Skip to main content

Table 1 Absolute frequencies, outcome prevalences, exposure prevalences, crude and pooled prevalence ratio (PR) estimates, and relative confounding for the analysis of the original data using underweight (weight for age Z-score < -2) as the outcome, previous hospitalization as the risk factor and low birth weight as confounder (situation 1 original).

From: Alternatives for logistic regression in cross-sectional studies: an empirical comparison of models that directly estimate the prevalence ratio

First stratum: Normal birth weight

 

Underweight

Normal

All

 
 

N

Prev.

N

N

Exp. prev. = 19.2%

Ever in hospital

8

4.7%

163

171

PR = 2.40

Never

14

1.9%

704

718

M-H weight = 2.69

All

22

2.5%

867

889

 

Second stratum: Low birth weight

 

Underweight

Normal

All

 
 

N

Prev.

N

N

Exp. prev. = 31.1%

Ever in hospital

16

13.4%

103

119

PR = 2.54

Never

14

5.3%

250

264

M-H weight = 4.35

All

30

7.8%

353

383

 

Combined strata: Normal and low birth weight

 

Underweight

Normal

All

Exp. prev. = 22.8%

 

N

Prev.

N

N

PR (crude) = 2.90

Ever in hospital

24

8.3%

266

290

PR (M-H) = 2.48

Never

28

2.9%

954

982

Confounding = -14.4%

All

52

4.1%

1220

1272

P-value(het)*= 0.9

  1. * P-value for testing heterogeneity of the prevalence ratios across strata.