Skip to main content

Table 3 Socio-demographic, behavioural and health factors associated with changes in mean TV viewing time (in hours) between baseline and 2-year follow-up ( N= 6,090)

From: Sociodemographic, behavioural and health factors associated with changes in older adults’ TV viewing over 2 years

Variable

N

Model 1 B (95% CI)

Model 2 B (95% CI)

Socioeconomic status*

   

Managerial/ professional

2226

Reference

Reference

Intermediate

1580

0.41 (0.16, 0.66)

0.36 (0.11, 0.60)

Manual/routine

2236

1.31 (1.08, 1.54)

1.12 (0.89, 1.36)

Other

48

-

-

p-trend

 

<0.001

<0.001

Depression (CES-D ≥ 4)

   

No

5378

Reference

Reference

Yes

712

0.81 (0.51, 1.11)

0.43 (0.12, 0.74)

p-trend

 

<0.001

0.007

Disability

   

No

4817

Reference

Reference

Yes

1273

0.69 (0.45, 0.94)

0.25 (−0.01, 0.52)

p-trend

 

<0.001

0.06

Chronic illness

   

No

2924

Reference

Reference

Yes

3166

0.27 (0.08, 0.46)

−0.04 (−0.24, 0.17)

p-trend

 

0.007

0.73

Body mass index

   

15 - 25

1631

Reference

Reference

≥25 < 30

2559

0.37 (0.13, 0.61)

0.43 (0.19, 0.66)

≥30

1900

0.91 (0.65, 1.16)

0.82 (0.56, 1.08)

p-trend

 

<0.001

<0.001

Physical activity

   

Inactive

1085

Reference

Reference

Moderate

3017

−0.95 (−1.22, −0.68).

−0.65 (−0.92, −0.37)

Vigorous

1988

−1.07 (−1.36, −.077)

−0.57 (−0.78, −0.26)

p-trend

 

<0.001

<0.001

Smoking

   

Never

2503

Reference

Reference

Ex-smoker

2849

0.09 (−0.12, 0.30)

0.02 (−0.19, 0.23)

Current

738

0.93 (0.61, 1.25)

0.71 (0.39, 1.03)

p-trend

 

<0.001

<0.001

  1. Model 1 adjusted for age, sex and baseline TV viewing. Model 2 adjusted for age, sex, baseline TV viewing and mutually for all variables presented. Β (95% Confidence Interval) coefficients reflect increases in hours/day of TV viewing between baseline and follow-up. * Coefficients are not reported for the ‘other’ socioeconomic status category due to small sample size and heterogeneity of employment statuses captured by this category. All available data were however entered into the regression model, in that coefficients for other socioeconomic status categories were calculated using dummy variables that contrasted the focal category with all three other categories.