A new study published in Clinical and Experimental Allergy found that there is scant evidence that allergic diseases cause mental disorders or vice versa.
That conclusion may settle a long-simmering question arising from high rates of comorbidity between the two sets of conditions.
The study was motivated by “the great as well as increasing burden of disease posed by both sets of disorders,” said lead author Ashley Budu-Aggrey, PhD, senior research associate at Bristol Medical School, Bristol, United Kingdom. She and her co-authors sought to find out whether atopic dermatitis (eczema), hay fever, or asthma is causally related — in either direction — to mental disorders such as anxiety, depression, neuroticism, bipolar disorder, and schizophrenia.
The authors’ ability to infer or refute causation hinged on their use of an epidemiologic technique known as Medelian randomization. The technique is used when “it is impractical or unethical to perform a randomized controlled trial,” said Budu-Aggrey in an interview with Medscape Medical News.
Mendelian randomization utilizes genetic information, which in this case came from genome-wide association studies, to decide on matters of causality.
Before using Mendelian randomization, the authors first confirmed the observational associations between the two sets of disorders. They relied on the UK Biobank, a database with information on hundreds of thousands of people. Biobank is a larger database than was used in previous studies.
The strongest observational association that was uncovered using Biobank’s data was between atopic dermatitis and anxiety. Atopic dermatitis increased the odds of developing anxiety by 63%. Another robust association was between asthma and bipolar disorder. Asthma increased the risk of developing bipolar disorder by 52%. Were these associations causal?
Mendelian randomization using gene variant data could resolve the question. The technique is an epidemiologic cousin to a randomized clinical trial. Its preeminent advantages are that it is free of confounding factors and of reverse causation.
“The authors took genetic variations that are associated with increased risk of allergic disease and tested [by Mendelian randomization] whether those with a higher genetic susceptibility to allergic disase also have a higher risk of mental health disorders,” wrote Stephen Burgess, PhD, in an email to Medscape. Burgess is medical research council investigator, Biostatistics Unit, University of Cambridge,Cambridge, United Kingdom.
The analyses using Mendelian randomization could not replicate the observational associations from Biobank. The investigators could not infer causality. “The Mendelian randomization analyses suggest that allergic disease is not the causal risk factor in this case,” wrote Burgess, who was not involved in the study.
The clinical implications of the absence of causality were immediately obvious to the authors. Steps to prevent the onset of allergic diseases are not likely to prevent the onset of mental health conditions, and vice versa, they write.
Still, academics never say never. First, the authors speculated that interventions aimed at improving — as opposed to preventing — allergic disorders may yet improve mental health, and vice versa.
Second, they found a weak but tantalizing causal relationship between hay fever and bipolar disorder in the Mendelian randomization analyses, despite finding no association between the two in the Biobank dataset. Evidence of a weak causal effect of hay fever genetic risk upon bipolar disorder would suggest that preventing one may prevent the other.
Study limitations include the fact that the genetic studies used for Mendelian randomization analyses were stronger for allergic diseases than for mental disorders. In addition, the Biobank dataset is largely self-reported, which could lead to the misreporting of diagnoses. Also, both the genetic studies and the Biobank dataset generally are limited to White Europeans.
After concluding that there was no causality, a question remained: why is there so much comorbidity in observational studies? Budu-Aggrey attributed this to confounders. “Socioeconomic status, level of education, and smoking are examples of confounders linked to both mental disorders and allergic diseases,” she said.
The study provides yet another example of the time-honored axiom: correlation is not causation.
The study was funded by the UK Medical Research Council, the European Research Council, and the National Institute for Health Research Bristol Biomedical Research Center, among others. Budu-Aggrey and Burgess report no relevant financial relationships.
Clin Exp Allergy. Published online October 6, 2021. Full text
Miriam Davis, PhD, is an award-winning freelance medical writer who has written more than 1000 articles over a 25-year career.
Content Source: https://www.medscape.com/viewarticle/960757?src=rss