Skip to main content

Table 3 Comparison of class belongingness across covariates

From: Sociodemographic characteristics associated with alcohol consumption and alcohol-related consequences, a latent class analysis of The Norwegian WIRUS screening study

 

Age (S.E.; p-value)

Gender (S.E.; p-value)

Education (S.E.; p-value)

Occupational levela (S.E.; p-value)

Full-time employment (S.E.; p-value)

Income (quintiles)b (S.E.; p-value)

Unadjusted

 Class 2 (ref) vs 1

1.38 (.044; p < 0.001)

1.71 (.139; p < 0.001)

1.13 (.049; p = .009)

0.84 (.070; p = .024)

0.79 (.070; p = .002)

0.95 (.025; p = .031)

 Class 3 (ref) vs 1

1.65 (.080; p < 0.001)

3.30 (.357; p < 0.001)

1.38 (.083; p < 0.001)

0.98 (.121; p = .849)

0.70 (.094; p = .002)

1.13 (.045; p = .005)

 Class 3 (ref) vs 2

1.19 (.059; p = .001)

1.94 (.210; p < 0.001)

1.22 (.075; p = .003)

1.16 (.146; p = .273)

0.89 (.126; p = .397)

1.19 (.050; p < 0.001)

Fully adjustedc

 Class 2 (ref) vs 1

1.51 (.054; p < 0.001)

1.87 (.169; p < 0.001)

1.28 (.064; p < 0.001)

0.83 (.075; p = .022)

0.86 (.088; p = .102)

0.90 (.026; p < 0.001)

 Class 3 (ref) vs 1

1.77 (.096; p < 0.001)

3.45 (.425, p < 0.001)

1.44 (.103, p < 0.001)

0.87 (.122; p = .278)

0.76 (.123; p = .046)

1.06 (.051; p = .227)

 Class 3 (ref) vs 2

1.18 (.065; p = .006)

1.85 (.223; p < 0.001)

1.13 (.081; p = .117)

1.05 (.148; p = .749)

0.88 (.148; p = .420)

1.18 (.059; p = .002)

  1. Bold indicates statistical significance at p < .05
  2. S.E.: standard error
  3. Ref: Reference (base) class for comparison of two classes (multinomial logistic regression)
  4. Reference categories for covariates: age (18–29 yrs), gender (male), education (primary/lower seconday), occupational level (employee), full-time employment (less than full-time employment), income (1st quintile)
  5. aN = 173 deleted observations due to missing specific information regarding occupational level
  6. bN = 228 deleted observations due to missing information regarding family income
  7. cModel include age, gender, education, occupational level, full-time employment, income and company in same multinomial logistic regression (N = 3925)