Prenatal antidepressant exposure and the risk of decreased gestational age and lower birthweight: A polygenic score approach to investigate confounding by indication

Anna-Sophie Rommel; Birgitte Dige Semark; Xiaoqin Liu; Kathrine Bang Madsen; Esben Agerbo; Trine Munk-Olsen; Liselotte Vogdrup Petersen; Veerle Bergink

doi:10.1111/acps.13636

Prenatal antidepressant exposure and the risk of decreased gestational age and lower birthweight: A polygenic score approach to investigate confounding by indication

Anna-Sophie Rommel, Birgitte Dige Semark, Xiaoqin Liu, Kathrine Bang Madsen, Esben Agerbo, Trine Munk-Olsen, Liselotte Vogdrup Petersen, Veerle Bergink

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaper › Journal article › Research › peer-review

2 Citations (Scopus)

Abstract

INTRODUCTION: Prenatal antidepressant exposure has been associated with lower gestational age and birthweight. Yet, unmeasured residual confounding may inflate this association. We explored if maternal genetic liability for major depression explains part of the association of antidepressant use in pregnancy with lower gestational age and birthweight.

MATERIAL AND METHODS: We employed the maternal polygenic score (PGS) for major depression as a measure of genetic liability. We used generalised linear models to estimate the differences in gestational age and birthweight at each PGS quintile between children whose mothers continued antidepressant use during pregnancy (continuation group), children whose mothers discontinued antidepressant use during pregnancy (discontinuation group) and unexposed children.

RESULTS: After adjusting for confounders, we found significant differences in birthweight between PGS quintiles in the continuation and unexposed group. Yet, this relationship was not linear. Furthermore, at the lowest and highest PGS quintiles, the continuation group had significantly reduced mean gestational ages (adjusted β ranges: 1.7-4.5 days, p < 0.001-0.008) and lower mean birthweights (adjusted β ranges: 58.6-165.4 g, p = 0.001-0.008) than the discontinuation and unexposed groups.

CONCLUSION: We confirmed that antidepressant use in pregnancy was associated with small reductions in gestational age and birthweight but found that genetic liability for depression was not linearly associated with this risk. The causality of the observed associations could not be established due to the observational nature of the study. Residual confounding linked to the underlying disease was likely still present.

Original language	English
Journal	Acta Psychiatrica Scandinavica
Volume	150
Issue	5
Pages (from-to)	344-354
Number of pages	11
ISSN	0001-690X
DOIs	https://doi.org/10.1111/acps.13636
Publication status	Published - Nov 2024

Keywords

antidepressants
low birthweight
neonatal outcomes
perinatal depression
preterm birth

Access to Document

10.1111/acps.13636

Cite this

Rommel, A.-S., Semark, B. D., Liu, X., Madsen, K. B., Agerbo, E., Munk-Olsen, T., Petersen, L. V., & Bergink, V. (2024). Prenatal antidepressant exposure and the risk of decreased gestational age and lower birthweight: A polygenic score approach to investigate confounding by indication. Acta Psychiatrica Scandinavica, 150(5), 344-354. https://doi.org/10.1111/acps.13636

@article{ffb32de8fe1a48829067748d0cea62f2,

title = "Prenatal antidepressant exposure and the risk of decreased gestational age and lower birthweight: A polygenic score approach to investigate confounding by indication",

abstract = "INTRODUCTION: Prenatal antidepressant exposure has been associated with lower gestational age and birthweight. Yet, unmeasured residual confounding may inflate this association. We explored if maternal genetic liability for major depression explains part of the association of antidepressant use in pregnancy with lower gestational age and birthweight.MATERIAL AND METHODS: We employed the maternal polygenic score (PGS) for major depression as a measure of genetic liability. We used generalised linear models to estimate the differences in gestational age and birthweight at each PGS quintile between children whose mothers continued antidepressant use during pregnancy (continuation group), children whose mothers discontinued antidepressant use during pregnancy (discontinuation group) and unexposed children.RESULTS: After adjusting for confounders, we found significant differences in birthweight between PGS quintiles in the continuation and unexposed group. Yet, this relationship was not linear. Furthermore, at the lowest and highest PGS quintiles, the continuation group had significantly reduced mean gestational ages (adjusted β ranges: 1.7-4.5 days, p < 0.001-0.008) and lower mean birthweights (adjusted β ranges: 58.6-165.4 g, p = 0.001-0.008) than the discontinuation and unexposed groups.CONCLUSION: We confirmed that antidepressant use in pregnancy was associated with small reductions in gestational age and birthweight but found that genetic liability for depression was not linearly associated with this risk. The causality of the observed associations could not be established due to the observational nature of the study. Residual confounding linked to the underlying disease was likely still present.",

keywords = "antidepressants, low birthweight, neonatal outcomes, perinatal depression, preterm birth",

author = "Anna-Sophie Rommel and Semark, {Birgitte Dige} and Xiaoqin Liu and Madsen, {Kathrine Bang} and Esben Agerbo and Trine Munk-Olsen and Petersen, {Liselotte Vogdrup} and Veerle Bergink",

note = "{\textcopyright} 2023 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd.",

year = "2024",

month = nov,

doi = "10.1111/acps.13636",

language = "English",

volume = "150",

pages = "344--354",

journal = "Acta Psychiatrica Scandinavica",

issn = "0001-690X",

publisher = "John Wiley & Sons Inc.",

number = "5",

}

Prenatal antidepressant exposure and the risk of decreased gestational age and lower birthweight: A polygenic score approach to investigate confounding by indication. / Rommel, Anna-Sophie; Semark, Birgitte Dige ; Liu, Xiaoqin et al.
In: Acta Psychiatrica Scandinavica, Vol. 150, No. 5, 11.2024, p. 344-354.

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaper › Journal article › Research › peer-review

TY - JOUR

T1 - Prenatal antidepressant exposure and the risk of decreased gestational age and lower birthweight

T2 - A polygenic score approach to investigate confounding by indication

AU - Rommel, Anna-Sophie

AU - Semark, Birgitte Dige

AU - Liu, Xiaoqin

AU - Madsen, Kathrine Bang

AU - Agerbo, Esben

AU - Munk-Olsen, Trine

AU - Petersen, Liselotte Vogdrup

AU - Bergink, Veerle

PY - 2024/11

Y1 - 2024/11

N2 - INTRODUCTION: Prenatal antidepressant exposure has been associated with lower gestational age and birthweight. Yet, unmeasured residual confounding may inflate this association. We explored if maternal genetic liability for major depression explains part of the association of antidepressant use in pregnancy with lower gestational age and birthweight.MATERIAL AND METHODS: We employed the maternal polygenic score (PGS) for major depression as a measure of genetic liability. We used generalised linear models to estimate the differences in gestational age and birthweight at each PGS quintile between children whose mothers continued antidepressant use during pregnancy (continuation group), children whose mothers discontinued antidepressant use during pregnancy (discontinuation group) and unexposed children.RESULTS: After adjusting for confounders, we found significant differences in birthweight between PGS quintiles in the continuation and unexposed group. Yet, this relationship was not linear. Furthermore, at the lowest and highest PGS quintiles, the continuation group had significantly reduced mean gestational ages (adjusted β ranges: 1.7-4.5 days, p < 0.001-0.008) and lower mean birthweights (adjusted β ranges: 58.6-165.4 g, p = 0.001-0.008) than the discontinuation and unexposed groups.CONCLUSION: We confirmed that antidepressant use in pregnancy was associated with small reductions in gestational age and birthweight but found that genetic liability for depression was not linearly associated with this risk. The causality of the observed associations could not be established due to the observational nature of the study. Residual confounding linked to the underlying disease was likely still present.

AB - INTRODUCTION: Prenatal antidepressant exposure has been associated with lower gestational age and birthweight. Yet, unmeasured residual confounding may inflate this association. We explored if maternal genetic liability for major depression explains part of the association of antidepressant use in pregnancy with lower gestational age and birthweight.MATERIAL AND METHODS: We employed the maternal polygenic score (PGS) for major depression as a measure of genetic liability. We used generalised linear models to estimate the differences in gestational age and birthweight at each PGS quintile between children whose mothers continued antidepressant use during pregnancy (continuation group), children whose mothers discontinued antidepressant use during pregnancy (discontinuation group) and unexposed children.RESULTS: After adjusting for confounders, we found significant differences in birthweight between PGS quintiles in the continuation and unexposed group. Yet, this relationship was not linear. Furthermore, at the lowest and highest PGS quintiles, the continuation group had significantly reduced mean gestational ages (adjusted β ranges: 1.7-4.5 days, p < 0.001-0.008) and lower mean birthweights (adjusted β ranges: 58.6-165.4 g, p = 0.001-0.008) than the discontinuation and unexposed groups.CONCLUSION: We confirmed that antidepressant use in pregnancy was associated with small reductions in gestational age and birthweight but found that genetic liability for depression was not linearly associated with this risk. The causality of the observed associations could not be established due to the observational nature of the study. Residual confounding linked to the underlying disease was likely still present.

KW - antidepressants

KW - low birthweight

KW - neonatal outcomes

KW - perinatal depression

KW - preterm birth

UR - http://www.scopus.com/inward/record.url?scp=85177444720&partnerID=8YFLogxK

U2 - 10.1111/acps.13636

DO - 10.1111/acps.13636

M3 - Journal article

C2 - 37990478

SN - 0001-690X

VL - 150

SP - 344

EP - 354

JO - Acta Psychiatrica Scandinavica

JF - Acta Psychiatrica Scandinavica

IS - 5

ER -

Prenatal antidepressant exposure and the risk of decreased gestational age and lower birthweight: A polygenic score approach to investigate confounding by indication

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this