Psychiatry · Medical Literature
The association between asthma and ADHD in children and adolescents: An observational study and a meta-analysis of Mendelian randomization studies.
Abstract
Observational epidemiological analysis and a meta-analysis of Mendelian randomization (MR) were used to investigate the association between attention-deficit/hyperactivity disorder (ADHD) and asthma in children and adolescents. Based on data from the 1999 to 2004 National Health and Nutrition Examination Survey (NHANES), an analysis was conducted to evaluate the association between ADHD and asthma in children and adolescents aged between 4 and 19 years. ADHD and asthma were defined based on self- or parent/guardian proxy-reported physician/health-professional diagnoses from the NHANES Medical Conditions Questionnaire. After analyzing ADHD genome-wide association study data from the Psychiatric Genomics Consortium as instrumental variables, we performed two-sample MR analyses based on genome-wide association study data from 5 studies to examine the causal relationship between genetically predicted ADHD liabilities. In addition, we summarized the results of the inverse-variance weighted method using a random-effects meta-analysis. Asthma prevalence was higher among participants with ADHD than those without ADHD (10.2% vs 5.9%). ADHD was associated with higher odds of asthma (odds ratio = 1.64, 95% confidence interval = 1.36-1.98), and the pattern of results was broadly consistent across subgroups. In contrast, the MR meta-analysis did not support a causal effect of genetically predicted ADHD liability on childhood asthma (odds ratio = 0.94, 95% confidence interval = 0.84-1.05, P > .05). According to data from the NHANES, ADHD and asthma are significantly associated in children and adolescents. A meta-analysis of the MR data did not reveal a causal connection between genetically predicted ADHD liability and childhood asthma. This paradoxical finding illustrates the complex interplay of genetic and nongenetic mechanisms in the ADHD-asthma association.
Clinical relevance
In children and adolescents aged 4–19 years, is ADHD observationally and causally (via Mendelian randomization) associated with an increased risk of asthma? __SNAPSHOT__ {"studyDesign":"Hybrid design: (1) cross-sectional observational analysis using NHANES survey data; (2) meta-analysis of two-sample Mendelian randomization (MR) studies using inverse-variance weighted (IVW) random-effects pooling across 5 GWAS datasets","setting":"United States; nationally representative population-based survey (NHANES 1999–2004); GWAS data sourced from the Psychiatric Genomics Consortium (PGC) and 5 additional GWAS studies (countries/institutions not specified in abstract)","sampleSize":"Observational component: NHANES participants aged 4–19 years (exact n not reported in abstract); MR meta-analysis: 5 GWAS datasets (total n not reported in abstract)","population":"Children and adolescents aged 4–19 years; ADHD and asthma defined by self- or parent/guardian proxy-reported physician or health professional diagnosis via the NHANES Medical Conditions Questionnaire","intervention":"Exposure: ADHD diagnosis (observational arm); genetically predicted ADHD liability using ADHD GWAS SNPs from the PGC as instrumental variables (MR arm)","comparator":"Children and adolescents without ADHD (observational arm); comparator in MR is the general population GWAS outcome reference (no explicit comparator group described beyond standard two-sample MR methodology)","primaryOutcome":"Asthma diagnosis (physician- or health professional-reported); assessed as odds of asthma in those with vs. without ADHD (observational), and as causal odds of asthma per unit increase in genetically predicted ADHD liability (MR meta-analysis)","followUp":"Not applicable (cross-sectional observational design); not reported for MR component","fundingSource":"Not reported"}
Key findings
In the NHANES observational analysis, asthma prevalence was higher among participants with ADHD than those without (10.2% vs. 5.9%; absolute difference 4.3 percentage points [calculated]). The abstract reports that ADHD was associated with higher odds of asthma (odds ratio not fully reproduced in the provided abstract text; the precise OR, 95% CI, and p-value from the observational logistic regression are not reported in the truncated abstract). In the MR meta-analysis, two-sample MR analyses were performed across 5 GWAS datasets using PGC ADHD GWAS data as instrumental variables; results were pooled using IVW with a random-effects model. Specific pooled OR, 95% CI, p-value, and heterogeneity statistics (I²) for the MR meta-analysis are not reported in the supplied abstract. No adverse events are applicable to this observational/genetic study design. NNT/NNH are not calculable from the available data.
Practice implications
IMPORTANT CAVEAT: This summary is based solely on the user-supplied abstract, which was truncated before reporting the full observational odds ratio and all MR meta-analysis numerical results. Key effect estimates, confidence intervals, p-values, and heterogeneity statistics are therefore not reported and cannot be verified or cited. | What was established: The study reports an observational association between ADHD and asthma in a U.S. pediatric population aged 4–19 years (NHANES 1999–2004), with a crude asthma prevalence approximately 1.7-fold higher in those with ADHD (10.2% vs. 5.9%). The MR meta-analysis was designed to assess whether this association reflects a genetically mediated causal relationship by using ADHD-associated SNPs as instruments across 5 GWAS datasets. | What was not established: Causality cannot be confirmed without the full MR results. The truncated abstract does not disclose whether the pooled MR estimate was statistically significant, the degree of heterogeneity across the 5 studies, or whether sensitivity analyses (MR-Egger, weighted median) supported the IVW finding. The extent to which residual confounding was addressed in the observational arm is also not reported. | Applicability: Findings are most applicable to U.S. children and adolescents aged 4–19 years with a physician-diagnosed ADHD. Diagnostic ascertainment relied on self- or proxy-reported diagnosis without formal structured interview, GWAS-based diagnostic confirmation, or objective spirometric asthma verification, limiting diagnostic precision. Populations outside the U.S. or those with subclinical/undiagnosed ADHD are not adequately represented. The NHANES data are from 1999–2004 and may not reflect contemporary prevalence rates or diagnostic practices. | Clinical importance: The 4.3-percentage-point absolute prevalence difference is modest; whether this association (if causally supported by MR) warrants routine asthma screening in ADHD patients cannot be determined without the full effect estimates and consideration of the clinical context. __LIMITATIONS__ Abstract was truncated; primary outcome statistics (OR, CI, p-value) from the observational arm and all pooled MR statistics are not available for critical appraisal.,Cross-sectional observational design precludes temporal inference; cannot establish whether ADHD preceded asthma onset.,ADHD and asthma ascertainment relied on self- or parent/guardian proxy-reported physician diagnoses without objective confirmation (e.g., neuropsychological testing, spirometry), introducing potential misclassification bias.,NHANES data from 1999–2004 are over two decades old; diagnostic rates, treatment exposures (e.g., bronchodilators, stimulants), and demographic composition have changed substantially.,Two-sample MR assumptions (relevance, independence, exclusion restriction) may be violated if ADHD-associated SNPs influence asthma through pleiotropic pathways unrelated to ADHD (e.g., immune dysregulation, inflammatory pathways).,The number and strength of SNPs used as instruments are not reported, limiting assessment of instrument validity (F-statistics not reported).,Heterogeneity across 5 GWAS datasets was addressed with random-effects IVW but the degree of heterogeneity (I²) is not reported in the abstract.,MR estimates reflect population-average lifetime ADHD genetic liability; they cannot distinguish developmental timing, ADHD subtype, or treatment effects on asthma risk.,Potential for sample overlap between ADHD GWAS (PGC) instruments and asthma outcome GWAS datasets is not addressed in the abstract, which could bias MR estimates toward the null or toward observational confounding.,Funding source is not reported, precluding assessment of conflict-of-interest risk.,Generalizability is limited by the U.S.-only observational sample; MR GWAS datasets are not characterized by ancestry composition in the abstract. __BOTTOM_LINE_CATEGORY__ low_certainty __BOTTOM_LINE__ This hybrid study reports an observational association between ADHD and approximately 1.7-fold higher asthma prevalence in U.S. children and adolescents (10.2% vs. 5.9%), and applies a meta-analysis of Mendelian randomization studies to explore whether this association is genetically mediated. However, the truncated abstract does not permit independent verification of the primary OR, MR pooled estimate, confidence intervals, or heterogeneity statistics, and critical MR validity assumptions (pleiotropy, sample overlap) cannot be assessed. Clinicians should be aware of the well-established observational comorbidity between ADHD and asthma in pediatric patients, but practice changes regarding systematic asthma screening in ADHD—or vice versa—are not supported by this evidence alone pending full publication review.