Commentary: Cat exposure and asthma outcomes in a cohort of children with asthma and allergy
Year: 2026
Type: editorial
Abstract: Putri et al. [1] reported that among 30,277 Swedish children with established asthma and airway allergy, parental cat ownership was not associated with asthma exacerbation, medication-defined severity, asthma control, or lung function. The study is unusually comprehensive: validated diagnostic algorithms, a national register of cats, five outcome dimensions, and a probabilistic bias analysis for exposure misclassification. Its central message -that blanket advice to avoid cats may not be uniformly beneficial -is well supported and clinically important.We wish to develop an issue the authors raise and then, understandably, leave. Our central point is that their exposure variable, cat ownership, is a decision, not a dose: it reflects which families keep a cat rather than how much allergen a child receives. Once this is taken seriously, the behavior of families ceases to be background noise and becomes part of what the estimate is made of. We describe three consequences -selection into ownership, misclassification of exposure, and the analytic role of behavioral constructs -indicate the likely direction of bias, distinguish hypotheses from findings, and argue that the remedy is measurement rather than caution.Putri et al. note that residual confounding by indication cannot be excluded, since parents may avoid keeping a cat because of a child's allergy. The authors' data are consistent with this, with severe allergic rhinitis at baseline being less common among cat-exposed children (11.5% vs 14.2%), despite comparable initial asthma severity and parental atopy.The direction of this selection is not random. When physicians advise removal of a pet, most families decline. In a 1997 survey of 341 adults with diagnosed cat or dog allergy who had been advised specifically to stop sharing living quarters with their animal, 21.4% removed it or excluded it from the home; among a subset of 122 whose animal had since died, 70.5% had acquired another [2]. Though dated and single-sample, this estimate is consistent with recent reviews describing avoidance as frequently insufficient [3]. Families who relinquish a pet appear to be a minority and may represent a subgroup in whom allergic symptoms or medical advice have become sufficiently consequential to alter ownership. Cat-owning households may therefore represent a selected population, in which the presence of a cat has already proved acceptable, or at least tolerable, to the family. We hypothesize that this biases the association toward the null, by an amount register data cannot estimate.The relevant comparator is not "no cat exposure" but "no cat in the household," a substantially different construct. Children classified as living without a cat may nevertheless encounter substantial cat allergen through schools, childcare, relatives, and friends [3,4]. This was demonstrated in the same country by the senior author of the present study: among more than 400 asthmatic, cat-sensitized children with no cat at home, those attending classes where more than 18% of classmates owned cats showed decreased peak expiratory flow, more symptom days, and greater medication use once term began [5]. If unrelated to outcome, this misclassification would attenuate the estimate toward the null.A related issue concerns how ownership was ascertained. The National Cat Register was established in 2023, and the authors appropriately model incomplete ascertainment in their probabilistic bias analysis. Nevertheless, the cat-exposed group was markedly enriched for Nordic-born parents (mothers 92.2% vs 78.9%; fathers 89.9% vs 77.7%), illustrating that cat ownership is socially patterned. This does not demonstrate differential misclassification, but it raises the possibility that completeness of registration or linkage varies with sociodemographic or behavioral characteristics. Were these also related to asthma outcomes, misclassification could be differential, which can bias estimates in either direction [6] -a possibility that a bias analysis assuming non-differential ascertainment does not address. We raise this not to dispute the null finding, which was consistent across all five outcomes, but because the assumption is substantive and merits explicit interrogation.The authors list "cat-keeping practice" and "cultural attitudes toward cats" as limits on generalizability. These are measurable, but their analytic use depends on their causal position (Table 1).Attachment to the animal determines ownership. It is a confounder only if it also affects asthma independently of cat exposure; otherwise it characterizes selection. Validated instruments include the Lexington Attachment to Pets Scale [7] and two brief versions [8,9].Cat-keeping and avoidance practices lie between ownership and allergen dose, so adjusting for them would overadjust; recorded by structured inventory and environmental Fel d 1 [10], they can instead grade exposure by dose. Symptom attribution [11] may itself reflect the child's asthma and is not a simple covariate. This Research Topic notes that avoidance of animal allergens is in most cases not possible because of the emotional impact. Putri et al. provide unusually thorough evidence that, for cats at least, this concern may be less consequential for asthma outcomes than has been feared [1].We propose that the field stop treating the human-animal bond as an unmeasured limitation and begin recording it. A brief validated scale, administered alongside sensitization testing, would make selection into ownership measurable rather than assumed, and would allow the field to ask a question that ownership data alone cannot answer: not simply whether cat-owning families differ in outcome, but whether any difference reflects allergen dose or selection into ownership -a question for explicit mediation analysis rather than covariate adjustment.
Source: Frontiers in Allergy
Institutions: University of Zurich, University Hospital Zurich, CK-CARE
Language: English
Subfield: Pulmonary and Respiratory Medicine
Field: Medicine
Domain: Health Sciences
Open Access status: gold
Estimated APC paid: $3,077
OpenAlex ID: W7222519713