who take care of children are recognizing disordered eating attitudes and behaviors in children as young as 2 years, and these occur well into the geriatric years. With the publication of the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), clarification of definitions has made it easier for clinicians to recognize the spectrum of eating disorders in their patients, with early intervention leading to improved outcomes.
Updated Definitions: What Does DSM-5 Actually Change?
Challenges to prior versions (DSM-IV and DSM-IV, Text Revision, or DSM-IV-TR) stemmed from the narrowness of criteria for anorexia nervosa and bulimia nervosa, and broadness of the definition for the former category entitled, Eating Disorder Not Otherwise Specified (ED-NOS). In one study, more than 50% of patients did not meet the DSM-IV or DSM-IV-TR criteria for either anorexia nervosa or bulimia nervosa. The process of “lumping” a diverse group of patients poses challenges for performing useful outcome studies because what works therapeutically and psychopharmacologically in patients with anorexia nervosa may not be identical to what works best in patients with bulimia nervosa, for example. Broad categorization also presented challenges for reimbursement in the United States, in which some insurers covered anorexia nervosa, but not necessarily ED-NOS, thereby limiting days of needed treatment for some patients in past years. Earlier editions also did not recognize that very young children as well as adolescent boys might develop eating disorders. The new definitions for the various eating disorders are found in Box 15-1 .
Restriction of energy intake relative to requirements leading to a significantly low body weight in the context of age, sex, developmental trajectory, and physical health. Significantly low weight is defined as a weight that is less than minimally normal, or, for children and adolescents, less than that minimally expected.
Intense fear of gaining weight or becoming fat or persistent behavior that interferes with weight gain, even though at a significantly low weight.
Disturbance in the way in which one’s body weight or shape is experienced, undue influence of body weight or shape on self-evaluation, or persistent lack of recognition of the seriousness of the current low body weight.
Specify current type:
Restricting Type: during the last 3 months, the person has not engaged in recurrent episodes of binge eating or purging behavior (i.e., self-induced vomiting or the misuse of laxatives, diuretics, or enemas).
Binge-Eating/Purging Type: during the last 3 months, the person has engaged in recurrent episodes of binge eating or purging behavior (i.e., self-induced vomiting or the misuse of laxatives, diuretics, or enemas).
Recurrent episodes of binge eating, characterized by both:
Eating, in a discrete period of time (for example, within any 2-hour period), an amount of food larger than most people would eat during a similar period of time and under similar circumstances.
A sense of lack of control over eating during the episode.
Recurrent inappropriate compensatory behavior in order to prevent weight gain, such as self-induced vomiting; misuse of laxatives, diuretics, or other medications; fasting; or excessive exercise.
The binge eating and inappropriate compensatory behaviors both occur, on average, at least once a week for 3 months.
Self-evaluation is unduly influenced by body shape and weight.
The disturbance does not occur exclusively during episodes of anorexia nervosa.
Avoidant/Restrictive Food Intake Disorder
Eating or feeding disturbance (including but not limited to apparent lack of interest in eating or food; avoidance due to sensory characteristics of food; or concern about aversive consequences of eating) as manifested by persistent failure to meet appropriate nutritional and/or energy needs associated with one or more of the following:
Significant weight loss (or failure to gain weight or faltering growth in children);
Significant nutritional deficiency;
Dependence on enteral feeding;
Marked interference with psychosocial functioning.
No evidence of lack of available food or an associated culturally sanctioned practice.
Feeding and Eating Conditions Not Elsewhere Classified
Atypical anorexia nervosa (not yet underweight).
Purging disorder (no binges).
Sub-threshold bulimia nervosa (<1x/week or <3 months).
Sub-threshold binge eating disorder (<1x/week or <3 months).
Night eating syndrome (nocturnal eating disorder).
For anorexia nervosa (AN), the illness is still defined by self starvation in order to get to and maintain an abnormally low body weight combined with an intense fear of becoming fat. Specific changes include the elimination of amenorrhea as a criterion, as it does not necessarily predict medical risk or treatment outcome. Furthermore, it does not apply to premenarchal girls or males, or postmenopausal women. In prior versions, the definition included only those individuals at less than 85% of expected weight. DSM-5 changes this criterion to low weight in context of age, sex, developmental trajectory, and physical health. In plainer words, if a child falls off a growth curve or crosses percentiles, that change should ring the clinician’s alarm bells for an eating disorder.
In DSM-5, bulimia nervosa continues to be characterized by episodes of binge eating (bulimia stems from the Greek phrase, “appetite like a bull”), followed by inappropriate compensatory behaviors to avoid weight gain. These compensatory behaviors can include vomiting, laxative abuse, diuretic or diet pill abuse, and hyperexercising (also called overexercising; in layperson’s terms, these individuals have “orthorexia”). Individuals with diabetes mellitus and anorexia nervosa or bulimia nervosa might precipitate weight loss by withholding insulin or misusing it, called in layperson’s terms, “diabulimia.” These colloquial terms may pop up in the literature but are not officially found in DSM-5, which also eliminated the specific subtypes (purging and nonpurging). The new version of DSM reduced the requirement for binges to only once weekly for three months, instead of two times or more in 3 months. The lowered frequency stemmed from the lack of difference in psychopathology or treatment outcomes between patients with once a week cycles of binges and purges versus more binge-purge episodes.
New subcategories in DSM-5 were designed to help distinguish different treatment needs and eventual outcomes between various subtypes. Binge eating disorder is defined by binges without inappropriate compensatory behaviors. Wonderlich et al. described this group as more obese, with greater functional impairment. They were also more likely than obese patients without eating disorders to progress to metabolic syndrome.
Another new category is called ARFID, or Avoidant/Restrictive Food Intake Disorder. These children, adolescents, and adults fail to meet nutritional needs for reasons other than weight control. They may begin to avoid food after a choking or gagging incident, with fear of repeating that sensation. Another child or adolescent may develop ARFID after a vomiting illness, with fear of ever vomiting again. Abdominal pain, dislike of sensory characteristics of food, and other factors may keep a child from needed intake for growth. Unlike the child with anorexia nervosa, these children do not state that they fear weight gain or do not want to grow; rather, they fear the consequences of eating (abdominal pain, eating a food the picky eater would rather avoid, or other challenge). The prior DSM version used the term “feeding disorder of infancy or early childhood.”
In a multicenter study of 215 consecutive patients (mean age 15.4 ± 3.3 years) who presented to adolescent medicine physicians for initial evaluation of an eating disorder, DSM-IV-TR and DSM-5 diagnostic categories were used to distinguish how patients would be categorized with the proposed DSM-5 criteria. Utilizing the updated criteria, 40% of patients were categorized as anorexia nervosa (increased from 30% with DSM-IV-TR), 11.8% as bulimia nervosa (increased from 7.3% using DSM-IV-TR), and 15% were categorized as ARFID using DSM-5 criteria. Cases of ED-NOS decreased from 62.3% using DSM-IV-TR to 32.6% (p < 0.001). At the time, ARFID was being considered as a change that was useful for the next iteration, or DSM-6; this study distinguished this category as a separate entity, especially because children can run into medical trouble as they fall off the growth curve or lose even a small bit of weight, and failure to recognize this drop can cause delays in recognition and needed treatment. Furthermore, this study had more pronounced findings than those found in a community sample of adolescents or in adults. The hope is that re-categorization into the new DSM-5 subtypes will help clinicians and therapists choose the right interventions in a timely fashion to optimize treatment outcomes.
Recognizing Eating Disorders in Children and Adolescents
Eating disorders may often be missed or diagnosis delayed by primary care providers and subspecialists. Pediatricians need to take seriously any concern from a parent or other caring adult that a child or adolescent may be developing an eating disorder; many of these patients either already have or are in the process of hardwiring maladaptive eating attitudes and behaviors that could be prevented or treated with earlier intervention. Eating disorders are affecting youth of all races, ethnicities, and socioeconomic backgrounds; children living in poverty with these conditions continue to be underrecognized, with challenges including lack of access to care as well as assumptions that eating disorders only occur in upper middle class communities. Eating disorders may also be “invisible” in obese populations, where individuals may be cheered on for decreasing weight from 220 pounds to 180 pounds, even if they have done so with significantly disordered eating; they may be at as much risk medically as a teen decreasing weight from 120 pounds to 80 pounds, but in the latter case, the family and physician is more likely to recognize a problem and react appropriately.
Disordered eating and the female athlete triad are not just contained to the “visual sports” such as ballet, figure skating, and gymnastics; individuals can develop the triad who are C students, do P90X or find their favorite gym over exuberantly. Children and adolescents have been found to run in place or in school bathrooms to the point of developing blistering heels and toes, stress fractures, and other physical findings.
How Can You Tell? Predictive Factors for Eating Disorders
Allen et al. found in an Australian population-based sample that being female as well as a parent’s perception that the child is overweight both strongly predicted the development of an eating disorder. Other predictive factors include having a mother who is overweight (defined by elevated body mass index, BMI), low social-related self-efficacy in the child or teen, or the presence of social problems and/or neurocognitive difficulties in the child. Box 15-2 outlines potential risk factors. Diane Neumark-Sztainer suggests avoiding “weighty conversations,” or having home or dinner table talk centered around dieting or weight problems. In fact, parental concerns about their own weight, shape, or eating habits are predictive of the child’s abnormal eating attitudes and behaviors. Furthermore, comments from parents or other “caring” adults on the child’s weight or shape have not improved outcomes in children with eating disorders. As Dr. Neumark-Sztainer states, if a child is a poor reader, calling them stupid will not help them read better and in fact may delay the process. Similarly, chiding a child for being overweight or creating derogatory weight-related nicknames will not help them better manage their weight, exercise, or meal planning.
Pregnancy with birth complications
Weight-related teasing of the child
Low degree of parental contact
High parental expectation
Parental alcoholism or drug use
Being perceived as overweight
Abuse (sexual, physical, emotional)
In adolescent medicine, we prefer to treat “diet” as a four-letter word; adolescents who “diet” are more likely to end up at a higher weight, with short losses replaced by sustained weight gain. For the adolescent with disordered eating attitudes and behaviors, the clinician can use epigenetics as an explanation; “if you fast or have insufficient energy intake, your body thinks it has to store up nuts for the winter, or have enough energy in the form of fat or other tissues in order to last through the next perceived famine, or fasting period.” Thus, the chronic dieter may end up continuing to gain, above where their natural set point should be. This rationale makes sense to many patients and families and can help them start to break a cycle of abnormal eating attitudes and behaviors. Similarly, in patients with anorexia nervosa who state that they are just “fine,” yet who have significant medical findings such as bradycardia, orthostatic hypotension, and amenorrhea, epigenetics can serve as a rational explanation of why these changes happen. The clinician might say, “You were supposed to be one of Darwin’s fittest, able to survive an internment camp on next to nothing nutritionally. Now allow your body to perceive that the war is past.” For the female athlete triad patient, “You were designed to be able to outrun the saber-toothed tiger and then migrate to the fertile plain without conceiving along the way; your body adapts to its current state of undernutrition yet does not prefer it, with consequences to brain, bone, and heart that are reversible with refeeding.”
Taking a History
Detailed aspects of the history can be seen in Box 15-3 . If a child falls off the growth curve or crosses percentiles, the astute clinician should assess immediately for disordered eating attitudes and behaviors, while simultaneously doing a review of systems that could tease out newly manifesting celiac disease, inflammatory bowel disease, or other systemic process. Child abuse—physical, emotional, or sexual—may manifest as inappropriate weight gain or loss. Establishing confidentiality is essential in obtaining an accurate diagnosis.
What is the most you ever weighed? How tall were you then? When was that?
What is the least you ever weighed? At what height? When was that?
What do you think you should weigh? How much time and/or energy do you put into that?
What weight feels too high? What feels too low?
For girls: When was your last period? What weight were you then? When was the period before that? Have you missed any periods? How old were you when you started periods? Have they ever been regular?
What do you do for exercise? How intense are you when you exercise?
How stressed are you if you miss a workout/exercise?
What did you eat yesterday? Assess quantity and quality for breakfast, snack, lunch, snack, dinner, and snack. Also, assess fluids consumed, and dairy/calcium/vitamin D servings.
What foods do you not eat? Are there any foods you used to eat that no longer feel “safe” or taste right to you? What foods?
Do you count calories and/or fat grams? If so, how many do you allow yourself?
Do you ever make yourself vomit? How often?
Do you ever use diet pills, laxatives, diuretics, other methods? If yes, how many of which kinds, how often, over what time period?
Has anyone ever bullied or teased you? About weight/body, or other teasing? Were the bullies at school? At home? Adults, peers, other?
How much of your day is food or body thoughts?
Depression? Have you ever wished you didn’t exist? If yes, how often does that happen? Any thoughts, “I am going to kill myself?” If yes, how often? Have you ever had a specific plan? What were those plans? Did you ever try? When was that? Who knew about that, if anyone?
Anxiety? Panic attacks? How often?
Obsessive-compulsive tendencies? Can you describe?
HEADS (Home, Education, Activies, Drugs, Sex) questions:
Home: Who lives at home? What happens when there is an argument in the home?
Education: What grade are you in school? How are your grades this year? How were they last year?
Activities: How do you fill your time?
Drugs: Do your friends smoke cigarettes? Do you smoke cigarettes? How much, how often? Do your friends use drugs? Do you use drugs? Over the counter, prescription, marijuana, other? Do your friends drink alcohol? Do you drink alcohol? What do you drink? How much, how often? Any pre-gaming (chugging/doing shots of alcohol before going out)?
Sex: Are you attracted to guys, girls, or both? Have you ever had sex? With guys, girls, or both? If heterosexual and has had sex, what do you use for your two methods of contraception? Do you use condoms sometimes, most of the time, or all the time? Has anyone ever done anything to you sexually that made you uncomfortable?
Clinicians can start with parent(s) and child together, asking the child or adolescent why they are seeing you today. How the child and parent responds can be telling; does the teen say, “I don’t know,” and look angrily at the mom? Does the teen think they are there just for a checkup? Is loss of periods a concern to the mother but not the child? Listen to what they say, and do not say, with both parent and child in the room. The clinician may need to referee, or curtail the intrusive parent who will not let the teen get a word in edgewise, or, for the quiet child or teen who will not say a word, ask their permission to elicit information from the parent, by stating, “Is it okay if the peanut gallery (your parent) answers this one?” This can help direct the questioning to assess the assumptions of both what they have shared and not shared with each other. The clinician can ask how the child feels about his/her body. Do they feel overweight, underweight, or just right? Has he/she ever tried to change his/her weight or shape, and why? Any trouble spots that particularly bother him/her? (If the clinician gets a blank stare, he or she might state that some kids worry a lot about stomach, or hips, or arm, or hair, and see if that elicits a response.) How much effort he/she puts into food and body can be telling; the child with an eating disorder may reveal that it is on their mind all day, and it gives the clinician a chance to recognize otherwise hidden abnormal eating attitudes and behaviors while simultaneously acknowledging the stress he or she may feel about weight or body issues. In contrast, the child with ARFID may experience no body stress but may talk more about fear of abdominal pain, or challenges with eating new foods, or other processes. A child with obsessive-compulsive disorder may reveal a fear of having foods touch on the plate, or of eating foods not made at home by a trusted source. Children with underlying anxiety disorder may fear a perceived risk, such as vomiting unintentionally or other aversive conditioning.
Asking about diet history and exercise can be particularly useful with both parent and child together, as “the peanut gallery” may use body language or words that can be queried to clarify whether the child is eating just bites of cereal or half a box. Can the teen eat at restaurants or friends’ houses? What are family meals like? Near holidays, asking about what happens at the family dinner and how stressful or easy it is for the child can be revealing. Who is in charge of food buying and meal prep? How often do family dinners occur? What patterns have the family noticed? Many patients laud the new interest a child or adolescent may have in eating healthy, especially if weight has been a family concern or a parent/family member has had a heart attack, diabetes, or other complication of obesity. Cutting out junk food, engaging in more exercise, and other actions may be reinforced at school and home as “being healthy,” with the family unaware that the child has cut out all or most of a food group or has been skipping meals in or out of the parents’ sight. Parents faced with a new vegetarian or vegan may not understand how to feed their child appropriately yet may be trying to be respectful of their desire not to eat meat after viewing content about animal rights in the classroom, or other scenario. The shift may be so gradual that the parent may be caught unaware until shopping for clothes with their child, or a spring break or family vacation when the body changes are more noticeable, or in the pediatrician’s office seeing the growth curve.
It is useful to obtain a family history with child and parents present, educating about risk for the adolescent, in particular. A parent’s prior eating disorder history may never have been shared; some parents will save that history until they get private time with the clinician, thus making it important to carve out time with both parent and child separately. Helping to open the lines of communication can also diminish the negative power of “family secrets.”
In establishing confidentiality, the clinician can state that he/she wishes to take time alone with the child/adolescent and also with the parents, in order for the child/adolescent to be able to ask private questions, seek information, and to reinforce skills of being able to speak up for themselves/be a good health care consumer. The clinician needs to state clearly the rules of confidentiality and its exceptions, for instance, “Everything we talk about without your parent in the room is private or confidential, meaning I am not going to share your private concerns with your parent unless you tell me something life-threatening or dangerous. In that case, I would need to tell your family in order to help keep you safe. Similarly, everything I talk about privately with your parent will be kept confidential, meaning I will not share with you their private thoughts or concerns, unless they tell me something life-threatening or dangerous.” Adolescents given the definition of confidentiality as well as its limits are more likely to share confidential information and may be more receptive to interventions.
The HEADS (home, education, activity, drugs/depression, sex/suicide/safety) questions can be used to assess level of school and family functioning (see Box 15-3 ). The child or adolescent with an eating disorder may be avoiding situations where food is involved, isolating themselves socially from family or friends. Obsessive-compulsive behaviors may have been missed; asking if the child is a perfectionist can be illuminating. Does he/she take excessive time to do homework (e.g., taking 10 hours on a project that ought to take 1 hour), or have to keep books/homework a certain way? Did she/he sort all the purple crayons from the green ones as a young child? Any germ phobia, counting a certain way, or other rituals? If rituals are only about food and body, this finding may not represent true OCD, but rather starvation-induced brain changes that may fade without being replaced when the brain is better fed. Eating disorder questions may need to be asked privately, as some patients go to great lengths to cover or hide their behaviors. Does he or she make him or herself vomit? If so, using a finger? Spoon? Tooth brush or other object? Use of motivational interviewing can elicit history while also helping to manage change and reducing morbidity. For instance, in discussing vomiting habits, the clinician can ask how often they purge, and in what situations. If it is always after school at home alone, is there a place the teen could go at that time that he or she could keep from vomiting (e.g., going to the library to study with a finite snack rather than an open kitchen ready for a binge followed by purge). Does he brush his teeth after purging, or gargle and spit? Teaching the purging adolescent to gargle with mouthwash, water with baking soda, or even just plain water is safer than brushing, and the teen can be taught that the latter helps acid reach surfaces it may otherwise not have reached. Most teens will begin behavior modifications such as these, in simple steps, if given the opportunity.
Review of systems can include general questions as found in Box 15-4 . Signs of chronic disease such as inflammatory bowel disease, thyroid symptoms, or other process can be useful.
General: low energy, fatigue, exercise intolerance?
Cardiovascular: chest pain? Dyspnea on exertion? Palpitations? Lightheaded?
Gastrointestinal: early satiety, bloating, fullness, abdominal pain, reflux, constipation, diarrhea?
Genitourinary: oligomenorrhea, amenorrhea, low libido, erectile dysfunction?
Endocrine: cold intolerance?
Skin changes: hair loss, poor wound healing, acrocyanosis, dry skin?
Psychosocial: depression, anxiety, OCD, changes in relationships and school/work performance?
The physical examination should include weight in a gown after voiding, facing backward on a scale (“blinded” weight), height, and orthostatic pulse and blood pressure measurements ( Table 15-1 ). Blinded weights can be useful for the child or adolescent who is numbers obsessed or to help refocus on health rather than the scale. Some patients do better knowing the numbers, while others fall to pieces and escalate their behaviors before and after visits to manipulate the scale. Until you know how your patient will react, proceed with caution on sharing numbers rather than trends. Nursing staff may need to be trained not to react to what they see on the scale; comments such as, “Oh that’s great, honey, well done!” can trigger an adolescent with anorexia nervosa to return 10 pounds lighter at the next visit. Nursing staff can help orient the patient to the process, with words such as, “It is our policy to weigh patients backward on the scale after voiding, and screen for diabetes, protein in the urine, or other problems, and we will save all discussion of findings for you and your doctor.” Before weighing, certain patients may require “pat down” or weighing with no underwear on, as a significant amount of creativity can be discovered in the patient motivated to manipulate weight, with weights sewn into bra linings, in socks, or in underwear. Urinalysis with specific gravity of 1.005 or less may suggest water loading, whereas a specific gravity of 1.030 may suggest fluid restricting. Ketosis and proteinuria may also be detected. In the face of extreme dehydration and/or use of caffeine/stimulants, patients who should be bradycardic from starvation may actually have a normal heart rate or relative tachycardia. Certain patients with anxiety or an anxiety disorder may also have a “fight or flight” response, with resultant tachycardia masking a loss of heart mass.
|Skin||Dry; atrophic with restricting; Russell sign with purging using a finger; easy bruising|
|Mouth||Dental erosions on lingual and occlusal surfaces from vomiting|
|Face||Parotitis in those who purge|
|Thyroid||Not enlarged; but euthyroid sick syndrome notable with dry skin, cold intolerance, constipation|
|Cardiac||Bradycardia; orthostatic hypotension; mitral valve prolapse (one-third of patients)|
|Gastrointestinal||Palpable loops of stools when constipated, diffuse abdominal discomfort|
|Extremities||Acrocyanosis, Raynaud’s syndrome, edema with third spacing|
|Hair||Loss of shine or thickness; hair loss on head; lanugo on trunk/extremities|
|Central nervous system||Nerve compression/sciatica from lack of padding|
|Musculoskeletal||Stress fractures; osteopenia|
|Genitourinary||Atrophic vaginitis; loss of libido|
On head to toe physical, the astute clinician may observe parotid gland swelling in patients who vomit; this finding tends to worsen before it gets better, similar to smoker’s cough after cessation of smoking. Adolescents should be warned about this phenomenon when working on quitting purging behaviors. If they know it is not forever, they may tolerate it better in the short term. It is important to warn them that while quitting, each purge may cause a flare of parotitis—and that it is not just their face getting fatter, as many fear. Dental erosions on the lingual and occlusal surfaces can also be found with vomiting. Russell sign, or callous on the knuckle from hitting it with teeth during self-induced vomiting, can be seen in Figure 15-1 . Hair thinning, lack of shine or fullness, can occur with insufficient protein and fat intake. Lanugo hair may be found on the trunk and extremities when starvation has occurred of sufficient duration for the body to sense the need for an extra-induced layer of warmth. Carotinemia can occur in those patients eating more orange vegetables than anything else. Acrocyanosis, peripheral edema, and Raynaud’s can be explained to the patient as the body trying to preserve needed energy for the major organs such as heart and brain, shunting nutrition and circulation from the periphery in order to preserve the core. Telltale effects of purging can also include subconjunctival hemorrhages (red spots in the white of the eye) as well as unexplained nosebleeds without associated nasal pathology.
Cardiac findings include bradycardia and orthostatic hypotension, the latter defined as diastolic blood pressure dropping 10 points or pulse rising 20 points or more when going from a lying position to a standing position. Orthostatics should be checked in that order, lying then standing; if abnormal when done immediately with standing, the patients should stand for 2 minutes and have the standing vital signs rechecked, as a percentage of normal individuals may take up to 2 minutes for values to normalize when shifting position from horizontal to vertical. Orthostatic findings can be explained to the patient and family as the body having blood and the oxygen it carries rushing away from the head with standing, and the heart trying to beat faster to get it back up to the head again. A mitral valve click may be heard, as approximately one-third of patients with anorexia nervosa lose cardiac mass and develop a floppy mitral valve. Flow murmurs may be heard in those with a thin chest wall.
Tanner staging of breasts and pubic hair for girls, and pubic hair and genitals for boys, should be performed. Girls may have atrophy of breasts with starvation, and when an eating disorder occurs early or in mid-puberty, a patient may be Tanner 4 or 5 for a prolonged period without experiencing menarche. Asking an amenorrheic patient when they started noticing pubic hair and breast buds, and when there were too many pubic hairs to count (indicating Tanner 4) can help determine whether puberty has been delayed or has been progressing. A delay of 3 years from pubarche to menarche is abnormal and deserves evaluation. Abdominal masses may include palpable loops of stool from constipation; hepatosplenomegaly can occur in the context of refeeding syndrome with congestive heart failure. Any significant neurologic findings such as blurred optic disc margins warrant neuroimaging of brain, as a brain tumor may mimic an eating disorder. A summary of physical findings can be found in Box 15-4 .
Medical Complications from Restricting Energy Intake
Medical effects of starvation affect every organ system. Cardiac effects can kill, but quality of life can be affected by other aspects of starvation, especially effects on brain, bone, and the gastrointestinal tract.
Effects on the Heart: Precipitous or prolonged starvation can lead to myocardial wasting, arrhythmias including prolonged QTc syndrome, and sudden cardiac death, especially in the face of electrolyte abnormalities. In a prospective case control study of 41 consecutive patients with AN, several patients were found to have a prolonged QTc interval, with two of these patients dying suddenly, having QTc intervals greater than 450 msec (normal is <440 msec). Swenne demonstrated prolongation of QTc intervals after a period of rapid weight loss in patients with anorexia nervosa, which normalized after restoration of healthy weight. Patients with anorexia nervosa have been found to have decreased cardiac chamber size with thinning of the left ventricle, as well as hypotension and reduced cardiac output. Sinus bradycardia and sinus arrhythmias occur commonly and may be viewed as a protective adaptation to prolonged starvation. All cardiac changes are fully reversible, with the exception of the irreversible direct myocardial damage that occurs with chronic emetine use, the active ingredient in ipecac. Emetine can also bind to noncardiac muscle, causing a diffuse myositis. Although ipecac is no longer available in the United States, it still may be found in other countries, with risks of misuse and fatalities.
Effects on the Brain: Starvation can cause reductions in gray and white matter, a change that can occur within months. Although cortical volumes may increase with weight gain, gray matter volume changes may not be completely reversible. Recent studies have shown a difference in pain perception and response to food in patients with anorexia nervosa versus healthy controls; women with anorexia nervosa showed a blunted pain response combined with increased activity in the right anterior insula and prefrontal cortex. It is not clear whether these changes predate or postdate the development of the eating disorder. What is seen clinically with prolonged starvation is psychomotor retardation, or a slowing of thought, speech, and movement, all reversible relatively quickly (over a period of days to weeks) with refeeding. This transiently diminished cognitive function supports use of family-based refeeding, or the Maudsley method, where parents are put in full control of the child’s eating until the child can maintain a healthy weight, with a gradual transition of responsibility of meals back to the child, with close supervision.
Effects on Bone: Starvation lowers estrogen levels and suppresses the hypothalamic-pituitary-ovarian axis, with associated findings of elevated serum cortisol levels, and lowered adrenal dehydroepiandrosterone (DHEA) and insulin-like growth factor (IGF-1). These findings appear causally linked to osteoporosis and an increased risk of stress fractures, particularly in patients who are underweight. Osteopenia occurs when there is a negative net balance between bone deposition and bone resorption; in layperson’s terms, the adolescent is making more withdrawals than deposits into his or her “bone bank.” In boys, the lowered androgen levels are associated with osteopenia, although fewer studies track bone density in boys. Because 40% to 60% of bone density is deposited in early adolescence (ages 11 to 14 years) in girls and late adolescence in boys (ages 15 to 17 years), disordered eating during those crucial times can pose substantial risk for osteopenia. Even a 5% to 10% decrease in bone deposition during these crucial times can result in significant fracture risk in the third decade of life. Adolescents with anorexia nervosa in particular have inappropriately low bone formation rates. Questions remain as to why skeletal losses in adolescents and adults with anorexia nervosa do not fully reverse with weight gain.
Hematologic Effects of Starvation: Children and adolescents who restrict intake long enough will first show neutropenia, followed usually by a normochromic or hypochromic anemia, and finally pancytopenia (platelet formation is initially preserved). With profound restricting, they may get a low absolute neutrophil count (ANC), although relatively few patients get more than the average amount of illnesses. All hematologic changes resolve with refeeding. Of interest, the bone marrow in anorexia nervosa appears to be replaced by a fatty marrow. Gimble et al. and Okazaki et al. proposed that creation of fat cells may compete with osteogenesis, resulting in differentiation of mesenchymal stem cells into adipocytes rather than osteoblasts. Higher fat in the marrow may lead to diminished structural bone integrity, with diminished bone mass and lowered skeletal strength. In healthy marrow, the red cells or red marrow helps with biomechanical strength, whereas the adipocytes provide a fatty cushion, causing greater vertebral compressibility. Thus, lower red marrow and higher yellow marrow may lead to the observed increased fracture risk.
Other endocrine changes include a euthyroid sick syndrome picture, with hypothermia, slowed metabolism, and perception of always being cold. Delayed puberty and amenorrhea can occur due to suppression of gonadotropin releasing hormone (GnRH) secretion by the hypothalamus. Delayed puberty is defined as no secondary sexual characteristics by age 14 years or menarche that has not occurred within 3 years of onset of thelarche/adrenarche. Primary amenorrhea means no menses by age 15 years. In fact, most girls achieve menarche within 2 years of occurrence of both thelarche and adrenarche, with 3 years representing more than 2 standard deviations from normal. Secondary amenorrhea means the absence of menstruation after cycles have been previously established, usually defined as no periods in at least 3 months. Amenorrhea can precede weight loss in 50% to 75% of patients with eating disorders.
Osteopenia results from a negative net balance between bone deposition and bone resorption. In children and adolescents with eating disorders, failure to deposit bone at age appropriate times as well as increased resorption set the stage for osteopenia. For girls, a critical time of bone deposition should occur between ages 11 and 14 years, with 40% to 60% of peak bone mass deposited in healthy girls at that time; for boys, peak growth spurt with increased bone deposition occurs slightly later ( Box 15-5 ). An additional 5% accrues in the third decade, with even that low amount accounting for a significant decrease in fracture risk. Even a 10% decrease in adult bone mineral density is associated with a two- to threefold increase in fracture risk. Net gains in bone mass positively correlate with physical activity, weight gain, and calcium and protein intake; net loss correlates with age. Inadequate calcium and protein intake lead to decreased bone formation, as well. Excess glucocorticoids decrease calcium absorption from the gut and inhibit bone formation through direct, receptor-mediated osteoblast effects. Acquisition of bone mineral normally continues through the second decade of life, with peak bone mass reached in late adolescence or early adulthood.