Do reductions of daily activities mediate the relationship between COVID-19 restrictions and mental ill-health among older persons in Europe?

Figures & data

Figure 1. Sample selection flowchart.

Table 1. Descriptive statistics, data: Wave 8 SHARE Corona Survey.

Measures	Frequency (%)
	Female n = 23,351	Male n = 18,058	Total n = 41,409
Reduction visiting family	19,959 (85.5)	14,838 (82.2)	34,797 (84.0)
Reduction meeting people	20,934 (90.0)	15,577 (86.3)	36,511 (88.2)
Reduction going shopping	17,275 (74.0)	11,271 (62.4)	28,546 (68.9)
Reduction going walking	11,436 (49.0)	7,751 (42.9)	19,187 (46.3)
Age ≥70 yrs	10,345 (44.3)	8,754 (48.4)	19,099 (46.1)
Employed*	5,590 (24.0)	4,768 (26.4)	10,358 (25.0)
Single household	6,515 (27.9)	2,627 (14.6)	9,142 (22.1)
Fair/Poor prepand health	6,628 (28.4)	4,948 (27.4)	11,576 (28.0)

*Employment status when COVID-19 broke out.

Table 2. Regressing stringency on activities while adjusting for covariates and other activities.

	Walking			Shopping			People			Family
	Coef	SE	p	Coef	SE	p	Coef	SE	p	Coef	SE	p
Intercept	−1.34	0.12	<.001	−1.16	0.06	<.001	−0.03	0.07	.675	−0.42	0.05	<.001
Activity:
Walking	–			0.70	0.02	<.001	0.62	0.02	<.001	0.45	0.02	<.001
Shopping	0.71	0.02	<.001	–			0.41	0.02	<.001	0.41	0.02	<.001
People	0.47	0.03	<.001	0.72	0.02	<.001	–			1.24	0.02	<.001
Family	0.46	0.02	<.001	0.49	0.02	<.001	1.18	0.02	<.001	–
Covariate:
Infection rate	−0.10	0.12	.395	0.10	0.05	.053	0.14	0.06	.014	0.07	0.04	.086
Stringency	0.41	0.11	<.001	−0.11	0.05	.030	0.04	0.05	.421	0.00	0.04	.986
Gender	0.09	0.01	<.001	0.32	0.01	<.001	0.05	0.02	.013	0.03	0.02	.050
Age > 70	0.09	0.02	<.001	0.09	0.02	<.001	0.00	0.02	.872	0.12	0.02	<.001
Employed	−0.02	0.02	.252	−0.12	0.02	<.001	−0.47	0.02	<.001	0.08	0.02	<.001
Fair/poor prepand health	−0.33	0.02	<.001	−0.17	0.02	<.001	−0.02	0.02	.518	−0.11	0.02	<.001
Single husehold	0.11	0.02	<.001	−0.16	0.02	<.001	−0.07	0.02	.006	−0.15	0.02	<.001

Table 3. Regressing activities on mental ill-health while adjusting for covariates and other activities.

	Coef	SE	p
Intercept	0.032	0.008	<.001
Activity:
Walking	0.032	0.003	<.001
Shopping	0.040	0.003	<.001
People	0.010	0.004	.014
Family	0.020	0.004	<.001
Covariates:
Infection rate	0.026	0.005	<.001
Gender	0.057	0.002	<.001
Age > 70	−0.011	0.003	<.001
Employed	0.008	0.003	.008
Fair/poor prepand health	−0.080	0.003	<.001
Single Household	0.039	0.003	<.001

Figure 2. Mediation results from the fully adjusted model. The results are presented as (i) the indirect or mediating (med.Eff) of activity reduction (walking, shopping, people or family) in the association between stringency levels and mental-ill health, (ii) the direct effect (dir.Eff) of stringency levels on mental ill-health while controlling for activity, and (iii) the total effect (tot.Eff) of stringency levels on mental ill-health, corresponding to the sum of the direct and the mediating effect.

Data availability statement

This paper uses data from the Survey of Health, Ageing and Retirement in Europe (SHARE): Wave 8, release version 8.0.0, DOI: 10.6103/SHARE.w8.800 and Wave 8 COVID-19 Survey 1, release version 8.0.0 DOI: 10.6103/SHARE.w8ca.800, https://share-eric.eu/data/.