Season of birth and reproductive performance: An analysis of family reconstitutions of 800 women born in the Netherlands at the end of the 19th century
Lj. Smits et al., Season of birth and reproductive performance: An analysis of family reconstitutions of 800 women born in the Netherlands at the end of the 19th century, CHRONOBIO I, 18(3), 2001, pp. 525-539
Background: Several studies have reported associations between season of bi
rth and reproductive characteristics such as menarcheal age, fecundability,
and twinning, but the results are inconsistent with respect to the locatio
n of high- and low-risk seasons. To assess whether this disagreement could
be due to the use of populations from different geographic areas and time f
rames instead of different etiologic pathways, we investigated the season-o
f-birth dependency of a variety of reproductive outcomes within one time- a
nd area-limited population. Methods: In a historic follow-up study, the rec
onstituted families of 800 women born between 1873 and 1887 in or near Rott
erdam, The Netherlands, were used to determine eight types of reproductive
outcome: childlessness, interval to first pregnancy, pregnancy interval, st
illbirth, neonatal death, postneonatal death, multiple birth, and gender of
offspring. The relation of these outcomes with season of birth was modeled
using cosinor functions with periods of 1 year or a half year. Data were a
nalyzed by use of logistic regression or general estimation equations (GEE)
, dependent on whether outcomes could occur more than once per woman. Resul
ts: Peaks in the model-based risks of reproductive failure were found withi
n two small temporal ranges, January 1 to February 11 and July 1 to August
11 for all outcomes except gender. The picture did not change after control
ling fur known and possible risk factors, including age, offspring's birth
cohort, and some social variables. Conclusions: This study reconfirms the i
dea that seasonal factors around conception or birth influence later reprod
uctive characteristics. Observing the consistency of the location of high-r
isk seasons across a variety of outcomes, the explanation of season-of-birt
h dependency of different reproductive outcomes need not involve multiple e
tiological pathways.