Questioning the Causal Link between Maternal Smoking during Pregnancy and Offspring Use of Psychotropic Medication: A Sibling Design Analysis
Merlo J (2013) Questioning the Causal Link between Maternal Smoking during Pregnancy and Offspring Use of
Psychotropic Medication: A Sibling Design Analysis. PLoS ONE 8(5): e63420. doi:10.1371/journal.pone.0063420
Questioning the Causal Link between Maternal Smoking during Pregnancy and Offspring Use of Psychotropic Medication: A Sibling Design Analysis
Lovisa So derstro m 0
Raquel Perez-Vicente 0
Sol Jua rez 0
Juan Merlo 0
Cuilin Zhang, National Institutes of Health - National Institute of Child Health and Human Development, United States of America
0 1 Unit for Social Epidemiology, Department of Clinical Sciences, Faculty of Medicine, Lund University , Malmo , Sweden , 2 Centre for Economic Demography, Lund University , Malmo , Sweden
A recent population-based, longitudinal study from Finland observed a dose-response association between smoking during pregnancy (SDP) and use of psychotropic medications in exposed children and young adults. However, this association may be confounded by unmeasured familial characteristics related to both SDP and offspring mental health. Consequently, we aim to investigate the effect of SDP by means of a sibling design that to some extent allows controlling for unknown environmental and genetic confounders. Using the Swedish Medical Birth Register (1987-1993), which was linked to the Swedish Prescribed Drugs Register (July 2005-December 2008), we investigated 579,543 children and among them 39, 007 were discordant for use of psychotropic medication and 4,021 siblings discordant for both use of psychotropic medication and for smoking exposure. Replicating the Finnish study using traditional logistic regression methods we found an association between exposure to $10 cigarettes per day during pregnancy and psychotropic drug use (odds ratio = 1.61, 95% confidence interval 1.56, 1.66). Similar in size to the association reported from Finland (odds ratio = 1.63; 95% confidence interval 1.53, 1.74). However, in the adjusted sibling analysis using conditional logistic regression, the association was considerably reduced (odds ratio 1.22; 95% confidence interval 1.08, 1.38). Preventing smoking is of major public health importance. However, SDP per se appears to have less influence on offspring psychotropic drug use than previously suggested.
-
Funding: This work was supported by the Centre for Economic Demography at Lund University http://www.ed.lu.se/EN/AboutCED/default.asp. The funders had
no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
Several studies have reported an association between maternal
smoking during pregnancy (SDP) and offspring psychological
disorders. These include mainly externalizing behavioral
disorders, such as attention deficit hyperactivity disorder (ADHD),
conduct disorder, and oppositional defiant disorder, but also
internalizing psychopathology (i.e., depression and anxiety
disorders) [1,2,3,4]. Ekblad et al have recently observed a
dosedependent association between maternal SDP and offspring use of
psychotropic medication during childhood and up to young
adulthood [5]. Their hypothesis was that prenatal smoking
exposure interferes with the development of the fetal brain and,
thus, increases psychiatric morbidity, leading to increased risk for
use of psychotropic medications. Compared to children unexposed
to SDP, the authors found odds ratios (ORs) of 1.36 and 1.63,
respectively, for exposure to less than, and more than, 10 cigarettes
per day. Their analysis was adjusted for sex, maternal age,
obstetric characteristics, as well as psychiatric diagnosis of the
mother.
However, almost simultaneously with Ekblad and coauthors [5],
another study by Lavigne et al questioned the role of maternal
SDP as a risk factor for psychopathology in young children [6].
The analysis by Lavigne et al controlled for a more detailed set of
potential confounders, such as socioeconomic status, life stress,
family conflict, maternal depression, maternal scaffolding skills,
motherchild attachment, and other variables that were not
available in the large record linkage databases used by Ekblad et al
[5].
Investigating causality through observational studies is
problematic and the study of long-term causal effects of prenatal
smoking exposure presents specific difficulties [7,8,9]. In fact,
psychological problems of the mother, like depression and
neuroticism, as well as socioeconomic disadvantage may be
common causes of both maternal SDP [10,11] and offspring
psychiatric morbidity. Therefore, it is possible that the results
obtained by Ekblad et al [5] are due to residual confounding
through environmental circumstances in these families or through
an inherited risk for psychiatric morbidity.
Recently, in an attempt to reduce residual confounding, several
studies applied a design comparing siblings differently exposed to
materna (...truncated)