TABLE 1.
Univariate and multivariate ordinal regression analysis on caregiving stress
| Univariate regression result | Multivariate regression result | |||
|---|---|---|---|---|
| OR (95% CI) | p | OR (95% CI) | p | |
| Age | ||||
| CG | 1.05 (1.02–1.08) | <0.001 a | 1.04 (1.00–1.07) | 0.036* |
| CR | 1.01 (0.97–1.04) | 0.701 | – | – |
| Gender | ||||
| CG (male) | 0.76 (0.35–1.68) | 0.501 | – | – |
| CR (male) | 1.57 (0.84–2.93) | 0.161 | – | – |
| CG relationship with CR (reference: others) | ||||
| Spouse | 1.34 (0.26–6.94) | 0.724 | – | – |
| Child/Child‐in‐law | 0.33 (0.07–1.59) | 0.167 | – | – |
| CG worked at home during pandemic | 1.24 (0.51–3.00) | 0.630 | – | – |
| CG spent more time with CR during pandemic | 3.76 (1.93–7.32) | <0.001 a | 4.02 (1.67–9.66) | 0.002* |
| CR went out more during pandemic | 2.07 (0.44–9.78) | 0.360 | – | – |
| Observed CR deterioration | ||||
| Cognition (worse) | 5.33 (2.63–10.82) | <0.001 a | 2.19 (0.90–5.34) | 0.084 |
| Mobility (worse) | 3.22 (1.64–6.32) | 0.001 a | 1.12 (0.47–2.64) | 0.799 |
| Physical health (worse) | 3.99 (2.07–7.69) | <0.001 | 1.60 (0.70–3.68) | 0.264 |
| Emotion (worse) | 7.45 (3.70–15.01) | <0.001 | 5.61 (2.48–12.67) | <0.001* |
| Communication (worse) | 4.44 (2.31–8.50) | <0.001 a | 2.33 (1.01–5.41) | 0.048* |
| Transportation (worse) | 0.96 (0.43–2.13) | 0.919 | – | – |
| Shopping (worse) | 0.95 (0.35–2.55) | 0.911 | – | – |
| CR living status during pandemic | ||||
| Lived with CG | 1.44 (0.75–2.75) | 0.268 | – | – |
| Lived with spouse | 1.43 (0.78–2.61) | 0.246 | – | – |
| Lived with child/child‐in‐laws | 0.43 (0.22−0.81) | 0.009 a | 0.24 (0.10−0.54) | 0.001* |
| Lived with other relatives | 0.57 (0.26–1.23) | 0.151 | – | – |
| Lived with maid | 0.54 (0.29–1.00) | 0.052 a | 0.58 (0.28–1.21) | 0.146 |
| Lived alone | 0.15 (0.01–1.52) | 0.108 | – | – |
| Observed CR dementia stage (reference: late) | ||||
| Early | 0.95 (0.33–2.72) | 0.920 | – | – |
| Moderate | 0.89 (0.39–2.01) | 0.775 | – | – |
Abbreviations: CG, Participant; CI, Confidence level; CR, Care recipient; OR, Odds ratio.
a
Variables with p < 0.1 in univariate regression were included in the multivariate regression: participant age; time spent with care recipient during the pandemic; observed care recipient deterioration of cognitive functioning, mobility, physical health, emotion, and communication; and whether the care recipient lived with child and maid; these variables also had good univariate model fit (test of parallel lines p > 0.05). The final model met the proportional odds assumption (χ 2 = 9.379, df = 18, p = 0.950).
*p < 0.05.