Skip to main content

Table 3 Multinomial logistic regression, factors associated with sexual abuse, before and after age 12, Adjusted Odds Ratios (AOR) values.

From: Child sexual abuse in southern Brazil and associated factors: a population-based study

  Sexual abuse
(Before age 12)
Sexual abuse
(Age 12 to 18)
p-values
Heterogeneity test§
  AOR 95% CI p-value AOR 95% CI p-value (Likelihood ratio test)#
Block I (Socio-demographic)        
Sex        
   Male 1.0    1.0    0.0002
   Female 3.6 (1.8; 7.3) 0.001 5.1 (1.6; 15.9) 0.006 (0.001)
Age at interview        
   Lower tertile (up to 30 y.o.) 3.2 (1.2; 8.4) 0.02 3.0 (1.0; 8.7) 0.05 0.01
   Intermediate tertile (31–46 y.o.) 1.9 (0.7; 5.2) 0.19 3.9 (1.3; 11.2) 0.01 (0.01)
   Upper tertile (above 46 y.o.) 1.0    1.0    
Education (complete years)        
   0–3 years 1.0    1.0    
   4–8 years 1.2 (0.4; 4.2) 0.68 0.9 (0.3; 2.2) 0.64 0.29
   9–11 years 1.9 (0.6; 5.4) 0.25 0.6 (0.2; 1.6) 0.22 (0.03)
   12 or more years 1.9 (0.4; 7.3) 0.39 0.2 (0.1; 1.1) 0.06  
Income        
   Lower tertile (up to $43) 1.0    1.0    
   Intermediate tertile ($44–$143) 0.9 (0.4; 2.4) 0.80 1.5 (0.7; 3.2) 0.27 0.80
   Upper tertile (above $143) 0.9 (0.3; 2.6) 0.88 1.8 (0.7; 4.7) 0.23 (0.45)
Block II        
Sexual partners (past 2 months)        
   None 1.0    1.0    
   One 1.3 (0.6; 2.7) 0.56 1.4 (0.6; 3.7) 0.43 0.12
   Two or more 1.9 (0.6; 6.4) 0.29 4.5 (1.2; 16.2) 0.02 (0.35)
Condom use (last intercourse)        
   Yes 0.8 (0.3; 2.2) 0.64 1.4 (0.7; 2.9) 0.35 0.63
   No 1.0    1.0    (0.5)
Stressfull events past 12 months (lost job/divorce/robbery)        
   None 1.0    1.0    
   One 1.1 (0.4; 2.7) 0.84 1.4 (0.5; 3.8) 0.61 0.85
   Two or more 0.4 (0.1; 3.6) 0.43 1.4 (0.2; 9.5) 0.34 (0.08)
Physical abuse        
   No 1.0    1.0    < 0.001
   Yes 5.6 (2.5; 12.3) < 0.001 9.4 (4.5; 18.7) < 0.001 (< 0.001)
Lifetime consultation for mental health issue        
   No 1.0    1.0    0.10
   Yes 2.2 (1.1; 4.6) 0.04 0.9 (0.3; 2.6) 0.84 (0.003)
Health perception        
   Good 1.0    1.0    0.04
   Average/bad 2.0 (1.1; 3.8) 0.03 1.9 (1.1; 3.6) 0.04 (0.005)
  1. Reference category is 'no abuse'. Hierarchical model with backwards elimination (final model n = 1874)
  2. Note: Bolded variables are in the final model and are adjusted to each other. Estimates in italics were obtained from each variable's last entry in the model, before it was removed from the model (backwards elimination). Variables from subsequent levels are adjusted for bolded variables from the same level and from preceding levels.§ Likelihood ratio test for combining alternatives. P-value corresponds to the probability that the odds ratio of a given variable with CSA before age 12 and from ages 12 to 18 is the same, indicating that both levels of the dependent variable could be collapsed. # Probability that the association of each independent variable with each of the levels of the dependent variable (child sexual abuse) is simultaneously null according to the likelihood ratio test.