Asthma

2026-04-05 · Chen E et al., combining Global Biobank Meta-analysis Initiative and Million Veteran Program data (over 158,000 asthma cases, 1.6 million+ controls)

Deep learning and statistical methods identify novel asthma risk variants in Europeans

A meta-analysis combining the Global Biobank Meta-analysis Initiative and the Million Veteran Program (over 158,000 asthma cases, 1.6 million+ controls) used deep learning -- a Transformer-based framework called InsightGWAS -- alongside standard statistical methods to find 69 previously unreported genome-wide significant loci for asthma risk, one of the largest asthma genetics studies to date. The paper's scale, not a small number of headline genes, is the finding: functional follow-up implicated genes across immune regulation, airway remodeling and metabolic processes.

2026-09-01 · Genome-wide association study of asthma with high treatment burden and/or worse outcomes defined using electronic healthcare data in UK Biobank. ERJ Open Research. 2026. DOI:10.1183/23120541.00327-2026

Using medical records instead of self-reported diagnosis, a UK Biobank study finds BACH2 hits harder in difficult-to-treat asthma

Around 10% of asthma patients remain poorly controlled despite maximal treatment, and it was unclear whether this hard-to-treat subtype has distinct genetic risk factors. Rather than relying on self-reported asthma diagnosis, this study used linked electronic healthcare records in UK Biobank to identify 7,681 patients specifically with high treatment burden and/or worse outcomes -- a group enriched for adult-onset asthma, female sex, and higher BMI. A GWAS of this subtype against 38,405 controls found 21 reproducible association signals, replicated across four independent studies (12,152 cases, 32,316 controls total). Notably, all 21 signals had been previously linked to asthma generally, but showed a larger effect size specifically in this high-burden group -- evidence that known asthma genes matter more, not differently, in the hardest-to-treat cases. Variant-to-gene mapping highlighted 85 candidate genes, with 5 rated high-confidence: BACH2, D2HGDH, IL1RL1, RPS26 and SMAD3. Four of these five (D2HGDH, IL1RL1, RPS26, SMAD3) are already on this site's own asthma page; BACH2 -- a transcription factor gene with a broader role in regulating immune cell differentiation -- is the one genuinely new addition here. This entry is recorded as context on the existing genes plus the new BACH2 signal, reflecting that the real news is the stronger effect size in difficult-to-treat asthma specifically, verified through a novel use of electronic health records rather than self-report.

2025-08-01 · Genomic Correlations, Shared Loci, and Drug Targets Between Polycystic Ovary Syndrome and Asthma: Insights From Genome-wide Association Analysis. Journal of Clinical Endocrinology and Metabolism. 2025. DOI:10.1210/clinem/dgaf276

PCOS and asthma co-occur more than chance predicts -- a shared gene, ERBB3, may explain why

Observational studies had noted that polycystic ovary syndrome (PCOS) and asthma-related traits co-occur, but it was unclear whether this reflects shared genetics or just observational bias (e.g., obesity affecting both independently). Combining 10,074 PCOS cases with asthma-related traits from UK Biobank and FinnGen, this study found a genuine positive genetic correlation between PCOS and asthma. Along the way, it found 3 novel genome-wide significant loci for asthma subtypes specifically (2 for allergic asthma, 1 for childhood asthma) -- a real addition to asthma genetics discovered as a byproduct of the cross-condition analysis. Pleiotropy analysis identified 5 shared causal regions spanning 13 genes, with ERBB3 emerging as a candidate central gene for the shared biology between the two conditions. Tissue enrichment analysis found relevant shared gene expression in the hypothalamo-pituitary, endocrine, respiratory and urogenital systems -- a genuinely cross-system biological story, not just statistical coincidence. Drug-repurposing analysis flagged anakinra (an existing anti-inflammatory drug) and artenimol as candidates worth investigating for PCOS-asthma comorbidity specifically. This site's polycystic ovary syndrome page carries 23 variants and asthma page carries variants from earlier 2026 findings; ERBB3 is not currently on either.

Source. Moffatt et al. (NEJM 2010), the GABRIEL consortium, genotyped 10,365 persons with physician-diagnosed asthma and 16,110 unaffected persons, matched for ancestry, with random-effects pooled analysis across the whole population and within childhood-onset (before 16 years), later-onset, severe and occupational subgroups. Genome-wide significant associations: rs3771166 (chromosome 2, IL1RL1/IL18R1, p = 3 × 10⁻⁹); rs9273349 (chromosome 6, HLA-DQ, p = 7 × 10⁻¹⁴); rs1342326 (chromosome 9, flanking IL33, p = 9 × 10⁻¹⁰); rs744910 (chromosome 15, SMAD3, p = 4 × 10⁻⁹); rs2284033 (chromosome 22, IL2RB, p = 1.1 × 10⁻⁸). Association at the ORMDL3/GSDMB locus on 17q21 (rs2305480, p = 6 × 10⁻²³) was specific to childhood-onset disease.

Total serum IgE. Only HLA-DR showed genome-wide significant association with total serum IgE concentration, and the loci strongly associated with IgE levels were not associated with asthma. The authors concluded that elevation of total serum IgE has a minor role in the development of asthma — a conclusion about the trait as a causal intermediate, not about allergen-specific IgE in an individual patient's assessment.

The two positions listed here. rs1342326 is the reported IL33-flanking signal from this study. rs3894194 was recorded by the collector against GSDMA, adjacent to the 17q21 ORMDL3/GSDMB region whose association was childhood-onset-specific; the study's own index SNP for that region, rs2305480, has since been collected and is now directly on this page.

Clinical use. None. Diagnosis rests on a compatible history with demonstrated variable expiratory airflow limitation — typically spirometry with bronchodilator reversibility, peak-flow variability or a bronchial challenge. No genotype directs the choice of inhaled corticosteroid, bronchodilator or biologic. Biologic therapy is selected on phenotype and biomarkers such as blood eosinophils and FeNO, not on genotype.