Demographic analysis and clinical characteristics in subtypes of ED.
The 42 female outpatients aged 12–45 years (mean 26.2 ± 7.8 years) were analysed. The sample comprised the following diagnoses: 23 with BN (54.7%), 8 with AN-BP (19.0%), 6 with AN-R (14.3%), 5 with BED (11.9%). Among these patients, only 11 did not have self-induce vomiting (i.e., had AN-R and BED) at the time of visiting the hospital. The clinical and demographic characteristics are summarised in Table 1.
As shown in Table 1, the groups were not significantly different in their age, there was a significant difference in illness duration between AN-BP and AN-R (AN-BP: 10.0 ± 6.7, AN-R: 1.9 ± 1.7).
However, the AN groups (AN-BP and AN-R) had a significantly lower BMI compared with the other groups (BN and BED) (AN-R: 15.6 ± 1.5, AN-BP: 17.0 ± 0.7; BN: 20.4 ± 2.4, BED: 24.3 ± 7.0) (H = 26.3, p < 0.01). There were also statistically significant differences between the AN groups and BN in terms of the EDE-Q global scores for clinical severity (AN-R: 15.2 ± 11.5; BN: 74.5 ± 40.8; and AN-BP: 78.8 ± 39.4). AN-BP tended to be higher than the other subtypes in terms of restricting, eating, and weight, that were the sub-items of EDE-Q, and there was a significant difference in the comparison with AN-R (Restricting: AN-BP 5.2 ± 1.0, AN-R 2.4 ± 1.5; Eating: AN-BP 4.9 ± 0.6, AN-R 2.0 ± 1.4, Weight: AN-BP 5.4 ± 0.6, AN-R 2.6 ± 1.9).
Table 1 also shows that the average of AQ total score was highest for patients with BED (32.4 ± 6.2), followed by those with AN-R (26.0 ± 7.5), AN-BP (25.3 ± 4.8), and BN (22.1 ± 6.4). The difference between BN and BED was significant for the attention to detail score (BN: 3.2 ± 2.0, BED: 7.6 ± 2.5) (H = 9.4, p < 0.05).
Clinical characteristics, EDE-Q, and AQ score by the presence or absence of self-induced vomiting
As observed in Table 2, two groups were formed: 31 patients with self-induced vomiting and 11without self-induced vomiting. Although there were no significant differences in age or BMI, there was a significant difference in illness duration between vomiting present group and vomiting absent group (BN, AN-BP: 7.3 ± 6.2, AN-R, BED: 3.1 ± 3.0).
There were no significant differences in the EDE-Q, except for the “diagnosis” category, which is expected to be affected by the frequency of self-induced vomiting. The AQ total scores of those who did not self-induced vomiting were significantly higher than for those who self-induced vomiting. In particular, the scores for social and communication skills—which are subscales of the AQ—were significantly higher in the group that did not have self-induced vomiting.
There were also statistically significant differences between the vomiting and non-vomiting groups in the EDE-Q global scores for clinical severity. Two bulimic vomiting subtypes (BN and AN-BP) scored higher than non-vomiting subtypes (U = 67.5, p < 0.01, r = 0.46).
However, BMI was also considered to be a confounder of EDE-Q and AQ scores; therefore, the patients were divided by anorexic and bulimic, so only AN-BP to AN-R and BN to BED were analysed (Table 3).
Despite the significant differences in illness duration, BMI, and age between AN-BP and AN-R, there were no differences in any of the AQ sub-items. In contrast, there was no significant difference in the illness duration, BMI, and age between BN and BED. However, there was a significant difference in the social skills, attention to detail, and communication in the AQ score.
In addition, illness duration could be a confounding factor for the EDE-Q and AQ scores. Therefore, we performed the multivariate analysis with 4 groups: illness duration of <4 years for patients with and without vomiting and illness duration of ≥4 years for patients with and without vomiting. A Kruskal–Wallis analysis was conducted to confirm these groups, because illness duration of ≤4 years or more has been the common cut-off used for determining acute vs chronic eating disorders [21] .
There were 22 patients (illness duration ≥ 4 years) and 9 patients (illness duration <4 years) in the vomiting group and 3 patients (illness duration ≥ 4 years) and 8 patients (illness duration <4 years or less) in the non-vomiting group. The results are shown in Table 4. Illness duration resulted in significant differences in AQ communication, EDE-Q diagnosis, and age. The communication score of AQ was highest in the non-vomiting group with illness duration <4 years, followed by the non-vomiting group with illness duration ≥ 4 years. The communication score of AQ in the non-vomiting group with illness duration < 4 years was significantly higher than that in the vomiting with illness duration ≥ 4 years.
AQ score cut-off value by ED subtype
Lastly, the ratio of the subtypes and the number of patients who exceeded the cut-off value of the AQ score were examined, and the difference in the ratio of the number of patients was examined using Fisher’s exact test.
Three of eight patients of BED had an AQ score above 33, which well exceeded expected value in 60% of patients with BED. By contrast, one patient with BN had an AQ ≥33 (4.3%), and this amount was below the expected value. The difference between the BN and BED groups was significant (p = 0.02) (Table.5).