Regression Results of CDI and Three-Year Mortality on ALI and Lagged ALI
| Panel A: CDI | Panel B: Three-Year Mortality | |||
|---|---|---|---|---|
| (1) | (2) | (1) | (2) | |
| ALI | 0.0642*** | 0.0005*** | ||
| (0.0018) | (0.0001) | |||
| ALIa−t | 0.0714*** | 0.0004*** | ||
| (0.0021) | (0.0001) | |||
| Controls | Yes | Yes | Yes | Yes |
| Observations | 120,083 | 120,083 | 92,369 | 92,369 |
| R-squared | 0.158 | 0.158 | 0.097 | 0.096 |
Notes: The analysis includes individuals who participated continuously in Lifelines for two waves, in order to obtain lagged health outcomes. The sample size in Panel B is reduced due to partly missing three-year mortality information in Lifelines wave 3a (see Section III.C for details). Panel A is estimated using a linear regression model, while Panel B employs a logistic regression model. Accordingly, we report average marginal effects for the latter. We present the R-squared for Panel A and the pseudo R-squared for Panel B. Robust standard errors are reported in parentheses. Control variables include age, age-squared, age group, gender, cohort, urban residency, province, and survey year. *p < 0.1, **p < 0.05, ***p < 0.01.