Table. Model Parameter Search Bounds and Calibrated Values.
| Parameter | Prior Search Bounds, Uniform Distribution | Calibrated Values, Mean (95% UI) | Source |
|---|---|---|---|
| Naive incidencea | |||
| AN | Age- and sex-specific | NAb | Global Burden of Disease Study 20176 |
| BED | Age- and sex-specific | NAb | Stice et al7 report 8-y cumulative incidence of 2.6% for female patients; age-specific estimates were imputed according to Global Burden of Disease Study 20176 curves |
| BN | Age- and sex-specific | NAb | Global Burden of Disease Study 20176 |
| OSFED | Age- and sex-specific | NAb | Stice et al7 report 8-y cumulative incidence of 24.2% for female patients; age-specific estimates were imputed according to Global Burden of Disease Study 20176 curves |
| First-year relapse probability, %a | |||
| AN | 5-40 | Male: 21.3 (2.9-52.7) | Berends et al11 report overall relapse rate of 31% |
| Female: 20.4 (2.3-55.6) | |||
| BED | 5-40 | Male: 9.8 (0.7-28.2) | Hudson et al10 report relapse rate of 32.1% within 6 mo |
| Female: 25.2 (3.6-52.0) | |||
| BN | 5-40 | Male: 25.0 (2.2-56.9) | Olmsted et al8 report relapse rate of 31% at 2 y; Olmsted et al9 report relapse rate of 27.6% within 6 mo of treatment completion |
| Female: 17.7 (1.4-47.0) | |||
| OSFED | 5-40 | Male: 12.2 (0.1-50.6) | Stice et al7 report relapse rate of 20%-30% for EDNOS subtypes |
| Female: 13.1 (0.1-52.1) | |||
| Relapse decay parametera | |||
| AN | 0-1 | Male: 0.52 (0.07-0.94) | Uniform prior |
| Female: 0.52 (0.05-0.94) | |||
| BED | 0-1 | Male: 0.43 (0.01-0.95) | |
| Female: 0.57 (0.07-0.95) | |||
| BN | 0-1 | Male: 0.49 (0.03-0.97) | |
| Female: 0.49 (0.03-0.89) | |||
| OSFED | 0-1 | Male: 0.49 (0.02-0.93) | |
| Female: 0.42 (0.05-0.90) | |||
| Annual probability of remission without treatment, %a | |||
| AN | 0-10 | Male: 7.8 (0.2-20.1) | Bergh et al12 and Hay et al13; based on control groups in trial |
| Female: 8.5 (0.9-20.3) | |||
| BED | 0-10 | Male: 16.3 (6.4-27.3) | |
| Female: 7.0 (0.3-21.5) | |||
| BN | 0-10 | Male: 8.5 (1.4-18.9) | |
| Female: 10.0 (2.1-21.7) | |||
| OSFED | 0-10 | Male: 10.9 (1.6-24.6) | |
| Female: 13.6 (3.6-27.2) | |||
| Treatment proportion, %a | |||
| AN | Male: 0-50 | Male: 28.2 (2.4-49.0) | Hudson et al4; assumed annual proportion is 50% of lifetime |
| Female: 0-30 | Female: 14.0 (1.9-28.9) | ||
| BED | Male: 0-51 | Male: 30.1 (4.8-50.2) | Hudson et al4; prior search bounds based on 95% CI |
| Female: 10-53 | Female: 29.7 (11.0-52.2) | ||
| BN | Male: 0-30 | Male: 15.7 (1.3-28.5) | Hudson et al4; for male patients, assumed annual proportion is 50% of lifetime, and for female patients, the proportion is based on 95% CI |
| Female: 0-37 | Female: 17.7 (1.9-35.8) | ||
| OSFED | Male: 0-30 | Male: 16.5 (2.1-28.5) | Same prior search bounds as for AN used |
| Female: 0-30 | Female: 16.7 (1.4-28.8) | ||
| Treatment efficacy, relative rate of remissiona | |||
| AN | 1.5-10 | Male: 7.59 (1.32-19.83) | Bergh et al12 report risk ratio of 7.69 (95% CI, 1.67-33.3) |
| Female: 8.00 (1.73-18.73) | |||
| BED | 1.2-2.0 | Male: 2.08 (1.36-3.37) | Hay et al13 report risk ratio of 1.59 (95% CI, 1.20-2.08) for BED and BN |
| Female: 1.81 (1.10-2.86) | |||
| BN | 1.2-2.0 | Male: 1.75 (1.07-2.92) | |
| Female: 1.74 (1.10-3.13) | |||
| OSFED | 1.2-2.0 | Male: 1.80 (1.07-3.01) | Same prior search bounds as for BED and BN used |
| Female: 1.83 (1.05-3.01) | |||
| Annual eating disorder crossovers, %a | |||
| AN to BED | Male: 0-1.5 | Male: 1.2 (0.1-3.0) | Relative probabilities informed by PEDSnet analysis by Rodriguez et al14 |
| Female: 0-1.5 | Female: 1.2 (0.1-3.1) | ||
| AN to BN | Male: 0-0.03 | Male: 0.0 (0.0-0.1) | |
| Female: 0-0.4 | Female: 0.3 (0.0-0.9) | ||
| AN to OSFED | Male: 0-4.5 | Male: 4.0 (0.2-10.3) | |
| Female: 0-4.5 | Female: 3.8 (0.4-10.4) | ||
| BED to AN | Male: 0-1.3 | Male: 1.1 (0.2-2.9) | |
| Female: 0-1.2 | Female: 0.5 (0.0-1.6) | ||
| BED to BN | Male: 0-0.2 | Male: 0.2 (0.0-0.4) | |
| Female: 0-0.4 | Female: 0.3 (0.0-0.9) | ||
| BED to OSFED | Male: 0-4.9 | Male: 5.4 (0.2-12.1) | |
| Female: 0-4.7 | Female: 3.4 (0.2-8.2) | ||
| BN to AN | Male: 0-0.9 | Male: 0.7 (0.1-1.9) | Assumed that BED estimates accounted for 25% of OSFED prevalence on the basis of lifetime prevalence estimates from Hudson et al4 and Stice et al7 |
| Female: 0-2.1 | Female: 1.3 (0.1-3.4) | ||
| BN to BED | Male: 0-1.7 | Male: 1.4 (0.1-3.7) | |
| Female: 0-1.7 | Female: 1.4 (0.1-3.1) | ||
| BN to OSFED | Male: 0-5.2 | Male: 4.5 (0.5-9.9) | |
| Female: 0-5.0 | Female: 4.5 (0.4-10.1) | ||
| OSFED to AN | Male: 0-4.0 | Male: 0.4 (0.0-1.6) | |
| Female: 0-3.6 | Female: 0.1 (0.0-0.6) | ||
| OSFED to BED | Male: 0-1.6 | Male: 0.6% (0.1-1.6) | |
| Female: 0-1.6 | Female: 0.7 (0.0-2.2) | ||
| OSFED to BD | Male: 0-0.6 | Male: 0.1 (0.0-0.4) | |
| Female: 0-1.3 | Female: 0.4 (0.0-1.2) | ||
| Mortalitya | |||
| Background mortality | Age- and sex-specific | US life tables (Arias et al15) | |
| AN, SMR | 4.17-8.26 | Male: 6.71 (1.43-13.12) | Arcelus et al3 report SMR of 5.86 (95% CI, 4.17-8.26) |
| Female: 6.31 (1.33-12.44) | |||
| BED, SMR | 1.46-2.52 | Male: 2.01 (1.08-3.53) | Ackard et al16 report suicide ideation or attempt odds ratios; Arcelus et al3 report SMR of 1.92 (95% CI, 1.46-2.52) for EDNOS |
| Female: 2.15 (1.10-3.68) | |||
| BN, SMR | 1.44-2.59 | Male: 2.26 (1.19-3.49) | Arcelus et al3 report SMR of 1.93 (95% CI, 1.44-2.59) |
| Female: 2.31 (1.09-3.64) | |||
| OSFED, SMR | 1.46-2.52 | Male: 2.17 (1.15-3.81) | Arcelus et al3 report SMR of 1.92 (95% CI, 1.46-2.52) for EDNOS |
| Female: 2.07 (1.12-3.67) | |||
Abbreviations: AN, anorexia nervosa; BED, binge eating disorder; BN, bulimia nervosa; EDNOS, eating disorder not otherwise specified; NA, not applicable; OSFED, other specified feeding and eating disorder; SMR, standardized mortality ratio; UI, uncertainty interval.
a
See eAppendix 2 in the Supplement for more details on how the parameters were calculated.
b
See eFigure 1 in the Supplement for more details on how the parameters were calculated.