Objective: The purpose of this retrospective study was to replicate and to
extend previous work indicating that bulimia nervosa is associated with ele
vated serum levels of cholesterol and to examine the association between ch
olesterol and other clinical and serum measures. Methods: Clinical characte
ristics and laboratory data including nonfasting cholesterol and thyroid in
dices were collected from charts of 119 women with bulimia nervosa and 42 f
emale controls. Results: There was no significant difference between the tw
o groups in mean age or body mass index. The mean serum cholesterol level o
f the patient group was statistically significantly greater than that of th
e control group (patients: 194 +/- 36 mg/dl; controls: 176 +/- 34 mg/dl; t
= 2.77; df = 159; p = .006). The mean calculated free T4 of the bulimia ner
vosa patient group was statistically lower than that of the control group (
patients: 1.89 +/- 0.32; controls: 2.09 +/- 0.26; t = 2.86; df = 141; p = .
005). There was no correlation between calculated free T4 and serum cholest
erol within the combined group (r = .04, p > .5, N = 143) or within the pat
ient group (r = .08, p > .3, N = 118). Conclusions: The average serum level
of cholesterol is elevated in the patients with bulimia nervosa. The mecha
nism for and the consequences of this abnormality are uncertain. (C) 2001 b
y John Wiley & Sons, Inc.