Table 1.

Mixed effects models for repeated measures for the relative reductions in the median radius of gyration with respect to the baseline.

independent variables [estimates (s.e.)]	(A)	(B)	(C)
dependent variable: relative reduction of rg in pre-lockdown
attack rate	1.73* (0.77)	1.88* (0.81)	1.95* (0.82)
population density	1.04* (0.51)	1.51** (0.49)	1.36* (0.6)
old age index	0.72 (0.55)	—	0.97 (0.6)
high education	0.09 (0.49)	—	0.29 (0.52)
no-profit org.	−1.15* (0.57)	—	−0.90 (0.84)
unemployment	—	0.35 (0.68)	0.40 (0.84)
commuters	—	−0.56 (0.64)	−0.26 (0.67)
agriculture ratio	—	0.53 (0.57)	0.70 (0.63)
industry ratio	—	0.51 (0.66)	0.83 (0.7)
intercept	14.21*** (3.04)	14.21*** (3.02)	14.21*** (3.00)
no. of days	16	16	16
AIC	12286.70	12289.29	12291.74
dependent variable: relative reduction of rg in lockdown
attack rate	−0.23 (0.28)	0.61 (0.37)	1.03** (0.37)
population density	2.92*** (0.2)	2.38*** (0.18)	2.96*** (0.22)
old age index	0.69*** (0.20)	—	0.62** (0.22)
high education	0.92*** (0.18)	—	0.96*** (0.19)
no-profit org.	0.14 (0.25)	—	1.13*** (0.31)
unemployment	—	0.76* (0.31)	2.50*** (0.4)
commuters	—	−0.61* (0.24)	−0.20 (0.25)
agriculture ratio	—	−1.57*** (0.21)	−0.86*** (0.24)
industry ratio	—	−0.26 (0.25)	0.30 (0.26)
intercept	64.68*** (1.09)	64.68*** (1.07)	64.68*** (1.07)
no. of days	38	38	38
AIC	25428.28	25413.20	25367.58
dependent variable: relative reduction of rg in phase 2
attack rate	−1.24** (0.46)	1.31* (0.62)	1.75** (0.64)
population density	2.29*** (0.42)	2.6*** (0.38)	3.00*** (0.46)
old age index	1.21** (0.43)	—	0.80 (0.46)
high education	0.94* (0.38)	—	1.31*** (0.41)
no-profit org.	−1.99*** (0.52)	—	0.63 (0.65)
unemployment	—	2.95*** (0.66)	4.68*** (0.87)
commuters	—	−0.65 (0.49)	−0.15 (0.51)
agriculture ratio	—	−0.25 (0.48)	0.51 (0.48)
industry ratio	—	−1.26* (0.51)	−0.57 (0.53)
intercept	38.92*** (1.36)	38.92*** (1.37)	38.92*** (1.37)
no. of days	14	14	14
AIC	9952.55	9922.64	9912.82

Significance levels: *p < 0.05; **p < 0.01; ***p < 0.001.