PURPOSE OF THE STUDY.
To identify novel genetic variations imputing asthma risk.
Pediatric and adult Australians of European descent consisting of 2669 asthmatic subjects (28% diagnosed by clinical examination, 54% childhood onset, 59% atopic) and 4528 controls (40% with unknown asthma status) produced candidate asthma genetic variations. Meta-analysis of 12 475 physician-diagnosed asthmatic subjects and 19 967 controls provided a prioritization cohort. Four additional individual cohorts totaling 3322 asthmatic subjects and 22 036 controls provided replication.
Microarray-derived genotypes of Australian asthmatic subjects and controls underwent a genome-wide association study to produce asthma candidate loci. Meta-analysis using unique adult Australian and previously published GABRIEL genotypes prioritized candidate novel genetic loci. The prioritized, novel genetic loci endured replication analysis via 4 individual cohorts and a combined meta-analysis. A genomic phenotype prediction model for asthma was produced and subsequently tested. Secondary analysis considered asthma and other immune disease associations.
Two novel loci were significantly associated with asthma risk. The interleukin-6 receptor (IL6R) loci on chromosome 1q21.3 associated more significantly (odds ratio [OR] = 1.09, P = .0033) than the indeterminate loci (rs7130588) at chromosome 11q13.5 (OR = 1.07, P = .328) in the replication cohorts. However, when considering all available data, 1q21.3 and 11q13.5 both demonstrated very significant asthma association (OR = 1.09, P = 2.3 × 10−8 and OR = 1.09, P = 1.8 × 10−8, respectively). Secondary analyses of 11q13.5 revealed an association with atopy and no asthma risk association in nonatopic individuals. Genomic phenotypic predictive modeling significantly predicted asthma cases, but demonstrated poor discriminative ability. Finally, associations between asthma and other immune diseases were noted.
Two novel loci (1q21.3 and 11q13.5) significantly associate with asthma risk with the use of large genome-wide discovery and replication cohorts.
In a “big data” world, sifting reality from statistical chaff presents a significant challenge. By the use of genome-wide association study on several independent cohorts, the researchers give weight to their novel candidate loci. Although not previously associated with asthma risk, IL6R’s feasible mechanistic connection to asthma encourages further investigation. Without a known function and with only loose atopy and Crohn associations, the proposed 11q13.5 locus carries much less biological heft. Despite the inclusion of known, novel, and marginally statistically significant genetic variances, the study’s phenotypic predictive models lack discrimination value, likely because of data imprecision, missing environmental covariates, and other unappreciated asthma risk variables. Even so, more refined phenotypic predictive modeling holds great promise in translating ever-expanding data into improved patient care.
- Copyright © 2012 by the American Academy of Pediatrics