Re. Glasgow et al., PERSONAL-MODEL BELIEFS AND SOCIAL-ENVIRONMENTAL BARRIERS RELATED TO DIABETES SELF-MANAGEMENT, Diabetes care, 20(4), 1997, pp. 556-561
OBJECTIVE - The specific aims of the present study were to report on t
he level of personal beliefs and social and environmental barriers acr
oss different regimen areas and patient subgroups and on the relations
hip of personal models and perceived barriers to the level of selfmana
gement RESEARCH DESIGN AND METHODS - This study focused on several iss
ues related to personal models (representations of illness) and percei
ved barriers to diabetes self-management among a large heterogeneous s
urvey sample of 2,056 adults throughout the U.S. RESULTS - Respondents
felt that diabetes was a serious disease and that their self-manageme
nt activities will control their diabetes and reduce the likelihood of
long-term complications. Most frequently reported barriers were relat
ed to dietary adherence, followed by exercise and glucose testing barr
iers. Both personal models and barriers significantly predicted level
of self-management in all three regimen areas studied (diet, exercise,
and glucose testing) after controlling for the influence of demograph
ic and medical history factors. Regimen-specific models and barriers p
roved to be stronger predictors than more global measures. Differences
on personal models and barriers were observed among different patient
groups (e.g., age, health insurance, and insulin-taking status). Poss
ible reasons for these differences and implications for intervention a
nd future research are discussed. CONCLUSIONS - Both the personal-mode
l and barriers scales had good internal consistency and predicted vari
ance in each of the self-management variables after controlling for de
mographic and medical history factors. These brief self-report persona
l-model scales demonstrated good internal reliability and were as pred
ictive of self-management as the lengthier interview-based measures in
previous studies. The assessment of the treatment effectiveness compo
nent of personal models may be sufficient for most clinical purposes.