Predicting Diagnostic Crossover

It isn’t unusual for a patient with an eating disorder to migrate between diagnoses, particularly after the initial diagnosis is made. One of the most common crossovers is from anorexia nervosa (AN) to bulimia nervosa (BN), particularly among patients first diagnosed with restricting-type AN (AN-R). Being able to predict diagnostic crossover from AN-R to BN at an early state would be useful for treatment because bulimia and purging predict an unfavorable prognosis and increased time to remission.

Study seeks crossover risks

A team of Japanese researchers recently sought to isolate psychological factors that might predict crossover among patients with eating disorders (Biopsychosocial Medicine 2008; 2:5), and to examine the relationship between these factors and depression. Eighty patients (1 man and 79 women) undergoing treatment for AN and BN participated in the study. The mean age was 24; patients had been ill for an average of nearly 5 years; and the current mean body mass index (BMI, kg/m2) was 15.27. The patients were placed in three groups: those with AN-R only (n =44); patients who had crossed over from AN-R to AN-BN (n =22); and those who had crossed over from AN-R to BN (n =14).

The patients were evaluated with several questionnaires, including the Multidimensional Perfectionism Scale (MPS), which measures six dimensions of perfectionism, such as concern over mistakes, parental expectations, and doubt about actions. They also used the Temperament and Character Inventory-125-4 (TCI), which includes novelty-seeking, harm avoidance, reward dependence, persistence, self-directedness, and cooperativeness. Each item is scored on a four-point scale. The Beck Depression Inventory-II (BD-II) was also used.

Significant differences emerged among the three groups

Significant differences emerged among the three groups in duration of illness, current BMI scores, maximum BMI in the past, and BDI-II scores. Patients who crossed over from AN-R to AN-BN had been ill much longer than had those with AN-R only. Patients who crossed over from AN-R to BN had significantly higher current BMIs, maximum BMIs in the past, and higher BDI-II scores than did those with AN-R only. Those who developed BN after AN-R scored significantly higher in “parental criticism” and lower in “self-directedness,” but both of these factors disappeared when depression scores were used as a co-variant.

Past obesity was a factor

A tendency toward obesity also was a factor. Dr. Hiroki Nishimura and colleagues found that patients with AN-R at onset who were overweight in the past were inclined to gain weight and to develop BN over the course of their illness. Consistent with previous studies, high parental criticism and low self-directedness were associated with crossover from AN-R to BN among the Japanese patients.

The researchers compared four patients whose diagnoses changed five years or more after the onset of illness and 32 who had different diagnoses within five years of the onset of illness (32). They found that the duration of AN-R had little effect on the psychological characteristics of patients in the two groups.

Investigating a tendency toward obesity or depression might help predict crossover from AN-R to BN. And, examining patient characteristics such as parental criticism and self-directedness may predict crossover and also, at least partly, the outcome of treatment.