Skip to main content

Table 4 Interaction effects of self-harm on bullying behaviour and gender, school well-being, parental-, and friend support

From: The relationship between self-harm and bullying behaviour: results from a population based study of adolescents

 

Model 1

OR (95% CI)

Bivariate effect

Model 2

OR (95% CI)

Multivariate effect

Model 3

OR (b)

Interaction effects

Neither bullied nor bullies

1.0 (ref.)

 

Gender*

Bullied

5.04 (4.48–5.68)***

4.94 (4.36–5.59)***

1.32 (0.28)*

Bullies

3.19 (2.48–4.10)***

4.43 (3.40–5.77)***

0.84 (−0.18)

Bully-victims

5.97 (4. 86–7.35)***

8.29 (6.65–10.34)***

0.68 (− 0.39)

Gender

3.34 (3.02–3.71)***

3.73 (3.34–4.17)***

 

Neither bullied nor bullies

1.0 (ref.)

 

Parental support*

Bullied

5.04 (4.48–5.68)***

4.38 (3.85–4.97)***

1.35 (0.30)*

Bullies

3.19 (2.48–4.10)***

2.40 (1.83–3.15)***

1.24 (0.22)

Bully-victims

5.97 (4.86–7.35)***

4.83 (3.85–6.07)***

1.08 (0.08)

Parental support

0.20 (0.18–0.22)***

0.23 (0.21–0.26)***

 

Neither bullied nor bullies

1.0 (ref.)

 

School well-being*

Bullied

5.04 (4.48–5.68)***

2.60 (2.28–2.97)***

1.06 (0.06)**

Bullies

3.19 (2.48–4.10)***

1.98 (1.50–2.62)***

1.01 (0.01)

Bully-victims

5.97 (4.86–7.35)***

2.63 (2.08–3.32)***

1.09 (0.09)**

School well-being

0.73 (0.72–0.74)***

0.76 (0.75–0.78)***

 

Neither bullied nor bullies

1.0 (ref.)

 

Friend support*

Bullied

5.04 (4.47–5.68)***

4.93 (4.36–5.56)***

1.68 (0.52)**

Bullies

3.19 (2.48–4.09)***

3.10 (2.40–4.01)***

0.78 (−0.25)

Bully-victims

5.97 (4.86–7.35)***

6.13 (4.96–7.59)***

0.70 (−0.35)

Friend support

0.63 (0.55–0.71)***

0.80 (0.70–0.92)**

 
  1. Note. *p < 0.05 **p < 0.01 ***p < 0.001. Self-harm = Self harmed last 12 months. Model 1 shows the crude OR, the bivariate associations with no controls, Model 2 shows adjusted OR with control for gender or the potential protective variables: parental support, school well-being and friend support. Model 3 shows the regression analysis with the interaction terms: bullying behaviour multiplied by gender, and bullying behaviour multiplied by the potential protective variables (parental support, school well-being and friend support). Only significant results reported