The Relationship Between Internet Use and Mental Health Among Older Adults in China: The Mediating Role of Physical Exercise

Figures & data

Table 1 Descriptive Statistics

Variable	Mean	SD	Min	Max	N
Mental health	11.838	2.177	3	15	1674
Internet use	1.738	1.343	1	5	1674
Physical exercise	2.468	1.712	1	5	1674
Internet use * Physical exercise	4.912	6.170	1	25	1674
Gender	0.492	0.500	0	1	1674
Age	71.573	8.042	60	107	1674
Education	0.484	0.500	0	1	1674
Income	0.060	0.237	0	1	1674
Social class	4.091	1.778	1	10	1674

Note: Internet use *Physical exercise means the interaction item of internet use and physical exercise.

Table 2 Results of Global Regression Analysis

Variable	Full Sample	Urban Sample	Rural Sample
	Model 1	Model 2	Model 3
Internet use	0.114*** (2.69)	0.108** (2.18)	0.019 (0.18)
Gender	−0.061 (−0.60)	0.067 (0.47)	−0.045 (−0.30)
Age	0.025*** (3.95)	0.027*** (3.02)	0.014 (1.42)
Education	0.632*** (5.45)	0.250 (1.43)	0.604*** (3.36)
Income	0.290 (1.37)	0.187 (0.67)	0.431 (1.35)
Social class	0.319*** (11.16)	0.247*** (6.13)	0.358*** (8.78)

Notes: ***p<0.01, **p<0.05. The values inside the brackets are t-values.

Table 3 Regression Results of the Effect of the Combination of Internet Use and Physical Exercise on the Mental Health of the Elderly

Variable	Model 4
Internet use	−0.091 (−1.10)
Physical exercise	0.039 (0.79)
Internet use * Physical exercise	0.055*** (2.48)
Gender	−0.050 (−0.50)
Age	0.025*** (3.89)
Education	0.562*** (4.80)
Income	0.255 (1.21)
Social class	0.304*** (10.61)

Notes: ***p<0.01. The values inside the brackets are t-values. Internet use *Physical exercise means the interaction item of internet use and physical exercise.

Table 4 Mediating Role of Physical Exercise Between Internet Use and Mental Health

	Effect	Standard Error	LLCI	ULCI	Ratio
Total effect (c)	0.114	0.043	0.031	0.198
Direct effect (c’)	0.084	0.041	0.004	0.163
Indirect effect (ab)	0.031	0.009	0.012	0.048
Ratio of indirect to total effect mediated					0.272
Ratio of indirect to direct effect					0.369

Notes: c path represents the total effect of internet use on mental health directly. a path represents the effect of the independent variable of internet use on the intermediary variable of physical exercise. b path represents the effect of the intermediary variable of physical exercise on the dependent variable of mental health, while c’ represents the effect of the independent variable of internet use on the dependent variable of mental health after considering or controlling for intermediary variable physical exercise.

Abbreviations: LLCI, lower level for confidence interval; ULCI, upper level for confidence interval.

Table 5 Regression Results of the Heterogeneity of the Psychological Effects of Internet Use on Older Adults

Variable	Model 5
Internet use	0.318*** (3.22)
Internet use * Social class	−0.470** (−2.28)
Gender	−0.049 (−0.48)
Age	0.025*** (3.93)
Education	0.620*** (5.35)
Income	0.332 (1.57)
Social class	0.404*** (8.64)

Notes: ***p<0.01, **p<0.05. The values inside the brackets are t-values.

Table 6 Results of Sensitivity Analysis

Variable	Full Sample	Urban Sample	Rural Sample
	Model 9	Model 10	Model 11
Internet use	0.057*** (2.75)	0.051** (2.02)	−0.040 (−0.84)
Gender	0.084* (1.68)	0.172** (2.36)	0.075 (1.09)
Age	0.002 (0.49)	−0.001 (−0.11)	−0.003 (−0.66)
Education	0.243*** (4.28)	0.037 (0.41)	0.213** (2.52)
Income	0.111 (1.07)	0.122 (0.85)	0.078 (0.52)
Social class	0.110*** (7.88)	0.092*** (4.44)	0.112*** (5.87)

Notes: ***p<0.01, **p<0.05, *p<0.1. The values inside the brackets are t-values.