Table 3 Logistic regression analysis and multiplicative interaction: results of poor self-rated health regressed on problematic social media use (PSMU) and covariates. Odds ratios (OR) and 95% confidence intervals (95% CI) with robust standard errors. (n = 3 274)

Problematic social media use

 Low risk of PSMU (ref.)

1.00

-

1.00

-

1.00

-

1.00

-

 Moderate risk of PSMU

2.77***

1.98, 3.85

2.32***

1.67, 3.22

1.84**

1.29, 2.63

2.45***

1.49, 4.03

 PSMU

4.23***

2.45, 7.31

3.58***

2.09, 6.11

2.53**

1.44, 4.43

2.22

0.81, 6.14

Gender

 Boys (ref.)

1.00

-

1.00

-

1.00

-

1.00

-

 Girls

2.29***

1.66, 3.16

2.00***

1.45, 2.70

1.93***

1.38, 2.67

1.93***

1.39, 2.69

Age

 11 years (ref.)

1.00

-

1.00

-

1.00

-

1.00

-

 13 years

1.97**

1.22, 3.17

1.77**

1.09, 2.86

1.63*

1.09, 2.63

1.62*

1.01, 2.63

 15 years

2.58***

1.68, 3.95

2.20***

1.43, 3.39

1.93**

1.24, 3.00

1.91**

1.23, 2.96

Family affluence

 Lowest 20ptc

1.62

0.93, 2.82

1.53

0.87, 2.69

1.24

0.71, 2.20

1.24

0.70, 2.19

 Medium 60ptc

1.54*

1.02, 2.33

1.40

0.92, 2.15

1.28

0.82, 1,97

1.26

0.82, 1.95

 Highest 60ptc (ref.)

1.00

-

1.00

-

Family support

 High support (ref.)

1.00

-

1.00

-

1.00

-

 Moderate support

4.53***

3.29, 6.23

3.74***

2.66, 5.26

4.37***

2.84, 6.71

 Low support

6.03***

3.91, 9.29

4.74***

2.99, 7.51

5.05***

2.86, 8.95

PSMU # Family support

p = 0.61