Table 4.

The Effect Estimates of COVID-19 Occurrence on Frailty and Mental Health in Older Population.

Models	Variables	UCLA-Loneliness				GDS				GAI				Frailty Phenotype Index
		β	95% CI		P-value	OR	95% CI		P-value	OR	95% CI		P-value	OR	95% CI		P-value
Model 1
	COVID-19 occurrence	−0.33	−0.39	−0.26	<.001	4.61	3.77	5.63	<.001	1.85	1.52	2.24	<.001	1.42	1.17	1.72	<.001
Model 2
	COVID-19 occurrence	−0.33	−0.39	−0.26	<.001	4.61	3.78	5.64	<.001	1.85	1.52	2.24	<.001	1.42	1.17	1.72	<.001
	Agea	−0.07	−0.13	−0.01	.01	1.07	0.89	1.29	.41	0.93	0.78	1.10	.42	1.06	0.89	1.27	.47
	COVID-19 occurrence × age	.05	−0.01	0.12	.09	0.97	0.79	1.19	.83	1.10	0.90	1.33	.32	0.88	0.73	1.08	.24
Model 3
	COVID-19 occurrenceb	−0.20	−0.32	−0.07	.001	5.54	4.03	7.61	<.001	2.25	1.66	3.05	<.001	1.36	1.00	1.85	.04
	Gender	.14	0.01	0.26	.02	0.96	0.67	1.39	.85	1.37	0.96	1.96	.07	0.98	0.68	1.40	.92
	COVID-19 occurrence × gender	−0.21	−0.36	−0.07	.002**	0.74	0.49	1.11	.15	0.71	0.48	1.06	.09	1.07	0.72	1.59	.72
Model 4
	COVID-19 occurrence	−0.29	−0.39	−0.19	<.001	3.64	2.67	4.97	<.001	1.94	1.43	2.63	<.001	1.46	1.07	1.98	.01
	Education	.09	−0.01	0.21	.10	0.77	0.53	1.11	.17	0.97	0.68	1.38	.88	1.06	0.74	1.52	.72
	COVID-19 occurrence × education	−0.06	−0.19	0.06	.32	1.48	0.99	2.23	.05**	0.92	0.62	1.36	.68	0.95	0.64	1.41	.80
Model 5
	COVID-19 occurrence	−0.44	−0.68	−0.19	<.001	4.47	2.45	8.13	<.001	1.09	0.60	1.98	.76	0.92	0.49	1.71	.79
	Comorbidities	−0.11	−0.34	0.12	.34	0.86	0.49	1.52	.62	0.56	0.31	1.00	.05	0.56	0.30	1.02	.05
	COVID-19 occurrence × comorbidities	.12	−0.13	0.37	.35	1.03	0.54	1.95	.91	1.79	0.95	3.36	.06**	1.62	0.84	3.13	.14
Model 6
	COVID-19 occurrence	−0.34	−0.41	−0.26	<.001	4.46	3.54	5.62	<.001	2.05	1.64	2.56	<.001	1.36	1.08	1.70	.007
	Living alone	−0.01	−0.14	0.11	.80	0.87	0.57	1.31	.51	1.36	0.91	2.03	.13	0.89	0.59	1.34	.59
	COVID-19 occurrence × living alone	.02	−0.12	0.17	.73	1.13	0.71	1.80	.58	0.66	0.42	1.03	.06**	1.18	0.76	1.85	.45

Abbreviation: IQR, interquartile range.

^aAge was z-scored, and COVID-19 occurrence is a binary variable. COVID-19 occurrence is coded as before (0) or during (1) COVID-19 occurrence. Variables in mixed model, including gender, education, comorbidities, and living alone, were binary variables. Coefficients for these variables in linear mixed model, that is, for loneliness outcomes, can be interpreted mean group differences. For example, gender = 0.14 shows when COVID-19 accounting.

^bFor COVID-19 occurrence, on average, females scored 0.14 higher than male. Since loneliness outcome standardized (M = 0, SD = 1), the interpretation of regression coefficients for binary variables are like a standardized mean difference.

^**Sig. ≤ .01.