Objectives: The purpose of this study was to clarify the relationship of the annual transition of implementation of nonsmoking at eating and drinking establishments with indices of population/household and economy/labor by prefecture.
Methods: The prefectural rates of eating and drinking establishments implementing nonsmoking (hereafter, nonsmoking rate) were computed in a year using the data from "Tabelog®". Forty-seven prefectures were classified by hierarchical cluster analysis into "prefecture clusters" 1 to 5 in descending order of the median of nonsmoking rates. The indices of population/household (e.g., percentage of the population aged 65 years and over and percentage of nuclear family household) and economy/labor (e.g., prefectural income per capita and percentage of construction and mining workers) were classified by hierarchical cluster analysis into 11 "index clusters", and the representative index in each index cluster was extracted from the results of the Jonckheere-Terpstra test. An ordinal logistic regression analysis was performed using the numbers 1 to 5 of prefecture clusters as dependent variables and the indices representing the index clusters as independent variables.
Results: The percentage of the population aged 65 years and over and the percentage of construction and mining workers were positively related to the order of prefectural clusters.
Conclusions: To promote implementation of nonsmoking in eating and drinking establishments in prefectures especially in those with larger numbers of elderly people and construction and mining workers, it is important to inform the persons in charge that implementation of nonsmoking does not affect the number of customers.
Keywords: cluster analysis; eating and drinking establishments; smoking cessation; trend in a year.