Table 2.
Repartition of the misclassification of the observed silhouette, according to the “Expected” silhouette, for the women of the E3N cohort (n=91,815).
“Expected” body silhouette | N | Lower
|
Realistic
|
Higher
|
|||
---|---|---|---|---|---|---|---|
N | % | N | % | N | % | ||
1 | 632 | 24 | 3.8 | 608 | 96.2 | 0 | 0.0 |
1.5b | 1,259 | 98 | 7.8 | 1,161 | 96.2 | 0 | 0.0 |
2 | 6,991 | 67 | 1.0 | 6,867 | 98.2 | 57 | 0.8 |
2.5 | 8,405 | 182 | 2.2 | 7,899 | 94.0 | 324 | 3.8 |
3 | 34,509 | 532 | 1.5 | 33,677 | 97.6 | 300 | 0.9 |
3.5 | 9,733 | 432 | 4.4 | 9,014 | 92.6 | 287 | 3.0 |
4 | 15,024 | 210 | 1.4 | 14,545 | 96.8 | 269 | 1.8 |
4.5 | 1,049 | 17 | 1.6 | 890 | 84.8 | 142 | 13.6 |
5 | 5,925 | 18 | 0.3 | 5,255 | 88.7 | 652 | 11.0 |
5.5 | 3,233 | 27 | 0.8 | 2,033 | 62.9 | 1,173 | 36.3 |
6 | 4,419 | 11 | 0.3 | 3,518 | 79.6 | 890 | 20.1 |
<6 | 636 | 0 | 0.0 | 0 | 0.0 | 636 | 100 |
Total | 91,815 | 1,618 | 1.8 | 85,467 | 93.1 | 4,730 | 5.1 |
Expected body silhouette was found fitting the distribution of the silhouette estimated by technicians, on the validation sample, to the whole cohort sorted by increasing BMI.
intermediate silhouettes in case of two adjacent choices.