Fig. 2From: Association between health risks and frailty in relation to the degree of housing damage among elderly survivors of the great East Japan earthquakea Age-adjusted changes in lifestyle and psychosocial factors by degree of housing damage†in men (n = 679). †Degree of housing damage: extensive damage (solid line); partial damage (dashed line); and no damage (dotted line). All factors a) through i), p < .001 by Cochran–Mantel–Haenszel tests for trends adjusted for age. A significant result on this test means that after controlling for age and time of year, the proportion of people who reported a certain factor was different depending on the degree of housing damage. b Age-adjusted changes in lifestyle and psychosocial factors by degree of housing damage†in women (n = 850). †Degree of housing damage: extensive damage (solid line); partial damage (dashed line); and no damage (dotted line). Factors c) through i), p < .001 by Cochran–Mantel–Haenszel tests for trends adjusted for age. A significant result on this test means that after controlling for age and time of year, the proportion of people who reported a certain factor was different depending on the degree of housing damageBack to article page