Abstracts Health Sciences

Development and evaluation of the cancer and cardiovascular disease health belief scale and the use of the health belief scale in determining the adoption of a lowfat dietary pattern

by Nancy Carolyn Vizenor

Objectives. To develop and evaluate an instrument (Modified Health Belief Model Scale, MHBMS) to measure Health Belief Model constructs (perceived susceptibility to, and severity of, cancer and cardiovascular disease, perceived benefits of, and barriers to, decreasing dietary fat behavior), health motivation, self-efficacy, and social norms; to evaluate Fat-Related Diet Habits Questionnaire (DHQ) psychometric properties; to determine if MHBMS constructs are related to dietary fat behavior.Design. This auxiliary study was part of a randomized, controlled intensive dietary intervention trial, the Women's Health Initiative (WHI).Subjects/setting. 192 post-menopausal women (intervention = 79, comparison = 113) enrolled in the WHI Dietary Modification arm.Main outcome measures. Baseline and 6 months post-randomization dietary fat behavior.Statistical analyses/results. MHBMS and DHQ construct validity was shown using exploratory and confirmatory factor analysis. DHQ convergent validity was exhibited by assessing the correlation between dietary fat behavior and percent energy from fat (Food Frequency Questionnaire) at baseline (r = -.35, p < .01). Internal consistency (Cronbach's Alpha) ranged from .67 to .84 (MHBMS) and .42 to .78 (DHQ). Test-retest correlations ranged from .37 to .77 (MHBMS) and .36 to .75 (DHQ). At baseline, all constructs were significantly correlated with baseline dietary fat behavior score (r = -.16 to -.33) with the exceptions of self-efficacy when under stress and social norms. The best fitting regression model predicting baseline fat behavior included baseline perceived susceptibility, severity, benefits, barriers, and health motivation (R2 = .20, F = 22.73, p < .001). Baseline MHBMS constructs were not correlated with dietary fat behavior change. With one exception, perceived susceptibility (r = .24, p < .05), changes in constructs were unrelated to dietary fat behavior change.Conclusions. The newly developed MHBMS was valid and reliable. There was no evidence to support an association between MHBMS constructs and change in dietary fat behavior yet there was an association at baseline. Respondents who felt most susceptible and felt the diseases to be severe had the highest dietary fat behavior at baseline, contrary to basic HBM theory. Respondents who perceived benefits to be high, were motivated toward health, and perceived barriers to be low, had the lowest dietary fat behavior at baseline, consistent with HBM theory. These findings, especially the unexpected positive association seen with perceived susceptibility and severity, provide interesting insight into dietary fat behavior.