Send to

Choose Destination
See comment in PubMed Commons below
Eat Behav. 2001 Summer;2(2):151-68.

Exercise and eating disorders: a 3-year inpatient hospital record analysis.

Author information

  • 1Eat-N-Play Nutritional Fitness, Inc, Lawton, OK 73505, USA.


The purpose of this study was to describe the exercise behavior across the DSM-IV eating disorder diagnosis (anorexia nervosa (AN), bulimia nervosa (BN), and eating disorder not otherwise specified (EDNOS)) and determine if differences exist among exercise category (high level and low level), type of exercise (aerobic, strength, and total), Eating Attitudes Test (EAT), Eating Disorders Inventory (EDI-2), and the length of inpatient hospitalization. This study was a retrospective analysis of 199 inpatient hospital records of female patients (age M=20.6, S.D.=7.03) in the following diagnostic categories: AN (n=115, 58%), BN (n=38, 19%), and EDNOS (n=46, 23%). Patients were also grouped by a median split into low- and high-level exercise categories based on their total amount of weekly exercise. No significant relationship was found between diagnosis and exercise category (high and low) or diagnosis and exercise type (total, aerobic, and strength exercise). This is valuable because exercise specialist may be a worthwhile addition to treatment teams working with all eating disorder diagnostic groups. The high-level exercisers showed significantly greater scores on the EAT [F(2,1,110)=5.117, P<.05] and weight preoccupation [F(1,1010)=4.861, P<.05] than the low-level exercise groups. There were significant relationships between total exercise and EAT (r=.354, P<.01), drive for thinness (r=.262, P<.01), and body dissatisfaction (r=.312, P<.001). Total aerobic exercise was significantly related to EAT (r=.352, P<.001), drive for thinness (r=.272, P<.01), and body dissatisfaction (r=.268, P<.01). Finally, total strength exercise was related to body dissatisfaction (r=.249, P<.05). A significant difference was found among exercise group and length of hospitalization [F(1,155)=14.384, P<.000]. The high-level exercisers were in treatment for 3 weeks longer than their low-level exercisers. A forced stepwise multiple linear regression was calculated to predict the length of hospitalization. At Step 1, diagnosis and Body Mass Index (BMI) on admission explained 27% of the variance in length of inpatient hospitalization [F(4,154)=18.08, P<.004, r(2)=.273]. When two exercise variables were included, aerobic activity explained an additional 5% of variance, whereas strength activity explained an additional 0.3% of the variance. Three-fourths of the predictors explained significant components of the total 32% variance. The present study demonstrates that patients with greater total exercise and a greater total aerobic exercise show a significantly greater drive for thinness and require a longer length of hospitalization. This association was not present with total strength exercise levels. Effective clinical methods to modify total aerobic exercise levels need to be determined with inpatient populations.

PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for Elsevier Science
    Loading ...
    Support Center