Skip to main content
. 2006 Feb;12(2):323–325. doi: 10.3201/eid1202.050268

Table. Climate and socioeconomic variables associated with the number of cases of verocytotoxin-producing Escherichia coli* by multiple linear regression†.

Climate‡ and socioeconomic§ variable t p value
Average air temperature of the day (°C) 9.72 <0.001
Wind speed (m/s) 4.69 <0.001
No. sunny days –1.91 Not significant
Average no. persons in a household¶ 6.30 <0.001
Population density 8.61 <0.001
% children (<15 years of age) 2.69 0.007
% elderly (>65 years of age) 20.70 <0.001
Average income# –10.43 <0.001
Beef cattle/population** 2.71 0.007
Chicken/population –3.36 0.001

*5,580 of 13,489 weeks (287 weeks × 47 prefectures) or 41.4 % of the weeks were included in the analysis as no cases of E. coli were reported during 7,909 weeks (58.6%).
†R2 = 0.31: calculated based on the multiple linear regression model using the 7 socioeconomic variables, 3 climate variables, and calendar months.
‡Data for the 2-week period before the week E. coli was reported were used to approximate the period between infection and diagnosis.
§Annual data in each prefecture were used.
¶Correlation between average no. persons in a household and population density was –0.4.
#Correlation between average income and population density, average no. persons in a household, and percentage of elderly was 0.9, –0.4, and –0.6, respectively.
**Beef cattle/population had a strong correlation with hog/population.