Clinical Study

Excess Early Postnatal Weight Gain Leads to Increased Abdominal Fat in Young Children

Table 2

Associations between both size at birth and early infancy growth parameters and fat distribution indices at 5 years.

Body mass index (kg/m2)Waist circumference (cm)Intra-abdominal fat (mm)Subcutaneous fat (mm)
Linear regression coefficient (95%-CI) valueLinear regression coefficient (95%-CI) valueLinear regression coefficient (95%-CI) valueLinear regression coefficient (95%-CI) value

-score birth size*
 Crude0.24 (0.080–0.39)0.0030.42 (0.007–0.83)0.0460.081 (−0.64–0.81)0.830.15 (−0.21–0.51)0.42
 Model 10.24 (0.084–0.40)0.0030.45 (0.040–0.85)0.0310.048 (−0.67–0.77)0.900.16 (−0.20–0.51)0.38
 Model 20.40 (0.029–0.77)0.0340.050 (−0.67–0.77)0.890.13 (−0.21–0.48)0.45
 Model 30.12 (−0.54–0.78)0.720.14 (−0.20–0.49)0.42

Weight gain (gr/day)
 Crude0.13 (0.089–0.16)<0.0010.34 (0.24–0.43)<0.0010.16 (−0.018–0.34)0.0780.096 (0.009–0.18)0.031
 Model 10.13 (0.087–0.16)<0.0010.33 (0.23–0.43)<0.0010.11 (−0.075–0.30)0.240.16 (0.071–0.25)<0.001
 Model 20.21 (0.11–0.31)<0.0100.14 (−0.062–0.34)0.180.10 (0.004–0.20)0.040
 Model 30.12 (−0.067–0.30)0.210.10 (0.004–0.19)0.041

Length gain (mm/day)
 Crude0.79 (−2.6–4.2)0.658.8 (0.005–17.6)0.0503.0 (−12.3–18.3)0.70−1.3 (−8.9–6.2)0.73
 Model 10.48 (−2.9–3.8)0.787.7 (−1.0–16.4)0.0832.1 (−13.2–17.3)0.79−0.50 (−8.0–7.0)0.90
 Model 2
 Model 31.5 (−12.5–15.5)0.84−0.55 (−8.0–6.9)0.89

Z-score weight for length gain**
 Crude0.51 (0.37–0.66)<0.0011.2 (0.82–1.6)<0.0010.43 (−0.29–1.1)0.250.68 (0.33–1.0)<0.001
 Model 10.51 (0.37–0.66)<0.0011.2 (0.83–1.6)<0.0010.42 (−0.29–1.1)0.250.68 (0.34–1.0)<0.001
 Model 20.84 (0.46–1.2)<0.0010.49 (−0.26–1.2)0.200.50 (0.15–0.85)0.006
 Model 30.44 (−0.24–1.1)0.210.50 (0.15–0.85)0.005

A -score of +1 SD WLG indicates that the weight gain of a certain child is one standard deviation larger than the mean weight gain in the population based on the length gain of that child.
*Birth weight adjusted for birth length, gestational age and gender.
**Weight gain rate adjusted for length gain rate and gender.
All results are linear regression coefficients with 95%-confidence intervals.
Model 1: adjusted for age and gender.
Model 2: adjusted for age, gender and current height (the analyses with rate of length gain are not adjusted for current height).
Model 3: adjusted for age, gender, current height and observer (for intra-abdominal fat and subcutaneous fat).