• binge eating disorder;
  • obesity;
  • ethnicity;
  • Black women;
  • bias



The current study investigated sampling bias as it affects recruited clinic samples of Black and White women with binge eating disorder (BED).


Clinical characteristics of a recruited clinic sample (35 Black and 302 White consecutively evaluated women) with BED were compared with a community sample of Black and White women with BED drawn from the New England Women's Health Project. The clinic and community groups met the same definition of BED and were assessed with identical methods.


Among White women, the clinic and community samples differed on some features (higher body mass index [BMI], and greater eating concerns and shape concerns in the clinic sample) but these differences reflected small to moderate effects sizes. In contrast, among Black women, the clinic sample had substantially higher levels (large effect sizes) of several features of eating disorders (eating concerns, dietary restraint, and shape concern), higher (moderate effect size) BMI, but lower frequency of binge eating (moderate effect size) than the community sample. A comparison of Black and White women within the clinic sample revealed little difference in clinical presentation, except for the significantly higher BMI among Black women.


A sampling bias appears to exist between both White and Black recruited clinic samples of women with BED, although the bias appears to be substantially greater for Black women. Black women with BED who seek treatment have significantly different characteristics than those who do not. © 2005 by Wiley Periodicals, Inc.