Skip to main content
Download PDF

An epigenome-wide study of DNA methylation profiles and lung function among American Indians in the Strong Heart Study

Clinical Epigenetics volume 14, Article number: 75 (2022) Cite this article

Abstract

Background

Epigenetic modifications, including DNA methylation (DNAm), are often related to environmental exposures, and are increasingly recognized as key processes in the pathogenesis of chronic lung disease. American Indian communities have a high burden of lung disease compared to the national average. The objective of this study was to investigate the association of DNAm and lung function in the Strong Heart Study (SHS). We conducted a cross-sectional study of American Indian adults, 45–74 years of age who participated in the SHS. DNAm was measured using the Illumina Infinium Human MethylationEPIC platform at baseline (1989–1991). Lung function was measured via spirometry, including forced expiratory volume in 1 s (FEV1) and forced vital capacity (FVC), at visit 2 (1993–1995). Airflow limitation was defined as FEV1 < 70% predicted and FEV1/FVC < 0.7, restriction was defined as FEV1/FVC > 0.7 and FVC < 80% predicted, and normal spirometry was defined as FEV1/FVC > 0.7, FEV1 > 70% predicted, FVC > 80% predicted. We used elastic-net models to select relevant CpGs for lung function and spirometry-defined lung disease. We also conducted bioinformatic analyses to evaluate the biological plausibility of the findings.

Results

Among 1677 participants, 21.2% had spirometry-defined airflow limitation and 13.6% had spirometry-defined restrictive pattern lung function. Elastic-net models selected 1118 Differentially Methylated Positions (DMPs) as predictors of airflow limitation and 1385 for restrictive pattern lung function. A total of 12 DMPs overlapped between airflow limitation and restrictive pattern. EGFR, MAPK1 and PRPF8 genes were the most connected nodes in the protein–protein interaction network. Many of the DMPs targeted genes with biological roles related to lung function such as protein kinases.

Conclusion

We found multiple differentially methylated CpG sites associated with chronic lung disease. These signals could contribute to better understand molecular mechanisms involved in lung disease, as assessed systemically, as well as to identify patterns that could be useful for diagnostic purposes. Further experimental and longitudinal studies are needed to assess whether DNA methylation has a causal role in lung disease.

Introduction

Between 1980 and 2014, the mortality rate for chronic respiratory disease, including chronic obstructive pulmonary disease (COPD) and interstitial lung disease (ILD), increased by 29.7% in the U.S [1]. COPD is defined by airflow limitation that is not fully reversible [2], whereas ILD is defined by the presence of cellular proliferation, infiltration and/or fibrosis of the lung not due to infection or neoplasia and resembles a restrictive spirometry pattern [3]. The development of chronic lung disease is associated with both environmental and genetic risk factors. Although cigarette smoking is one of the main risk factors for chronic lung disease development, not every smoker will develop chronic lung disease and many patients with chronic lung disease have never smoked.

Epigenetic modifications, including DNA methylation (DNAm), are often related to environmental exposures, and are increasingly recognized as key processes in the pathogenesis of chronic lung disease 4,5,6,7,8]. In a systematic review examining the association of lung function with global, epigenome-wide, and locus-specific DNAm in peripheral blood from population-based studies, five of the six included studies showed evidence that DNAm profiles were differentially associated with lung function, including loci associated with the SERPINA1, ORC4, WT1, and FXYD1 genes [9]. SERPINA1, for example, encodes alpha-1-antitrypsin, and alpha-1-antitrypsin deficiency has been shown to cause degenerative pulmonary disease through unregulated tissue breakdown [10]. Evidence suggests that DNAm alterations could play a role in the predisposition to or pathogenetic mechanism of lung disease. While there is a growing number of studies that evaluate the association of lung disease and differential DNAm profiles, epidemiologic studies examining lung disease-related DNAm profiles of American Indian communities are scarce.

The objective of this study was to investigate the association of DNAm with lung function and spirometry-defined lung disease in the Strong Heart Study (SHS). We used elastic-net models to select relevant CpGs, and conducted a bioinformatic analysis to evaluate the biological plausibility of the findings.

Methods

Study population

The SHS is a prospective cohort study funded by the National Heart, Lung and Blood Institute and the National Institute of Environmental Health Sciences to investigate cardiovascular disease and its risk factors in American Indian adults [11]. In 1989–1991, 4549 men and women aged 45–75 years, members of 13 tribes based in Arizona, Oklahoma, and North Dakota and South Dakota who were free of cardiovascular disease enrolled in the study. DNAm was measured in 2351 participants at the SHS baseline visit (1989–1991). Details regarding inclusion criteria for blood DNAm measurements have been described elsewhere [12]. Among eligible participants with DNAm data, participants without a valid spirometry test at visit 2 (1993–1995) were excluded (N = 648), as were individuals missing relevant covariate information, leaving a total of 1677 participants in this study (Fig. 1).

Fig. 1
figure 1

Flowchart of included participants. *5 participants missing education data, 2 smoking status, 11 BMI, 52 LDL cholesterol, 14 hypertension treatment, 111 eGFR, 30 diabetes

Full size image

Participant characteristics

At baseline, trained and certified nurses and medical examiners administered a standardized questionnaire and physical examination including collecting information on sociodemographic (age, sex, study region, education level), lifestyle (smoking status), medical history (prior tuberculosis infection) and anthropometric (height and weight) factors. A fasting blood sample was also collected during the physical examination.

Spirometry and self-reported lung disease

Pre-bronchodilator spirometry testing was conducted by centrally trained and certified nurses and technicians. Maneuvers were considered acceptable according to then-current American Thoracic Society recommendations [13]. Spirometry reference values for SHS participants have been previously derived [14]. Spirometry endpoints include absolute measures of forced expiratory volume in 1 s (FEV1) and forced vital capacity (FVC), FEV1/FVC, and fixed ratio-defined airflow limitation (FEV1/FVC < 0.70) and restriction (FVC < 80% predicted, FEV1/FVC > 0.70). For fixed-ratio defined lung disease endpoints, participants with FEV1/FVC > 0.7 and FVC > 80% predicted served as the reference group.

Blood DNA methylation determinations

Details of microarray DNAm measurements at the baseline visit of the SHS have been published elsewhere [12]. Briefly, DNAm from white blood cells was measured using the Illumina MethylationEPIC BeadChip (850 K). CpGs with a p-detection value greater than 0.01 in more than 5% of the individuals (6159 CpGs) were removed. In addition, cross-hybridizing probes, probes located in sex chromosomes and Single Nucleotide Polymorphisms (SNPs) with minor allele frequency > 0.05 were excluded. Single sample snoob normalization and regression on correlated probes normalization were conducted following Illumina’s recommendations for preprocessing (minfi and Enmix R packages) [15]. Blood cell proportions (CD8T, CD4T, NK cells, B cells, monocytes and neutrophils) were estimated using the FlowSorted.Blood.EPIC R package [16]. Beta values, which range from 0 to 1 and represent the proportion of unconverted cytosines (Cs) in bisulfite-converted DNA at specific locations, were calculated using the R package minfi. [15] We used all cell types except neutrophils (the most common cell type) as adjustment variables in regression models. We detected and corrected for potential batch effects by sample plate, sample row, and DNA isolation time with the combat function (sva R package) [17]. We conducted annotation of CpGs to the nearest gene according to the Infinium MethylationEPIC Manifest File v1.0b4 [18, 19]. CpG sites that were not annotated to any gene according to Illumina’s manifest files were annotated to the closest gene using the matchGenes function from the bumphunter R package. The preprocessing resulted in data from 1677 individuals and 788,368 CpG sites in our analyses.

Statistical analysis

Differentially methylated positions (DMPs) analysis by elastic net

We examined five outcomes: (1) FEV1 (in Liters) as a continuous variable, (2) FVC (in Liters) as a continuous variable, (3) FEV1/FVC (%) as a continuous variable, (4) airflow limitation versus normal lung function as a dichotomous variable, and (5) restrictive versus normal lung function as a dichotomous variable. Given that many smoking-related genes were found to be DMPs for airflow limitation, we repeated the analysis among never smokers, both as self-reported and as identified by the EpiSmokEr tool [20], which predicts smoking status using DNAm data. In contrast to traditional one-by-one linear regression CpG modeling approaches, which are limited in accounting for large numbers of predictors or correlated data, we used elastic-net. Elastic-net methods have recently become very popular in Epigenome-Wide and Genome-Wide Association Studies 21,22,23] as the elastic-net method is robust to limitations of the Lasso method such as dealing with multicollinearity in very high-dimensional settings [24, 25]. Specifically, when the correlations among predictors are high, the elastic-net method exceeds the predictive accuracy of the Lasso [26]. Elastic-net has previously shown to be able to select relevant predictors in differential DNAm analysis and has been used to construct methylation-based risk-scores that have shown great promise for disease prediction based in epigenetic data. [21, 27, 28]

We used elastic-net to select DMPs (simultaneously modeled independent variables) that were associated with lung function and spirometry-defined lung disease (dependent variables). Among the DMPs selected by elastic-net, we then ran traditional linear regression models (for continuous outcomes) and logistic regression models (for dichotomous outcomes) for each CpG separately to obtain effect estimates and 95% CI-s.

Elastic-net, linear and logistic models were adjusted for smoking status (never, former, current), cumulative smoking (cigarette pack-years), age, squared age, sex, BMI, study center (Arizona, Oklahoma or North Dakota and South Dakota), prior tuberculosis diagnosis [29] and cell counts (CD8T, CD4T, NK, B cells and monocytes). For continuous, absolute measures of lung function, we also adjusted for height. To account for population stratification, models were additionally adjusted for five genetic principal components (PCs) [30]. Of 2562 genotyped SHS participants as part of the CALiCo/PAGE Study, we identified 644 unrelated individuals (either founders of pedigrees or unrelated spouses of their descendants). Of 162,718 autosomal SNPs that passed quality control, we selected 15,158 based on the following criteria: minor allele frequency ≥ 0.05, minimum physical separation of 1 kb, and pairwise correlation of genotype scores ≤ 0.1 within a 100 kb sliding window. We performed PC analysis on the genotype scores within unrelated individuals using the R function prcomp. The first five PCs were kept as adjustment variables, as they explained most of the variance. Multiple comparisons were accounted separately using the Benjamini and Hochberg method for false discovery rates (FDR).

To assess whether results were affected by family relatedness, we ran a sensitivity analysis and repeated the linear and logistic models for each of the five lung function measures restricted to unrelated individuals (i.e., selecting only one individual within each family). In this sensitivity analysis, we additionally excluded individuals with mismatches in reported sex vs sex predicted using DNA methylation data as computed by the getSex function from the minfi R package. [15]

Protein–protein interaction network

From the DMPs selected in the elastic-net models, we created two sets of protein-coding genes. The first set represents the airflow limitation phenotype using the following 3 outcomes: FEV1, FEV1/FVC, and airflow limitation vs. normal lung function. The second set represents a restrictive phenotype using the following 3 outcomes: FEV1, FVC, and restrictive vs. normal lung function. The protein interaction information was obtained from the STRING database v11.0 [31]. The STRING database provides a confidence score (from 0 to 1) obtained from the estimated likelihood of each annotated interaction between a given pair of proteins being biologically meaningful, specific and reproducible [31]. The protein interaction networks were analyzed and displayed using the yfiles Organic layout by Cytoscape v. 3.7.2 [32]. In the resultant networks, we only kept connections obtained from experimental studies with a minimum confidence score of 0.4. The unconnected nodes were excluded from the network. We also conducted PPI network enrichment analysis in the resultant networks.

Enrichment analyses

We used the EWAS Toolkit [33] to test for trait enrichment for each of the five lung endpoints. The CpGs selected by the elastic-net models were introduced in the EWAS Toolkit separately for airflow limitation, restrictive pattern, FEV1, FVC and FEV1/FVC. In addition, we used the ToppCluster tool [34] for comparative enrichment among gene clusters selected by elastic-net for the five endpoints. CpG sites were annotated to the closest gene using the matchGenes function from the R package bumphunter, and then were introduced in the Toppcluster tool in five clusters (one per endpoint). Enriched Gene Ontology terms within and between clusters were checked at a Bonferroni-corrected p-value of 0.05. In addition, we introduced the top genes selected by elastic-net as well as the most connected nodes in the protein–protein interaction network into the GWAS Catalog [35] to test whether they had been identified in previous genome-wide association studies.

Results

1677 participants were included (Fig. 1). Participant’s characteristics are presented in Table 1. At baseline (time of blood collection), all participants were 44–75 years of age (mean 55 years old). 61% of participants were female, and 32% had never smoked. The elastic-net model selected 838 DMPs for FEV1, 762 DMPs for FVC, 545 DMPs for FEV1/FVC, 1118 DMPs for airflow limitation and 1385 for restrictive pattern lung function. 328 of the DMPs selected for FEV1 (38.8%) overlapped with selected DMPs for FVC elastic-net models, whereas 26 DMPs (3.1%) overlapped with FEV1/FVC (Fig. 2). Airflow limitation shared 36 DMPs with FEV1 (3.2%), 143 DMPs with FEV1/FVC (12.7%) and 11 DMPs overlapped together with FEV1, FEV1/FVC and airflow limitation (Fig. 2). Restrictive pattern lung function shared 50 DMPs with FEV1 (2.9%), 67 DMPs with FVC (4.8%) and 32 DMPs overlapped between FEV1, FVC and restrictive pattern lung function (Fig. 2), while 12 DMPs overlapped between airflow limitation and restrictive pattern, no DMPs overlapped with all 5 outcomes.

Table 1 Participant characteristics
Full size table
Fig. 2
figure 2

Venn diagram of elastic-net selected DMPs for lung function outcomes, airflow limitation, and restrictive pattern

Full size image

Table 2 shows the top five DMPs selected by the elastic-net models and the mean differences (95% CIs) for continuous lung function measures (FEV1, FVC, FEV1/FVC) comparing percentile 90 to percentile 10 of DNA methylation calculated using linear regression models. Table 3 shows the top five DMPs selected by the elastic-net models and the Odds Ratios (95% CIs) for airflow limitation and restrictive pattern comparing percentile 90 to percentile 10 of DNA methylation calculated using logistic regression models. A list of all DMPs selected by elastic-net models for each of the five lung function outcomes studied are included in Additional file 1: Tables S1–S5. Additional file 2: Fig. S1 shows the distribution of DNA methylation proportions by lung disease status of the top five DMPs for restrictive pattern and the top five DMPs for airflow limitation.

Table 2 Top five differentially methylated positions for continuous lung function measures (FEV1, FVC, and FEV1/FVC)
Full size table
Table 3 Top five differentially methylated positions for airflow limitation and restrictive pattern lung function
Full size table

Of the 1677 individuals included in this study, 1142 were from unique families (i.e., unrelated). When comparing self-reported sex vs sex predicted using DNA methylation, seven additional individuals presented sex mismatch and were excluded in sensitivity analysis, leaving 1135 participants. Among those, 155 had restrictive pattern disease and 262 had airflow limitation. The OR-s and mean differences when excluding sex-mismatched and related participants were very similar to those of the main analysis (Additional file 1: Tables S1–S5).

In the protein–protein interaction networks, the obstructive phenotype network (FEV1, FEV1/FVC and airflow limitation vs normal lung function) included 1965 unique genes associated with 2326 DMPs identified by elastic-net models. Of these, 1467 non-coding RNA genes or unconnected nodes were discarded (Fig. 3, network 1). The protein–protein interaction network for airflow limitation included 498 nodes and 829 interactions (Additional file 3: Fig. S2 and Additional file 1: Tables S6, S7). EGFR, MAPK1 and PRPF8 were the most connected nodes in the network with 32, 22 and 19 interactions, respectively. The restrictive phenotype network (FEV1, FVC, and restrictive pattern vs normal lung function) included 2156 unique genes associated with 2583 DMPs identified by elastic-net models. Of these, 1551 ncRNA genes or unconnected nodes were discarded (Fig. 3, network 2). The protein–protein interaction network for restrictive pattern included 605 nodes and 1101 interactions (Additional file 4: Fig. S3 and Additional file 1: Tables S8, S9). UBA52, CREBBP, SRC and EGFR were the most connected nodes with 38, 34, 29 and 27 interactions, respectively. The PPI network enrichment analysis identified a total of 204 and 355 Gene Ontology (GO) terms significantly enriched (FDR < 0.05) for airflow limitation (network from Additional file 3: Fig. S2, Additional file 1: Table S10) and for restrictive pattern (network from Additional file 4: Fig. S3, Additional file 1: Table S11), respectively.

Fig. 3
figure 3

Workflow for protein–protein interaction networks

Full size image

Figure 4 shows the top enriched traits for each of the five endpoints. Several lung-related traits were enriched: lung function for FEV1/FVC and airflow limitation; smoking, smoking cessation or maternal smoking for all five endpoints, and lung carcinoma for FEV1/FVC and airflow limitation. Additional file 1: Table S12 shows the results from the Toppcluster algorithm. No Gene Ontology terms were commonly enriched across all five endpoints. The Gene Ontology term “animal organ morphogenesis” was enriched for all endpoints except for FEV1/FVC. In addition, the terms “head development", “neuron projection development", “brain development” and “synaptic signaling” were enriched for three endpoints. Other 95 Gene Ontology terms were enriched at Bonferroni 0.05 significance level for two or one lung endpoints. Among the top genes annotated to DMPs identified in our elastic-net models, only the gene EGFR was present in the GWAS Catalog, associated with lung adenocarcinoma.

Fig. 4
figure 4

Top enriched traits for FEV1, FVC, FEV1/FVC, airflow limitation and restrictive pattern from the EWAS Toolkit

Full size image

Given that several smoking-related genes (AHRR, F2RL3, PRSS23, RARA) were found to be DMPs for airflow limitation, we repeated that analysis restricted to self-reported never smokers and restricted to those that were classified as never-smokers by the EpiSmokEr tool. There were N = 531 self-reported never smokers in our study, of which 92 presented airflow limitation. Eight hundred and sixty-nine CpGs were selected by elastic-net as DMPs. The elastic-net model selected two CpGs annotated to AHRR; however, it did not select any CpG annotated to F2RL3, PRSS23 or RARA (Additional file 1: Table S13). There were N = 848 participants classified by EpiSmokEr as never-smokers. Of those, 139 presented airflow limitation. 756 CpGs were selected by elastic-net. No CpGs annotated to any of the smoking-related genes were selected (Additional file 1: Table S14). The number of overlapping CpGs between the two never-smoker models was 90. The number of overlapping CpGs between the overall population model and the model restricted to never-smokers as classified by EpiSmokEr was 136.

For restrictive pattern, the model was only run for never-smokers as classified by the EpiSmokEr tool, due to lack of power for running it in self-reported never-smokers. There were 121 restrictive pattern cases. 1070 CpGs were selected by elastic-net. The number of overlapping CpGs between the overall population model and the model restricted to never-smokers was 253 (data not shown). Two of the top CpG sites in the overall population model (annotated to genes ADARB2 and ZNF540) were also selected for the model restricted to never-smokers.

Table 4 shows the effect estimates and p-values of the CpGs identified in a meta-analysis [36] that were replicated in the SHS (annotated to genes AHRR, F2RL3, ALPPL2, IER3, GPR15, SOCS3, TMEM184B and CDKN1B).

Table 4 Effect estimates and p values of the CpGs identified in the meta-analysis by Imboden et al. that were replicated in the SHS
Full size table

Discussion

We conducted an epigenome-wide association study investigating the association between DNA methylation and lung function and explored common epigenetic signatures between lung function and disease. Using robust methods for high-dimensional correlated data, we found 1118 DMPs associated with airflow limitation and 1385 associated with restrictive pattern lung function. A total of 12 DMPs overlapped between airflow limitation and restrictive pattern. The biological functions of the top genes, as well as the most connected nodes in the protein–protein interaction network, were related to biological processes associated with lung disease.

Several top genes and highly connected nodes in the protein–protein interaction networks for FEV1 (PIM1), FVC (CDK5), FEV1/FVC (NTRK2), airflow limitation (CPPED1) and restrictive pattern (EGFR, MAPK1) are protein kinases. In addition, the GO term “Positive regulation of transmembrane receptor protein serine/threonine kinase signaling pathway” (GO:0090100) was found to be significantly enriched (FDR = 0.0119) for the restrictive lung function phenotype. Protein kinases play a role in many key pulmonary cellular responses, including mediating inflammatory signals and airway remodeling. Thus, they have been proposed as therapeutic targets for several lung diseases such as chronic obstructive pulmonary disease and asthma [37, 38]. PIM1 and CKD5, top genes associated with FEV1 and FVC, respectively, are serine/threonine protein kinases. An animal study reported evidence of high-tidal volume ventilation increasing pulmonary fibrosis in acute lung injury via the serine/threonine protein kinase B [39]. Also, the mitogen-activated protein kinase 1 gene (MAPK1) was a highly connected node in the airflow limitation protein–protein interaction network. Lung endothelial barrier function is regulated by multiple signaling pathways, including mitogen-activated protein kinases (MAPK) [40]. MAPK kinases might contribute to ameliorate the lung endothelial barrier-disruptive effects. [41]

The EGFR gene, which was among the most connected nodes in the protein–protein interaction networks for both airflow limitation and restrictive pattern, is a protein kinase. It plays an essential role in pulmonary physiology by regulating key cellular processes such as self-renew, wound-healing, proliferation, survival adhesion, migration and differentiation [42]. EGFR inhibitors have been widely used in treatment of non-small cell lung cancer, in fact, it was the first biomarker identified as a potential therapeutic target for personalized treatment in lung cancer. EGFR was also identified as associated with lung adenocarcinoma in the GWAS Catalog [35]. DNA methylation in EGFR has been proposed as a predictive biomarker for lung adenocarcinoma. Our results suggest that DNA methylation levels in EGFR might be predictive of airflow limitation and restrictive pattern as well. Prospective studies of DNA methylation changes and lung disease are needed to assess the potential predictive ability of EGFR in lung disease.

On the other hand, the CRISPLD2 gene, which was a DMP associated with both FEV1 and FVC in our study, was identified in a whole genome sequencing study in children with asthma as associated with FEV1/FVC [43]. Many experimental and human studies have highlighted the importance of the CRISPLD2 gene in fetal lung regulation, branching morphogenesis and alveologenesis, among other lung function related biological processes 44,45,46]. Other genes identified in our study are also associated with biological processes relevant for lung function. The Ubiquitin A-52 Residue Ribosomal Protein Fusion Product 1 gene (UBA52), for instance, was a highly connected node in the restrictive pattern protein–protein interaction network. Ubiquitination regulates the proteins that modulate the alveolocapillary barrier and the inflammatory response, therefore playing an important role in acute lung injury [47]. Also, the Adenosine Deaminase RNA Specific B2 gene (ADARB2) was the top differentially methylated position for restrictive pattern. An animal model showed that adenosine deaminase deficiency might lead to pulmonary fibrosis [48]. Our work provides further evidence that these biological processes are involved in lung disease and can be measured systemically. However, experimental studies are needed to disentangle whether DNAm changes influence these biological pathways or, conversely, alterations in these pathways lead to DNAm dysregulations.

Of note, there was very little overlap between the DMPs associated with airflow limitation and with restrictive pattern (only 12 CpGs), and there was no overlap between the five lung function measures. This fact as well as the fact that the top DMPs associated with restrictive pattern and with airflow limitation (Table 3) had opposite directions of association with DNA methylation, might point to different biological pathways being involved in airflow limitation and restrictive pattern. Importantly, hypomethylation of several smoking-related genes was associated with airflow limitation in our study (AHRR, F2RL3, PRSS23, RARA), whereas none of those was associated with restriction. Previous literature have pointed out that hypomethylation in the gene AHRR, the most well known smoking-related gene, might be associated with lower lung function and respiratory symptoms [49, 50]. Also, DNAm dysregulation in AHRR was associated with lung function in two multi-cohort epigenome-wide association studies in adults [36, 51]. When running the airflow limitation analysis only among self-reported never smokers, elastic-net selected CpGs annotated to the AHRR gene, but not to the other three smoking-related genes. However, when running the airflow limitation analysis among never-smokers as classified by the EpiSmokEr tool, no smoking-related genes were selected. Second-hand smoke exposure might be responsible for the effect of AHRR in airflow limitation among self-reported never smokers. Further studies are needed to investigate the role of the AHRR gene in lung disease beyond smoking.

This is, to our knowledge, the first epigenome-wide association study with the main focus on lung function conducted in a population of American Indians. We found four previous epigenome-wide association studies of lung function with spirometry measurements in other adult populations. One did not report any significant associations (N = 1091) [52]. The second one was conducted in a population of female twins in 2012, and only found one DMR associated with FEV1 and FVC annotated to the gene WT1, which was not replicated in our population [53]. In 2018, another EWAS of lung function was conducted in two cohorts. Three CpG sites associated with lung function were consistently found in the two cohorts [51]. Only one (cg05575921, annotated to AHRR) was replicated in our study. Last, another EWAS was conducted in eight cohorts (three discovery cohorts and five replication cohorts) in 2019 [36], our results are highly consistent with the findings of this recent EWAS, with nine CpGs associated with lung function being replicated in our population (Table 4), and many more at the gene level. Although several other epigenome-wide association studies in lung function have been conducted, they were conducted in specific populations such as children [54], individuals with chronic obstructive pulmonary disease [55], individuals with HIV [56] or never smokers [57]. Findings for these specific populations might not be generalizable. Nevertheless, many differentially methylated positions found in these studies overlapped with our findings. For instance, eight of the top sites found in the never-smokers EWAS of lung function were replicated in our population, which might indicate that the epigenomic signature of lung function is also stable across different population groups. A meta-analysis was conducted among Mexican American and Puerto Rican children [58]. Among the genes identified, only the gene TBC1D16 was replicated in our population as associated with restrictive pattern. In addition, another meta-analysis conducted by Machin et al. did not find any consistent CpG sites across the six articles assessed for either chronic obstructive pulmonary disease or lung function, which suggests that part of the epigenomic signature of lung function might also be specific to populations. [9]

This work has some limitations. First, only 1677 of the SHS participants were included, which might induce some bias among those who were excluded due to not meeting spirometry quality standards. We were also unable to use the lower limit of normal to classify airflow limitation and restriction due to sample size contraints. In addition, we only have one measure of spirometry and we lack other clinical information, therefore, we cannot discard potential measurement errors. DNA methylation is highly cell-type specific and results from blood cells might not be comparable to DNA methylation in other tissues such as lung. However, the biological plausibility of the findings suggests that blood DNA methylation might be relevant for chronic lung disease. Longitudinal and experimental studies are needed to assess the directionality of the findings. Strengths of this work include using one of the largest methylation arrays currently available in microarray technology (the EPIC array), the large sample size in an indigenous population, measurement of spirometry-defined lung disease, and innovative statistical methods that allow evaluating the effect of methylation sites jointly instead of individually.

In conclusion, we found several differentially methylated positions for FEV1, FVC, FEV1/FVC, obstructive pattern and restrictive pattern, with several genes pointing to biological pathways related to lung disease including protein kinases, which are therapeutic targets for lung disease. Further studies are needed to investigate the potential mechanistic role of DNAm in lung disease.

Availability of data and materials

The data underlying this article cannot be shared publicly in an unrestricted manner due to limitations in the consent forms and in the agreements between the Strong Heart Study tribal communities and the Strong Heart Study investigators. The data can be shared to external investigators following the procedures established by the Strong Heart Study, available at https://strongheartstudy.org/. All analyses were conducted in R version 3.6.2, and all packages used are freely available in the CRAN repository.

References

  1. Dwyer-Lindgren L, Bertozzi-Villa A, Stubbs RW, et al. Trends and patterns of differences in chronic respiratory disease mortality among US counties, 1980–2014. JAMA. 2017;318(12):1136. https://0-doi-org.brum.beds.ac.uk/10.1001/jama.2017.11747.

    Article  PubMed  PubMed Central  Google Scholar 

  2. Vestbo J, Hurd SS, Agustí AG, et al. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2013;187(4):347–65. https://0-doi-org.brum.beds.ac.uk/10.1164/rccm.201204-0596PP.

    Article  CAS  PubMed  Google Scholar 

  3. Rosas IO, Dellaripa PF, Lederer DJ, Khanna D, Young LR, Martinez FJ. Interstitial lung disease: NHLBI workshop on the primary prevention of chronic lung diseases. Ann Am Thorac Soc. 2014;11(Supplement 3):S169–77. https://0-doi-org.brum.beds.ac.uk/10.1513/AnnalsATS.201312-429LD.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Yang IV, Schwartz DA. Epigenetic control of gene expression in the lung. Am J Respir Crit Care Med. 2011;183(10):1295–301. https://0-doi-org.brum.beds.ac.uk/10.1164/rccm.201010-1579PP.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Mortaz E, Masjedi MR, Barnes PJ, Adcock IM. Epigenetics and chromatin remodeling play a role in lung disease. 2011;10(4):7–16.

  6. Helling BA, Yang IV. Epigenetics in lung fibrosis: from pathobiology to treatment perspective. Curr Opin Pulm Med. 2015;21(5):454–62. https://0-doi-org.brum.beds.ac.uk/10.1097/MCP.0000000000000191.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Adcock IM, Tsaprouni L, Bhavsar P, Ito K. Epigenetic regulation of airway inflammation. Curr Opin Immunol. 2007;19(6):694–700. https://0-doi-org.brum.beds.ac.uk/10.1016/j.coi.2007.07.016.

    Article  CAS  PubMed  Google Scholar 

  8. Berndt A, Leme AS, Shapiro SD. Emerging genetics of COPD. EMBO Mol Med. 2012;4(11):1144–55. https://0-doi-org.brum.beds.ac.uk/10.1002/emmm.201100627.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Machin M, Amaral AFS, Wielscher M, et al. Systematic review of lung function and COPD with peripheral blood DNA methylation in population based studies. BMC Pulm Med. 2017;17(1):54. https://0-doi-org.brum.beds.ac.uk/10.1186/s12890-017-0397-3.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Laurell C-B, Eriksson S. The electrophoretic α 1-globulin pattern of serum in α 1 -antitrypsin deficiency. Scand J Clin Lab Invest. 1963;15:132–40. https://0-doi-org.brum.beds.ac.uk/10.3109/15412555.2013.771956.

    Article  CAS  Google Scholar 

  11. Lee ET, Welty TK, Fabsitz R, et al. The strong heart study a study of cardiovascular disease in American Indians: design and methods. Am J Epidemiol. 1990;132(6):1141–55. https://0-doi-org.brum.beds.ac.uk/10.1093/oxfordjournals.aje.a115757.

    Article  CAS  PubMed  Google Scholar 

  12. Domingo-Relloso A, Riffo-Campos AL, Haack K, et al. Cadmium, smoking, and human blood DNA methylation profiles in adults from the Strong Heart Study. Environ Health Perspect. 2020;128(6): 067005. https://0-doi-org.brum.beds.ac.uk/10.1289/EHP6345.

    Article  CAS  PubMed Central  Google Scholar 

  13. ATS. Standardization of Spirometry, 1994 Update. American Thoracic Society. Am J Respir Crit Care Med. 1995;152(3):1107–36. https://0-doi-org.brum.beds.ac.uk/10.1164/ajrccm.152.3.7663792.

  14. Marion MS, Leonardson GR, Rhoades ER, Welty TK, Enright PL. Spirometry reference values for American Indian adults: results from the Strong Heart Study. Chest. 2001;120(2):489–95. https://0-doi-org.brum.beds.ac.uk/10.1378/chest.120.2.489.

    Article  CAS  PubMed  Google Scholar 

  15. Aryee MJ, Jaffe AE, Corrada-Bravo H, et al. Minfi: a flexible and comprehensive bioconductor package for the analysis of Infinium DNA methylation microarrays. Bioinformatics. 2014;30(10):1363–9. https://0-doi-org.brum.beds.ac.uk/10.1093/bioinformatics/btu049.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Salas LA, Koestler DC, Butler RA, et al. An optimized library for reference-based deconvolution of whole-blood biospecimens assayed using the Illumina HumanMethylationEPIC BeadArray. Genome Biol. 2018;19(1):64. https://0-doi-org.brum.beds.ac.uk/10.1186/s13059-018-1448-7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Leek JT, Johnson WE, Parker HS, Jaffe AE, Storey JD. The sva package for removing batch effects and other unwanted variation in high-throughput experiments. Bioinformatics. 2012;28(6):882–3. https://0-doi-org.brum.beds.ac.uk/10.1093/bioinformatics/bts034.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Illumina Inc. Infinium MethylationEPIC Product Files. Available at https://emea.support.illumina.com/downloads/infinium-methylationepic-v1-0-product-files.html. Accessed 6 June 2022.

  19. Fortin J-P, Triche TJ, Hansen KD. Preprocessing, normalization and integration of the Illumina HumanMethylationEPIC array with minfi. Bioinformatics. 2016;33(4):btw691. https://0-doi-org.brum.beds.ac.uk/10.1093/bioinformatics/btw691.

    Article  CAS  Google Scholar 

  20. Bollepalli S, Korhonen T, Kaprio J, Anders S, Ollikainen M. EpiSmokEr: a robust classifier to determine smoking status from DNA methylation data. Epigenomics. 2019;11(13):1469–86. https://doi.org/10.2217/epi-2019-0206.

    Article  CAS  PubMed  Google Scholar 

  21. Benton MC, Sutherland HG, Macartney-Coxson D, Haupt LM, Lea RA, Griffiths LR. Methylome-wide association study of whole blood DNA in the Norfolk Island isolate identifies robust loci associated with age. Aging (Albany NY). 2017;9(3):753–68. https://0-doi-org.brum.beds.ac.uk/10.18632/aging.101187.

    Article  CAS  Google Scholar 

  22. Abraham G, Kowalczyk A, Zobel J, Inouye M. SparSNP: fast and memory-efficient analysis of all SNPs for phenotype prediction. BMC Bioinform 2012;13:88. https://0-doi-org.brum.beds.ac.uk/10.1186/1471-2105-13-88.

    Article  Google Scholar 

  23. Waldmann P, Mészáros G, Gredler B, Fuerst C, Sölkner J. Evaluation of the lasso and the elastic net in genome-wide association studies. Front Genet. 2013. https://0-doi-org.brum.beds.ac.uk/10.3389/fgene.2013.00270.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Fan J, Lv J. Sure independence screening for ultrahigh dimensional feature space. J R Stat Soc Ser B Stat Methodol. 2008;70(5):849–911. https://0-doi-org.brum.beds.ac.uk/10.1111/j.1467-9868.2008.00674.x.

    Article  Google Scholar 

  25. Hasinur M, Khan R, Ewart J, Shaw H. Variable selection for survival data with a class of adaptive elastic net techniques. Stat Comput. 2016;26(3):725–41. https://0-doi-org.brum.beds.ac.uk/10.1007/s11222-015-9555-8.

    Article  Google Scholar 

  26. Zou H, Hao A, Zhang H. On the adaptive elastic-net with a diverging number of parameters. Ann Stat. 2009;37(4):1733–51. https://0-doi-org.brum.beds.ac.uk/10.1214/08-AOS625.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Liu J, Liang G, Siegmund KD, Lewinger JP. Data integration by multi-tuning parameter elastic net regression. BMC Bioinform. 2018;19:369. https://0-doi-org.brum.beds.ac.uk/10.1186/s12859-018-2401-1.

    Article  CAS  Google Scholar 

  28. Zhang Q, Vallerga CL, Walker RM, et al. Improved prediction of chronological age from DNA methylation limits it as a biomarker of ageing. bioRxiv. 2018. https://0-doi-org.brum.beds.ac.uk/10.1101/327890.

    Article  Google Scholar 

  29. Powers M, Sanchez TRTR, Welty TKTK, et al. Lung function and respiratory symptoms after tuberculosis in an American Indian population. The Strong Heart Study. Ann Am Thorac Soc. 2020;17(1):38–48. https://0-doi-org.brum.beds.ac.uk/10.1513/AnnalsATS.201904-281OC.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Barfield RT, Almli LM, Kilaru V, et al. Accounting for population stratification in DNA methylation studies. Genet Epidemiol. 2014;38(3):231. https://0-doi-org.brum.beds.ac.uk/10.1002/GEPI.21789.

    Article  PubMed  PubMed Central  Google Scholar 

  31. Szklarczyk D, Gable AL, Lyon D, et al. STRING v11: protein–protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res. 2019;47(D1):D607–13. https://0-doi-org.brum.beds.ac.uk/10.1093/nar/gky1131.

    Article  CAS  PubMed  Google Scholar 

  32. Shannon P, Markiel A, Ozier O, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13(11):2498–504. https://0-doi-org.brum.beds.ac.uk/10.1101/gr.1239303.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Xiong Z, Yang F, Li M, et al. EWAS Open Platform: integrated data, knowledge and toolkit for epigenome-wide association study. Nucleic Acids Res. 2022;50(D1):D1004–9. https://0-doi-org.brum.beds.ac.uk/10.1093/nar/gkab972.

    Article  CAS  PubMed  Google Scholar 

  34. Kaimal V, Bardes EE, Tabar SC, Jegga AG, Aronow BJ. ToppCluster: a multiple gene list feature analyzer for comparative enrichment clustering and network-based dissection of biological systems. Nucleic Acids Res. 2010;38(Web Server issue):W96-102. https://0-doi-org.brum.beds.ac.uk/10.1093/nar/gkq418.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Buniello A, MacArthur JAL, Cerezo M, et al. The NHGRI-EBI GWAS Catalog of published genome-wide association studies, targeted arrays and summary statistics 2019. Nucleic Acids Res. 2019;47(D1):D1005–12. https://0-doi-org.brum.beds.ac.uk/10.1093/nar/gky1120.

    Article  CAS  PubMed  Google Scholar 

  36. Imboden M, Wielscher M, Rezwan FI, et al. Epigenome-wide association study of lung function level and its change. Eur Respir J. 2019;54:1900457 https://0-doi-org.brum.beds.ac.uk/10.1183/13993003.00457-2019.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Defnet AE, Hasday JD, Shapiro P. Kinase inhibitors in the treatment of obstructive pulmonary diseases. Curr Opin Pharmacol. 2020;51:11–8. https://0-doi-org.brum.beds.ac.uk/10.1016/j.coph.2020.03.005.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  38. Dempsey EC, Cool CD, Littler CM. Lung disease and PKCs. Pharmacol Res. 2007;55(6):545–59. https://0-doi-org.brum.beds.ac.uk/10.1016/j.phrs.2007.04.010.

    Article  CAS  PubMed  Google Scholar 

  39. Li L-F, Liao S-K, Huang C-C, Hung M-J, Quinn DA. Serine/threonine kinase-protein kinase B and extracellular signal-regulated kinase regulate ventilator-induced pulmonary fibrosis after bleomycin-induced acute lung injury: a prospective, controlled animal experiment. Crit Care. 2008;12(4):R103. https://0-doi-org.brum.beds.ac.uk/10.1186/cc6983.

    Article  PubMed  PubMed Central  Google Scholar 

  40. Birukova AA, Birukov KG, Gorshkov B, Liu F, Garcia JGN, Verin AD. MAP kinases in lung endothelial permeability induced by microtubule disassembly. Am J Physiol Lung Cell Mol Physiol. 2005;289(1):L75-84. https://0-doi-org.brum.beds.ac.uk/10.1152/ajplung.00447.2004.

    Article  CAS  PubMed  Google Scholar 

  41. Schweitzer KS, Hatoum H, Brown MB, et al. Mechanisms of lung endothelial barrier disruption induced by cigarette smoke: role of oxidative stress and ceramides. Am J Physiol Lung Cell Mol Physiol. 2011;301(6):L836–46. https://0-doi-org.brum.beds.ac.uk/10.1152/ajplung.00385.2010.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Li C, Wei R, Jones-Hall YL, Vittal R, Zhang M, Liu W. Epidermal growth factor receptor (EGFR) pathway genes and interstitial lung disease: an association study. Sci Rep. 2014;4:4893. https://0-doi-org.brum.beds.ac.uk/10.1038/srep04893.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Kachroo P, Hecker J, Chawes BL, et al. Whole genome sequencing identifies CRISPLD2 as a lung function gene in children with asthma. Chest. 2019;156(6):1068–79. https://0-doi-org.brum.beds.ac.uk/10.1016/j.chest.2019.08.2202.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Oyewumi L, Kaplan F, Gagnon S, Sweezey NB. Antisense oligodeoxynucleotides decrease LGL1 mRNA and protein levels and inhibit branching morphogenesis in fetal rat lung. Am J Respir Cell Mol Biol. 2003;28(2):232–40. https://0-doi-org.brum.beds.ac.uk/10.1165/rcmb.4877.

    Article  CAS  PubMed  Google Scholar 

  45. Lan J, Ribeiro L, Mandeville I, et al. Inflammatory cytokines, goblet cell hyperplasia and altered lung mechanics in Lgl1+/− mice. Respir Res. 2009;10:83. https://0-doi-org.brum.beds.ac.uk/10.1186/1465-9921-10-83.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Nadeau K, Montermini L, Mandeville I, et al. Modulation of Lgl1 by steroid, retinoic acid, and vitamin D models complex transcriptional regulation during alveolarization. Pediatr Res. 2010;67(4):375–81. https://0-doi-org.brum.beds.ac.uk/10.1203/PDR.0b013e3181d23656.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  47. Magnani ND, Dada LA, Sznajder JI. Ubiquitin-proteasome signaling in lung injury. Transl Res. 2018;198:29–39. https://0-doi-org.brum.beds.ac.uk/10.1016/j.trsl.2018.04.003.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Chunn JL, Molina JG, Mi T, Xia Y, Kellems RE, Blackburn MR. Adenosine-dependent pulmonary fibrosis in adenosine deaminase-deficient mice. J Immunol. 2005;175(3):1937–46. https://0-doi-org.brum.beds.ac.uk/10.4049/jimmunol.175.3.1937.

    Article  CAS  PubMed  Google Scholar 

  49. Kodal JB, Kobylecki CJ, Vedel-Krogh S, Nordestgaard BG, Bojesen SE. AHRR hypomethylation, lung function, lung function decline and respiratory symptoms. Eur Respir J. 2018;51(3):1701512. https://0-doi-org.brum.beds.ac.uk/10.1183/13993003.01512-2017.

    Article  CAS  PubMed  Google Scholar 

  50. Bojesen SE, Timpson N, Relton C, Davey Smith G, Nordestgaard BG. AHRR (cg05575921) hypomethylation marks smoking behaviour, morbidity and mortality. Thorax. 2017;72(7):646–53. https://0-doi-org.brum.beds.ac.uk/10.1136/thoraxjnl-2016-208789.

    Article  PubMed  Google Scholar 

  51. Carmona JJ, Barfield RT, Panni T, et al. Metastable DNA methylation sites associated with longitudinal lung function decline and aging in humans: an epigenome-wide study in the NAS and KORA cohorts. Epigenetics. 2018;13(10–11):1039–55. https://0-doi-org.brum.beds.ac.uk/10.1080/15592294.2018.1529849.

    Article  PubMed  PubMed Central  Google Scholar 

  52. Marioni RE, Shah S, McRae AF, et al. The epigenetic clock is correlated with physical and cognitive fitness in the Lothian Birth Cohort 1936. Int J Epidemiol. 2015;44(4):1388–96. https://0-doi-org.brum.beds.ac.uk/10.1093/ije/dyu277.

    Article  PubMed  PubMed Central  Google Scholar 

  53. Bell JT, Tsai P-C, Yang T-P, et al. Epigenome-Wide Scans Identify Differentially Methylated Regions for Age and Age-Related Phenotypes in a Healthy Ageing Population. PLoS Genet. 2012;8(4):e1002629. https://0-doi-org.brum.beds.ac.uk/10.1371/journal.pgen.1002629.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Mukherjee N, Arathimos R, Chen S, et al. DNA methylation at birth is associated with lung function development until age 26 years. Eur Respir J. 2021;57(4):2003505. https://0-doi-org.brum.beds.ac.uk/10.1183/13993003.03505-2020.

    Article  CAS  PubMed  Google Scholar 

  55. Casas-Recasens S, Noell G, Mendoza N, et al. Lung DNA methylation in chronic obstructive pulmonary disease: relationship with smoking status and airflow limitation severity. Am J Respir Crit Care Med. 2021;203(1):129–34. https://0-doi-org.brum.beds.ac.uk/10.1164/rccm.201912-2420LE.

    Article  PubMed  Google Scholar 

  56. Hernandez Cordero AI, Yang CX, Obeidat M, et al. DNA methylation is associated with airflow obstruction in patients living with HIV. Thorax. 2021;76(5):448–55. https://0-doi-org.brum.beds.ac.uk/10.1136/thoraxjnl-2020-215866.

    Article  PubMed  Google Scholar 

  57. de Vries M, Nedeljkovic I, van der Plaat DA, et al. DNA methylation is associated with lung function in never smokers. Respir Res. 2019;20(1):268. https://0-doi-org.brum.beds.ac.uk/10.1186/s12931-019-1222-8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  58. Herrera-Luis E, Li A, Mak ACY, et al. Epigenome-wide association study of lung function in Latino children and youth with asthma. Clin Epigenetics. 2022;14(1):9. https://0-doi-org.brum.beds.ac.uk/10.1186/s13148-022-01227-5.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We thank the dedication of the Strong Heart Study participants, investigators and staff, without whom this work would not have been possible.

Funding

This work was supported by Grants from the National Heart, Lung, and Blood Institute (NHLBI) (Contract Numbers 75N92019D00027, 75N92019D00028, 75N92019D00029 and 75N92019D00030) and previous Grants (R01HL090863, R01HL109315, R01HL109301, R01HL109284, R01HL109282, and R01HL109319 and cooperative agreements: U01HL41642, U01HL41652, U01HL41654, U01HL65520 and U01HL65521); by the National Institute of Environmental Health Sciences (Grant Numbers R01ES021367, R01ES025216, P42ES033719, P30ES009089), by ANID—Millennium Science Initiative Program—No NCS2021_013—SocioMed (ALRC), by the Maria Zambrano grant Nº ZA21-063 for the requalification of the Spanish university system—NextGeneration EU (ALRC) and by a fellowship from “la Caixa” Foundation (ID 100010434) (fellowship code “LCF/BQ/DR19/11740016”) (ADR). The funders had no role in the planning, conducting, analysis, interpretation, or writing of this study. The content of this manuscript is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health (United States) or the National Health Institute Carlos III (Spain).

Author information

Authors and Affiliations

  1. Integrative Epidemiology Group, Department of Chronic Diseases Epidemiology, National Center for Epidemiology, Carlos III Health Institute, 28029, Madrid, Spain

    Arce Domingo-Relloso & Maria Tellez-Plaza

  2. Department of Environmental Health Sciences, Columbia University Mailman School of Public Health, New York, USA

    Arce Domingo-Relloso, Ana Navas-Acien & Tiffany R. Sanchez

  3. Department of Statistics and Operations Research, University of Valencia, Valencia, Spain

    Arce Domingo-Relloso

  4. Millennium Nucleus on Sociomedicine (SocioMed) and Vicerrectoría Académica, Universidad de La Frontera, Temuco, Chile

    Angela L. Riffo-Campos

  5. Department of Computer Science, ETSE, University of Valencia, Valencia, Spain

    Angela L. Riffo-Campos

  6. United States Environmental Protection Agency, Washington, DC, USA

    Martha Powers

  7. Population Health Program, Texas Biomedical Research Institute, San Antonio, TX, USA

    Karin Haack & Shelley A. Cole

  8. Department of Environmental Health and Engineering, Johns Hopkins Bloomberg School of Public Health, Baltimore, USA

    Robert H. Brown

  9. MedStar Health Research Institute, Hyattsville, MD, USA

    Jason G. Umans

  10. Georgetown-Howard Universities Center for Clinical and Translational Science, Washington, DC, USA

    Jason G. Umans

  11. Departments of Mental Health and Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, USA

    M. Daniele Fallin

Authors
  1. Arce Domingo-Relloso
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Angela L. Riffo-Campos
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Martha Powers
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Maria Tellez-Plaza
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Karin Haack
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Robert H. Brown
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Jason G. Umans
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. M. Daniele Fallin
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Shelley A. Cole
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Ana Navas-Acien
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Tiffany R. Sanchez
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

ADR conducted data analysis, planning and writing of the article. ALRC conducted data analysis. KH and SC obtained and processed the blood DNA methylation data in the Strong Heart Study. MP, MTP, RHB, JU and MDF helped with the planning of the study and provided feedback, suggestions and careful review. ANA and TRS were responsible for the design and planning of the study and conducted writing and review. All authors have read and approved the final version of the manuscript.

Corresponding author

Correspondence to Arce Domingo-Relloso.

Ethics declarations

Ethics approval and consent to participate

This study was approved by Institution Review Boards of the academic organizations, tribal communities, and the Indian Health Service for the Strong Heart Study.

Consent for publication

Informed consent from participants was obtained for the Strong Heart Study.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1. Table S1

. DMPs selected by elastic-net for FEV1. Table S2. DMPs selected by elastic-net for FVC. Table S3. DMPs selected by elastic-net for FEV1/FVC. Table S4. DMPs selected by elastic-net for airflow limitation. Table S5. DMPs selected by elastic-net for restrictive pattern. Table S6. Protein-protein interaction network nodes for airflow limitation. Table S7. Protein-protein interaction network edges for airflow limitation. Table S8. Protein-protein interaction network nodes for restrictive pattern. Table S9. Protein-protein interaction network edges for restrictive pattern. Table S10. Protein-protein interaction network enrichment analysis for airflow limitation. Table S11. Protein-protein interaction network enrichment analysis for restrictive pattern. Table S12. Toppcluster enrichment for FEV1, FVC, FEV1/FVC, airflow limitation and restrictive pattern. Table S13. DMPs selected by elastic-net for airflow limitation restricted to self-reported never smokers. Table S14. DMPs selected by elastic-net for airflow limitation restricted to never smokers as reported by EpiSmokEr.

Additional file 2. Figure S1.

Distribution of DNA methylation proportions by lung disease status of the top five differentially methylated positions for restrictive pattern and the top five DMPs for airflow limitation.

Additional file 3. Figure S2.

Protein-protein interaction network for airflow limitation phenotype: FEV1, FEV1/FVC and airflow limitation vs normal lung function.

Additional file 4. Figure S3.

Protein-protein interaction networks for restrictive lung function phenotype: FEV1, FVC and restrictive vs normal lung function.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Domingo-Relloso, A., Riffo-Campos, A.L., Powers, M. et al. An epigenome-wide study of DNA methylation profiles and lung function among American Indians in the Strong Heart Study. Clin Epigenet 14, 75 (2022). https://0-doi-org.brum.beds.ac.uk/10.1186/s13148-022-01294-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://0-doi-org.brum.beds.ac.uk/10.1186/s13148-022-01294-8

Keywords

Clinical Epigenetics

ISSN: 1868-7083

Contact us