Table 4.

Risk factors for food insecurities in two rural areas with different demographic, economic, and food environment factors.

Risk Factors	Odds Ratios	95% Confidence Limits
Mountain area
Objective economic indices
Average monthly income for last 1 year, $	0.995	0.991	0.999
Subjective economic indices
Experience of reducing food expenditure resulting from burden of heating costs during the winter (reference ‘no’)	5.664	1.518	21.136
Household food availability
Farming or home gardening (reference ‘no’)	0.022	0.002	0.283
Community food accessibility, perception
Having difficulties in food purchasing due to food stores far from home (reference ‘no’)	48.58	4.83	488.69
Land Area
Objective economic indices
Percent proportion of housing fee	1.023	1.01	1.036
Household food accessibility
Household cook’s physical disability (reference ‘no’)	2.846	1.074	7.538
Community food accessibility, perception
Having difficulties in food purchasing due to food stores far from home (reference ‘no’)	4.675	2.05	10.66

Odds ratios and 95% confidence limits of risk factors for households’ food insecurity. A stepwise approach was applied to select the most explainable risk factors in a model with potential independent variables that were selected based on their association with food insecurities in each area (α = 0.15). Potential independent variables in multivariate models for the mountain area were the education years of the households’ cook (<6 years or ≥6 years), average monthly income, experience of reducing food expenditure resulting from the burden of non-food expenses, such as housing fees or heating costs in winter, food acquisition by farming/ home gardening (yes or no), participating in a public food assistance program (yes or no), purchasing foods in a transitional market (yes or no), or difficulties in food purchasing due to food stores being far from home (yes or no); potential independent variables in multivariate models for the land area were the education years of the households’ cook (<6 years or ≥6 years), households’ income with earnings ($), average monthly expenditure ($), percent proportions of housing fees, percent proportions of heating costs, percent proportions of medical expenditures, experience of reducing food expenditure resulting from the burden of non-food expenses, such as housing fees (yes or no) or heating costs in winter (yes or no), food acquisition by farming/ home gardening (yes or no), physical disabilities of the households’ cook, no variety of foods in the nearest food store, difficulties in food purchasing due to food stores being far from home (yes or no), or inconvenient bus route (yes or no).