There has been a marked increase in the prevalence of atopy-related diseases during the last 40 years (1, 2). There has also been a documented increase in specific sensitization to pets (2). There seem to be geographic differences; for example, children in Sundsvall, Sweden, have a higher prevalence of sensitization to pets than children in Konin, Poland, although it was as common to keep dogs in Poland as in Sundsvall, Sweden (3). The differences in the prevalence of hay fever and allergic sensitization between the former East Germany and West Germany have decreased without change in exposure (4). It is not clear what mechanisms are behind the increase in prevalence of atopic diseases, including sensitization to common airborne allergens. Exposure to environmental tobacco smoke, the length of breast-feeding, damp houses, and exposure to allergens during the first years of life have been claimed to be important causes of the increased prevalence (5–10). A lower prevalence of atopic diseases and specific allergy, including allergy to pets, has been reported among children living on farms and children in families with an anthroposophic lifestyle (11–13). It has been debated whether a low-dose allergen exposure would increase the risk of sensitization while a higher dose of the same allergen would promote development of tolerance. All in all, it is quite unclear to what extent pet exposure during the first years of life is important for sensitization or disease development later in life.

Recently, a panel of 13 Swedish pediatric allergologists were asked to give their views on whether child exposure to pets should be avoided until 2 years of age in families with different risk backgrounds, and they were also asked to give scientific evidence for their views. There was no consensus for any of the risk categories. Seventeen references were given as support for their positions, of which two were editorials, 14 were original articles, and only 10 dealt with the question of exposure to pets and sensitization. Of these, seven were transectional studies with a retrospective design, and several of them were not designed to answer the question of exposure to pet animals. The results of the different studies varied considerably. Some of the most interesting studies were those by Burr et al. (14) and Wahn et al (15), since they both included prospective cohorts. The study by Burr et al. was not primarily designed to answer the question of exposure to pets and the risk of sensitization. They found a negative relationship between pet exposure and sensitization. In this study, there was a lack of description and details of exposure and outcome regarding specific sensitization. Furthermore, there was a very small group of sensitized children in the study as a whole. Wahn and coworkers performed a study primarily designed to answer the question about early exposure to pets and sensitization. Exposure to cat allergen during the first years of life increased the risk of specific cat sensitization. Recently, the same authors have not been able to show any relationship between exposure to cat allergen at 6 and 18 months of age and asthma at 7 years (16). Selection before birth, lack of precision of exposure, and low allergen exposure may have influenced the results. However, the authors obviously restricted themselves to calculation of the odds ratios for exposure to cat allergen at the age of 6 months and asthma at 7 years of age, and no analysis was done of the rather clear dose-response relation for early exposure to cat allergen and health effect.

Lindfors et al. (17) measured cat allergen in house dust in acute asthma among children and found a significant relationship between exposure to cat allergen and sensitization. Other studies (18, 19) in the European Community Health Respiratory Health Survey found a negative relationship between exposure to cat or dog during childhood and sensitization measured at the adult age.

It is difficult to perform strictly prospective intervention studies randomized with respect to pet exposure. Furthermore, there is always a risk of methodological problems in epidemiologic studies. The basic principles are to collect information on exposure that is truly representative for the test subjects. It is equally important to use preset, preferably test-based criteria for outcome, since there is a risk of denial of symptoms that could be related to exposure; in this case, keeping a pet. We urgently need high-quality studies regarding these important questions. There are not enough data available today to give firm, evidence-based advice regarding exposure to pets during children's first years of life and prevention of allergy. The study by Nafstad et al. (20) in the present issue of Allergy constitutes a good attempt to add important new information, and also illustrates some methodological considerations. This study is a population-based study in a cohort of 2531 children born in Oslo, Norway, and followed from birth to the age of 4 years. The exposure was defined as keeping dogs or cats at birth. The exposure estimate was validated in a subsample of 251 patients and controls, and was found to give a valid estimate of exposure at home. Follow-up was conducted by a parental self-administered questionnaire at the age of 4 years. The outcomes used were bronchial obstruction at 0–2 years of age, and current asthma and current rhinitis at the age of 4 years. Data were analyzed by logistic regression adjusted for potentially important characteristics such as parental atopy, the child's sex, birth weight, birth order, mother’s age at delivery, maternal education, family income per year, environmental tobacco smoke exposure, length of breast-feeding, and episodes of low respiratory tract infections during the first year of life. These factors have all been reported to have an influence on the risk of atopy. Several of them also seem to have a relationship to exposure to pet allergen in the present study. Standardizing for these factors introduces the risk that effects from true differences in pet exposure might be erroneously attributed to other known risk factors related to the pet exposure of interest. However, the adjusted odds ratios were very similar to the crude ones, and the authors claimed that no convincing confounding was found. Another problem in the present study was the risk of prebirth selection; that is, the groups with pets at home at birth were not comparable to those not having pets at home at birth in the risk of having an allergic child due to factors they were aware of, but that were not measured in the study. Of course, it is also possible that the negative associations could indicate a protective effect of early-life exposure to pets. However, it would be difficult to advise parents to keep pets to avoid sensitization since this would increase exposure for those already sensitized in the same family, and when does one decide to change from exposure to no exposure when allergy has developed? It is still an open question whether early-life exposure to pets is good or bad for induction of atopy-related diseases, since several factors, such as design and methodological problems (prebirth selection with regard to keeping pets) might explain the divergent results. More good studies are needed.


  1. Top of page
  2. References