Skip to main content

Table 1 Associations between overweight at four years of age and the candidate predictors

From: Overweight at four years of age in a Swedish birth cohort: influence of neighbourhood-level purchasing power

   Univariate models Multivariate modele
Candidate predictors n (% overweight)a OR (95 % CI)d P OR (95 % CI)d P
Child variables      
Gender    0.07   0.083
 Female 1013 (13.2) 1.28 (0.98–1.67)   1.32 (0.97–1.83)  
 Male 1013 (10.7) 1.00 (reference)   1.00 (reference)  
Large for gestational age (LGA)    <0.001   0.003
 Nob 1954 (11.5) 1.00 (reference)   1.00 (reference)  
 Yes 66 (27.3) 2.90 (1.66–5.07)   2.74 (1.42–5.28)  
Parental variablesc      
Mother smoking    0.05   Not included
 No 1836 (11.6) 1.00 (reference)    
 Yes 99 (18.2) 1.69 (1.00–2.88)    
Mother’s BMI before pregnancy    <0.001   0.01
 ≤25 1407 (10.0) 1.00 (reference)   1.00 (reference)  
 25–30 334 (16.8) 1.81 (1.29–2.53)   1.56 (1.06–2.29)  
 30+ 148 (19.6) 2.19 (1.41–3.40)   1.87 (1.14–3.09)  
Father’s BMI    <0.001   <0.001
 ≤25 866 (7.4) 1.00 (reference)   1.00 (reference)  
 25–30 701 (13.0) 1.87 (1.34–2.62)   1.84 (1.30–2.60)  
 30+ 130 (25.4) 4.26 (2.67–6.82)   3.49 (2.11–5.75)  
Parental educational level    0.001   0.03
 None post-secondary 641 (15.0) 2.03 (1.41–2.92)   1.69 (1.13–2.53)  
 One post-secondary 578 (11.4) 1.48 (1.00–2.19)   1.29 (0.84–1.98)  
 Both post-secondary 600 (8.0) 1.00 (reference)   1.00 (reference)  
  1. aStudy children categorised according to each candidate predictor. The total number of children in the birth cohort with data on overweight at four years of follow-up = 2,026. The number of missing values for candidate predictors: LGA, n = 6; mother smoking, n = 91; mother’s BMI, n = 137; father’s BMI, n = 329; parental educational level, n = 207
  2. bThis category includes 23 SGA (small for gestational age) children (none of these children was overweight at four years)
  3. cQuestionnaire data obtained at recruitment
  4. dEstimated odds ratio with 95 % confidence interval
  5. eBackward selection of the candidate predictors (P for exclusion >0.10). No significant interaction effects between the selected variables. Number of children with complete data for the selected variables = 1,643 [818 boys (82 with overweight, 10.0 %) and 825 girls (101 with overweight, 12.2 %)]