Epidemiology of anorexia nervosa in Japanese adolescents

Background No epidemiologic survey examining eating disorders in Japan has been done at a national level since 1992. The prevalence of anorexia nervosa, as assessed by questionnaires to hospitals, is thought to be underestimated because patients with anorexia nervosa tend to avoid consultations. In conformity with the School Health and Safety Act of Japan, schools are required to have physicians perform a medical examination of students every year. The teachers in charge of health education and school physicians determine the height, weight, and health condition, and examine the medical records of each student. Therefore, we as members of the Survey Committee for Eating Disorders of the Japanese Ministry of Health, Labour, and Welfare conducted an epidemiologic survey using questionnaires sent to schools in seven prefectures to determine the current prevalence of anorexia nervosa among adolescents. Methods We sent a questionnaire to elementary, junior high, and senior high schools. Questionnaires contained items on the number of students, patients with anorexia nervosa in each grade who were diagnosed by specialists, and students who the school physician strongly suspected to have anorexia nervosa but who did not undergo a clinical examination in a medical institution. Results We found patients of both sexes with anorexia nervosa aged 9–10 years in elementary schools. The point prevalence of anorexia nervosa for girls, including strongly suspected cases, in the three grades of junior high school and three grades of senior high school were 0–0.17 %, 0–0.21 %, 0.17-0.40 %, 0.05-0.56 %, 0.17-0.42 % and 0.09-0.43 %, respectively. We also confirmed a prominent sex difference in the prevalence of anorexia nervosa. The prevalence of boys was one third that of girls in some prefectures. One third to one half of diagnosed and strongly suspected students with anorexia nervosa had not received medical consultation or treatment. Conclusions Although the prevalence of anorexia nervosa had regional differences in Japan, it has reached levels comparable to those in Western societies. Because no eating disorder center exists and the treatment environment is poor, national action to address this disease is a pressing need in Japan.


Background
Anorexia nervosa is associated with various types of medical morbidity due to malnutrition [1] and a significant mortality rate in adolescents [2]. The last survey performed in Japan at a national level was done in 1992 by the Survey Committee for Eating Disorders of the Japanese Ministry of Health, Labour and Welfare who sent questionnaires to hospitals with 300 or more beds [3]. The prevalence of anorexia nervosa of female patients aged 10 to 29 years was 0.025 % to 0.031 %. Based on a questionnaire sent to hospitals and clinics in Niigata prefecture, which faces the Sea of Japan and had the 14 th largest population in Japan in 1997, the prevalence of anorexia nervosa of female residents aged 15 to 29 years was 0.017 % [4]. In contrast, the prevalence of anorexia nervosa as assessed by questionnaires returned by schools was higher than that obtained from medical facilities [3]; in that study, the prevalences of anorexia nervosa for the junior and senior high school girls of Kyoto were 0.24 % and 0.15 %, respectively. It was thought that the prevalence obtained based on questionnaires sent to medical facilities might underestimate disease prevalence because patients with anorexia nervosa may refuse or not seek consultation [5].
In conformity with the School Health and Safety Act of Japan, school physicians have to perform a medical examination for students every year. The teachers in charge of health education in Japan complete a curriculum for adolescent health and some have a school-nurse license. Every school has a school physician, whose specialty is internal medicine, pediatrics, or gynecology. The teachers in charge of health education and school physicians determine height, weight, and health condition and review the medical records of each student.
School physicians and teachers in charge of health education list students who show poor weight gain, lose weight in comparison with the last year, or have a body weight lower than 80 % of the expected body weight in this annual medical examination. The expected weight of a child is based on height and indexes prescribed to age and gender. Growth curves are also drawn to identify growth retardation. These health care professionals usually see students with very low weight to identify the clinical characteristics of anorexia nervosa, including amenorrhea, lanugo, bradycardia, constipation, hypotension, hypothermia, and edema, which are described in the diagnostic criteria of The Survey Committee for Eating Disorders of the Japanese Ministry of Health, Labour and Welfare, and also ask the students questions according to the Diagnostic and Statistical Manual fourth edition (DSM-IV) criteria. When a student is a strongly suspected case of anorexia nervosa, the school physician advises the parents or guardians to consult with a specialist.
In this study, the Survey Committee performed an epidemiologic survey of anorexia nervosa by sending questionnaires to schools in seven prefectures in Japan to obtain updated prevalence data.

Methods
A questionnaire was sent to principals and the teachers in charge of health education at junior and senior high schools in Tokyo, Hokkaido, Hiroshima, Nagano, Kumamoto, Yamaguchi, and Miyazaki, which had respective population rankings during this survey of 1, 8, 12, 16, 23, 25, and 36 of the 47 Japanese prefectures, and to elementary schools in six prefectures (same as the above except for Hokkaido). Tokyo is the capital of Japan. Hokkaido is the northernmost prefecture. Nagano is a mountainous prefecture in the center of the country. Hiroshima and Yamaguchi are located in the west of Honshu, the main island. Kumamoto and Miyazaki prefectures are located on the island of Kyushu. The questionnaire gathered background on each school, and the number of students with anorexia nervosa, including those with strongly suspected anorexia nervosa, in each grade at the time of the survey. We used the DSM-IV criteria for anorexia nervosa in addition to the criteria for anorexia nervosa put forth by the Survey Committee for Eating Disorders of the Japanese Ministry of Health, Labour, and Welfare (Table 1) [6]. The build of Japanese is smaller than that of Caucasian. Based on the DSM-IV criteria, many healthy girls with regular menstruation who are at 85 % of normal weight are selected for screening for anorexia nervosa. In this study, we used a body weight less than 80 % of the ideal body weight based on the Japanese criteria to identify patients with anorexia nervosa. In the present study, we may have overlooked patients with anorexia nervosa who met only the DSM-IV criteria.
The questionnaire did not ask the teachers in charge of health education or the school physicians about the types of anorexia nervosa because they may not always recognize the binge eating/purging-type without a report from the student.
Japanese compulsory education begins when children are six to seven years old. They study for six years at an elementary school and then three years at a junior high school. The percent of junior high school students who go on to senior high school in Japan is more than 97 %. Japanese senior high schools use a three-year system. A senior high school student's age ranges from 15 to 18 years. We provided a written explanation of this study to the parents of the students in each school. Based on the judgment of the principal, personal data can be provided to a research team when its goal is to benefit students and the data are provided anonymously. The study protocol was approved by the institutional review board of each author. This study was also carried out with the permission of the Board of Education in each of the seven prefectures. Table 2 shows data on the total number of students enrolled in the seven prefectures according to the School Table 1 The diagnostic criteria of anorexia nervosa put forth by The Survey Committee for Eating Disorders of Japanese Ministry of Health, Labour and Welfare Basic Survey by the Japanese Ministry of Health, Labour and Welfare, the number of subjects who participated in this study, the participation rate, and the number of students diagnosed with anorexia nervosa, including those with strongly suspected anorexia nervosa, for each grade. Participation rates for the questionnaire varied by school and by prefecture. In Tokyo, elementary and junior high schools in a limited area participated in the study and those subjects represented only about 5 % of the total number of children. Although the Education Bureau of Tokyo prefecture requested schools to cooperate with our survey and we sent a collection letter to schools that did not provide answers, some schools seemed to resist information disclosure. In contrast, about 60 % of students in every grade in Yamaguchi and Kumamoto participated in the research.

Results and discussion
We confirmed that anorexia nervosa was predominant in girls. This is the first study in Japan that examined the prevalence of anorexia nervosa among children 12 years old or younger. We found patients of both sexes in the fourth grade in elementary schools (aged 9-10 years) with anorexia nervosa. For junior high school girls, the prevalence increased in a straight line and became highest among third-year students. For senior high school girls, the peak prevalence was observed in the second or third year in all prefectures except for Hiroshima. The point prevalences of anorexia nervosa for girls, including strongly suspected cases in the three grades of junior high school and three grades of senior high school were 0-0.17 %, 0-0.21 %, 0.17-0.40 %, 0.05-0.56 %, 0.17-0.42 %, and 0.09-0.43 %, respectively. We also confirmed a prominent sex difference in the prevalence of anorexia nervosa, with the prevalence of boys one third that of girls in senior high schools in Yamaguchi. One third to one half of those diagnosed with anorexia nervosa or with strongly suspected anorexia did not receive medical treatment (data not shown). We demonstrated the recent prevalence of anorexia nervosa, including strongly suspected cases of anorexia nervosa, among adolescents in seven prefectures in Japan. We cannot show evidence to explain the regional differences in the prevalence at the present stage. Although there were regional differences among prefectures, the prevalence of anorexia nervosa in this study was similar or higher than the rates reported in a previous epidemiologic study using the same survey method as the present study [3]. Our report is also the first on the prevalence of anorexia nervosa among elementary school children. Because anorexia nervosa before menarche can retard growth and lower peak bone mass [7,8] preventing this disorder, early detection, early consultation, and communication between parents, physicians, and schools are pressing needs.
Since the 1960s, eating disorders such as anorexia nervosa began to be recognized as an important health problem among adolescent girls and young women in Western societies, and their prevalence has increased over time [9]. Although the survey methods and ages of the subjects in other studies differed from those in our study, the prevalence of anorexia nervosa was shown to be 0.3 % of young females in the Netherlands [10], 0.5 % of teenaged girls in Germany [11] 0.3 % of secondary school students in Hungary [12], and 0.2-0.3 % of girls aged 13 to 18 years in the United States [13]. Although it was previously reported that the prevalence of anorexia nervosa in Japan is lower than that for European Caucasian populations [4], we showed that the prevalence among adolescents in Japan is near that of developed European countries and the United States.
As shown in other studies [14,15], adolescent girls with eating disorders often refuse consultations with a physician, never seek treatment, or are untreated. One third to one half of the students with anorexia nervosa in the present study had not consulted physicians. Unfortunately, there is no eating disorder center in Japan, which leads to a poor treatment environment. The establishment of an eating disorder center can contribute at a national level to education for physicians and paramedical professionals, research support, and a public awareness campaign on eating disorders. Medical administration policies should be considered based on the results of the current study. Few hospitals employ specialists in eating disorders. These two factors may contribute to the low consultation rates found in the present study.
There are several critical limitations to this study. First, prevalence was estimated using indirect information obtained from schools. It is possible that teachers in charge of health education and school physicians were unable to identify milder cases of anorexia nervosa. Second, only seven prefectures participated. Many prefectural Boards of Education refused to participate because of the pressure of additional work, and some schools refused to participate due to privacy concerns. Third, participation rates for the questionnaire varied by school and by prefecture.

Conclusions
In conclusion, medical facilities and treatment environments conductive to the treatment of for eating disorders need to be developed in Japan because the prevalence of anorexia nervosa among adolescents in this survey was similar to that found in European countries and the United States. It is clear that this health concern needs to be addressed.