Skip to main content
. 2012 Mar 2;2(2):e000425. doi: 10.1136/bmjopen-2011-000425

Table 2.

Adjusted prefecture-level variance for all-cause premature mortality, Japan, 1970–2005*

Men
Women
Variance (on logit scale) Range of OR Variance (on logit scale) Range of OR
Estimate (95% CI) Estimate (95% CI)
Overall 0.003 (0.001 to 0.004) 0.87–1.13 0.005 (0.003 to 0.007) 0.84–1.11
1970 0.003 (0.002 to 0.005) 0.89–1.12 0.007 (0.004 to 0.010) 0.79–1.14
1975 0.003 (0.001 to 0.004) 0.88–1.09 0.007 (0.004 to 0.010) 0.82–1.19
1980 0.004 (0.002 to 0.005) 0.82–1.11 0.005 (0.003 to 0.008) 0.85–1.15
1985 0.003 (0.001 to 0.004) 0.85–1.09 0.005 (0.002 to 0.007) 0.86–1.13
1990 0.003 (0.002 to 0.004) 0.89–1.11 0.004 (0.002 to 0.006) 0.88–1.10
1995 0.006 (0.003 to 0.009) 0.85–1.22 0.008 (0.004 to 0.012) 0.80–1.15
2000 0.007 (0.004 to 0.010) 0.84–1.25 0.010 (0.005 to 0.015) 0.76–1.15
2005 0.011 (0.007 to 0.016) 0.81–1.27 0.012 (0.007 to 0.017) 0.75–1.18
*

We adjusted for age (5 year categories) and occupations. We further adjusted for time-trends in the overall model.

The range of adjusted ORs for mortality in each prefecture is shown. The reference is the grand mean of all the prefectures.

The variance between 46 prefectures is shown because the data for Okinawa prefecture were not available in 1970.