ABSTRACT
【Objective】 To describe the BMI growth trajectories from birth to 24 months and examine the effects of fully breastfeeding time and breastfeeding duration on the BMI growth trajectories. 【Methods】 We conducted a follow-up study of a cohort of 1 388 eligible births. Height and weight were measured at birth, every 3 months to 12 months, and every 6 months to 24 months. Detailed demographic, feeding, activity and diseases data were also collected prospectively. Latent growth mixture modeling was conducted to model the BMI growth trajectories from birth to 24 months. Mixed-effects logistic regression analyses were made to examine the impacts of breastfeeding on this outcome. 【Results】 We identified six BMI trajectory patterns in the 1 285 infants analyzed and labeled them as follows: "normative growth" (26.6%), "normal-accelerating growth" (8.7%), "normal-decelerating growth" (11.7%), "lower-persistent growth" (6.6%), "rapid growth" (26.8%), and "rapid-accelerating growth" (19.5%). Mixed-effects logistic regression analyses revealed that fully breastfeeding ≥ 3 months could significantly decrease the risk of "rapid growth" (OR=0.71, 95% CI 0.51-0.98) and "rapid-accelerating growth" (OR=0.53, 95% CI 0.37-0.75). After adjustment for confounding factors, the relationship remained significant (OR=0.61, 95% CI 0.41-0.89; OR=0.46, 95% CI 0.31-0.68). Weaning before 6 months could significantly increase the risk of "rapid growth" (OR=1.76, 95% CI 1.19-2.61) and "rapid-accelerating growth" (OR=2.08, 95% CI 1.38-3.14). These associations persisted after adjustment for confounding factors (OR=1.69, 95% CI 1.08-2.71; OR=1.88, 95% CI 1.18-2.99). 【Conclusion】 In rural Western China, the main BMI growth trajectory patterns include "normative growth" , "rapid growth" , and "rapid-accelerating growth" . Prolonging breastfeeding may reduce excess weight gain during infancy.
ABSTRACT
Dados de peso de animais da raça Nelore de 90 a 450 dias de idade das regiões Sudeste (SE) e Centro-Oeste (CO) do Brasil foram utilizados para comparar estruturas de (co)variância de efeitos aleatórios em função de idade estimadas para as duas regiões por meio de modelos de regressão aleatória. Componentes de (co)variância referentes aos coeficientes de regressão aleatória foram estimados por EMREML por meio do programa REMLF90. Os efeitos fixos de grupo contemporâneo e os efeitos aleatórios genético aditivo direto, genético aditivo materno e permanente de ambiente foram modelados por polinômios quadráticos de Legendre. As comparações envolveram estruturas de covariância e de correlação dos efeitos aleatórios, herdabilidades direta e materna e a razão entre variâncias genéticas de diferentes regiões. As herdabilidades e estruturas de covariância e de correlação apresentaram comportamento semelhante nas duas regiões. A variância residual e as variâncias de efeito permanente de ambiente foram menores no CO, bem como a variância genética aditiva materna dos 150 aos 400 dias de idade. Trajetórias dos efeitos fixos em função de idade de diferentes grupos contemporâneos apresentaram diferentes formas, sugerindo a necessidade de estimar um conjunto de coeficientes de regressão específico para cada grupo contemporâneo. A variância do efeito genético aditivo materno apresentou maior heterogeneidade entre regiões do que a variância genética aditiva direta.
Body weight records from 90 to 450 days of age of Nellore animals from the Southeast and Center West of Brazil were used do estimate covariance structures of age dependent random effects for each region using random regression models. Covariance components of the regression coefficients were estimated by EMREML using the software REML90. The fixed effects of contemporary groups and additive genetic, additive maternal and permanent environment random effects were modeled by quadratic Legendre polynomials. The comparisons included structures of covariance and correlation of random effects, direct and maternal heritabitability and the ratio between genetic variances from different regions. The heritability, covariance and correlation structures showed similar patterns for both regions. Residual variance and permanent environment variances were smaller for Center West region as well as the maternal genetic additive from 150 to 400 days of age. Fixed effect trajectories in function of age of different contemporary groups showed different patterns, suggesting the necessity of specific set of regression coefficient estimates for each contemporary group. The maternal additive genetic variance showed higher heterogeneity between regions than the direct additive genetic variance.