Skip to main content

Table 3 Multivariable logistic and multinomial regression models for past-year household violence and changes in frequency relative to COVID-19 (referent = decreased household violence), per country

From: Violence against women from partners and other household members during COVID-19 in Burkina Faso and Kenya

 

Burkina Faso

Kenya

Multivariable Logistic

Multivariable Multinomial

Multivariable Logistic

Multivariable Multinomial

Any Household Violence (n = 710; 16.2%)

Decreased (ref) (n = 426; 58.8%)

Increased (n = 87; 14.3%)

Sustained (n = 173; 26.9%)

Any Household Violence (n = 701; 11.0%)

Decreased (ref) (n = 313; 44.6%)

Increased (n = 179; 26.2%)

Sustained (n = 207; 29.3%)

%*

OR

(95% CI)

%*

%*

RRR

(95% CI)

%*

RRR

(95% CI)

%*

OR

(95% CI)

%*

%*

RRR

(95% CI)

%*

RRR

(95% CI)

Age and Household Structure

 Age

  15–19

18.1

2.17

(1.27, 3.70)**

59.3

14.6

1.48

(0.30, 7.38)

26.1

2.14

(0.55, 8.34)

17.0

 

40.9

28.0

 

31.1

 

  20–29

16.9

2.17

(1.33, 3.54)**

53.9

17.0

2.22

(0.62, 7.99)

29.1

3.45

(1.10, 10.76)*

10.1

 

47.4

22.8

 

29.8

 

  30–39

16.8

1.73

(1.01, 2.95)*

56.8

13.1

1.77

(0.42, 7.41)

30.2

3.35

(1.32, 8.45)*

8.9

 

40.9

29.8

 

29.3

 

  40–49

11.4

ref

81.0

8.0

ref

11.1

ref

10.9

 

49.9

24.7

 

25.4

 

 Marital Status

  Married

15.3

 

58.3

14.9

ref

26.8

ref

8.7

ref

49.1

25.3

 

25.6

 

  Living with partner

14.5

 

63.2

14.8

0.89

(0.30, 2.70)

22.0

1.02

(0.25, 4.18)

8.1

0.88

(0.54, 1.43)

56.5

25.0

 

18.5

 

  Divorced/separated or widower

18.7

 

79.3

3.3

0.23

(0.05, 1.12)

17.5

0.85

(0.21, 3.37)

16.5

2.02

(1.47, 2.76)***

39.4

35.6

 

25.0

 

  Never married

21.0

 

54.7

14.6

0.84

(0.29, 2.45)

30.8

2.63

(1.18, 5.86)*

15.1

1.73

(1.21, 2.45)**

39.2

24.2

 

36.6

 

 Lives with extended family

  No

14.8

 

61.0

11.4

 

27.6

 

9.4

ref

46.6

24.3

 

29.0

 

  Yes

18.1

 

56.2

17.7

 

26.1

 

14.2

1.34

(1.09, 1.67)**

41.9

28.5

 

29.6

 

Socioeconomic

 Education

  None

15.4

 

60.5

10.2

 

29.4

 

13.2

1.18

(0.73, 1.92)

54.6

29.0

 

16.3

 

  Primary

18.4

 

56.7

19.6

 

23.8

 

12.9

ref

45.3

27.5

 

27.2

 

  Secondary or Higher

16.7

 

56.5

19.9

 

23.6

 

9.5

0.73

(0.58, 0.92)**

42.4

24.6

 

33.0

 

 Household Wealth Tertile

  Lowest

14.8

 

63.2

5.1

ref

31.7

ref

14.2

ref

42.2

29.2

 

28.6

 

  Middle

17.1

 

56.7

17.2

3.64

(1.17, 11.34)*

26.1

0.97

(0.48, 1.96)

11.3

0.85

(0.68, 1.07)

47.8

22.6

 

29.6

 

  Highest

16.7

 

56.8

19.8

3.57

(0.73, 17.42)

23.4

0.82

(0.35, 1.92)

7.2

0.67

(0.49, 0.91)*

44.3

25.7

 

30.1

 

 Financial knowledge

  None/low

15.5

 

63.1

12.9

 

24.0

 

14.4

 

38.0

30.6

ref

31.4

ref

  Medium

17.6

 

47.1

17.7

 

35.1

 

10.6

 

43.0

28.4

0.86

(0.40, 1.85)

28.6

0.95

(0.53, 1.70)

  High

20.8

 

47.1

19.2

 

33.7

 

9.1

 

53.3

18.9

0.47

(0.25, 0.87)*

27.8

0.75

(0.41, 1.37)

 Income loss, past year

  None

13.5

ref

64.4

13.1

ref

22.4

ref

7.8

ref

52.1

23.2

 

24.7

 

  Partial

19.9

1.64

(1.16, 2.30)**

51.9

16.2

1.23

(0.50, 3.00)

31.9

1.80

(1.08, 2.98)*

10.5

1.82

(1.30, 2.54)***

45.8

26.5

 

27.8

 

  Complete

16.8

1.37

(0.78, 2.42)

63.9

11.3

0.73

(0.19, 2.79)

24.8

1.17

(0.46, 3.00)

14.7

2.58

(1.82, 3.67)***

39.6

27.0

 

33.4

 
  1. Multivariable logistic regression compares any household violence experience to no household violence experience (referent). Multivariable multinomial regression compares household violence trajectory since COVID-19 restrictions (increased, sustained) to decreased violence experience (referent). Model adjusted for all variables with p < 0.1 from bivariate models (not presented). Associations that remained significant (p < 0.05) after adjustment are presented. Burkina Faso logistic model adjusted for age, marital status, parity, household composition, and income loss in the last 12 months. Burkina Faso multinomial model adjusted for age, marital status, education, household wealth tertile, level of financial knowledge, and income loss in the last 12 months. Kenya logistic model adjusted for age, marital status, education, residence, parity, number of household members, household composition, household wealth tertile, savings, mobile money account, level of financial knowledge, knows where to go for financial advice, and income loss in the last 12 months. Kenya multinomial model adjusted for marital status, parity, level of financial knowledge, knows where to go for financial advice, and has financial goal
  2. *=p<0.05; **=p<0.01; ***=p<0.001
Back to article page

BMC Public Health

ISSN: 1471-2458

Contact us