Abstract
Objective: According to research on multiple health behavior changes, co-variation is defined as a phenomenon in which one behavior increases the probability of another behavior. This study examined the co-variation of multiple eating behaviors using readiness for behavioral changes.
Methods: In this cross-sectional study, questionnaire data were collected from 4,861 health insurance union members in Japan from August to September 2010. The data comprised information on the prevalence of five healthy eating behaviors (eating breakfast, eating in moderation, not eating between meals, not eating 2 hours before sleeping, and chewing well when eating), sex, and age. Each behavior was categorized into the pre-action or action/maintenance stage based on the transtheoretical model. The relationship between each behavior and the stage of change of other behaviors was assessed by multinomial logistic regression analysis.
Results: A total of 3,571 respondents (1,532 men, 42.9%; 2,039 women, 57.1%) answered the questionnaire (response rate, 73.5%). Among men, co-variation was observed mainly between the action of “eating breakfast” and other behaviors (“not eating 2 hours before sleeping”: odds ratio [OR]=2.2; 95% confidence interval [95%CI]=1.41-3.49). Among women, the action of “not eating 2 hours before sleeping” led to co-variation with “not eating between meals” which led to negative co-variation with “eating breakfast” (OR=0.41; 95%CI=0.19-0.89). All behaviors co-occurred, happened at the same time, with others both among men and women.
Conclusion: Co-variation with other behaviors was observed mainly in “eating breakfast” among men and “not eating 2 hours before sleeping” among women. Among women, “not eating between meals” led to negative co-variation.
