Causal effects of lung function and COPD on functional outcome following ischemic stroke: a univariable and multivariable Mendelian randomization study

Weiwei Sun, Juan Wang, Yao Zhang, Fang Pan, Zhending Gan, Yun Yang, Jun Lü, Long‐yun Zhou, Guangxu XumedRxiv 2025 · DOI

Abstract

While clinical observations frequently link impaired respiratory health with poor recovery after an ischemic stroke (IS), the causal nature of this relationship remains elusive. This study utilizes univariable and multivariable Mendelian randomization (MR) to investigate whether lung function indices and Chronic Obstructive Pulmonary Disease (COPD) genetically influence functional outcomes post-stroke. Using data from large-scale GWAS repositories, including the FinnGen project and the GISCOME Network, the researchers found no significant genetic causal relationship in univariable analyses. However, multivariable models adjusting for lifestyle factors like smoking and insomnia revealed significant associations. These findings suggest that the observed clinical link may be driven by shared risk factors rather than a direct causal pathway from lung function to stroke recovery.

Introduction

Ischemic stroke is a leading cause of long-term disability, and optimizing recovery is a primary goal of post-stroke care. Clinicians have long observed that patients with pre-existing respiratory issues—such as COPD or reduced lung capacity—tend to have worse functional outcomes. Specifically, metrics like Forced Expiratory Volume in one second (FEV1) and Forced Vital Capacity (FVC) are often correlated with higher disability scores months after the event.

However, it is unclear if poor lung function directly impairs neurological recovery (perhaps through systemic inflammation or chronic hypoxia) or if both conditions are simply consequences of shared risk factors like smoking, obesity, and advanced age. Distinguishing between correlation and causation is critical: if the relationship is not causal, intensive respiratory interventions may not yield the expected improvements in stroke prognosis. This study employs Mendelian randomization to clarify these relationships.

Methods

To minimize the impact of confounding factors and reverse causation, the researchers used Two-Sample Mendelian Randomization (MR). This method uses naturally occurring genetic variants (Single Nucleotide Polymorphisms, or SNPs) as instrumental variables to proxy for specific exposures, effectively mimicking a randomized controlled trial.

Data Sources

  • Lung Function: GWAS summary data for FVC, FEV1, FEV1/FVC ratio, and Peak Expiratory Flow (PEF) were sourced from the IEU OpenGWAS project, primarily utilizing UK Biobank data.
  • COPD: Genetic data were obtained from the FinnGen project, encompassing over 433,000 Finnish participants.
  • Stroke Outcomes: Functional outcome data (measured by the modified Rankin Scale at three months) were provided by the Genetics of Ischemic Stroke Functional Outcome (GISCOME) Network, involving 6,021 patients of European ancestry.

Statistical Analysis

The primary analysis used the Inverse Variance Weighting (IVW) method. To ensure robustness, the authors also applied MR-Egger, weighted median methods, and two-sample Maximum Likelihood. Sensitivity analyses were conducted to detect pleiotropy (where a single gene affects multiple traits) and heterogeneity. Finally, Multivariable MR (MVMR) was used to adjust for potential confounders, including smoking initiation, insomnia, depression, and waist-to-hip ratio (WHR).

Results

Univariable MR: The Null Hypothesis

The univariable MR analysis found no evidence of a direct genetic causal relationship between lung function indices (FVC, FEV1, FEV1/FVC, PEF) or COPD and functional outcomes after ischemic stroke (all $P > 0.05$). Sensitivity checks confirmed the absence of significant directional pleiotropy or heterogeneity, suggesting that the null results are statistically robust.

Multivariable MR: The Role of Confounders

Interestingly, the relationship changed when adjusting for external factors. In the multivariable models:

  • COPD was associated with a higher risk of adverse functional outcomes (OR: 1.34, 95% CI: 1.09–1.65) after adjusting for insomnia, smoking, and depression.
  • PEF showed a significant association with outcomes (OR: 0.27) when adjusted for waist-to-hip ratio.

These results suggest that while lung function may not be a primary driver of stroke recovery in isolation, it interacts significantly with a patient's broader metabolic and lifestyle profile.

Discussion

The findings challenge the conventional clinical assumption that "improving lung function itself" will directly enhance stroke prognosis. Instead, the data suggest that lung function indices often serve as proxies for an aggregate of pervasive confounders. In the clinical setting, the poor outcomes observed in COPD patients are likely the result of the same lifestyle and metabolic factors that contributed to both their respiratory decline and their poor stroke recovery.

Clinical Implications

This study suggests that rehabilitation strategies should focus on a comprehensive management of risk factors—such as smoking cessation, weight management, and mental health—rather than focusing narrowly on respiratory therapy as a means to improve neurological outcomes.

Limitations and Future Directions

The study was limited to populations of European ancestry, and the results may not be generalizable to other ethnic groups. Furthermore, while MR is a powerful tool for establishing causality, it does not elucidate the specific biological mechanisms at play. Future research should involve high-quality, multicenter randomized controlled trials to validate these genetic findings in a clinical environment and explore other potential causal pathways in stroke rehabilitation.

Conclusion

At the level of genetic causality, lung function and COPD do not appear to be independent determinants of functional recovery following an ischemic stroke. These results emphasize the complexity of stroke prognosis and the necessity of addressing a broad spectrum of metabolic and lifestyle factors to improve patient outcomes.

Summary updated

Abstract

Abstract Background and Objectives Studies have shown that lung function and chronic obstructive pulmonary disease (COPD) are associated with functional outcome for several diseases, as well as the onset of stroke. However, the causal relationship between lung function, COPD and the functional outcome of ischemic stroke (IS) is unclear. The aim of this study was to investigate the causal relationship between lung function, COPD and IS functional outcomes using Mendelian randomization (MR). Methods Lung function data were derived from the Genome-wide Association Study (GWAS) summary database. Data of COPD were derived from the national genetic factors data repository of the FinnGen project in Finland. The outcome data of IS function were obtained from the Genetics of Ischemic Stroke Functional Outcome (GISCOME) Network. Univariate MR analysis mainly used the inverse variance weighting (IVW) method, MR-Egger, weighted median method and two-sample Maximum Likelihood to evaluate the causal effects of lung function and COPD on the prognosis of IS. Sensitivity analyses were conducted by evaluating heterogeneity and potential pleiotropy. Additionally, multivariable MR analyses were employed to ascertain whether these effects were independent. Results In univariate MR analyses, the estimates of IVW, MR-Egger, weighted median method and two-sample Maximum Likelihood analysis all showed that there was no genetic causal relationship between COPD, FVC, FEV1, FEV1/FVC, PEF and functional outcome after IS (P>0.05). Sensitivity analyses yielded no evidence of directional pleiotropy or substantial heterogeneity. Multivariable MR analyses demonstrated that, after adjustment for insomnia, smoking initiation, waist-to-hip ratio(WHR) or depression, both COPD and PEF exhibited a causal association with adverse functional outcome following IS. Conclusion Our research indicated that at the level of genetic causality, no significant associations were observed between lung function indices (FEV1, FVC, FEV1/FVC, PEF) or COPD and functional outcome following IS. Our discovery may challenge the conventional notion that “improving lung function itself can efficiently promote the prognosis of IS function”. Future research needs to further explore other possible causal pathways and consider various factors comprehensively to develop more effective rehabilitation strategies after IS.
BibTeX
@misc{sun2025causal,
  author = {Weiwei Sun and Juan Wang and Yao Zhang and Fang Pan and Zhending Gan and Yun Yang and Jun Lü and Long‐yun Zhou and Guangxu Xu},
  title = {{Causal effects of lung function and COPD on functional outcome following ischemic stroke: a univariable and multivariable Mendelian randomization study}},
  year = {2025},
  doi = {10.1101/2025.09.03.25335062},
  howpublished = {medRxiv},
}
Figure