Skip to main content

Table 3 Logistic regression analyses for the associations between demographic variables and female genital mutilation/cutting

From: Health care-seeking patterns for female genital mutilation/cutting among young Somalis in Norway

Female genital mutilation/cutting

Demographic variables

Univariate analysis (unadjusted)

Model 1 (Adjusted)

Model 2 (Adjusted)

  

aAIC = 118.34

aAIC = 116.70

OR (95% CI)

P -value

OR (95% CI)

P -value

OR (95% CI)

P -value

Age

 16 to 20 years

1

 

1

 

1

 

 21 to 25 years

1.04 (0.53–2.06)

0.89

0.80 (0.23, 2.70)

0.71

1.20 (0.37, 3.84)

0.76

Age at migration to Norway

 0 to 11 years

1

 

1

 

1

 

  ≥ 12 years

12.16 (4.99–29.62)

<  0.01

4.78 (1.53, 15.00)

0.01

4.84 (1.57, 14.91)

0.01

Marital status

 Single

1

 

1

 

1

 

 Married

1.07 (0.46–2.51)

0.86

0.86 (0.23, 3.21)

0.82

0.71 (0.20, 2.51)

0.60

 Divorce

0.92 (0.05–15.07)

0.95

–

–

–

–

Support of FGM/C practice

 No

1

 

1

 

1

 

 Yes

5.15 (1.41–18.72)

0.01

2.06 (0.38, 11.10)

0.40

2.10 (0.39–11.43)

0.39

Education

 University

1

 

1

 

1

 

 Secondary

1.51 (0.57–4.02)

0.40

1.44 (0.32, 6.42)

0.64

1.73 (0.40, 7.43)

0.46

 Primary

2.53 (0.86–7.48)

0.09

2.72 (0.48, 15.51)

0.26

3.06 (0.57, 16.38)

0.19

 No formal education

3.25 (0.48–21.9)

0.22

10.34 (0.17, 618.48)

0.26

10.19 (0.46, 227.76)

0.14

Place of birth of women

 Born out of Norway

1

 

1

 

1

 

 Born in Norway

0.02 (0.01, 0.07)

< 0.01

0.01 (0.001, 0.10)

<  0.01

0.02 (0.005, 0.12)

<  0.01

Stigmatized of FGM/C practice

 No

1

 

1

   

 Yes

1.53 (0.69–3.38)

0.28

2.58 (0.59, 11.28)

0.21

  
  1. Model 1 is a full adjustment for all main demographic variables
  2. Model 2 is based on variables that are statistically significant from the univariate analysis and variables that have been shown to be associated with FGM/C in the literature
  3. aModel selection was based on the Akaike Information Criterion (AIC) which states that a model with a smaller AIC estimate fits the data better. Therefore, based on the AIC, model 2 was selected