Harris Center Research - Prevention

Working with Dr. Ruth Striegel-Moore, Ph.D., at Wesleyan University, the Harris Center is learning more about what factors contribute to a healthy lifestyle among adolescents. One such study involves meal frequency and its relationship to body mass index (a measure of height for weight) in adolescent girls. The research team asked over 2,000 girls for information about their eating, exercise and TV-watching patterns and found that those who ate at least three regular meals a day tended to maintain or move toward a healthy weight. These results, which are consistent with previous data, are valuable because they counter the popular misconception that “eating less is best” (1) .

The Harris Center also collaborates with Dr. Striegel-Moore to study how family environment influences adolescent eating habits. Does eating meals together as a family during childhood have an effect on adolescent health concerns? Does family cohesion (emotional connection, positive communication between family members) play a role in the development of healthy eating behaviors? To shed light on these questions, the team examined longitudinal data from the 10-year National Heart Lung and Blood Institute Growth and Health Study (which followed over 2,000 girls through adolescence) and found that frequent family meals during childhood were associated with greater family cohesion and healthier coping skills in later years (2). And adolescent girls who reported greater family cohesion – when compared to those who reported less family cohesion—ate breakfast more regularly, drank less soda, and showed a bit more tendency to choose fruits, vegetables and milk (3).

College Prevention
"Food, Mood, and Attitude" was developed as a 2-hour CD-ROM prevention program to reduce eating disorder risk in college students. In collaboration with researchers at the University of Missouri and Inflexxion, Inc., the Harris Center recently assessed the effectiveness of the program and found that it had a positive impact on knowledge, attitudes, weight and shape concerns, overeating and excessive exercise behaviors in first semester college women (4).

Harris Center Research - Etiology

Genes and Eating Disorders
David B. Herzog, M.D., is currently a consultant for principal investigator Walter Kaye, M.D. on an NIMH-funded international study exploring the genetics of anorexia nervosa. The project seeks to identify which areas of the human genome contain genes that increase the risk of this illness. For further details on this study, please log onto www.angenetics.org.

Neuroimaging and Cognitive Dysfunction in Eating Disordered Women
Recently we collaborated on a neuroimaging (PET) study which indicated that there are distinct areas of brain hypometabolism in women with anorexia compared to controls (5). Our research on cognitive functioning in anorexia shows that nonverbal memory deficits and impairments of strategic organization in anorexia are consistent with findings from studies of OCD (obsessive compulsive disorder) populations. This suggests that these disorders share strategic memory deficits and have certain similarities in patterns of dysregulation (6).

Psychosocial Risk Factors
A characteristic that increases the likelihood of developing a condition or an illness is known as a risk factor. Working with Ruth Striegel-Moore, Ph.D., at Wesleyan University, the Harris Center researches psychosocial risk factors for eating disorders. Body dissatisfaction, particularly in adolescent girls, appears to be a strong risk factor for eating-disorder behaviors and for other psychological issues, such as depression and low self-esteem (7). Self-esteem is an important issue for adolescent girls. Studies conducted in collaboration with Dr. Striegel-Moore have found that self-esteem was greater in black than in white adolescent girls and greater with lower body mass index in both races (8) .

Harris Center Research - The Course of Eating Disorders

• The majority of treatment-seeking eating disorder patients substantially improve symptomatically over time (9).
• Less than half of patients with anorexia fully recover (9).
• Although more than half of patients with bulimia recover, nearly one-third relapse (9).
• Mortality is extraordinarily high; it is 12 times that expected for gender and age (10) (11)
(13).
• The suicide rate is 56 times that expected for gender and age (9) (11) (12).
• Coexisting substance abuse is the strongest predictor of mortality (10) (13).
• Alcohol and drug abuse occur commonly in patients with eating disorders; in the longitudinal sample, about 17% of patients struggled with substance abuse (14) (15).

The Longitudinal Study is also making significant inroads into the nosology, or diagnostic classification, of eating disorders. Published by the American Psychiatric Association and used by clinicians across the country, The Diagnostic and Statistical Manual of Mental Disorders (DSM) categorizes the various psychiatric illnesses and defines criteria for each diagnosis. Our team is studying how well the current DSM, which includes anorexia nervosa, bulimia nervosa, and eating disorder not otherwise specified (EDNOS) "fits" individuals with abnormal eating patterns. This topic is fundamental to all aspects of eating disorders research; if we haven't correctly defined the disorders that we are studying or treating, our results might be invalid or limited. Diagnostic classification is integral to clinical research and practice, influencing etiological, preventive, and treatment models of eating disorders. Data from research on diagnostic classification can help inform revisions to the DSM.

The current classification system recognizes two subtypes of anorexia nervosa: the restricting type and the binge eating-purging type. Both of these are distinguished from bulimia nervosa, which is characterized by binge eating and compensatory behaviors (self-induced vomiting, misuse of laxatives or diuretics, fasting, or excessive physical activity) in individuals of generally healthy weight. Our current research examines whether these distinctions are useful: Is eating disorder diagnosis stable? Does diagnosis predict longitudinal outcome? In our research, most women with anorexia nervosa experienced periods of pure restricting as well as periods of bingeing and/or purging, effectively "crossing over" between the restricting type and the binge eating-purge type over a period of seven years. Further, a third of those with anorexia nervosa transitioned to bulimia nervosa but tended to fall back into anorexia. In contrast, women with bulimia nervosa were generally unlikely to crossover to anorexia nervosa (16).

Among women with bulimia nervosa, it was those with a past history of anorexia nervosa who were likely to suffer a longer course of illness with resurgences of anorexia, whereas those without a history of anorexia tended to move forward in recovery (17). Taken together, the longitudinal course differences between anorexia nervosa and bulimia nervosa do support their distinction in the DSM; however, less support was found for the distinction between the restricting and binge/purge anorexia nervosa subtypes.

While anorexia nervosa and bulimia nervosa receive the most research attention, eating disorder not otherwise specified (EDNOS) is actually the most common eating disorder among adult clinical samples. EDNOS consists of combinations of eating disorder characteristics that do not quite meet the criteria for anorexia nervosa or bulimia nervosa. To examine the EDNOS diagnosis in youth, we studied adolescents who presented for outpatient treatment and found that, as in adults, more of them had EDNOS than either anorexia nervosa or bulimia nervosa. All three groups of adolescents--those with anorexia, those with bulimia and those with EDNOS--were tested for self-esteem, depression, and eating disorder severity. Results indicated no differences between patients with anorexia and those with EDNOS variants of anorexia, perhaps suggesting that the criteria for anorexia be broadened in youth. Compared to adolescents with bulimia, those with EDNOS variants of bulimia tended to be slightly younger and to have somewhat higher self-esteem and lower eating disorder severity. These findings suggested that EDNOS is a mixed diagnostic category that may be meaningfully subgrouped (18).

Harris Center Research - Treatment

Treatment of Osteopenia/Osteoporosis in Anorexia Nervosa

A loss of as much as 25-50% of total bone mass occurs frequently in women with anorexia nervosa, increasing the risk of debilitating bone fractures. For over ten years, the Harris Center and the Neuroendocrine Unit at Massachusetts General Hospital have worked together on research aimed at understanding and treating osteopenia and the more serious condition, osteoporosis.

It is believed that bone loss in anorexia nervosa stems in part from a combination of dietary and endocrine factors. When an individual does not take in enough food, the chemical messenger system that helps support normal bone growth and maintenance is disrupted. In order to learn more about these complex mechanisms, the research team has examined a number of hormones. Estrogen, long recognized for its impact on the female reproductive system, helps to build and maintain bone strength. As a consequence, the amenorrhea (absence of menstrual periods) and low estrogen levels that typify anorexia nervosa contribute to bone loss (19). Also important is growth hormone, which is secreted by the pituitary gland, a small but crucial structure at the base of the brain. In healthy individuals, growth hormone stimulates bone formation. In anorexia nervosa, growth hormone is high in concentration but impaired in its ability to increase bone formation and therefore bone density (20).

Although some people think of testosterone as existing only in men, small amounts of this hormone are also produced by women, mostly in the ovaries. Testosterone levels are low in women with anorexia nervosa (as compared to normal controls) and this deficiency is associated with abnormal bone density and fat-free body mass (21). Interestingly, reduced testosterone levels may also contribute to the anxiety, depression and unhealthy eating behaviors that are characteristic of anorexia nervosa (22).

Hormones that are regulated by nutrition, such as IGF-1, cortisol, ghrelin and PYY, are likely to influence bone mass (23). Synthesized by the liver, IGF-1 (insulin-like growth factor) promotes bone formation; levels of IGF-1 decline with undernutrition in individuals with anorexia nervosa, and are an important cause of low bone density (24). As a result of insufficient food intake, levels of the stress hormone cortisol increase, further contributing to the problem of bone loss (25). Ghrelin and PYY, which help regulate eating behavior and appetite, also play a role in bone density changes (26); our studies demonstrate higher concentrations of these two hormones in individuals with anorexia nervosa than in healthy controls (27). Furthermore, adiponectin – a hormone produced by fat cells – and insulin, which is secreted by the pancreas to control blood sugar levels, appear to be involved in bone density changes in this disorder (28). So is leptin, a protein that is found at lower concentrations in individuals with anorexia nervosa than in normal controls (29).

Recent findings suggest that bone density tends to stabilize in adolescent girls with anorexia nervosa who gain weight and resume their menstrual periods in contrast to those who do not gain weight or bring back their menses, in whom bone density continues to drop (30) (31). Research into effective treatments for anorexia nervosa-induced osteopenia has suggested that IGF-1 and biphosphonates such as risedronate (actonel) are promising, at least in adults (32) (33), and that the hormone testosterone may also hold potential (34). Birth control pills have not been effective in treating bone loss in this population (35).

This page was last updated on February 20, 2008.