92.07 The relationship between paternal age and birthweight: Evidence from Hong Kong 2000-2005

Wednesday, April 29, 2009
Sadrivaan A and B (The Hilton Istanbul Hotel )
William B. Goggins Chinese University of Hong Kong, Hong Kong
Emily Y. Y. Chan Chinese University of Hong Kong, Hong Kong
Sian Griffiths The Chinese University of Hong Kong, Hong Kong
Background: Many countries, including Hong Kong, have seen increasing numbers of parents delaying having children into middle age. While it’s well-known that infants born to teenage and older mothers have higher risk of low birthweight (LBW), the influence of paternal age on birthweight is less studied and results have been inconsistent, particularly in regard to the possibility of increased LBW risk for older fathers. Proper confounder control is important for studying this issue as many predictors of infant birthweight are highly correlated, especially maternal and paternal age. Many previous studies have treated maternal age as categorical which may have left residual confounding in the models.

Methods: Data were all 156,180 singleton live births to Hong Kong native parents <60 years old from 1998-2005. Generalized additive models, a flexible method for handling non-linear associations, were employed with birthweight as the outcome and maternal and paternal age (as continuous variables), infant gender, parity, and nature of parental relationship as predictors.

Results
After controlling for maternal age and other confounders, mean birthweight exhibited an 'r' shaped association with paternal age, with an increasing trend until age 30, plateau until age 45, thereafter declining slightly and non-significantly. In stratified analysis infant birthweight declined considerably more with advanced paternal age for fathers with blue collar jobs than for fathers with white collar jobs. Results from models with birthweight as a binary outcome (LBW vs. normal) were similar.  

 Conclusions
Infants born to teenage fathers had lower mean birthweight even after controlling for maternal age and an association of paternal age beyond age 45 with lower mean birthweight was observed only for men in lower socioeconomic groups. This finding is consistent with the idea of ‘weathering’, a more rapid deterioration of health with age for those exposed to harsher conditions.


Learning Objectives: 1. Recognize the importance of proper confounder control and proper accounting for non-linear associations when modeling the association between birthweight and paternal age. 2. Describe the association found between paternal age and infant birthweight after controlling for confounders. 3. Discuss the potential implications of the findings in regards to the increasing tendeny of parents in developed countries to delay having children into middle age.

Sub-Theme: The impact of changing demography on public health