
Figure 1
Distribution of floods in Thailand, 2017–2022.
Note: Khon Kaen (40) and Songkhla (90) are two provinces that experienced disasters with frequencies between 5 and 8, while all other provinces experienced frequencies of 4 or below. Details can be found in Appendix B.
Table 1
Descriptive statistics of participants.
| 2017 WAVE (N = 2,168) | 2020 WAVE (N = 3,088) | 2022 WAVE (N = 4,802) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| MEAN | SD | N | MEAN | SD | N | MEAN | SD | N | |
| Age | 65.9100 | 11.3297 | 2168 | 69.4090 | 11.3482 | 3088 | 66.7791 | 11.4521 | 4802 |
| Gender (Male = 1; Female = 0) | 0.4515 | 0.4978 | 2168 | 0.4061 | 0.4912 | 3088 | 0.3813 | 0.4858 | 4802 |
| Married (Yes = 1; No = 0) | 0.6752 | 0.4684 | 2168 | 0.5852 | 0.4928 | 3088 | 0.5796 | 0.4937 | 4802 |
| Working (Yes = 1; No = 0) | 0.3904 | 0.4880 | 2168 | 0.3277 | 0.4695 | 3088 | 0.4640 | 0.4988 | 4802 |
| Years of education | 6.8080 | 2.8391 | 2168 | 6.5855 | 2.7680 | 3088 | 7.2422 | 3.0787 | 4802 |
| Religion | |||||||||
| No religion (Yes = 1; No = 0) | 0.0000 | 0.0000 | 2168 | 0.0006 | 0.0254 | 3088 | 0.0002 | 0.0144 | 4802 |
| Buddhist (Yes = 1; No = 0) | 0.9180 | 0.2741 | 2168 | 0.9129 | 0.2820 | 3088 | 0.9263 | 0.2613 | 4802 |
| Christian (Yes = 1; No = 0) | 0.0020 | 0.0479 | 2168 | 0.0039 | 0.0622 | 3088 | 0.0050 | 0.0705 | 4802 |
| Muslims (Yes = 1; No = 0) | 0.0790 | 0.2705 | 2168 | 0.0823 | 0.2748 | 3088 | 0.0660 | 0.2483 | 4802 |
| Other (Yes = 1; No = 0) | 0.0000 | 0.0000 | 2168 | 0.0003 | 0.0180 | 3088 | 0.0025 | 0.0499 | 4802 |
| Living alone (Yes = 1; No = 0) | 0.2384 | 0.4262 | 2168 | 0.1308 | 0.3373 | 3088 | 0.1143 | 0.3182 | 4802 |
| Urban residents (Yes = 1; No = 0) | 0.4851 | 0.4999 | 2168 | 0.4812 | 0.4997 | 3088 | 0.5702 | 0.4951 | 4802 |
| Smoking (Yes = 1; No = 0) | 0.1410 | 0.3481 | 2168 | 0.1069 | 0.3090 | 3088 | 0.1056 | 0.3073 | 4802 |
| Drinking (Yes = 1; No = 0) | 0.2030 | 0.4023 | 2168 | 0.1253 | 0.3311 | 3088 | 0.1808 | 0.3849 | 4802 |
| Exercise (Yes = 1; No = 0) | 0.5526 | 0.4973 | 2168 | 0.5097 | 0.5000 | 3088 | 0.5092 | 0.5000 | 4802 |
| Medical examination (Yes = 1; No = 0) | 0.5613 | 0.4963 | 2168 | 0.5946 | 0.4911 | 3088 | 0.4536 | 0.4979 | 4802 |
| Logarithm of family assets | 12.7929 | 1.6424 | 2168 | 7.2202 | 6.3240 | 3088 | 4.1983 | 5.9692 | 4802 |
| Logarithm of GDP per capita | 11.8638 | 0.4507 | 2168 | 11.9013 | 0.5056 | 3088 | 11.9950 | 0.5409 | 4802 |
| Vegetation cover (NVDI) | 0.6409 | 0.0900 | 2168 | 0.5862 | 0.0925 | 3088 | 0.6188 | 0.0079 | 4802 |
[i] Note: SD = standard deviation. GDP = gross domestic product. NVDI = normalized difference vegetation index.

Figure 2
The effect of flooding on cognitive health among middle‑aged and older residents in Thailand.
Note: 90% confidence intervals were reported in the figure. All regression models were conducted in the same form as model (1). This figure reports standardized coefficients. Specific regression results were detailed in Appendix D. The standard errors clustered at the provincial level were in parentheses.

Figure 3
The effect of the Bangkok RCP on cognitive health.
Note: 90% confidence intervals were reported in the figure. Specific regression results were detailed in Appendix E1. This figure reports standardized coefficients. All regression models were conducted in the same form as model (2). RCP indicates resilient city policy. The standard errors clustered at the provincial level were in parentheses.

Figure 4
The effects of the Bangkok RCP on risk factors for cognitive health.
Note: 90% confidence intervals were reported in the figure. See Appendix E2 for specific regression results. All regression models were set up in the same form as model (2). RCP indicates resilient city policy. Missing results are omitted because of collinearity. The standard errors clustered at the provincial level are in parentheses.
