Light alcohol consumption in pregnancy and socioemotional difficulties and
cognitive performance in childhood.
(Yvonne Kelly, Amanda Sacker, Maria Iacavou)
The link between heavy alcohol consumption during pregnancy and health and
developmental problems in children is well established. An emerging literature suggests
that ‘light’ drinking during pregnancy is not linked to detrimental impacts
on socioemotional and cognitive development during early childhood. In this
context, randomised controlled trials are not feasible and the evidence base
relies on observational studies which typically use regression modelling to
account for the effects of confounding factors. The prroblem with this is
that drinking during pregnancy is socially patterned – ‘light’ drinkers
are more likely to be socially advantaged, and
Advantaged social position is linked to more favourable developmental
profiles in young children Therefore, we can not be sure that an apparent
lack of negative impacts of ‘light’ drinking on early child development is
‘real’.
Even though direct experimental studies can not be performed here, we can
simulate experimental conditions in an attempt to advance the evidence base.
In this paper we use propensity score matching, a technique in which
children of non-drinkers and ‘light’ drinkers are matched on a range of
observed factors, crreating matched “treatment”
and “control” samples, composed respectively of light drinkers and non
drinkers. We used data from theMillennium Cohort Study (MCS).
Presented by:
Yvonne Kelly (ISER)
Date & time:
May 16, 2012 12:00 pm
Internal seminars home