Table 2. Nine methods used in the recent literature for calculating the AN weight cut-off and the percentage (± standard error) of a population-based and treatment-seeking sample meeting the AN weight criterion under each method.
| Method for determining AN weight cut-off | Population-based sample (n = 1201) | Treatment-seeking sample (n = 189) | ||
|---|---|---|---|---|
| Percentage (s.e.) meeting AN weight criterion | Comparable methods (ppairwise diff > 0.001) | Percentage (s.e.) meeting AN weight criterion | Comparable methods (ppairwise diff > 0.001) | |
| BMI < 16.5 | 0.23 (0.05) | None | 48.1 (3.6) | None |
| BMI < 17.5 | 0.91 (0.11) | None | 70.9 (3.3) | BMI < 18.0 |
| Kemsley (1951/2) | ||||
| Fogarty (1975) | ||||
| BMI < 18.0 | 1.33 (0.13) | Kemsley (1951/2) | 76.2 (3.1) | BMI < 17.5 |
| Fogarty (1975) | Kemsley (1951/2) | |||
| Fogarty (1975) | ||||
| Met Life (1959) | ||||
| Weight < 85% of 1951/2 | 1.50 (0.11) | BMI < 18.0 | 71.0 (3.4) | BMI < 17.5 |
| Kemsley Average Body Weights | Fogarty (1975) | BMI < 18.0 | ||
| Met Life (1959) | Fogarty (1975) | |||
| Met Life (1959) | ||||
| Weight < 85% of Fogarty | 1.47 (0.11) | BMI < 18.0 | 71.4 (3.3) | BMI < 17.5 |
| 1975 Average Desirable Weight | Kemsley (1951/2) | BMI < 18.0 | ||
| Met Life (1959) | Kemsley (1951/2) | |||
| Met Life (1959) | ||||
| Weight < 85% of 1959 | 1.61 (0.14) | Kemsley (1951/2) | 77.0 (3.1) | BMI < 18.0 |
| Metropolitan Life Insurance Tables medium-frame mean | Fogarty (1975) | Kemsley (1951/2) | ||
| Fogarty (1975) | ||||
| Weight < 85% of 1983 | 3.48 (0.19) | None | 87.2 (2.4) | 10th centile BMI |
| Metropolitan Life Insurance Tables medium-frame mean | DHEW (1979) | |||
| BMI < age-adjusted BMI at 10th percentile | 6.64 (0.31) | None | 84.9 (2.6) | Met Life (1983) |
| DHEW (1979) | ||||
| Weight < 85% of DHEW | 10.1 (0.32) | None | 89.4 (2.3) | Met Life (1983) |
| 1979 age-adjusted 50th percentile | 10th centile BMI | |||
AN, Anorexia nervosa; BMI, body mass index; DHEW, Department of Health, Education, and Welfare; Met Life, Metropolitan Life Insurance Company.
Within each row, ‘comparable methods’ represent methods that did not exhibit significant differences in proportions based on pairwise McNemar tests with α set to 0.001 to control for family-wise error.