Page 47 - IJPS-11-1
P. 47
International Journal of
Population Studies Age, gender, internet and older adults’ cognition
Table 1. Sample characteristics
M (N) SD (%) Cognition t Men Women t (χ )
2
Internet use 4.98 2.59 4.93 5.01 −1.16
(dev) Internet 1.81e -16 2.59 -0.05 0.03 −1.16
Age 67.36 10.73 67.48 67.28 0.75
(dev) Age 2.39e -15 10.73 0.13 −0.09 0.75
Gender −2.70**
Men 2,429 40.38 21.83
Women 3,586 59.62 22.24
Race 11.25*** 1.96
White 4,339 72.38 22.53 73.36 71.71
Non-white 1,656 27.62 20.64 26.64 28.29
Married/partnered −8.98*** 371.37***
No 2,244 37.33 21.29 22.73 47.22
Yes 3,768 62.67 22.61 77.27 52.78
Years of education 13.09 3.01 13.17 13.04 1.66
Light activity 3.39 1.05 3.20 3.52 −11.63***
Moderate activity 3.03 1.29 3.17 2.95 6.22***
Vigorous activity 2.06 1.32 2.29 1.90 11.37***
Smoke now 2.76*** 0.79
No 5,503 87.18 22.15 86.9 87.68
Yes 809 12.82 21.49 13.1 12.32
Drink per week 2.72 6.51 4.12 1.75 14.47***
High blood pressure 9.76*** 8.34**
No 2,459 39.36 23.06 36.58 40.31
Yes 3,789 60.64 21.57 63.42 59.69
Diabetes 8.56*** 16.06***
No 4,655 74.08 22.47 71.29 75.93
Yes 1,629 25.92 21.06 28.71 24.07
Cancer 1.15 0.00
No 5,325 84.24 22.11 83.9 83.9
Yes 996 15.76 21.90 16.1 16.1
Depression 1.36 1.91 1.13 1.53 −8.02***
ADL 0.30 0.83 0.26 0.32 −2.50*
IADL 0.23 0.70 0.20 0.25 −2.32*
Self-rated health 3.13 1.02 3.16 3.11 2.03*
Notes: “dev” stands for mean deviation; “e” stands for exponent; *p<0.05, **p<0.01, ***p<0.001.
Abbreviations: ADL: Activities of daily living; IADL: Instrumental activities of daily living.
lower cognition. Covariates, such as race (β = −1.922, significant (β = 0.004, p < 0.10). In Model 3, older women
p < 0.001), years of education (β = 0.388, p < 0.001), had.685-unit cognition higher than older men (p < 0.001).
light activities (β = 0.201, p < 0.01), vigorous activities In Model 4, the two two-way interactions: Internet use ×
(β = −0.131, p < 0.05), smoke status (β = −0.541, p < 0.01), gender was marginally significant (β = 0.091, p < 0.10) and
IADL (β = −0.763, p < 0.001), and self-rated health status age × gender was significant (β = 0.033, p < 0.01). In Model
(β = 0.220, p < 0.001) were significantly associated with 5, the three-way interaction among Internet use, age, and
cognition among participants. In Model 2, the two-way gender was significant (β = 0.010, p < 0.05). The significant
interaction between Internet use and age was marginal three-way interaction is illustrated in Figure 1. As we can
Volume 11 Issue 1 (2025) 41 https://doi.org/10.36922/ijps.3119

