Down syndrome, temporal variation and fallout radiation revisited: statistical evidence

Fuad A Awwad ,Brian Francis ,Mohamed R Abonazel

doi:10.28919/cmbn/5409

Abstract

To revisit a study on the prevalence of Down Syndrome (DS) in the Fylde of Lancashire and ionizing radiation, using new birth data that allow better control for maternal age. Associations between ionizing radiation and DS prevalence have been controversial. Some studies link temporal variation in prevalence to ionizing radiation; others do not. Cases were ascertained in a prospective survey of major congenital malformations among residents in the Fylde of Lancashire between 1957 and 1991. New data on the birth maternal age distribution in the Fylde were obtained from the Office for National Statistics for most of the study period. Temporal clusters in prevalence rates were again detected using the Poisson log linear models used to measure the association between prevalence and ionizing radiation from atomic fallout. Significant effects of fallout radiation were found when maternal age was both controlled and not controlled for. Two DS prevalence peaks occurred during the study period. The first started in June 1958 and lasted for five months. The second lasted longer, starting in October 1962 and ending in August 1964. Although these peaks corresponded to peaks in fallout radiation dosage, they were not significant.

Highlights

A 1994 study conducted in the Fylde of Lancashire, UK, ascertained all cases of Down syndrome (DS) over a 35-year period from 1957 to 1991 and examined the sequence of cases for evidence of temporal clustering
The effect of ionizing radiation is assessed through a set of separate log-linear models, this time examining the yearly number of Down Syndrome (DS) counts cij and total births bij in a specific last menstrual period (LMP) year i and age group j, including ionizing radiation as an explanatory variable in the log linear models: log(μi) = log(bj) + β0 + β1RADIATIONi for the analysis uncontrolled for age and log(μij(g)) = log(bIj(g)) + [g ∈ 1]αj1 + [g ∈ 2]αj2 + [g ∈ 3]αj3 + β1 RADIATIONi for the data disaggregated by age group
A case-control study [17] found no association between parental occupational exposure to low-level ionizing radiation and DS risk, and an investigation of the impact of Chernobyl on the prevalence of congenital anomalies in 16 regions of Europe [18] concluded that the increasing prevalence of DS in the 1980s was probably unrelated to Chernobyl

Summary

INTRODUCTION

A 1994 study conducted in the Fylde of Lancashire, UK, ascertained all cases of Down syndrome (DS) over a 35-year period from 1957 to 1991 and examined the sequence of cases for evidence of temporal clustering. A significant association between DS conception prevalence rates and fallout radiation was detected both when maternal age was controlled for and when it was not. This study reanalyzes the effect of maternal age in the detection of temporal clusters using new data on normal births obtained from the Office of National Statistics (ONS). New information on the birth maternal age distribution in the Fylde was obtained for this study from the ONS for most of the study period. For 1981, data were available from the Consultant Hospital Registers (CHR) for Blackpool Victoria hospital, disaggregated into two age groups (

STATISTICAL METHODS

RESULTS

DISCUSSION AND CONCLUSION