Skip to main content
NCBI home page
Allergy, Asthma & Immunology Research logoLink to Allergy, Asthma & Immunology Research
. 2026 Feb 19;18(2):224–241. doi: 10.4168/aair.2026.18.2.224

A Risk Factor Atlas for Allergy-Related Airway Diseases: Evidence From Multi-Biobank Genetic Study

Tao Guo 1,2, Peiyu Luo 1, Guobing Jia 1, Dehong Liu 1, Hui Xie 2,✉
PMCID: PMC13047439  PMID: 41914530

Abstract

Purpose

Understanding factors contributing to allergy-related airway diseases (AADs).

Methods

We conducted a meta-analysis of Mendelian randomization (MR) estimates from multi-biobank to investigate the effects of 83 common factors on asthma, allergic rhinitis (AR), chronic rhinosinusitis (CRS), and nasal polyps (NPs). Data for AADs were obtained from multiple independent populations including FinnGen, UK Biobank and large consortia (asthma, 137,498 cases and 1,009,466 controls; AR, 51,712 cases and 592,784 controls; CRS, 24,688 cases and 760,073 controls; and NP, 11,796 cases and 760,075 controls). The meta-analyses combining the above multi-biobank MR results were employed as our main results.

Results

Genetically proxied social isolation, smoking time and frequency, sedentary time, global and central obesity, C-reactive protein (CRP), insomnia, major depressive disorder (MDD), anxiety, attention deficit hyperactivity disorder, neuroticism, posttraumatic stress disorder (PTSD), rheumatoid arthritis (RA), atopic dermatitis (AD), type 1 diabetes (T1D), gastroesophageal reflux disease (GERD), childhood obesity, early menarche, smoking around birth, and childhood maltreatment could increase the risk of asthma. Education, higher income, cheese intake, sleep duration, high density lipoprotein cholesterol, forced expiratory volume in 1 second/forced vital capacity, well-being and ulcerative colitis were associated with a decreased risk of asthma. Social isolation, CRP, Crohn’s disease, GERD, insomnia, MDD, neuroticism and AD could increase the risk of AR, whereas well-being could decrease the risk of AR. Sedentary time, MDD, anxiety, PTSD, opioid use disorder, RA, AD, T1D, hypothyroidism, and GERD could increase the risk of CRS, whereas coffee intake, apolipoprotein A1, and well-being were associated with a decreased risk of CRS. Smoking duration, RA, AD and T1D could increase the risk of NP, whereas fresh fruit intake could decrease the risk of NP.

Conclusions

This study offers potential causal evidence for AAD risk factors and provides actionable targets for primary prevention.

Keywords: Asthma, allergic rhinitis, rhinosinusitis, nasal polyp, FinnGen

INTRODUCTION

Allergy-related airway disease (AAD) refers to a spectrum of type 2-biased inflammatory conditions that impact both the upper and lower airways as well as the lung tissue; collectively, and these diseases have emerged as a global epidemic.1,2 Despite ongoing scientific research and advancements in understanding these conditions, their prevalence, diversity, and severity continue to rise.1 The epidemiology and etiology of allergic airway disease are still not fully understood. Certain controversies persist in the current investigations.3,4,5,6 Moreover, the evidence from observational studies has been mixed.

Genetics can contribute to elucidating cause and effect relationships, as inherited genetic risks are more difficult to influence by confounding factors and reverse causality.7 One of the most influential approaches is Mendelian randomization (MR), which uses independent single-nucleotide polymorphisms (SNPs) under genome-wide significance thresholds as instrumental variables (IVs) to attempt to quantify causal relationships between risk factors and disease outcomes.7 The MR design is based on the principle that genetic instruments need to show a strong correlation with exposure, remain unaffected by confounders, and influence outcomes exclusively via exposure.8 As a study design situated at a relatively high hierarchy of the clinical science evidence pyramid, MR effectively simulates randomized controlled trials.9,10

In this study, we considered the effects of 83 risk factors on asthma, allergic rhinitis (AR), chronic rhinosinusitis (CRS), and nasal polyps (NPs). Because MR results from a single outcome cohort may have limited generalizability and statistical power, we incorporated summary data from multiple genome-wide association study (GWAS) outcome cohorts related to these airway diseases for MR analysis and performed a meta-analysis on the results.

MATERIALS AND METHODS

We conducted an MR study using publicly available GWAS data. Because our study used summary data and did not involve individual-level data,11 and the original studies obtained ethical approval and informed consent from all participants, no additional ethical approval was needed. This study was reported in accordance with the STROBE-MR guidelines.12 The statistical analyses were conducted in the R (4.3.0; R Foundation for Statistical Computing, Vienna, Austria), utilizing the “TwoSampleMR” package (0.6.6), “ieugwasr” package (1.0.1), “MRlap” package (0.0.3.2), “RadialMR” package (1.1), “ldscr” package (0.1.0) and “meta” package (6.6-0). We present our research framework in Fig. 1.

Fig. 1. Research framework for MR analysis. Assumptions 1: a substantial association exists between genetic instruments and exposures. Assumptions 2: genetic instruments are not associated with any confounders. Assumptions 3: genetic instruments influence outcomes exclusively through their interaction with risk factors.

Fig. 1

Open in a new tab

MR, Mendelian randomization; AR, allergic rhinitis; CRS, chronic rhinosinusitis; NP, nasal polyp; MVPA, moderate to vigorous physical activity; ADHD, attention deficit hyperactivity disorder; OCD, obsessive-compulsive disorder; PTSD, post-traumatic stress disorder; BMI, body mass index; WHR, Waist-hip ratio; FEV1/FVC, forced expiratory volume in 1 second/forced vital capacity; eGFR, estimated glomerular filtration rate; TSH, thyroid stimulating hormone; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; COVID-19, coronavirus disease 2019; IVW, inverse-variance weighted; GERA, Genetic Epidemiology Research on Adult Health and Aging; FDR, false discovery rate.

Data source

Summary-level statistics for asthma (137,498 cases and 1,009,466 controls), AR (51,712 cases and 592,784 controls), CRS (24,688 cases and 760,073 controls) and NP (11,796 cases and 760,075 controls) were obtained from several nonoverlapping European population cohorts, including the UK Biobank (UKB), FinnGen, Genetic Epidemiology Research on Adult Health and Aging (GERA) and deCODE. We briefly summarize the four AAD phenotypes in Table 1. These original GWASs have been described in Supplementary Data S1.

Table 1. Genome-wide association study summary-level statistics for 4 allergy-related airway disease.

Allergic airway disease Population cohort Participant (cases/controls) Ancestry (%European) Covariates Year PMID
Asthma FinnGen 59,100/259,839 100% Sex, age, genotyping batch, PCs 2023 36653562
GERA 9,209/47,428 100% Sex, age, PCs 2021 33893285
UKB, deCODE 69,189/702,199 100% Sex, age, county of origin, PCs 2020 31959851
Allergic rhinitis FinnGen 15,569/474,650 100% Sex, age, genotyping batch, PCs 2023 36653562
GERA 13,936/42,701 100% Sex, age, PCs 2021 33893285
UKB 22,207/75,433 100% Sex, age, array, Townsend deprivation index, assessment center, PCs 2019 31427789
Chronic rhinosinusitis FinnGen 22,099/371,520 100% Sex, age, genotyping batch, PCs 2023 36653562
UKB 2,589/388,553 100% Sex, birth year, PCs 2018 30104761
Nasal polyp FinnGen 8,496/371,520 100% Sex, age, genotyping batch, PCs 2023 36653562
UKB 3,300/388,555 100% Sex, birth year, PCs 2018 30104761
Open in a new tab

PC, principal component; GERA, Genetic Epidemiology Research on Adult Health and Aging; UKB, UK Biobank.

We included 83 exposures, including socioeconomic, behavioral, physical, metabolic, mental, immune, and early-life factors; gastroesophageal reflux; oral diseases; and coronavirus disease 2019 (COVID-19). We searched the published literature (PubMed), public comprehensive GWAS repositories (GWAS Catalog [https://www.ebi.ac.uk/gwas/], IEU Open GWAS [https://gwas.mrcieu.ac.uk/], GWAS ATLAS, and various domain-specific GWAS consortia to obtain the largest-scale publicly available GWAS. Table 2 presents information on 83 GWASs.

Table 2. GWAS summary-level statistics for 83 risk factors.

Risk factors GWAS consortium Ancestry (%European) Unit Participant (cases/controls) Year PMID
Socioeconomic factors (n = 4)
Educational attainment SSGAC 100% SD 765,283 2022 35361970
Income Meta 100% SD 668,288 2025 39875632
Occupational attainment UKB 100% SD 248,847 2022 34613391
Social isolation UKB 100% SD 445,024 2018 29970889
Behavior traits (n = 15)
Cigarette per day GSCAN 100% SD 326,497 2022 36477530
Smoking initiation GSCAN 100% SD 805,431 2022 36477530
Drinks per week GSCAN 100% SD 666,978 2022 36477530
Coffee intake MRC-IEU 100% SD 428,860 2018 29846171
Tea intake MRC-IEU 100% SD 447,485 2018 29846171
Raw vegetable intake MRC-IEU 100% SD 435,435 2018 29846171
Fresh fruit intake MRC-IEU 100% SD 446,462 2018 29846171
Cheese intake MRC-IEU 100% SD 451,486 2018 29846171
Processed meat intake MRC-IEU 100% SD 461,981 2018 29846171
Leisure screen time Meta 100% SD 526,725 2022 36071172
Moderate to vigorous physical activity Meta 100% SD 606,820 2022 36071172
Sleep duration SDKP 100% SD 446,118 2019 30846698
Insomnia SDKP 100% Log odds 345,022/108,357 2019 30804566
Daytime sleepiness SDKP 100% SD 452,071 2019 31409809
Morning person SDKP 100% Log odds 252,287/150,908 2019 30696823
Physical Measurement (n = 6)
Body mass index GIANT 100% SD 681,275 2018 30124842
Waist-hip ratio GIANT 100% SD 694,649 2018 30239722
Hypertension CTG 100% Log odds 99,665/189,642 2019 31427789
FEV1/FVC Meta 100% SD 321,047 2019 30804560
Estimated glomerular filtration rate CKDGen 100% SD 567,460 2019 31152163
Bone mineral density UKB 100% SD 426,824 2019 30598549
Metabolic traits (n = 17)
Type 2 diabetes Meta 100% Log odds 62,892/596,424 2018 30054458
Thyroid stimulating hormone ThyroidOmics 100% SD 271,040 2024 38291025
Urate GUGC 100% SD 110,347 2013 23263486
C-reactive protein Meta 100% SD 575,531 2022 35459240
Serum 25 hydroxyvitamin D Meta 100% SD 443,734 2020 32059762
Plasma vitamin C Meta 100% SD 52,018 2021 33203707
High density lipoprotein cholesterol GLGC 100% SD 1,320,016 2021 34887591
Low density lipoprotein cholesterol GLGC 100% SD 1,320,016 2021 34887591
Triglyceride GLGC 100% SD 1,320,016 2021 34887591
Apolipoprotein A1 UKB 100% SD 393,193 2020 32203549
Apolipoprotein B UKB 100% SD 439,214 2020 32203549
Omega-3 fatty acids UKB 100% SD 94,420 2024 39278973
Omega-6 fatty acids UKB 100% SD 94,007 2024 39278973
Estradiol UKB 100% SD 206,927 2020 32042192
Total testosterone UKB 100% SD 425,097 2020 32042192
Sex hormone-binding globulin UKB 100% SD 370,125 2020 32042192
Age at menopause ReproGen 100% SD 201,323 2021 34349265
Mental health (n = 16)
Major depressive disorder PGC 100% Log odds 170,756/329,443 2019 30718901
Anxiety Meta 100% Log odds 87,517/1,008,941 2024 39294497
Attention deficit hyperactivity disorder PGC 100% Log odds 38,691/186,843 2023 36702997
Neuroticism CTG 100% SD 380,506 2018 29500382
Bipolar disorder PGC 100% Log odds 41,917/371,549 2021 34002096
Schizophrenia PGC 100% Log odds 52,017/75,889 2022 35396580
Anorexia nervosa PGC 100% Log odds 16,992/55,525 2019 31308545
Autism spectrum disorder* PGC 100% Log odds 18,381/27,969 2019 30804558
Tourette syndrome* PGC 100% Log odds 4,819/9,488 2019 30818990
Obsessive-compulsive disorder* PGC 100% SD 33,943 2024 38548983
Hoarding symptoms* PGC 100% SD 27,651 2022 36379924
Post-traumatic stress disorder PGC 100% Log odds 137,136/1,085,746 2024 38637617
Risk tolerance SSGAC 100% SD 466,571 2019 30643258
Opioid use disorder Gelernter Lab 87% Log odds 20,686/618,377 2022 35879402
Cannabis use disorder Gelernter Lab 100% Log odds 42,281/843,744 2023 37985822
Well-being Meta 100% Z-score 2,311,184 2019 30643256
Immune diseases (n = 11)
Rheumatoid arthritis Meta 100% Log odds 22,350/74,823 2022 36333501
Atopic dermatitis Meta 100% Log odds 60,653/804,329 2023 37794016
Type 1 diabetes Meta 100% Log odds 9,266/15,574 2020 32005708
Crohn’s disease Meta 100% Log odds 12,194/28,072 2017 28067908
Ulcerative colitis Meta 100% Log odds 12,366/33,609 2017 28067908
Celiac disease Meta 100% Log odds 4,533/10,750 2010 20190752
Systemic lupus erythematosus Meta 100% Log odds 5,201/9,066 2015 26502338
Multiple sclerosis IMSGC 100% Log odds 47,429/68,374 2019 31604244
Hypothyroidism CTG 100% Log odds 18,740/270,567 2019 31427789
Psoriasis Meta 100% Log odds 15,967/28,194 2021 34927100
Vitiligo Meta 100% Log odds 4,680/39,586 2016 27723757
Early life factors (n = 9)
Own birth weight EGG 100% SD 298,142 2019 31043758
Offspring birth weight EGG 100% SD 210,267 2019 31043758
Preterm delivery* EGG 100% Log odds 15,466/220,010 2023 37012456
Post-term delivery* EGG 100% Log odds 15,972/115,307 2023 37012456
Breastfed as a baby* CTG 100% Log odds 208,921/84,839 2019 31427789
Maternal smoking around birth CTG 100% Log odds 100,733/231,129 2019 31427789
Childhood maltreatment Meta 100% SD 185,414 2021 33740410
Childhood body mass index EGG 100% SD 61,111 2020 33045005
Age at menarche ReproGen 100% SD 252,514 2017 28436984
Reflux (n = 1)
Gastroesophageal reflux Meta 100% Log odds 78,707/288,734 2021 34187846
Oral diseases (n = 2)
Periodontitis* Meta 97% Z-score 506,594 2019 31235808
Dental caries Meta 100% Z-score 487,823 2019 31235808
COVID-19 (n = 2)
COVID-19 infection Meta 100% Log odds 122,616/2,475,240 2020 32404885
COVID-19 hospitalization Meta 100% Log odds 32,519/2,062,805 2020 32404885
Open in a new tab

GWAS, genome-wide association study; SSGAC, Social Science Genetic Association Consortium; SD, standard deviation; UKB, UK Biobank; GSCAN, genome-wide association study & Sequencing Consortium of Alcohol and Nicotine use; MRC-IEU, Medical Research Council Integrative Epidemiology Unit; SDKP, Sleep Disorder Knowledge Portal; GIANT, Genetic Investigation of Anthropometric Traits Consortium; CTG, Complex Trait Genetics Lab; CKDGen, Chronic Kidney Disease Genetics Consortium; GUGC, Global Urate Genetics Consortium; GLGC, Global Lipids Genetics Consortium; ReproGen, Reproductive Genetics Consortium; PGC, Psychiatric Genomics Consortium; IMSGC, International Multiple Sclerosis Genetics Consortium; EGG, Early Growth Genetics Consortium; COVID-19, coronavirus disease 2019.

*Indicated the instrumental variables threshold was relaxed to 1e-5.

Selection of genetic instruments

SNPs showing strong correlations (GWAS-correlated P < 5 × 10−8) with exposures and being independent of each other (link disequilibrium r2 < 0.001, window size = 10,000 kb) were selected as IVs.13 When fewer than 3 IVs were available,14 we relaxed the threshold to 1 × 10−5. The F-statistic was used to evaluate the strength of each IV with the formula F = (β2/se2).15 An F-statistic greater than 10 indicates the absence of weak instruments.16 For the missing SNPs, we did not look for proxies.17,18

Statistical analysis

We employed an inverse-variance weighted (IVW) with random effects model to combine the Wald ratios for each SNP to derive a composite estimate, which served as the primary analytical approach.19 The MR results are reported as odds ratios (ORs) and 95% confidence intervals (CIs). MR-Egger regression and weighted median served as complementary analyses that could offer more reliable estimates across a broader spectrum of situations.20,21 Cochran’s Q test was used to assess heterogeneity,22 and the MR-Egger intercept test was employed to detect horizontal pleiotropy.20 Upon detecting horizontal pleiotropy, radial MR was performed to identify outliers that might introduce horizontal pleiotropy.23 The MR and sensitivity analyses were re-executed after these outliers were excluded.24 Additionally, contemporary GWAS consortia frequently incorporate samples from various biobanks to increase their statistical power and identify a greater number of genome-wide significant loci. Avoiding potential sample overlap between exposure and AAD results in MR analyses utilizing summary-level statistics being challenging, which may increase the rate of type I errors.25 MRlap demonstrates robustness against biases arising from sample overlap, the winner’s curse, and weak instruments.26 We annotated the potential sample overlap between the 83 exposures and the cohort of AAD outcomes in Supplementary Table S1, and employed MRlap for sensitivity analysis.

We conducted a meta-analysis of the MR estimates for each of the four types of AADs. Before performing the meta-analysis, we assessed the genetic correlation between multiple cohort GWAS for each AAD using linkage disequilibrium score regression (LDSC)27 to ensure the validity of the meta-analysis. The heterogeneity among the results was quantified using the I2 statistic; when I2 was ≥ 50%, a random-effects model was employed for pooling; otherwise, a fixed-effects model was used.28 For 332 (83 × 4) estimates obtained from the meta-analysis, the false discovery rate (FDR) was employed to control for the type I errors.29 An FDR q value less than 0.05 indicated statistical significance.

RESULTS

In this study, we investigated the associations between 83 factors and 4 AADs. Given the limited number of SNPs associated with 8 exposures (as noted in Table 2), we relaxed their threshold for IVs to 1 × 10−5. All 83 IVs had F-statistics exceeding 10, indicating the absence of weak instruments (Supplementary Table S2). In total, we performed 830 MR analyses and multiple sensitivity analyses to ensure stability. Heterogeneity was detected in some MR estimates by Cochran’s Q test (Supplementary Tables S3, 4, 5, 6). However, heterogeneity among SNPs did not compromise our MR estimates, as we utilized the IVW method with a random-effects model. The full MR results are presented in Supplementary Tables S3, 4, 5, 6.

LDSC analysis indicated that there is a high genetic correlation (rg > 0.6) between cohorts for each type of AAD, suggesting that the application of meta-analysis is valid. The specific results are presented in Supplementary Table S7. In the subsequent 332 meta-analyses, we applied a fixed-effects model to the 281 meta-analyses with I2 values less than 50%, whereas the remaining 51 meta-analyses utilized a random-effects model (Supplementary Tables S3, 4, 5, 6). Overall, our results revealed 56 sets of evidence for potential causal associations between exposure factors and AADs after FDR correction (Supplementary Table S8). We present 332 results of the meta-analyses in Fig. 2. The detailed results are shown in Supplementary Tables S3, 4, 5, 6. The forest plots for the significant estimates are presented in Fig. 3 (detailed forest plots for each meta-analysis are shown in the Supplementary Fig. S1).

Fig. 2. Heatmap for results of the meta-analysis of all Mendelian randomization estimates.

Fig. 2

Open in a new tab

AR, allergic rhinitis; CRS, chronic rhinosinusitis; NP, nasal polyp; FEV1/FVC, forced expiratory volume in 1 second/forced vital capacity; COVID-19, coronavirus disease 2019; OR, odds ratio.

*Indicated P < 0.05, but q > 0.05; **Indicated q < 0.05.

Fig. 3. Forest plot for significant estimates after FDR correction. (A) Risk factors for asthma. (B) Risk factors for allergic rhinitis. (C) Risk factors for chronic rhinosinusitis. (D) Risk factors for nasal polyp.

Fig. 3

Open in a new tab

FDR, false discovery rate; OR, odds ratio; CI, confidence interval; FEV1/FVC, forced expiratory volume in 1 second/forced vital capacity; HDL-C, high-density lipoprotein cholesterol; ADHD, attention deficit hyperactivity disorder; PTSD, post-traumatic stress disorder; BMI, body mass index.

Socioeconomic factors and AADs

For socioeconomic factors, genetically proxied higher educational attainment (OR, 0.78; 95% CI, 0.67–0.90; q = 0.008, I2 = 89%) and income (OR, 0.73; 95% CI, 0.65–0.81; q = 3.26 × 10−7, I2 = 27%) were associated with a reduced risk of asthma. Social isolation was associated with a greater risk of asthma (OR, 1.77; 95% CI, 1.43–2.19; q = 3.02 × 10−6, I2 = 0%) and AR (OR, 1.57; 95% CI, 1.14–2.16; q = 0.032, I2 = 0%). The MR-Egger intercept test (Supplementary Tables S3, 4, 5, 6) and MRlap (Supplementary Table S9) analyses indicated that the estimates were not affected by horizontal pleiotropy or sample overlap. We failed to capture any potential effects of occupational attainment on AADs. In addition, the potential effects of educational attainment on CRS (OR, 0.88; 95% CI, 0.77–1.00; q = 0.066, I2 = 75%), educational attainment on NP (OR, 0.88; 95% CI, 0.77–1.00; q = 0.187, I2 = 30%), and income on AR (OR, 1.13; 95% CI, 1.01–1.28; q = 0.182, I2 = 49%) did not pass the FDR correction.

Behavior traits and AADs

In terms of the tobacco and alcohol consumption, a longer smoking duration (smoking initiation) was predicted to be genetically associated with a greater risk of asthma (OR, 1.38; 95% CI, 1.19–1.61; q = 1.96 × 10−4, I2 = 61%) and NP (OR, 1.21; 95% CI, 1.06–1.39; q = 0.03, I2 = 0%); and a higher smoking frequency (cigarette per day) was also associated with a greater risk of asthma (OR, 1.25; 95% CI, 1.15–1.37; q = 4.57 × 10−6, I2 = 24%). Additionally, the effect of cigarettes per day on CRS did not pass multiple correction (OR, 1.18; 95% CI, 1.01–1.38; q = 0.18, I2 = 0%). We did not observe any potential effects of smoking on AR or drinking on AADs. Interestingly, smoking initiation was associated with a greater risk of CRS in the UKB (OR, 2.20; 95% CI, 1.57–3.08; P = 3.84 × 10−6) and FinnGen (OR, 1.19; 95% CI, 1.03–1.38; P = 0.022) MR results separately. However, the results of the meta-analysis did not reveal a potential association (OR, 1.59; 95% CI, 0.87–2.90; I2 = 91%).

Among the eating habits, genetic predictions indicated that increased coffee intake was associated with a reduced risk of CRS (OR, 0.66; 95% CI, 0.50–0.87; q = 0.022, I2 = 0%), higher fresh fruit intake may lower the risk of NP (OR, 0.47; 95% CI, 0.27–0.81; q = 0.042, I2 = 0%), and elevated cheese consumption could diminish the risk of asthma (OR, 0.79; 95% CI, 0.69–0.91; q = 0.008, I2 = 0%).

Among sedentary and physical activities, longer sedentary time (leisure screen time) was predicted to be genetically associated with a higher risk of asthma (OR, 1.27; 95% CI, 1.21–1.33; q = 1.03 × 10−19, I2 = 0%) and CRS (OR, 1.22; 95% CI, 1.10–1.34; q = 0.001, I2 = 29%). The effect of moderate to vigorous physical activity on asthma was no longer significant after FDR correction (OR, 0.82; 95% CI, 0.69–0.96; q = 0.079, I2 = 0%). We did not observe potential effects of sedentariness on AR or NP or of physical activity on AR, CRS or NP.

For sleep characteristics, longer sleep duration was associated with a reduced risk of asthma (OR, 0.81; 95% CI, 0.71–0.92; q = 0.012, I2 = 0%). Insomnia was predicted to be genetically associated with asthma (OR, 1.42; 95% CI, 1.13–1.78; q = 0.015, I2 = 0%) and AR (OR, 1.70; 95% CI, 1.29–2.22; q = 0.001, I2 = 0%). The association between sleep duration and the risk of CRS was no longer significant after FDR correction (OR, 0.77; 95% CI, 0.62–0.95; q = 0.086, I2 = 0%). No evidence has indicated that daytime sleepiness or morning person behaviors were associated with changes in the risk of AADs.

The MR-Egger intercept provided no evidence of horizontal pleiotropy for the above results. MRlap tests indicated that the MR estimates of sleep duration and the risk of asthma in the UKB and deCODE populations, as well as the MR estimates of insomnia and the risk of AR in the UKB population, may be affected by sample overlap (Supplementary Table S9). However, subsequent exclusion analyses indicated that this potential bias did not substantially impact causal estimates (Supplementary Table S10).

Physical measurement and AADs

Among the physical measurement factors, genetic predictions indicated that both global obesity, as measured by body mass index (OR, 1.30; 95% CI, 1.18–1.43; q = 2.43 × 10−6, I2 = 82%), and central obesity, as assessed by the waist-to-hip ratio (OR, 1.23; 95% CI, 1.16–1.31; q = 3.46 × 10−11, I2 = 0%), were associated with an increased risk of asthma. Additionally, higher lung function (forced expiratory volume in 1 second/forced vital capacity) was predicted to be genetically associated with a lower risk of asthma (OR, 0.66; 95% CI, 0.64–0.68; q = 4.25 × 10−158, I2 = 8%). We failed to capture potential associations between hypertension, renal function, bone mineral density and the risk of AADs. Multiple sensitivity analyses provided no evidence of bias.

Metabolic traits and AADs

Among lipid traits, higher high-density lipoprotein cholesterol was found to be associated with a reduced risk of asthma (OR, 0.95; 95% CI, 0.92–0.98; q = 0.002, I2 = 28%), and higher apolipoprotein A1 levels were associated with a reduced risk of CRS (OR, 0.90; 95% CI, 0.85–0.96; q = 0.007, I2 = 0%). The potential effects of low-density lipoprotein cholesterol, triglycerides and apolipoprotein B with AADs were not observed. In addition, genetic predictions indicated that higher C-reactive protein levels were associated with an increased risk of asthma (OR, 1.11; 95% CI, 1.08–1.13; q = 1.45 × 10−13, I2 = 38%) and AR (OR, 1.06; 95% CI, 1.02–1.10; q = 0.039, I2 = 0%). We failed to capture significant effects of type 2 diabetes, thyroid-stimulating hormone, urate, serum 25-hydroxyvitamin D, plasma vitamin C, omega-3 fatty acids, omega-6 fatty acids, estradiol, total testosterone, sex hormone-binding globulin and age at menopause on AADs.

Sensitivity analyses provided no evidence that our results were affected by horizontal pleiotropy. MRlap tests indicated that the MR estimates of apolipoprotein A1 and the risk of CRS in the UKB population may be affected by sample overlap. However, the MR estimates for the remaining nonoverlapping populations validated the results.

Mental health and AADs

Among the mental health factors, major depressive disorder was found to be associated with a greater risk of asthma (OR, 1.36; 95% CI, 1.26–1.48; q = 2.54 × 10−12, I2 = 0%), AR (OR, 1.25; 95% CI, 1.14–1.36; q = 7.45 × 10−6, I2 = 0%) and CRS (OR, 1.38; 95% CI, 1.22–1.56; q = 4.60 × 10−6, I2 = 0%). Similarly, anxiety was predicted to be genetically associated with a greater risk of asthma (OR, 1.99; 95% CI, 1.30–3.05; q = 0.012, I2 = 73%) and CRS (OR, 2.06; 95% CI, 1.33–3.20; q = 0.009, I2 = 48%), whereas neuroticism was associated with a greater risk of asthma (OR, 1.43; 95% CI, 1.29–1.60; q = 1.66 × 10−9, I2 = 0%) and AR (OR, 1.22; 95% CI, 1.09–1.37; q = 0.007, I2 = 0%). Attention deficit hyperactivity disorder could increase the risk of asthma (OR, 1.17; 95% CI, 1.13–1.22; q = 4.17 × 10−13, I2 = 0%). Genetic predictions also revealed that posttraumatic stress disorder was associated with a greater risk of asthma (OR, 1.91; 95% CI, 1.63–2.22; q = 3.79 × 10−14, I2 = 34%) and CRS (OR, 2.20; 95% CI, 1.69–2.86; q = 1.03 × 10−7, I2 = 0%). Well-being was predicted to be a protective factor for asthma (OR, 0.56; 95% CI, 0.47–0.66; q = 5.10 × 10−10, I2 = 0%), AR (OR, 0.53; 95% CI, 0.41–0.70; q = 4.12 × 10−5, I2 = 0%) and CRS (OR, 0.45; 95% CI, 0.33–0.62; q = 7.65 × 10−6, I2 = 17%). For substance use disorders, we observed that opioid use disorders may causally increase CRS risk (OR, 1.51; 95% CI, 1.19–1.92; q = 0.007, I2 = 45%). The effects of attention deficit hyperactivity disorder, bipolar disorder, schizophrenia and opioid use disorder on NP were no longer significant after multiple correction. There was no evidence of significant effects of anorexia nervosa, autism spectrum disorder, Tourette syndrome, obsessive-compulsive symptoms, hoarding symptoms, risk tolerance, or cannabis use disorder on 4 AADs. In addition, the results of the meta-analysis of neuroticism and CRS were similar to those of smoking and CRS. Neuroticism was associated with a greater risk of CRS in the UKB (OR, 1.98; 95% CI, 1.25–3.14; P = 0.004) and FinnGen (OR, 1.23; 95% CI, 1.05–1.46; P = 0.013) MR results separately. However, the results of the meta-analysis did not reveal a potential association (OR, 1.48; 95% CI, 0.95–2.33, I2 = 72%).

The MR-Egger intercept test indicated potential horizontal pleiotropy in the effect of well-being on asthma within the UKB and deCODE populations. Furthermore, MRlap suggested that the effect of anxiety on asthma in the UKB and deCODE populations may be affected by sample overlap (Supplementary Table S9). Subsequent exclusion analyses indicated that these potential biases did not substantially impact causal estimates (Supplementary Table S10).

Immune diseases and AADs

Atopic dermatitis was predicted to be genetically associated with a greater risk of 4 AADs, including asthma (OR, 1.31; 95% CI, 1.11–1.55; q = 0.013, I2 = 92%), AR (OR, 1.34; 95% CI, 1.11–1.63; q = 0.018, I2 = 94%), CRS (OR, 1.14; 95% CI, 1.07–1.22; q = 8.17 × 10−4, I2 = 29%) and NP (OR, 1.46; 95% CI, 1.31–1.64; q = 5.87 × 10−10, I2 = 0%). Rheumatoid arthritis was found to be associated with a greater risk of asthma (OR, 1.07; 95% CI, 1.05–1.09; q = 1.35 × 10−10, I2 = 0%), CRS (OR, 1.12; 95% CI, 1.09–1.15; q = 8.04 × 10−13, I2 = 0%) and NP (OR, 1.16; 95% CI, 1.10–1.21; q = 2.88 × 10−8, I2 = 0%). Similarly, genetic predictions indicated that type 1 diabetes could increase the risk of asthma (OR, 1.04; 95% CI, 1.02–1.06; q = 0.002, I2 = 68%), CRS (OR, 1.05; 95% CI, 1.03–1.07; q = 1.24 × 10−7, I2 = 19%) and NP (OR, 1.10; 95% CI, 1.08–1.13; q = 1.73 × 10−13, I2 = 0%). Crohn’s disease was associated with an increased risk of AR (OR, 1.04; 95% CI, 1.03–1.06; q = 3.66 × 10−5, I2 = 0%), whereas ulcerative colitis was linked to protective effects against asthma (OR, 0.97; 95% CI, 0.95–0.99; q = 0.026, I2 = 56%). Additionally, hypothyroidism was predicted to be genetically associated with a greater risk of CRS (OR, 1.10; 95% CI, 1.05–1.15; q = 1.84 × 10−4, I2 = 0%). The effects of Crohn’s disease on asthma, vitiligo on AR, vitiligo on CRS, vitiligo on NP and hypothyroidism on NP were no longer significant after FDR correction. Potential causal effects of celiac disease, systemic lupus erythematosus, multiple sclerosis, and psoriasis on AADs were not observed.

The MR-Egger intercept test indicated potential horizontal pleiotropy in the association between rheumatoid arthritis and asthma risk within the GERA populations. However, the MR estimates for the remaining meta-analysis of nonoverlapping population validated the results. MRlap analyses indicated that the estimates were not affected by sample overlap.

Early-life factors and AADs

Among early-life factors, maternal smoking around birth (OR, 1.19; 95% CI, 1.06–1.33; q = 0.016, I2 = 0%), childhood maltreatment (OR, 1.36; 95% CI, 1.12–1.65; q = 0.015, I2 = 0%), a higher childhood body mass index (OR, 1.18; 95% CI, 1.12–1.24; q = 1.94 × 10−8, I2 = 0%) and earlier age at menarche (OR, 0.94; 95% CI, 0.91–0.97; q = 0.001, I2 = 0%) were all predicted to be genetically associated with a greater risk of asthma. The effect of preterm delivery on asthma risk did not pass FDR correction (OR, 1.04; 95% CI, 1.00–1.07; q = 0.12, I2 = 0%). We failed to observe potential causal associations between own birth weight, offspring birth weight, post-term delivery and being breastfed as a baby and the risk of 4 AADs. Sensitivity analyses provided no evidence of bias.

Other factors and AADs

Genetic predictions indicated that gastroesophageal reflux could increase the risk of asthma (OR, 1.40; 95% CI, 1.24–1.59; q = 9.80 × 10−7, I2 = 75%), AR (OR, 1.17; 95% CI, 1.09–1.25; q = 1.85 × 10−4, I2 = 0%) and CRS (OR, 1.26; 95% CI, 1.14–1.40; q = 7.20 × 10−5, I2 = 0%). The potential effects of dental caries on asthma (OR, 1.19; 95% CI, 1.04–1.38; q = 0.073, I2 = 13%), COVID-19 hospitalization on asthma (OR, 0.96; 95% CI, 0.93–1.00; q = 0.177, I2 = 30%) and COVID-19 hospitalization on CRS (OR, 0.93; 95% CI, 0.87–1.00; q = 0.185, I2 = 0%) were no longer significant after FDR correction. Potential causal effects of periodontitis and COVID-19 infection on AADs were not observed. Sensitivity analyses provided no evidence of bias.

DISCUSSION

This study systematically evaluated the potential causal relationships between 83 exposures and four AAD phenotypes using MR, leveraging large-scale GWAS summary statistics from diverse populations. By employing robust sensitivity analyses and meta-analytic approaches, we identified 56 inferred causal associations, highlighting risk factors and protective factors across asthma, AR, CRS, and NP. Importantly, our analysis revealed both shared and unique risk profiles for distinct AADs (Fig. 4). For example, major depressive disorder and gastroesophageal reflux had pan-AAD effects, whereas age at menarche had asthma-specific associations.

Fig. 4. Shared causal associations of risk factors with 4 allergy-related airway disease.

Fig. 4

Open in a new tab

CRS, chronic rhinosinusitis; NP, nasal polyp; AR, allergic rhinitis; FEV1/FVC, forced expiratory volume in 1 second/forced vital capacity.

With respect to socioeconomic factors, the findings regarding genetic associations between educational attainment, income and the risk of asthma derived from single outcome cohorts have been inconsistent in previous MR studies.30,31,32 Our study verified the protective effects of good education and higher income on asthma from a genetic perspective. Previous studies have indicated that socioeconomic inequality may influence asthma incidence through alterations in behavior and lifestyle, such as smoking and obesity.31 With the ongoing progress of population aging, social isolation has emerged as a significant public health issue. In alignment with prior observational and one-sample MR studies conducted in the UKB,33 we further confirmed the potential causal effect of genetically predicted social isolation on the increased risk of developing asthma. The increased asthma risk associated with loneliness may stem from heightened stress responses, such as overactivation of the neural alarm system, HPA axis dysregulation, and increased sympathetic nervous system activity.34,35 However, socioeconomic factors, as complex traits, are more prone to confounding influences than other variables, which may impact the accuracy of causal inference through MR. While we have statistically examined pleiotropy, residual bias from unknown or non-linear pleiotropic pathways remains a possibility. Therefore, a cautious interpretation of the results is still warranted.

Among unhealthy lifestyles, we primarily observed the risk effects of sedentary behavior and smoking on AADs. Smoking is a significant risk factor for respiratory system diseases. Our results indicate that smoking behavior, regardless of smoking frequency or duration, plays a significant role in the development of asthma. Previous studies have also emphasized the association between smoking and the risk of CRS36 and proposed several potential mechanisms.37 However, our results were not consistent. The smoking initiation phenotype demonstrated statistical significance in the MR estimates for both the UKB and FinnGen CRS outcome cohorts individually. However, there was substantial heterogeneity between the MR estimates during the meta-analysis (I2 = 90.8%), which may have obscured the true effect. A similar potential bias was present regarding the impact of neuroticism on CRS in our study. This heterogeneity may be attributed to differences between the UKB and FinnGen cohorts, such as the vastly divergent number of cases. We also observed associations between smoking duration and an increased risk of NP. Notably, CRS is a highly heterogeneous disease, with only those cases involving Th2-type inflammatory endotypes showing a stronger link to allergies.38 Consequently, recent updates to CRS guidelines and consensus documents recommend classifying CRS based on inflammatory endotypes.38,39 However, large-scale GWASs utilizing this classification framework have not yet been conducted, highlighting the need for further GWASs of CRS subtypes.40 This may also be a contributing factor to the significant heterogeneity observed in the MR meta-analysis regarding smoking and CRS, because the CRS cases defined by ICD codes in the UKB and FinnGen did not adequately clarify inflammatory endotypes.

Among early-life factors, childhood obesity, maltreatment, early menarche, and maternal smoking around birth were all potentially causally associated with an increased risk of asthma. These findings provide novel causal evidence that links early-life factors to asthma risk for previously reported studies41,42,43,44 and further underscore the importance of effective interventions targeting lifestyle, metabolic health, and mental health during early life for the management of AADs throughout the lifespan. However, our MR estimates regarding the effects of delivery timing and breastfeeding on AADs may be inaccurate. Owing to insufficient SNPs under the 5 × 10−8 threshold, the causal effect estimates derived from the MR method may approach zero when a more relaxed IV threshold is used,45 although we excluded weak IV bias by calculating the F-statistic.

The high comorbidity of mental disorders with allergic diseases results in a substantial reduction in health-related quality of life and poses a significant public health burden.46,47,48 In this study, we present potential causal evidence regarding the impact of major depressive disorder, anxiety, attention deficit hyperactivity disorder, neuroticism, posttraumatic stress disorder, opioid use disorder and well-being on allergic diseases, thereby enriching our understanding of this comorbidity. In addition, we found that the overlap of risk factors for the four AADs was associated primarily with mental health and immune-related diseases. Atopic dermatitis, which is also classified as an allergic disease, contributes to the genetic risk associated with all four AADs. In addition, major depressive disorder, well-being, rheumatoid arthritis, type 1 diabetes, and gastroesophageal reflux disease collectively increase the genetic risk for three types of AADs. A Venn diagram (Fig. 4) was generated to illustrate the common and independent risk factors associated with the four AADs. These findings suggest that different AADs may have partially distinct and synergistic preventive pathways, providing a potential foundation for establishing accurate risk evidence.

The primary strengths of this study are its MR design, which minimizes the potential for bias resulting from confounding factors and reverse causality, along with the utilization of multiple independent outcomes and meta-analyses to consolidate MR estimates for reliability. However, our study also has several limitations. First, causality must be interpreted with caution, as certain assumptions of the methods are insufficiently testable, and the heterogeneity of IVs and MR estimates of different AAD outcomes, along with potential horizontal pleiotropy, may still bias some results. However, we applied multiple sensitivity analyses to prevent weak IVs, horizontal pleiotropy and outliers from affecting our MR estimates. Second, to maximize the use of publicly available genetic resources, sample overlap inevitably occurred between the exposures and outcomes included in our study, which could introduce potential bias. However, we effectively mitigated this bias to an acceptable level through the application of MRlap. Third, our reliance on summary-level genetic statistics from public databases limits our ability to assess nonlinear causal relationships. Fourth, we were unable to assess some important AAD risk factors identified in previous observational studies, such as air pollution. The validity of address-based air pollution GWAS-derived MR studies has raised concerns.49 Besides, some risk factors still lack large-scale GWASs. Moreover, the heterogeneity in definitions between different cohorts of AADs may pose challenges for the meta-analysis; however, the high genetic correlations observed in our LDSC analysis suggest that such variability is acceptable. Furthermore, asthma, CRS, and NP exhibit multiple inflammatory endotypes, but due to the lack of relevant subtype data, further analyses are not possible at this study. This remains an area for future exploration. Finally, because the participants in this study were primarily of European ancestry, it is unlikely that population stratification biased our findings. However, the applicability of these findings to other populations should be approached with caution.

Our findings underscore the complex roles of socioeconomic, behavioral, metabolic, and psychosocial, immune, and early-life factors and reflux in AAD development. The identification of modifiable risks, such as reducing social isolation, managing mental health, and promoting healthier lifestyles, provides actionable targets for primary prevention. By establishing causal evidence through MR, this study advances our understanding of AAD etiology and informs public health policies aimed at mitigating the global burden of AADs.

ACKNOWLEDGMENTS

We thank all GWAS consortiums involved in this study, such as FinnGen, GERA, deCODE, SSGAC, UK Biobank, MRC-IEU, GSCAN, CTG, SDKP, GIANT, EGG, GUGC, CKDGen, ReproGen, GLGC, PGC, IMSGC, Gelernter Lab and all the researchers and participants who contribute to these studies.

Footnotes

Disclosure: There are no financial or other issues that might lead to conflict of interest.

Data Availability Statement: The data used in this study were obtained from public databases, and the sources of all data are indicated in Tables 1 and 2.

SUPPLEMENTARY MATERIALS

Supplementary Data S1

Description of GWAS studies of allergic airway diseases included in this study

aair-18-224-s001.doc (30KB, doc)
Supplementary Table S1

Main overlapping cohorts between 83 risk factors and 4 allergy-related airway disease

aair-18-224-s002.pdf (73.6KB, pdf)
Supplementary Table S2

Instrumental variables for 83 exposures

aair-18-224-s003.pdf (1.1MB, pdf)
Supplementary Table S3

Mendelian randomization results for asthma

aair-18-224-s004.pdf (162.7KB, pdf)
Supplementary Table S4

Mendelian randomization results for allergic rhinitis

aair-18-224-s005.pdf (160.6KB, pdf)
Supplementary Table S5

Mendelian randomization results for chronic rhinosinusitis

aair-18-224-s006.pdf (146.4KB, pdf)
Supplementary Table S6

Mendelian randomization results for nasal polyp

aair-18-224-s007.pdf (146.4KB, pdf)
Supplementary Table S7

LDSC results for every AAD

aair-18-224-s008.pdf (58.7KB, pdf)
Supplementary Table S8

FDR correction for meta-analysis results

aair-18-224-s009.pdf (99.4KB, pdf)
Supplementary Table S9

MRlap analysis of potential overlapping populations within the composition cohorts of significant meta-analysis results

aair-18-224-s010.pdf (76.9KB, pdf)
Supplementary Table S10

The remaining MR results after excluding cohorts with potential bias

aair-18-224-s011.pdf (85.4KB, pdf)
Supplementary Fig. S1

Forest plots for each significant Meta-analysis results after FDR correction.

aair-18-224-s012.doc (5.9MB, doc)

References

  • 1.Rothenberg ME. The climate change hypothesis for the allergy epidemic. J Allergy Clin Immunol. 2022;149:1522–1524. doi: 10.1016/j.jaci.2022.02.006. [DOI] [PubMed] [Google Scholar]
  • 2.Lan F, Zhang N, Gevaert E, Zhang L, Bachert C. Viruses and bacteria in Th2-biased allergic airway disease. Allergy. 2016;71:1381–1392. doi: 10.1111/all.12934. [DOI] [PubMed] [Google Scholar]
  • 3.Han YY, Forno E, Gogna M, Celedón JC. Obesity and rhinitis in a nationwide study of children and adults in the United States. J Allergy Clin Immunol. 2016;137:1460–1465. doi: 10.1016/j.jaci.2015.12.1307. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Bhattacharyya N. Associations between obesity and inflammatory sinonasal disorders. Laryngoscope. 2013;123:1840–1844. doi: 10.1002/lary.24019. [DOI] [PubMed] [Google Scholar]
  • 5.Lieberoth S, Gade EJ, Brok J, Backer V, Thomsen SF. Age at menarche and risk of asthma: systematic review and meta-analysis. J Asthma. 2014;51:559–565. doi: 10.3109/02770903.2014.903966. [DOI] [PubMed] [Google Scholar]
  • 6.Asrullah M, L’Hoir M, Feskens EJM, Melse-Boonstra A. Trend in age at menarche and its association with body weight, body mass index and non-communicable disease prevalence in Indonesia: evidence from the Indonesian Family Life Survey (IFLS) BMC Public Health. 2022;22:628. doi: 10.1186/s12889-022-12995-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Davey Smith G, Hemani G. Mendelian randomization: genetic anchors for causal inference in epidemiological studies. Hum Mol Genet. 2014;23:R89–R98. doi: 10.1093/hmg/ddu328. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Boef AG, Dekkers OM, le Cessie S. Mendelian randomization studies: a review of the approaches used and the quality of reporting. Int J Epidemiol. 2015;44:496–511. doi: 10.1093/ije/dyv071. [DOI] [PubMed] [Google Scholar]
  • 9.Arsenault BJ. From the garden to the clinic: how Mendelian randomization is shaping up atherosclerotic cardiovascular disease prevention strategies. Eur Heart J. 2022;43:4447–4449. doi: 10.1093/eurheartj/ehac394. [DOI] [PubMed] [Google Scholar]
  • 10.Sekula P, Del Greco M F, Pattaro C, Köttgen A. Mendelian randomization as an approach to assess causality using observational data. J Am Soc Nephrol. 2016;27:3253–3265. doi: 10.1681/ASN.2016010098. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Burgess S, Scott RA, Timpson NJ, Davey Smith G, Thompson SG EPIC- InterAct Consortium. Using published data in Mendelian randomization: a blueprint for efficient identification of causal risk factors. Eur J Epidemiol. 2015;30:543–552. doi: 10.1007/s10654-015-0011-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Skrivankova VW, Richmond RC, Woolf BAR, Yarmolinsky J, Davies NM, Swanson SA, et al. Strengthening the reporting of observational studies in epidemiology using Mendelian randomization: the STROBE-MR statement. JAMA. 2021;326:1614–1621. doi: 10.1001/jama.2021.18236. [DOI] [PubMed] [Google Scholar]
  • 13.Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D, et al. The MR-Base platform supports systematic causal inference across the human phenome. ELife. 2018;7:e34408. doi: 10.7554/eLife.34408. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Guo T, Jia G, Liu D, Deng X, Li J, Xie H. Understanding factors that cause benign paroxysmal positional vertigo, Ménière disease, and vestibular neuritis: a two-sample Mendelian randomization study. Ear Hear. 2025;46:305–314. doi: 10.1097/AUD.0000000000001574. [DOI] [PubMed] [Google Scholar]
  • 15.Bowden J, Del Greco M F, Minelli C, Zhao Q, Lawlor DA, Sheehan NA, et al. Improving the accuracy of two-sample summary-data Mendelian randomization: moving beyond the NOME assumption. Int J Epidemiol. 2019;48:728–742. doi: 10.1093/ije/dyy258. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Burgess S, Thompson SG CRP CHD Genetics Collaboration. Avoiding bias from weak instruments in Mendelian randomization studies. Int J Epidemiol. 2011;40:755–764. doi: 10.1093/ije/dyr036. [DOI] [PubMed] [Google Scholar]
  • 17.Ye F, Huang Y, Zeng L, Li N, Hao L, Yue J, et al. The genetically predicted causal associations between circulating 3-hydroxybutyrate levels and malignant neoplasms: a pan-cancer Mendelian randomization study. Clin Nutr. 2024;43:137–152. doi: 10.1016/j.clnu.2024.09.044. [DOI] [PubMed] [Google Scholar]
  • 18.Hong K, Hun M, Wu F, Mao J, Wang Y, Zhu J, et al. Association between Omega-3 fatty acids and autoimmune disease: evidence from the umbrella review and Mendelian randomization analysis. Autoimmun Rev. 2024;23:103651. doi: 10.1016/j.autrev.2024.103651. [DOI] [PubMed] [Google Scholar]
  • 19.Pierce BL, Burgess S. Efficient design for Mendelian randomization studies: subsample and 2-sample instrumental variable estimators. Am J Epidemiol. 2013;178:1177–1184. doi: 10.1093/aje/kwt084. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44:512–525. doi: 10.1093/ije/dyv080. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol. 2016;40:304–314. doi: 10.1002/gepi.21965. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Greco M FD, Minelli C, Sheehan NA, Thompson JR. Detecting pleiotropy in Mendelian randomisation studies with summary data and a continuous outcome. Stat Med. 2015;34:2926–2940. doi: 10.1002/sim.6522. [DOI] [PubMed] [Google Scholar]
  • 23.Bowden J, Spiller W, Del Greco M F, Sheehan N, Thompson J, Minelli C, et al. Improving the visualization, interpretation and analysis of two-sample summary data Mendelian randomization via the Radial plot and Radial regression. Int J Epidemiol. 2018;47:1264–1278. doi: 10.1093/ije/dyy101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Lu Y, Tang H, Huang P, Wang J, Deng P, Li Y, et al. Assessment of causal effects of visceral adipose tissue on risk of cancers: a Mendelian randomization study. Int J Epidemiol. 2022;51:1204–1218. doi: 10.1093/ije/dyac025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Burgess S, Davies NM, Thompson SG. Bias due to participant overlap in two-sample Mendelian randomization. Genet Epidemiol. 2016;40:597–608. doi: 10.1002/gepi.21998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Mounier N, Kutalik Z. Bias correction for inverse variance weighting Mendelian randomization. Genet Epidemiol. 2023;47:314–331. doi: 10.1002/gepi.22522. [DOI] [PubMed] [Google Scholar]
  • 27.Bulik-Sullivan B, Finucane HK, Anttila V, Gusev A, Day FR, Loh PR, et al. An atlas of genetic correlations across human diseases and traits. Nat Genet. 2015;47:1236–1241. doi: 10.1038/ng.3406. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Fang A, Zhao Y, Yang P, Zhang X, Giovannucci EL. Vitamin D and human health: evidence from Mendelian randomization studies. Eur J Epidemiol. 2024;39:467–490. doi: 10.1007/s10654-023-01075-4. [DOI] [PubMed] [Google Scholar]
  • 29.Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Ser B Stat Methodol. 1995;57:289–300. [Google Scholar]
  • 30.Xu H, Deng H, Wu Y, Yang Y, Zheng X. Household income unequally affects genetic susceptibility to pulmonary diseases: evidence from bidirectional Mendelian randomization study. Front Med (Lausanne) 2024;11:1279697. doi: 10.3389/fmed.2024.1279697. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Lan G, Xie M, Lan J, Huang Z, Xie X, Liang M, et al. Association and mediation between educational attainment and respiratory diseases: a Mendelian randomization study. Respir Res. 2024;25:115. doi: 10.1186/s12931-024-02722-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Peng L, Wen WP. Socioeconomic status and asthma: a bidirectional Mendelian randomization study. World Allergy Organ J. 2023;16:100790. doi: 10.1016/j.waojou.2023.100790. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Liang YY, Zhou M, He Y, Zhang W, Wu Q, Luo T, et al. Observational and genetic evidence disagree on the association between loneliness and risk of multiple diseases. Nat Hum Behav. 2024;8:2209–2221. doi: 10.1038/s41562-024-01970-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Eisenberger NI, Cole SW. Social neuroscience and health: neurophysiological mechanisms linking social ties with physical health. Nat Neurosci. 2012;15:669–674. doi: 10.1038/nn.3086. [DOI] [PubMed] [Google Scholar]
  • 35.Brown EG, Gallagher S, Creaven AM. Loneliness and acute stress reactivity: a systematic review of psychophysiological studies. Psychophysiology. 2018;55:e13031. doi: 10.1111/psyp.13031. [DOI] [PubMed] [Google Scholar]
  • 36.Tan CJ, Leow BHW, Tan BKJ, Tan SF, Teo NWY, Charn TC. Association between smoking and chronic rhinosinusitis: a systematic review and meta-analysis. Laryngoscope. 2024;134:2513–2524. doi: 10.1002/lary.31223. [DOI] [PubMed] [Google Scholar]
  • 37.Christensen DN, Franks ZG, McCrary HC, Saleh AA, Chang EH. A systematic review of the association between cigarette smoke exposure and chronic rhinosinusitis. Otolaryngol Head Neck Surg. 2018;158:801–816. doi: 10.1177/0194599818757697. [DOI] [PubMed] [Google Scholar]
  • 38.Orlandi RR, Kingdom TT, Smith TL, Bleier B, DeConde A, Luong AU, et al. International consensus statement on allergy and rhinology: rhinosinusitis 2021. Int Forum Allergy Rhinol. 2021;11:213–739. doi: 10.1002/alr.22741. [DOI] [PubMed] [Google Scholar]
  • 39.Fokkens WJ, Lund VJ, Hopkins C, Hellings PW, Kern R, Reitsma S, et al. European position paper on rhinosinusitis and nasal polyps 2020. Rhinology. 2020;58:1–464. doi: 10.4193/Rhin20.600. [DOI] [PubMed] [Google Scholar]
  • 40.Lal D, Brar T, Ramkumar SP, Li J, Kato A, Zhang L. Genetics and epigenetics of chronic rhinosinusitis. J Allergy Clin Immunol. 2023;151:848–868. doi: 10.1016/j.jaci.2023.01.004. [DOI] [PubMed] [Google Scholar]
  • 41.Malden S, Gillespie J, Hughes A, Gibson AM, Farooq A, Martin A, et al. Obesity in young children and its relationship with diagnosis of asthma, vitamin D deficiency, iron deficiency, specific allergies and flat-footedness: a systematic review and meta-analysis. Obes Rev. 2021;22:e13129. doi: 10.1111/obr.13129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Abajobir AA, Kisely S, Williams G, Strathearn L, Suresh S, Najman JM. The association between substantiated childhood maltreatment, asthma and lung function: a prospective investigation. J Psychosom Res. 2017;101:58–65. doi: 10.1016/j.jpsychores.2017.08.003. [DOI] [PubMed] [Google Scholar]
  • 43.Cai L, Li X, Qiu L, Wang Y, Wu L, Wu X, et al. Age at menarche and asthma onset among US girls and women: findings from NHANES, 2001-2018. Ann Epidemiol. 2023;87:25–30. doi: 10.1016/j.annepidem.2023.08.003. [DOI] [PubMed] [Google Scholar]
  • 44.Toppila-Salmi S, Luukkainen AT, Xu B, Lampi J, Auvinen J, Dhaygude K, et al. Maternal smoking during pregnancy affects adult onset of asthma in offspring: a follow up from birth to age 46 years. Eur Respir J. 2020;55:1901857. doi: 10.1183/13993003.01857-2019. [DOI] [PubMed] [Google Scholar]
  • 45.Hu X, Cai M, Xiao J, Wan X, Wang Z, Zhao H, et al. Benchmarking Mendelian randomization methods for causal inference using genome-wide association study summary statistics. Am J Hum Genet. 2024;111:1717–1735. doi: 10.1016/j.ajhg.2024.06.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Liu X, Plana-Ripoll O, McGrath JJ, Petersen LV, Dharmage SC, Momen NC. Bidirectional associations between asthma and types of mental disorders. J Allergy Clin Immunol Pract. 2023;11:799–808.e14. doi: 10.1016/j.jaip.2022.11.027. [DOI] [PubMed] [Google Scholar]
  • 47.Xu X, Li S, Chen Y, Deng X, Li J, Xiong D, et al. Association between allergic diseases and mental health conditions: an umbrella review. J Allergy Clin Immunol. 2025;155:701–713. doi: 10.1016/j.jaci.2024.10.030. [DOI] [PubMed] [Google Scholar]
  • 48.Guo T, Jia G, Wang J, Deng X, Liu D, Xie H. The lung-brain axis: genetic evidence for causal effects of asthma, allergic rhinitis, and chronic rhinosinusitis on brain structure. Int Arch Allergy Immunol. 2025;186:619–630. doi: 10.1159/000542718. [DOI] [PubMed] [Google Scholar]
  • 49.Au Yeung SL, Gill D. Concerns over using the Mendelian randomization design to investigate the effect of air pollution. Sci Total Environ. 2024;917:170474. doi: 10.1016/j.scitotenv.2024.170474. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data S1

Description of GWAS studies of allergic airway diseases included in this study

aair-18-224-s001.doc (30KB, doc)
Supplementary Table S1

Main overlapping cohorts between 83 risk factors and 4 allergy-related airway disease

aair-18-224-s002.pdf (73.6KB, pdf)
Supplementary Table S2

Instrumental variables for 83 exposures

aair-18-224-s003.pdf (1.1MB, pdf)
Supplementary Table S3

Mendelian randomization results for asthma

aair-18-224-s004.pdf (162.7KB, pdf)
Supplementary Table S4

Mendelian randomization results for allergic rhinitis

aair-18-224-s005.pdf (160.6KB, pdf)
Supplementary Table S5

Mendelian randomization results for chronic rhinosinusitis

aair-18-224-s006.pdf (146.4KB, pdf)
Supplementary Table S6

Mendelian randomization results for nasal polyp

aair-18-224-s007.pdf (146.4KB, pdf)
Supplementary Table S7

LDSC results for every AAD

aair-18-224-s008.pdf (58.7KB, pdf)
Supplementary Table S8

FDR correction for meta-analysis results

aair-18-224-s009.pdf (99.4KB, pdf)
Supplementary Table S9

MRlap analysis of potential overlapping populations within the composition cohorts of significant meta-analysis results

aair-18-224-s010.pdf (76.9KB, pdf)
Supplementary Table S10

The remaining MR results after excluding cohorts with potential bias

aair-18-224-s011.pdf (85.4KB, pdf)
Supplementary Fig. S1

Forest plots for each significant Meta-analysis results after FDR correction.

aair-18-224-s012.doc (5.9MB, doc)

Articles from Allergy, Asthma & Immunology Research are provided here courtesy of Korean Academy of Asthma, Allergy and Clinical Immunology and Korean Academy of Pediatric Allergy and Respiratory Disease

RESOURCES