Table 3.
Earthquake exposure and risk of depressive symptoms.
| OR (95% CI) | P value | |
|---|---|---|
| DID | ||
| Birth cohorts × earthquake areas | ||
| Birth cohort (1937–1951) and non-affected area (reference) | 1.00 | |
| Birth cohort (1952–1966) and affected area | 0.90 (0.82, 0.98) | 0.019 |
| Binary logistic regression | ||
| Birth cohort (1952–1966) | ||
| Earthquake areas | ||
| Not affected (reference) | 1.00 | |
| Affected | 0.48 (0.25, 0.94) | 0.031 |
| Multilevel logistic regression | ||
| Birth cohort (1952–1966) | ||
| Earthquake areas | ||
| Not affected (reference) | 1.00 | |
| Affected | 0.47 (0.23, 0.95) | 0.034 |
Note: The models adjusted for age, sex, marital status, ADLs, and life satisfaction. And the DID model also controlled for city fixed effects. DID = difference-in-difference; ADLs = activities of daily living. Birth cohort (1952–1966) indicates persons exposed to the earthquake during adolescence (10–24 years old), and birth cohort (1937–1951) indicates persons exposed to the earthquake at older ages (25–39 years old).