Objective: To examine the co-occurrence of multiple eating behaviors among Japanese adults by combination of presence and absence of healthy eating behaviors.
Methods: In this cross-sectional study, questionnaire data were collected from 4,861 health insurance union members of a company in Japan, from August to September 2010. The data comprised information on the prevalence of 5 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 categorized into 2 groups according to its presence or absence of healthy eating behaviors. Then, χ
2 tests were performed to examined the differences between sexes. We calculated the number and 2 proportions (observed, expected) of all possible 32 combinations. The ratio of the observed to expected prevalence was used to examine the co-occurrence of multiple eating behaviors.
Results: Among the 3,645 subjects who answered the questionnaire (response rate, 75.0%), 3,525 subjects (1,520 men, 2,005 women) were included in the analysis. In both men and women, the combination of the presence of all healthy eating behaviors was most frequently observed (men: 19.3%, women: 17.5%). The highest ratio of the observed to expected prevalence was demonstrated for the absence of the combination of all healthy eating behaviors (men: 7.6, women: 4.5); the second highest ratio was observed for the prevalence of only “eating breakfast” with the absence of the other eating behaviors (men: 3.1, women: 2.6).
Conclusion: In both men and women, the highest ratio of observed to expected was demonstrated for the absence of all healthy eating behaviors; followed by that for the prevalence of “eating breakfast” alone with the other healthy eating behaviors.
View full abstract