Table 2.
Association of depressive symptoms with community-level and individual-level volunteer participation by a multilevel Poisson regression.
| Model 1: Adjusted for Outcome, Individual-Level, Community-Level, Cross-Level, Age and Sex | ||
|---|---|---|
| Total, N = 37,552 | IRR b | 95% CI c |
| Fixed effects | ||
| Volunteer participation at the individual level | 0.70 | (0.63–0.78) |
| Proportion of volunteer participation at the community level 10% estimation a | 0.86 | (0.77–0.97) |
| Cross-level interaction | 1.06 | (0.79–1.42) |
| Age, y | ||
| 65–69 | 1.00 | |
| 70–74 | 1.13 | (1.04–1.22) |
| 75–79 | 1.35 | (1.23–1.47) |
| 80–84 | 1.71 | (1.54–1.90) |
| ≧85 | 2.12 | (1.83–2.46) |
| Sex | ||
| Man | 1.00 | |
| Woman | 1.04 | (0.98–1.11) |
| Intercept (SE) d | 0.08 | 1.807 |
| Random effect | ||
| Ωμ (SE) d | 0.005 | 0.022 |
| a. Cross-level interaction: community-level exposure 10% estimation × individual-level exposure. | ||
| b. IRR: incidence rate ratio. | ||
| c. CI: confidence interval. | ||
| d. SE: standard error. | ||
| Model 2: Model 1 Adjusted for Community-Level Covariates | ||
| Total, N = 37,552 | IRR b | 95% CI c |
| Fixed effects | ||
| Volunteer participation at the individual level | 0.70 | (0.63–0.77) |
| Proportion of volunteer participation at the community level 10% estimation a | 0.87 | (0.78–0.98) |
| Cross-level interaction | 1.06 | (0.79–1.43) |
| Population density, persons per kilometer squared of inhabitable area | ||
| Highest quartile (≥11,385) | 1.00 | |
| Second quartile (9028–11,378) | 1.00 | (0.86–1.16) |
| Third quartile (5194–8977) | 1.06 | (0.91–1.22) |
| Lowest quartile (<5158) | 1.21 | (1.06–1.38) |
| Total annual sunshine hours | ||
| Highest quartile (≥ 2076) | ||
| Second quartile (2047–2075) | 0.92 | (0.81–1.03) |
| Third quartile (1911–2047) | 0.97 | (0.83–1.13) |
| Lowest quartile (<1911) | 0.98 | (0.87–1.11) |
| Annual Rainfall (mm) | ||
| Highest quartile (≥1622) | 1.00 | |
| Second quartile (1521–1621) | 0.90 | (0.76–1.06) |
| Third quartile (1483–1520) | 0.98 | (0.84–1.14) |
| Lowest quartile (<1483) | 0.82 | (0.72–0.95) |
| Age, y | ||
| 65–69 | 1.00 | |
| 70–74 | 1.13 | (1.04–1.22) |
| 75–79 | 1.34 | (1.22–1.47) |
| 80–84 | 1.70 | (1.53–1.89) |
| ≧85 | 2.09 | (1.80–2.43) |
| Sex | ||
| Man | 1.00 | |
| Woman | 1.04 | (0.98–1.11) |
| Intercept (SE) d | 0.08 | 0.008 |
| Random effect | ||
| Community-level variance | ||
| Ωμ (SE) d | 0.036 | 0.008 |
| a. Cross-level interaction: community-level exposure 10% estimation × individual-level exposure. | ||
| b. IRR: incidence rate ratio. | ||
| c. CI: confidence interval. | ||
| d. SE: standard error. | ||
| Model 3: Model 2 Adjusted for Individual-Level Covariates | ||
| Total, N = 37,552 | IRR b | 95% CI c |
| Fixed effects | ||
| Volunteer participation at the individual level | 0.72 | (0.65–0.80) |
| Proportion of volunteer participation at the community level 10% estimation a | 0.90 | (0.80–1.01) |
| Cross-level interaction | 1.06 | (0.79–1.42) |
| Population density, persons per kilometer squared of inhabitable area | ||
| Highest quartile (≥11,385) | 1.00 | |
| Second quartile (9028–11,378) | 0.99 | (0.85–1.15) |
| Third quartile (5194–8977) | 1.02 | (0.88–1.19) |
| Lowest quartile (<5158) | 1.12 | (0.99–1.28) |
| Total annual sunshine hours | ||
| Highest quartile (≥2076) | 1.00 | |
| Second quartile (2047–2075) | 0.91 | (0.81–1.02) |
| Third quartile (1911–2047) | 0.96 | (0.83–1.12) |
| Lowest quartile (<1911) | 0.96 | (0.84–1.08) |
| Annual Rainfall (mm) | ||
| Highest quartile (≥1622) | 1.00 | |
| Second quartile (1521–1621) | 0.88 | (0.75–1.04) |
| Third quartile (1483–1520) | 0.98 | (0.84–1.14) |
| Lowest quartile (<1483) | 0.83 | (0.73–0.96) |
| Age, y | ||
| 65–69 | 1.00 | |
| 70–74 | 1.08 | (1.00–1.18) |
| 75–79 | 1.25 | (1.14–1.37) |
| 80–84 | 1.55 | (1.39–1.73) |
| ≧ 85 | 1.88 | (1.62–2.20) |
| Sex | ||
| Man | 1.00 | |
| Woman | 1.02 | (0.94–1.10) |
| Education, y | ||
| ≧6 | 1.00 | |
| <6 | 1.24 | (0.94–1.64) |
| Other and missing | 1.16 | (0.85–1.58) |
| Annual equivalent income, JPY | ||
| ≧4,000,000 | 1.00 | |
| <2,000,000 | 1.83 | (1.61–2.09) |
| 2,000,000-3,999,999 | 1.35 | (1.19–1.54) |
| Missing | 1.72 | (1.50–1.99) |
| Disease status in treatment | ||
| Stroke no | 1.00 | |
| Stroke yes | 1.13 | (0.93–1.36) |
| Hypertension no | 1.00 | |
| Hypertension yes | 1.04 | (0.98–1.11) |
| Diabetes no | 1.00 | |
| Diabetes yes | 1.14 | (1.04–1.30) |
| Ear disease no | 1.00 | |
| Ear disease yes | 1.32 | (1.16–1.50) |
| Family structure | ||
| No | 1.00 | |
| Yes (living alone) | 1.07 | (0.97–1.20) |
| Missing | 1.13 | (0.97–1.32) |
| Drinking status | ||
| None | 1.00 | |
| Past | 1.21 | (1.04–1.41) |
| Current | 0.86 | (0.79–0.93) |
| Missing | 0.78 | (0.51–1.18) |
| Smoking status | ||
| None | 1.00 | |
| Past | 1.10 | (0.99–1.22) |
| Current | 1.29 | (1.15–1.44) |
| Missing | 1.30 | (0.89–1.90) |
| BMI e (kg/m2) | ||
| ≧18.5 | 1.00 | |
| <18.5 | 1.24 | (1.10–1.41) |
| Missing | 1.24 | (1.06–1.44) |
| Intercept (SE) e | 0.06 | 0.007 |
| Random effect | ||
| Community-level variance | ||
| Ωμ (SE) e | 0.035 | 0.008 |
a. Cross-level interaction: community-level exposure 10% estimation × individual-level exposure. b. IRR: incidence rate ratio. c. CI: confidence interval. e. BMI: body mass index. SE: standard error.