Sage Journals: Discover world-class research

Abstract

Objective

Aging has been shown to be associated with adverse health outcomes in patients with interstitial lung disease. However, the causal relationship between them is not fully understood. In this study, two-sample Mendelian randomization was applied to analyze the causal relationship between aging phenotypes (facial aging and telomere length) and interstitial lung disease risk.

Methods

Data on single nucleotide polymorphisms were extracted from the pooled dataset of genome-wide association studies. Single nucleotide polymorphisms were used as instrumental variables. The causal association between aging phenotypes and interstitial lung disease risk was evaluated using inverse variance weighting, Bayesian weighted Mendelian randomization, Mendelian randomization–robust adjusted profile score, Mendelian randomization–Egger regression, and weighted median methods.

Results

Inverse variance weighting revealed that facial aging increased the risk of genetic susceptibility to interstitial lung disease (odds ratio: 2.336, 95% confidence interval: 1.256–4.342, p = 0.007). Telomere length was negatively correlated with interstitial lung disease risk (odds ratio: 0.632, 95% confidence interval: 0.523–0.765, p < 0.001). No reverse causality was found; interstitial lung disease had no significant effect on facial aging (odds ratio = 0.999, 95% confidence interval: 0.995–1.003, p = 0.664) or telomere length (odds ratio = 0.996, 95% confidence interval: 0.989–1.004, p = 0.328).

Conclusion

The results of this study show that facial aging significantly increases the risk of interstitial lung disease, while telomere length significantly reduces the risk. Anti-aging may be an effective strategy for the prevention and treatment of interstitial lung disease.

Keywords

Interstitial lung disease aging causal effect Mendelian randomization

Introduction

Interstitial lung disease (ILD) refers to a group of lung diseases characterized by interstitial fibrosis and/or inflammation. ILD can be idiopathic or associated with a known etiology. Idiopathic pulmonary fibrosis (IPF), fibrosing hypersensitivity pneumonitis, and connective tissue disease–related ILD (CTD-ILD) are reported to be the three most common subtypes of fibrosing ILDs, which can lead to progressive pulmonary fibrosis and respiratory failure.¹ ILD imposes a substantial burden on patients and healthcare systems. Common symptoms, including dyspnea on exertion, chronic cough, and fatigue, contribute to physical deconditioning and a subsequent reduction in the health-related quality of life. Moreover, fibrotic progression is often associated with early mortality, with patients with IPF exhibiting a median survival of only 3–5 years following diagnosis in a previous study.² Notably, the incidence and prevalence of several ILDs, especially IPF, increase significantly with aging, suggesting a strong link between biological aging and pulmonary fibrosis.^3,4 Concurrently, ILD is shown to cause age-related complications, such as cardiovascular diseases, malignant tumors, and metabolic disorders, which increase the complexity of clinical treatment and may affect disease progression and treatment outcome.⁵

Aging is a multifactorial biological process characterized by a gradual decline in cellular function, reduced tissue repair capacity, and increased susceptibility to chronic diseases, including ILD. The notable features of aging include genomic instability, telomere shortening, epigenetic changes, and cellular senescence, which have been increasingly observed in the lung tissues of patients with ILD.^6,7 Assessment of senescence features requires the use of relevant metrics such as telomere length (TL)⁸ and facial aging (FA).⁹ Telomeres are molecular biological indicators, and the shortening of telomeres is a microscopic indicator of cellular senescence and human aging.^10,11 FA is a phenotypic biomarker.¹² Despite significant progress in our understanding of the biology of aging based on previous studies, given the inherent limitations of observational studies and the potential bias in the association between aging and ILD risk introduced by insufficient adjustment for confounding variables, there is a need for a more rigorous approach to assess the evidence regarding their causality.

In this study, we employed a two-sample bidirectional Mendelian randomization (MR) framework to explore the potential causal relationship between the predicted aging phenotypes and ILD risk. By leveraging the aggregated data from large-scale genome-wide association studies (GWAS), we aimed to minimize the influence of residual confounding factors and reverse causality. The results of this study may deepen our understanding of the mechanisms related to aging in fibrotic lung diseases and provide a theoretical basis for future prevention and intervention strategies targeting the pathways associated with aging in ILD.

Methods

This study was conducted and reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology using MR (STROBE-MR) guidelines. The design of our study is shown in Figure 1, with a focus on adhering to three key assumptions to ensure the validity of the MR results. Assumption 1 (relevance): The genetic variants (instrumental variables (IVs)) must be strongly associated with the exposure. Assumption 2 (independence): The IVs must be independent of confounders that could affect both exposure and outcome. Assumption 3 (exclusion restriction): The IVs must influence the outcome only through the exposure, not via alternative pathways. The absence of direct paths from IVs to ILD, other than that through aging, supports the IV assumption of no horizontal pleiotropy.

Figure 1.

Three key assumptions of MR. MR: Mendelian randomization. Solid arrows indicate hypothesized causal pathways, while dashed arrows represent potential violations of MR assumptions that were tested and addressed using sensitivity analyses.

Data sources

The summary-level data for TL and FA were obtained from the OPEN GWAS database (https://gwas.mrcieu.ac.uk/), with sample sizes of 472,174 for TL (216,187 men and 255,987 women; average age: 56.1 ± 7.9 years) and 423,999 for FA (194,391 men and 229,601 women; age: 40–69 years). To prevent sample overlap, the GWAS dataset for ILD was obtained from the FinnGen database (R11, https://www.finngen.fi/en/access_results), which contained 5097 cases and 448,636 controls.¹³ This study was conducted in accordance with the ethical principles of the Declaration of Helsinki (1975), as revised in 2024. All data were obtained from publicly available databases and did not require ethical approval; the relevant details are shown in Table S1.

Selection of IVs

IVs were screened according to the following criteria. Initially, single nucleotide polymorphisms (SNPs) highly associated with exposure were selected based on a significance threshold of p < 5.0E-08. Subsequently, strict clustering was implemented to eliminate chain imbalance (window size = 10 MB, R2 < 0.001) based on data from the European 1000 Genomes Reference Group. In addition, the F-statistic (F = β²/(standard error)²) was calculated for each SNP, with F-statistics <10 indicating a weak IV.¹⁴ Finally, MR-Steiger test was performed to further refine the selection of SNPs to ensure the correct direction of causality.¹⁵

Statistical analyses

Five different MR analysis methods were used for this study. The multiplicative random-effect inverse variance weighted (IVW-MRE) method, which provides the most accurate and effective causal inference, was used as the primary method.¹⁶ Bayesian weighted MR,¹⁷ MR–robust adjusted profile score (MR-RAPS),¹⁸ MR-Egger,¹⁹ and weighted median²⁰ methods were used as supplementary analyses. A threshold value of p < 0.05 was considered to indicate a significant causal relationship.

A series of sensitivity analyses were conducted to ensure the robustness of the analytical results. Cochran Q-test was used to assess the presence of heterogeneity.²¹ MR-Egger regression intercept test¹⁹ and MR-pleiotropy residual sum and outlier (MR-PRESSO) global test²² were used to detect potential horizontal pleiotropy. A p value >0.05 indicated no heterogeneity or horizontal pleiotropy. The leave-one-out (LOO) method was used to identify influential SNPs.²³

Statistical analyses were performed using R (version 4.3.1). MR analysis was performed using TwoSampleMR and MRPRESSO software packages.

Results

The causal relationship between aging and ILD risk

In positive MR analysis estimating the causal effect of aging on ILD, the IVW-MRE method showed that FA increased ILD risk (odds ratio (OR) = 2.336, 95% confidence interval (CI): 1.256–4.342, p = 0.007) (Table 1). The direction of the causal effect estimated via other methods used as a complementary analysis was in line with that obtained using the IVW-MRE method. The results of the Cochran Q test showed that no significant heterogeneity existed (p_IVW = 0.88, p_Egger = 0.93), and there was no evidence of horizontal pleiotropy (p = 0.767) (Table S2); TL showed a negative causality with the risk of ILD (OR = 0.632, 95% CI: 0.523–0.765, p < 0.001) (Table 1), and the direction of the causal effect estimated via other methods used as a supplementary analysis was consistent with that obtained using the IVW-MRE method. The results of Cochran Q analysis showed significant heterogeneity (p_IVW = 0.011, p_Egger = 0.009); however, there was no evidence of horizontal pleiotropy (p = 0.938) (Table S2). LOO analysis supported the robustness of the results (Figure 2(a) and (b)).

Table 1.

MR results.

Exposure/Outcome	Method	nSNP	OR (95% CI)	p
FA/ILD	IVW-MRE	70	2.336 (1.256–4.342)	0.007
	BWMR	70	2.362 (1.222–4.566)	0.011
	MR-RAPS	70	2.292 (1.204–4.362)	0.012
	MR-Egger	70	1.883 (0.348–10.061)	0.468
	WM	70	1.467 (0.584–3.689)	0.415
TL/ILD	IVW-MRE	133	0.632 (0.523–0.765)	<0.001
	BWMR	133	0.633 (0.525–0.763)	<0.001
	MR-RAPS	133	0.630 (0.517–0.767)	<0.001
	MR-Egger	133	0.640 (0.446–0.918)	0.017
	WM	133	0.650 (0.492–0.860)	0.003
ILD/FA	IVW-MRE	12	0.999 (0.995–1.003)	0.664
	BWMR	12	0.999 (0.995–1.004)	0.739
	MR-RAPS	12	0.999 (0.994–1.002)	0.498
	MR-Egger	12	0.996 (0.989–1.003)	0.302
	WM	12	0.997 (0.993–1.001)	0.104
ILD/TL	IVW-MRE	8	0.996 (0.989–1.004)	0.328
	BWMR	8	0.996 (0.988–1.004)	0.362
	MR-RAPS	8	0.996 (0.987–1.005)	0.396
	MR-Egger	8	1.001 (0.989–1.013)	0.873
	WM	8	0.998 (0.991–1.005)	0.587

MR: Mendelian randomization; nSNP: number of single nucleotide polymorphisms; OR: odds ratio; CI: confidence interval; FA: facial aging; ILD: interstitial lung disease; TL: telomere length; IVW-MRE: inverse variance weighted multiplicative random effects; BWMR: Bayesian weighted Mendelian randomization; MR-RAPS: Mendelian randomization–robust adjusted profile score; MR-Egger: Mendelian randomization–Egger regression; WM: weighted median.

Figure 2.

Results of the leave-one-out method. (a) FA and ILD. (b) TL and ILD. (c) ILD and FA. (d) ILD and TL. FA: facial aging; ILD: interstitial lung disease; TL: telomere length.

For reverse analyses, the causal effect of ILD on aging was estimated, and the results of the IVW-MRE method showed no significant effect of ILD on FA (OR = 0.999, 95% CI: 0.995–1.003, p = 0.664) (Table 1); furthermore, the direction of the causal effect estimated via other methods used as complementary analyses was in line with that obtained using the IVW-MRE method. The results of the Cochran Q test suggested partial heterogeneity (p_IVW = 0.047, p_Egger = 0.056); however, there was no evidence of horizontal pleiotropy (p = 0.319) (Table S2). In addition, there was no significant effect of ILD on TL (OR = 0.996, 95% CI: 0.989–1.004, p = 0.328) (Table 1), and the direction of the causal effect estimated via other methods used as a supplementary analysis was consistent with that obtained using the IVW-MRE method. No heterogeneity existed (p_IVW = 0.88, p_Egger = 0.93), and there was no evidence of horizontal pleiotropy (p = 0.349) (Table S2). LOO analyses supported the robustness of the results (Figure 2(c) and (d)).

Discussion

ILD is a typical aging-related lung disease, most commonly seen in middle-aged and older adults. Its incidence and prevalence increase significantly with aging.²⁴ Patients with ILD usually present with dyspnea, cough, and progressive deterioration of lung function; these symptoms lead to a decline in physical function, lower quality of life, and higher risk of premature death.²⁵ Research has shown several associations between aging and ILD from both demographic and biological perspectives.²⁶

Several features of aging, such as telomere biology, inflammation, cellular senescence, and stem cell exhaustion, are particularly evident in ILD and have been shown to play an important role in disease onset and progression, supporting a strong link between premature aging, impaired regenerative capacity, and fibrosis.²⁷ Telomere shortening with aging is recognized as a major hallmark of aging, and molecular mechanisms have been identified to explain the poor prognosis observed for telomere shortening in ILDs such as IPF and CTD-ILD. Telomere shortening in alveolar epithelial type II (AT2) cells and alveolar stem cells can lead to senescence-associated phenotypes owing to increased cellular renewal, oxidative stress, and inflammation. Moreover, it can affect the natural microenvironment and local cellular signaling, including transforming growth factor beta, a key marker of senescence. This leads to the deposition of extracellular matrix filaments in the interstitium and loss of healthy alveolar epithelial cells.²⁸ This is consistent with our findings that TL is a protective factor against ILD and that telomere shortening contributes to ILD development through senescence and inflammatory mechanisms. In addition, Parimon et al. showed that senescence causes significant damage to epithelial cells in ILD, especially AT2 and basal cells.²⁹ Recent data from genetically engineered mouse models support the hypothesis that senescence acts as a major driver of pulmonary fibrosis rather than merely being an incidental phenomenon. In addition to senescent AT2 cells as a key player in lung fibrosis, other cell types, such as myofibroblasts and bone marrow-derived mesenchymal stromal cells, show significant senescence in IPF, exacerbating the repair response of damaged tissue.³⁰ Senescence is an emerging research focus in ILD, and understanding the mechanisms of senescence in ILD can facilitate more targeted therapies; furthermore, biomarkers of senescence may allow the evaluation of potential therapeutic approaches for lung anti-aging and regeneration.³¹ Experimental studies support a functional link between aerobic exercise and TL preservation, and regular and sustained physical activity (PA) over a long period may contribute to aging prevention.³² An MR study also reported a protective effect of PA against IPF; thus, enhanced PA may be an effective preventive strategy for IPF.³³

Rigorous sensitivity analyses were performed in this study to assess the validity of the MR hypothesis and minimize bias in the results. The presence of horizontal pleiotropy was indicated if IVs could directly influence the outcome without passing through the exposure. Compared with other MR methods, the IVW method provided higher statistical efficacy as a primary approach and yielded accurate causal estimates even in the presence of heterogeneity. In addition, this study fulfilled the prerequisite of consistent OR direction across all MR methods, thereby enhancing the reliability of the findings.

However, certain limitations warrant consideration. First, all participants were of European descent; therefore, the generalizability of our findings to other populations with different genetic backgrounds, cultures, and lifestyles may be limited. Second, due to the incomplete understanding of the biological functions of genetic variants, it remains difficult to fully exclude the possibility of horizontal pleiotropy. Third, although FA is a practical and quantifiable aging phenotype, it is an indirect marker and may not have fully captured the biological complexity of the aging process. Its use as a proxy for systemic aging could have introduced measurement bias, especially given the potential influence of environmental, lifestyle, or cosmetic factors unrelated to biological aging. Finally, although our findings offer preliminary evidence regarding a potential causal association between FA and ILD risk, the conclusions regarding anti-ageing interventions as preventative strategies should be interpreted with caution. Further mechanistic studies and interventional trials are needed to validate these implications before clinical translation.

Conclusion

This study investigated the potential causal relationship between aging phenotypes and ILD using an MR framework. The findings suggest that aging is a causal risk factor for ILD. Although these results provide preliminary evidence supporting the relevance of aging in ILD pathogenesis, further research is needed to elucidate the underlying mechanisms. Any consideration of anti-aging interventions as preventative strategies for ILD should be approached with caution until validated by mechanistic and clinical studies.

Supplemental Material

sj-pdf-1-imr-10.1177_03000605251384426 - Supplemental material for Causality between aging and interstitial lung disease: A bidirectional two-sample Mendelian randomization study

Supplemental material, sj-pdf-1-imr-10.1177_03000605251384426 for Causality between aging and interstitial lung disease: A bidirectional two-sample Mendelian randomization study by Wen Luo, Xiaoxia Tang, Xiaohan Zhou, Yuan Liu, Changfei Yuan, Wenyao Li, Yubo Xia, Songchuan Su, Chao Qian, Zongwu Li, Kaiyun Li, Libo Shang, Zhe Wang, Yuqing Xiang, Ziliang Ruan and Tao Wang in Journal of International Medical Research

Footnotes

Acknowledgments

This study would not have been possible without the GWAS data; we thank Dr Kurki and all study participants.

Author contributions

Wen Luo, Xiaohan Zhou, and Xiaoxia Tang designed the study and drafted the manuscript. Yuan Liu and Changfei Yuan performed data collection and statistical analysis. Wenyao Li, Yubo Xia, and Songchuan Su contributed to experimental implementation and validation. Chao Qian, Zongwu Li, Kaiyun Li, and Libo Shang participated in literature review and data interpretation. Zhe Wang and Yuqing Xiang assisted in manuscript revision and visualization. Tao Wang and Ziliang Ruan supervised the research, provided critical revisions, and approved the final version.

All authors have read and approved the final manuscript.

Wen Luo, Xiaoxia Tang, and Xiaohan Zhou have contributed equally to this work.

Consent for publication

Not applicable.

Clinical trial number

Not applicable.

Data availability statement

All data generated or analyzed during this study are included in this article and its supplementary information files.

Declaration of conflicting interests

The authors hereby declare that they have no conflicts of interest to disclose.

Funding

This study received funding from the Yunnan Provincial Department of Science and Technology Science and Technology Programme (Project No. 202401AZ070001-071).

ORCID iD

Wen Luo

References

Liu

Huang

Yang

, et al. Transcriptomic and network analysis identifies shared and unique pathways and immune changes across fibrotic interstitial lung diseases. Aging (Albany NY) 2024; 16: 3200–3230.

Guler

Kwan

Leung

, et al. Functional ageing in fibrotic interstitial lung disease: the impact of frailty on adverse health outcomes. Eur Respir J 2020; 55: 1900647.

Spagnolo

Guler

Chaudhuri

, et al. Global epidemiology and burden of interstitial lung disease. Lancet Respir Med 2025; 13: 739–755.

Ley

Collard

King

, et al. Clinical course and prediction of survival in idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 2011; 183: 431–440.

Jacob

Bartholmai

Rajagopalan

, et al. Functional and prognostic effects when emphysema complicates idiopathic pulmonary fibrosis. Eur Respir J 2017; 50: 1700379.

López-Otín

Blasco

Partridge

, et al. The hallmarks of aging. Cell 2013; 153: 1194–1217.

DePianto

Heiden

JAV

Morshead

, et al. Molecular mapping of interstitial lung disease reveals a phenotypically distinct senescent basal epithelial cell population. JCI Insight 2021; 6: e143626.

Chakravarti

LaBella

DePinho

, et al. Telomeres: history, health, and hallmarks of aging. Cell 2021; 184: 306–322.

Franco

Aveleira

Cavadas

, et al. Skin senescence: mechanisms and impact on whole-body aging. Trends Mol Med 2022; 28: 97–109.

10.

Rossiello

Jurk

Passos

, et al. Telomere dysfunction in ageing and age-related diseases. Nat Cell Biol 2022; 24: 135–147.

11.

Gao

Zhang

Mons

, et al. Leukocyte telomere length and epigenetic-based mortality risk score: associations with all-cause mortality among older adults. Epigenetics 2018; 13: 846–857.

12.

Zhang

Liu

Pan

, et al. Assessing causal associations between neurodegenerative diseases and neurological tumors with biological aging: a bidirectional Mendelian randomization study. Front Neurosci 2023; 17: 1321246.

13.

Kurki

Karjalainen

Palta

, et al. FinnGen provides genetic insights from a well-phenotyped isolated population. Nature 2023; 613: 508–518.

14.

Bowden

Del Greco M

Minelli

, et al. Improving the accuracy of two-sample summary-data Mendelian randomization: moving beyond the NOME assumption. Int J Epidemiol 2019; 48: 728–742.

15.

Hemani

Tilling

Davey Smith

, et al. Orienting the causal relationship between imprecisely measured traits using GWAS summary data. PLoS Genet 2017; 13: e1007081.

16.

Burgess

Scott

Timpson

; EPIC- InterAct Consortiumet al. Using published data in Mendelian randomization: a blueprint for efficient identification of causal risk factors. Eur J Epidemiol 2015; 30: 543–552.

17.

Zhao

Ming

, et al. Bayesian weighted Mendelian randomization for causal inference based on summary statistics. Bioinformatics 2020; 36: 1501–1508.

18.

Zhao

Wang

Hemani

, et al. Statistical inference in two-sample summary-data Mendelian randomization using robust adjusted profile score. Ann Statist 2020; 48: 1742–1769.

19.

Bowden

Davey Smith

Burgess

Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol 2015; 44: 512–525.

20.

Bowden

Davey Smith

Haycock

, et al. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol 2016; 40: 304–314.

21.

Burgess

Bowden

Fall

, et al. Sensitivity analyses for robust causal inference from Mendelian randomization analyses with multiple genetic variants. Epidemiology 2017; 28: 30–42.

22.

Verbanck

Chen

Neale

, et al. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases [published correction appears in Nat Genet. 2018 Aug(8):1196]. Nat Genet 2018; 50: 693–698.

23.

Xiao

Liu

, et al. Causality of genetically determined metabolites on anxiety disorders: a two-sample Mendelian randomization study. J Transl Med 2022; 20: 475.

24.

Huang

Chen

Shen

, et al. Klotho antagonizes pulmonary fibrosis through suppressing pulmonary fibroblasts activation, migration, and extracellular matrix production: a therapeutic implication for idiopathic pulmonary fibrosis. Aging (Albany NY) 2020; 12: 5812–5831.

25.

Raghu

Remy-Jardin

Myers

; American Thoracic Society, European Respiratory Society, Japanese Respiratory Society, and Latin American Thoracic Societyet al. Diagnosis of idiopathic pulmonary fibrosis. An official ATS/ERS/JRS/ALAT clinical practice guideline. Am J Respir Crit Care Med 2018; 198: e44–e68.

26.

Guler

Ryerson

CJ.

Frailty in patients with interstitial lung disease. Curr Opin Pulm Med 2020; 26: 449–456.

27.

Melo-Narváez

Stegmayr

Wagner

, et al. Lung regeneration: implications of the diseased niche and ageing. Eur Respir Rev 2020; 29: 200222.

28.

Tesolato

Vicente-Valor

Jarabo

, et al. Role of telomere length in survival of patients with idiopathic pulmonary fibrosis and other interstitial lung diseases. Biomedicines 2023; 11: 3257.

29.

Parimon

Yao

Stripp

, et al. Alveolar epithelial type II cells as drivers of lung fibrosis in idiopathic pulmonary fibrosis. Int J Mol Sci 2020; 21: 2269.

30.

Yao

Guan

Carraro

, et al. Senescence of alveolar type 2 cells drives progressive pulmonary fibrosis. Am J Respir Crit Care Med 2021; 203: 707–717.

31.

Lehmann

Königshoff

Regenerative medicine and the hope for a cure. Clin Chest Med 2021; 42: 365–373.

32.

Semeraro

Smith

Kaiser

, et al. Physical activity, a modulator of aging through effects on telomere biology. Aging (Albany NY) 2020; 12: 13803–13823.

33.

Lei

Yang

Yuan

, et al. The causal relationship between physical activity, sedentary time and idiopathic pulmonary fibrosis risk: a Mendelian randomization study. Respir Res 2023; 24: 291.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.10 MB