Does live birth reduce the rate of multiple sclerosis? Removing immortal time bias by alignment of time zero and exposure

Abstract Background Studies of pregnancy and multiple sclerosis report inconsistent associations suggesting a protective effect or no association. This study evaluates whether temporal misalignment between the start of follow-up (time zero) and exposure can explain the protective association between live birth and multiple sclerosis onset reported in some earlier studies. Methods Using Danish population‑based registry data, we assessed how different analytic strategies influence the estimated live birth-multiple sclerosis association. We reproduced two approaches used in prior literature: (i) fixing live birth exposure status at 15 years old (time zero); and (ii) treating live birth as a time‑varying exposure. Additionally, we used (iii) the clone-censor-weight method to align time zero with exposure and remove immortal time bias. Results When fixing live birth status at 15 years, we observed a protective association between live birth and multiple sclerosis (adjusted HR 0.30 [95% CI 0.28;0.33]). The association attenuated to the null, when live birth was treated as a time-varying exposure (adjusted HR 0.98 [95% CI 0.89;1.07]). Following the clone-censor-weight approach, the estimated effect yielded no association either (weighted HR 1.06 [95% CI 0.76;1.70]). Conclusions The protective association reported in prior studies is plausibly explained by immortal time bias. The live birth-multiple sclerosis association represents a case of temporal overlap between exposure and outcome, but the underlying issue applies broadly across epidemiological research. Aligning time zero with exposure and securing temporality is essential when temporal misalignment distorts effect estimates and hinders causal inference.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-14
DOI
https://doi.org/10.21203/rs.3.rs-11481993/v1
Primary Topic
Multiple Sclerosis Research Studies
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Does live birth reduce the rate of multiple sclerosis? Removing immortal time bias by alignment of time zero and exposure

Flemming Pociot, Julie Christine Antvorskov, Gitte Lindved Petersen, Tine Dreier Bille et al.
2 links
Research Square
Multiple Sclerosis Research Studies
preprint

Does live birth reduce the rate of multiple sclerosis? Removing immortal time bias by alignment of time zero and exposure

Flemming Pociot, Julie Christine Antvorskov, Gitte Lindved Petersen, Tine Dreier Bille, Tri-Long Nguyen
preprint en

Abstract

Abstract Background Studies of pregnancy and multiple sclerosis report inconsistent associations suggesting a protective effect or no association. This study evaluates whether temporal misalignment between the start of follow-up (time zero) and exposure can explain the protective association between live birth and multiple sclerosis onset reported in some earlier studies. Methods Using Danish population‑based registry data, we assessed how different analytic strategies influence the estimated live birth-multiple sclerosis association. We reproduced two approaches used in prior literature: (i) fixing live birth exposure status at 15 years old (time zero); and (ii) treating live birth as a time‑varying exposure. Additionally, we used (iii) the clone-censor-weight method to align time zero with exposure and remove immortal time bias. Results When fixing live birth status at 15 years, we observed a protective association between live birth and multiple sclerosis (adjusted HR 0.30 [95% CI 0.28;0.33]). The association attenuated to the null, when live birth was treated as a time-varying exposure (adjusted HR 0.98 [95% CI 0.89;1.07]). Following the clone-censor-weight approach, the estimated effect yielded no association either (weighted HR 1.06 [95% CI 0.76;1.70]). Conclusions The protective association reported in prior studies is plausibly explained by immortal time bias. The live birth-multiple sclerosis association represents a case of temporal overlap between exposure and outcome, but the underlying issue applies broadly across epidemiological research. Aligning time zero with exposure and securing temporality is essential when temporal misalignment distorts effect estimates and hinders causal inference.

Research Square
University of Copenhagen (DK), Steno Diabetes Center Copenhagen (DK)
Good health and well-being
Multiple Sclerosis Research Studies
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon