Relationships between 22 unique comorbidities and idiopathic pulmonary fibrosis (IPF) were assessed by a bidirectional Mendelian randomization (MR) approach in a retrospective study. Three of the comorbidities that were examined appeared causally associated with an increased risk of IPF.
Researchers used summary statistics of large-scale genomewide association studies (GWAS) obtained from the IPF Genetics Consortium. For replication, they used data from the Global Biobank Meta-Analysis Initiative.
Pulmonary or extrapulmonary illnesses are regularly observed to be comorbidities associated with IPF. Although randomized control trials can provide strong deductive evidence of causal relationships between diseases, they are also often subject to inherent practical and ethical limitations. MR is an alternative approach that exploits genetic variants of genes with known function as a means to infer a causal effect of a modifiable exposure on disease and minimizes possible confounding issues from unrelated environmental factors and reverse causation. Bidirectional MR extends the exposure-outcome association analysis of MR to both directions, producing a higher level of evidence for causality, Jiahao Zhu, of the Department of Epidemiology and Health Statistics, Hangzhou, China, and colleagues wrote.
Study details
In a study published in the journal Chest, the researchers reported on direction and causal associations between IPF and comorbidities, as determined by bidirectional MR analysis of GWAS summary statistics from five studies included in the IPF Genetics Consortium (4,125 patients and 20,464 control participants). For replication, they extracted IPF GWAS summary statistics from the nine biobanks from the GBMI (6,257 patients and 947,616 control participants). All individuals were of European ancestry.
The 22 comorbidities examined for a relationship to IPF were identified through a combination of a PubMed database literature search limited to English-language articles concerning IPF as either an exposure or an outcome and having an available full GWAS summary statistic. The number of patients in these studies ranged from a minimum of 3,203 for osteoporosis to a maximum of 246,363 for major depressive disorder.
To estimate causal relationships, single-nucleotide polymorphism selection for IPF and each comorbidity genetic instrument were based on a genomewide significance value (P < 5x10–8) and were clumped by linkage disequilibrium (r2 < 0.001 within a 10,000 kB clumping distance). Evidence from analysis associating each comorbidity with IPF was categorized as either convincing, suggestive, or weak. Follow-up studies examined the causal effects of measured lung and thyroid variables on IPF and IPF effects on blood pressure variables.
Convincing evidence
The bidirectional MR and follow-up analysis revealed “convincing evidence” of causal relationships between IPF and 2 of the evaluated 22 comorbidities. A higher risk of IPF was associated with gastroesophageal reflux disease (GERD). Importantly, a multivariable MR analysis conditioning for smoking continued to show the causal linkage between GERD and a higher risk of IPF. In contrast, the genetic liability of COPD appeared to confer a protective role, as indicated by an associated decrease in risk for IPF. The researchers suggest that this negative relationship may be caused by their distinct genetic architecture.
Suggestive evidence
“Suggestive evidence” of underlying relationships between IPF and lung cancer or blood pressure phenotype comorbidities was also found with this study. The MR results give support to existing evidence that IPF has a causal effect for a higher lung cancer risk. In contrast, IPF appeared to have a protective effect on hypertension and BP phenotypes. This contrasted with VTE: Evidence suggestive that genetic liability to hypothyroidism could lead to IPF was also found (International IPF Genetics Consortium, P < .040; MR PRESSO method; and GBMI, P < .002; IVW method).
Limitations
The primary strengths of the study was the ability of MR design to enhance causal inference, particularly when large cohorts for perspective investigations would be inherently difficult to obtain. Several noted limitations include the fact that causal estimates may not be well matched to observational or interventional studies and there was a low number of single-nucleotide polymorphisms available as genetic instruments for some diseases. In addition, there was a potential bias because of some sample overlap among the utilized databases, and it is unknown whether the results are applicable to ethnicities other than those of European ancestry.
Conclusion
Overall, this study showed that a bidirectional MR analysis approach could leverage genetic information from large databases to reveal causative associations between IPF and several different comorbidities. This included an apparent causal relationship of GERD, venous thromboembolism, and hypothyroidism with IPF, according to the researchers. These provide the basis to obtain “a deeper understanding of the pathways underlying these diverse associations” that may have valuable “implications for enhanced prevention and treatment strategies for comorbidities.”
The authors disclosed no relevant financial relationships.
A version of this article originally appeared on Medscape.com.