Association of Birth Weight with Health and Long-Term Survival up to Middle and Old Ages in China.

Published

Journal Article

This study is based on an unique dataset of birth records of 11,693 persons born in Beijing, China from 1921-54 and from clinical examinations (May 2003-April 2005) of 2,085 tracked surviving cohort members (aged 50-82) to diagnose seven major chronic diseases. Data were analyzed using the extended Fixed Attribute Dynamics (FAD) method and multivariate regressions. The results of our FAD analysis have shown that, as compared to the persons with low birth weight (<2,500 g), the probability of survival from age 0 to ages 50-56, 63-67 and 68-82 for persons with higher birth weight (2,500-2,999 g, 3,000-3499 g, or ≥3,500 g) was 16-31, 62-104, and 52-108% higher respectively. These estimates display similar patterns for both genders, with somewhat stronger effects in men. Of the 27 estimates of the Odds Ratio of Survival in the FAD analysis, 21 are statistically significant. Controlling for 14 confounding factors, multivariate binary logistic regressions have demonstrated that the risk of having cardiovascular disease and diabetes is negatively associated birth weight; ordinal logit regressions have shown that the number of major chronic diseases at ages 50-82 is significantly associated with birth weight: the lower the birth weight, the higher the risk of having more chronic diseases. We conclude that low birth weight is negatively associated with 1) long-term survival probability from age 0 to ages 50-82, and 2) overall health at middle and old ages. Further research is needed to understand the mechanisms of these effects.

Full Text

Duke Authors

Cited Authors

  • Zeng, Y; Zhang, Z; Xu, T; Fan, Z; Xiao, X; Chen, X; Wang, Z; Ma, E; Gu, D; Zhang, F; Corless, JM

Published Date

  • December 2010

Published In

Volume / Issue

  • 3 / 3-4

Start / End Page

  • 143 - 159

PubMed ID

  • 22468162

Pubmed Central ID

  • 22468162

International Standard Serial Number (ISSN)

  • 1874-7884

Digital Object Identifier (DOI)

  • 10.1007/s12062-011-9035-7

Language

  • eng

Conference Location

  • Netherlands