Rapunzel had magnificent long hair, fine as spun gold.
—The Grimm Brothers, Rapunzel(1)
Rapunzel’s hair fell to the ground like a rainbow. It was as strong as a dandelion and as strong as a dog leash.
—Anne Sexton, Rapunzel(2)
Emily (all names and identifiers have been changed to protect confidentiality) was a 7-year-old girl seen by her pediatrician for a routine annual examination. Although Emily had voiced no complaints, her mother had noted that she had appeared paler than usual. Upon physical examination, her pediatrician noted a nontender palpable mass in the child’s stomach, leading him to order an abdominal ultrasound. The study showed no evidence of any abnormalities, although the technician did note that Emily must have just eaten because her stomach was full. However, her mother reported that Emily had not eaten since breakfast, 4 hours earlier.
On reevaluation the next day, Emily’s pediatrician still palpated the same mass, leading him to order a computed tomography (CT) scan of the abdomen. The abdominal CT demonstrated a free-floating mass that filled the entire stomach and was surrounded by a thin rim of contrast material (F1). Emily’s pediatrician did not think the mass was consistent with a neoplasm but was concerned that it could be a conglomeration of something she had ingested. Upon further questioning, Emily reported no stomach pain, nausea, vomiting, reflux, diarrhea, flatulence, recent illnesses, or fever. Her mother reported no change in her bowel habits and stated that Emily never wanted to eat much in one sitting.
Given the large size and appearance of the mass on tomography, a decision was made to remove it surgically. Just as the elective surgery was scheduled, Emily was referred for evaluation by the child psychiatry consultation-liaison team.
Initial Psychiatric Consultation
The child psychiatry team (A.S.F. and A.M.) met with Emily and her mother the morning before surgery. By her own and her mother’s account, she was happy and well adjusted. She had many friends and was enrolled in a mainstream public school second-grade class that she enjoyed.
Emily lived in the home of a family friend with her mother (Kim), her older half sister, and her younger brother. Up until 2 years before, Emily had lived with her mother, siblings, and father in their own home. Kim felt that family life had been generally good during the marriage and that her husband had been a good father; Emily was especially close to him. Emily’s father, however, had a substance abuse problem that led to dissolution of the marriage 2 years before admission. After this, the parents’ relationship became strained, and Emily had not seen her father in a year.
Emily was briefly in counseling to help her deal with the changes in her life surrounding her parents’ separation, but the counselor felt that she was coping well and did not require further psychotherapy. Kim, however, expressed private concern that Emily was not dealing with the loss of her father. Meanwhile, Emily had started at a new school and was doing well academically and socially.
With this background information, further evaluation was deferred until after the surgery, when the identity of the mass would be elucidated.
Because the mass in Emily’s stomach appeared too large to be removed laparoscopically, a gastrotomy was performed instead through a 7-cm median xiphoumbilical incision. The mass was found to be a trichobezoar, or hairball, 45 cm long and 8 cm in diameter (F2). The hair mass was cast in the shape of the stomach, which it filled, and extended down through the distal end of the duodenum, past the ligament of Treitz.
The smooth surface of hair does not allow for its propagation through peristalsis. Thus, when hair is ingested, it gets trapped in the mucosa of the stomach. As more hair is added, the resulting mass causes the stomach to cease peristalsis completely. This mass is a trichobezoar (3), first described in 1779 by Baudomant (4).
Trichobezoars are nearly impossible to diagnose by plain film alone. Ultrasound has been shown to be effective in diagnosing bezoars in up to 88% of cases if a "clean" acoustic shadow, which represents a solid mass, can be visualized (4, 5). However, a calcified mass, neuroblastoma, aneurysm, abscess, or fecal material can appear similarly on an ultrasound (4). An abdominal CT with contrast has been shown to diagnose 97% of bezoars (5), which appear as free-floating filling defects in the stomach (3).
Although trichobezoars can be removed in several different ways, Emily’s was too large to avoid open surgery. Smaller trichobezoars can be removed by endoscopy after fragmentation by water pick, laser, or extracorporeal shock-wave lithotripsy (6, 7) or by enzymatically dissolving the mass. Finally, small bezoars can be removed laparoscopically through a small incision (6). However, none of these techniques are successful for very large bezoars (over 20 cm), which usually require removal through open surgery (3).
Trichotillomania was first described in the literature in 1889 (3) and first recognized by APA as a distinct disorder in 1987 (8). The prevalence of trichotillomania is 0.6%–1.6% if DSM-IV criteria are used (9). Among those who suffer from trichotillomania, only 30% will engage in trichophagia, or eating their hair, and of these, only some 1% will go on to eat their hair to the extent requiring surgical removal (10). In the case of Emily, the trichobezoar had a long tail that extended past the duodenum, a rare complication of trichotillomania referred to as Rapunzel syndrome in the surgical literature.
Trichotillomania has many different clinical presentations. Patients with trichotillomania usually pull hair from the scalp, but eyelashes, eyebrows, legs, armpits, and pubic regions are also targets for pulling. Although by definition, trichotillomania entails hair pulling to the point of alopecia, there is a much larger number of individuals who pull their hair but not to the point of noticeable thinning (11, 12). Not all hair pullers experience the sense of tension before pulling followed by relief afterward that is described in the DSM-IV definition; one study showed that among college students, although 1.5% of men and 3.4% of women had chronic hair pulling, only 0.6% met the full DSM criteria for trichotillomania (9, 13).
Clinical studies of trichotillomania delineate two types of the disorder, which may coexist in one individual (14). The "focused" type is trichotillomania in which time and attention are set aside specifically for the purpose of hair pulling. This type of hair pulling is associated with tension before and relief after pulling, and with negative affect. More common is the "automatic" or "sedentary" type of hair pulling, in which subjects pull their hair while engaged in other activities, such as lying in bed, driving, or watching television (15, 16). Patients with the latter type are unaware that they are pulling their hair at all or become aware only when a clump of hair "appears" in their hands (13). Many patients with trichotillomania have other associated habits; about 48%–70% have some form of oral behavior after pulling hair, such as running the hair across their lips, biting the root, or eating the hair (13, 16).
Trichotillomania is far more common in girls than boys. The age at onset of trichotillomania is bimodal; it begins either in early childhood or in adolescence (15).
Of the early-onset cases, one-third develop it before the age of 10, and 14% develop it before the age of 7. In children who begin hair pulling before the age of 8, the disorder is usually benign and self-limited, although not all cases disappear. Most of these cases are of the sedentary type, and the children, when verbal, often deny pulling (15). In one case, a 28-month-old girl was found to be pulling her own hair, and upon review of her history, it was found that she had pulled and eaten her dolls’ hair as early as 6 months of age. Thus, in early-onset trichotillomania, pulling and eating dolls’ hair can be a first symptom of trichotillomania (17). Although habits such as thumb sucking have been noted in early-onset trichotillomania, other comorbid features have not been found (15).
Trichotillomania occurring in the later school years and adolescence is more likely to be chronic and accompanied by comorbid psychiatric disorders, such as depression, anxiety, obsessive-compulsive disorder (OCD), body dysmorphic disorder, eating disorders, and alcohol and substance abuse (13, 15, 18). As is generally true, clinical samples of patients seeking treatment for trichotillomania appear to have higher rates of comorbidity than nonreferred community groups (11).
The physical complications of trichotillomania include rash or infection on the scalp, repetitive stress injury, and carpal tunnel syndrome (19). When trichophagia is present, complications can also include gingivitis and, secondary to formation of a trichobezoar, gastrointestinal blockage and perforation or secretory diarrhea and malabsorption of nutrients, leading to vitamin deficiencies (10, 19, 20). While rare, the formation of trichobezoars is the most serious complication of trichotillomania. Case reports of children with trichobezoars or Rapunzel syndrome are sparse, and many link the trichophagia either to early childhood deprivation or abuse (7, 10, 21), psychiatric comorbid conditions (21, 22), or mental retardation (23, 24). While there are case studies of individuals with trichobezoars who seemingly had none of these associated conditions (25–32), cases severe enough to require surgical removal are certainly rare.
The cutting of hair represents a restitution offered the mother, whom the subject fancies to have harmed in some fashion.
—Jeffrey J. Anderson (33)
Further Psychiatric Consultation
Upon discovering that Emily’s stomach mass was a trichobezoar, the psychiatric team reapproached her. Emily clearly had several nonspecific vulnerabilities putting her at risk for a psychiatric disorder: she had recently lost her father and had changed schools. Earlier, her father had been periodically absent and had a history of substance abuse. Furthermore, she had been witness to significant marital discord. However, these alone did not shed light on Emily’s trichophagia.
Kim was completely taken aback by the finding in her daughter’s stomach, since she had never seen Emily eat or pull her own hair. She remembered that when Emily was 3, she had had the habit of chewing on her hair and her dolls’ hair, and this had concerned her. Her pediatrician, however, had not been concerned. Subsequently, Emily appeared to have stopped these behaviors. And yet, Emily’s hair had seemed to grow very slowly, and she had needed few haircuts during her 7 years. Emily, when interviewed alone, agreed that she used to chew her hair when she was "little." Upon further questioning, she admitted that she had in the past sometimes twirled her hair around a finger and that hair would at times break off, and she would eat it. She denied a sense of tension or an urge before pulling or a sense of relief afterward. She said that she no longer pulled her hair but could remember doing it in the first grade. When asked why she had hair in her stomach, Emily was unable to explain how the hair had gotten there.
Kim and Emily’s older sister both had very long hair, reaching past their waists. Although Kim assumed that she did shed hair around the house, she doubted that Emily was eating her hair and stated that she always kept her hairbrush in her purse, where Emily would not have access to it.
Determining with certainty the provenance of the hair in the bezoar was a complicated process. The length of individual strands (over 25 inches for the longer ones) pointed to the mother’s and sister’s hair, as Emily’s own had never been over 12 inches long. The damage to the follicles and their cells did not permit DNA analysis that would have been confirmatory, and light microscopy did not help in differentiation. Scanning electron microscopy did offer useful information to distinguish between the daughter’s and mother’s hair. Specifically, the different widths and surface scaling patterns of the two allowed for ready differentiation of fresh hair clippings, as well as of hair samples taken from the bezoar itself (F3). In summary, the combination of gross pathology and scanning electron microscopy techniques confirmed that hair in the bezoar had derived from at least two sources: Emily and her mother.
Emily appeared to be an intelligent, cheerful child who reported no negative affect and was either very secretive about her trichophagia or entirely unaware of it. She had no alopecia or hair thinning, but had clearly had trichophagia of her own and her mother’s hair.
Emily seemed to be functioning well outside the area of her hair eating and possible hair pulling. The etiology of her trichophagia and apparent trichotillomania can be considered from multiple viewpoints, all informed by the social stressors in her life. Socially, there were factors to predispose her to psychopathology: an unstable family constellation, exposure to marital arguments, the loss of her father, and her change of school may all have contributed (14). From a dynamic perspective, her hair eating could reflect a desire to hold on to and keep inside what she had lost—perhaps her father. Rather than outwardly expressing her feelings of loss, Emily may be conceptualized as physically holding on to her past life by literally ingesting it and holding it within. Another explanation, from a cognitive psychological perspective, might posit that Emily ate her hair because she had unopposed impulses to do so. These internal feelings may have come to be because she felt a need for control in a world that was changing around her while she was powerless. Or these compulsive urges could have been related to a biological predisposition, maybe mediated through vulnerable serotonergic pathways. An argument can also be made from a behavioral perspective that perhaps having once tried pulling and eating her hair, Emily found it soothing and continued to maladaptively engage in it for the desired positive effect. Emily’s reasons for starting to pull and eat her hair probably have aspects of all of these explanations, as well as other idiosyncratic reasons.
Another way to conceptualize Emily’s trichotillomania accompanied by trichophagia is to group it with compulsive forms of disordered eating that may serve the function of self-soothing, namely bulimia, pica, and rumination. Both trichophagia and pica involve an urge to eat nonnutritive substances. The literature on pica distinguishes between voluntary pica (eating in response to what is available) and involuntary pica (eating in response to a compulsion or addiction) (34). The possibility that some cases of pica may respond to selective serotonin reuptake inhibitors (SSRIs) has led some to view it as a form of compulsion and to theorize that pica belongs on the same compulsivity-impulsivity spectrum that includes trichotillomania, trichophagia, and OCD (35).
Thus, a parsimonious view is that Emily probably had trichotillomania and trichophagia compounded by pica—eating her own hair, as well as her mother’s, just as she had earlier chewed on her dolls’ hair, similar to the case noted in the literature (17). As her onset was clearly before the age of 8, her trichotillomania can be characterized as an early-onset form. As is characteristic of that form, she apparently did not experience any urge or sense of relief in pulling, as required by the DSM-IV criteria for the disorder. Like many children with this form of trichotillomania, she appeared to have no other psychiatric comorbid conditions. Emily also experienced significant social stressors in her home, such as has been reported to often accompany the onset of childhood trichotillomania (14).
Rapunzel found the means of escaping her predicament with her own body.
—Bruno Bettelheim (36, p. 17)
But what more reliable source to recovery do we have than our own body?
—Bruno Bettelheim (36, p. 149)
There is no formal set of guidelines for the treatment of trichotillomania in children. Common treatment modalities include medication and cognitive behavior therapy (14), although their empirical support base for children is limited. Support groups, which appear useful and well accepted in adults, are also increasingly available to younger patients through advocacy groups, such as the Trichotillomania Learning Center in Santa Cruz, Calif. (37).
A study of 123 patients with trichotillomania with a mean age at onset of 11 years found that of the 42% who had been treated, there was no difference in responses to psychotherapy, behavior therapy, clomipramine, or fluoxetine. Furthermore, none of the treatments provided significant and lasting improvement of symptoms (38). In the absence of convincing evidence for the effectiveness of medication in childhood trichotillomania, many clinicians would opt for habit-reversal therapy as the initial, most benign intervention (38), with the possible subsequent addition of an SSRI or clomipramine, especially if there is comorbid depression, anxiety, or OCD.
Habit-reversal training is a form of behavioral therapy in which the focus is on eliminating an unwanted chronic habit. It may include a cognitive component, in which the patient and therapist address the thoughts associated with the behavior and replace these with more positive thoughts. Azrin et al. (39) demonstrated that habit-reversal training reduced the symptoms of trichotillomania in 74%, while negative practice (deliberately doing the behavior in mass practice) reduced the symptoms in only 33%. At the 22-month follow-up, nine of the 12 in the habit-reversal training group had achieved remission, while only two of the eight had in the negative practice group. Both this original and subsequent simplified versions of habit-reversal training have proven effective in achieving near or complete remission of trichotillomania in children (40–42). A study of 16 subjects comparing cognitive behavior therapy to clomipramine and placebo found that cognitive behavior therapy was significantly better than clomipramine or placebo in reducing the symptoms of trichotillomania, while clomipramine and placebo did not significantly differ from each other in producing clinically meaningful improvement (43). A study comparing fluoxetine to behavior therapy in 43 subjects also showed the best outcome in those using behavior therapy (44). Similar behavioral interventions have also been adapted successfully for the treatment of pica, particularly among individuals with developmental disabilities (45), among whom the prevalence of the condition has been reported to be as high as 25.8% (46).
Clinical Follow-Up and Prognosis
Emily was referred for outpatient psychiatric services after her discharge from the hospital and uneventful recovery from abdominal surgery. Seen at 1 and 2 months after discharge, she quickly thrived. She returned and readapted rapidly to school and family life. She appeared happy and carefree, reported no symptoms of depression or anxiety, and continued to deny pulling or eating of hair—her own or anyone else’s. Despite the watchful vigilance of her mother, sister, and grandparents, she was not found to engage in hair pulling or hair eating, and no areas of hair thinning or loss became evident. There was no evidence that trichotillomania, trichophagia, or pica had recurred.
In light of these facts, specific treatment as outlined was not instituted at a specialty clinic, as had originally been considered necessary. Rather, Emily was referred for supportive individual and family counseling at a more conveniently located clinic and for periodic monitoring through the surgery and specialized child psychiatric services. Given that her hair-eating behavior may have continued undetected or eventually recurred, minimally invasive follow-up tests were to be performed biannually to ensure that no hair reaccumulation took place. These included abdominal ultrasound or simple abdominal X-rays after swallowing barium. Endoscopy or CT would be avoided unless absolutely necessary, given their intrusive nature and higher radiation dose, respectively.
The overall prognosis for Emily is favorable. She has a supportive and loving family; she has many friends, does well academically, and has a positive affect. Moreover, the natural history of trichotillomania is that most girls who have onset before age 8 outgrow the disorder by adolescence (15). Nevertheless, she also has factors that will continue to remain challenging, and recurrent trichobezoars have been previously reported in the literature.
Overall, a hopeful outlook seems justified. A longer-term treatment goal will be to empower Emily to use alternative and more adaptive tools for dealing with the stressors in her life—in essence, to learn to use her mind, rather than her stomach or her hair, as an instrument for coping. With familial and professional help and support, we can hope that Emily will be able to overcome her trichotillomania, pica, and trichophagia, to reclaim control over her behavior and her body, and to expunge "Rapunzel" back into the fairy-tale books where it belongs.
Received June 30, 2004; accepted Aug. 11, 2004. From the Child Study Center, Yale University School of Medicine; and the Departments of Child Psychiatry and Pediatric Surgery, Yale-New Haven Hospital, New Haven, Conn. Address correspondence and reprint requests to Dr. Martin, Child Study Center, Yale University School of Medicine, 230 South Frontage Rd., New Haven, CT 06520–7900; firstname.lastname@example.org (e-mail). Supported in part by a Career Developmental Award (MH-01792) to Dr. Martin. The authors thank "Emily" and her family for their cooperation, and Drs. Parsia Vagefi, Allan Goldstein, Diane Findley, Ronald Salem, Paul Fiedler, Zeev Kain, Michael Kashgarian, and Miguel Reyes-Mugica for their comments on earlier drafts of this article.
Abdominal Computed Tomography Showing a Mass in the Stomach of a 7-Year-Old Childa
aOrally ingested contrast material has been displaced against the lining of the gastric wall.
Trichobezoar From a 7-Year-Old Childa
aSpecimen preparation and photograph courtesy of Dr. Jonathan Bennett, Department of Pathology, Yale-New Haven Hospital and Yale University School of Medicine.
Scanning Electron Micrographs of Gold-Coated Fresh Hair Clippings From a 7-Year-Old Child and Her Mother and From a Random Sample of the Bezoara
aThe distinctive diameter and surface scaling pattern of both hair strands are evident in a random sample taken from the trichobezoar. Images courtesy of Mr. Barry Piekos, Department of Molecular, Cellular, and Developmental Biology, Yale University.