Personal Weight Beliefs Predictive of Eating Behavior Patterns and Actual Weight in Young African-American Women

Sunday, 27 July 2014

Stephanie Pickett, PhD, MSN, BSN, RN
Center for Health Equity Research and Global Women's Health Research, University of Pennsylvania , School of Nursing, Philadelphia, PA


Over 80% of African American (AA) women in the United States (US) are overweight or obese.  Weight management in AA women is multifaceted and not completely understood.  However, cultural beliefs about weight suggest that many AA women hold beliefs about their weight that may interfere with a desire to engage in behaviors for weight management, it is not known whether personal weight beliefs are predictive of weight management behavior.  Personal weight beliefs were defined as a multidimensional concept consisting of three domains including the descriptive characteristics, causal attributions, and consequences of one’s personal weight.  The objective of this study was to determine whether personal weight beliefs were predictive of eating behavior patterns and weight in young AA women.


A sample of one hundred-fifty AA women, 18-40 years old (mean age = 28.4, mean years of education =13.78, mean BMI = 31.5) in Metropolitan Detroit completed a preliminary version of a newly developed Beliefs about Personal Weight Survey (BPW) based on the definition of personal weight beliefs.  Responses from BPW were analyzed by domain using principal component analysis, composites were developed from components.  The Eating Behavior Pattern Questionnaire (EBPQ) was used to measure eating behavior patterns (low fat diet, snacking on sweets, haphazard meal planning, and meal skipping and emotional eating).  BMI was calculated from measured weight and height.  Multiple linear regressions using stepwise entry was used to examine the ability of personal weight beliefs to predict eating behavior patterns and BMI.


Stepwise regression results show that the weight belief composites significantly predicted all five of the eating behavior patterns and BMI.  Two composites were most predictive.  The descriptive weight composite (describing personal weight as overweight) explained 51% of the variance in BMI (F = 154.16, p =.001).  Causal belief composites (unhealthy eating behaviors, not eating healthy behaviors, and stress factors) explained 34% of the variance in emotional eating (F = 24.91, p = .001).


Beliefs about the descriptive characteristics, and causal attributions, of one’s personal weight predicted eating behavior patterns and BMI in young AA women.