Discussion
There were three primary findings of this prospective cohort study on predictive factors for asthma in young adults after hospitalization for viral wheezing episodes aged younger than 24 months. First, early life wheezing, an asthma diagnosis one year after the index hospital admission and physician-diagnosed asthma at median ages of 4.0, 7.2 and 12.3 years were consistently associated with asthma at a mean age of 18.8 years (n = 49). Second, asthma in parents and atopic dermatitis in infancy were associated with asthma in early adulthood. Third, elevated blood eosinophils on the index hospital admission for viral wheezing episodes aged < 24 months were associated with asthma in early adulthood. No such associations were found for other markers of eosinophilic activity or laboratory markers of atopy. As expected, asthma in early adulthood was associated with current allergy and atopic sensitization. When the analyses were repeated and adjusted for sex, age on admission, contact with household pets in infancy, early-life exposure to tobacco smoke and current daily smoking, asthma in parents marginally lost its significance. As we reported previously,17 RSV and especially rhinovirus infections, linked to the index wheezing episode requiring hospitalization, were associated with asthma in early adulthood compared to population controls.
In the present study, an asthma diagnosis at any of the four follow-up visits until a mean age of 12.3 years was associated with an increased risk of asthma at a mean age of 18.8 years. In a Swedish post-bronchiolitis cohort, the risk of asthma in later life (i.e. 27 years) was 10-fold higher than in population-based controls.7 In the same study, although an asthma diagnosis at the age of 5 or 10 years did not predict asthma in adulthood, an asthma diagnosis at the age of 18 years increased the risk of asthma at the age of 27 years significantly to 6.5-fold.7 In the present cohort study, the risk of asthma aged 18.8 years was increased most (22-fold) among those with asthma aged 12.3 years. This finding is in line with the current consensus on asthma evolution and age,23 whereby wheezing in infants and asthma in toddlers seem to improve at school age.6,7 The latter does not imply total recovery but rather symptom remission, and relapses are common after puberty and in early adulthood.6,7 Such relapses may remain undiagnosed because adolescents and young adults are reluctant to consult a doctor for asthma-related symptoms due to denying their symptoms and negative perceptions of the disease among their peers4
In previous birth cohort and post-bronchiolitis studies, asthma in family members, especially mothers, was the most common risk factor for asthma in later life,1,4,7 as found in the present cohort. In a meta-analysis, which included 18 studies from good-quality systematic reviews, maternal asthma was associated with a 3.2-fold risk and paternal asthma with a 2.6 fold risk of physician-diagnosed asthma aged 5–18 years.11 Prenatal environmental tobacco smoke exposure and premature births, particularly very preterm births, were other significant early-life risk factors for later asthma.11 Although some studies found that atopic dermatitis in infancy was associated with asthma in later life, this association was not constant.4,7
Tobacco smoke exposure during prenatal and early childhood periods are well-known risk factors for bronchiolitis or viral wheezing in early life and asthma and lung function deficiency even in adulthood.23 However, neither maternal smoking nor passive smoke exposure in early childhood was associated with asthma in young adulthood in the present study. This finding may be explained by early exposure to tobacco smoke increasing the likelihood of hospitalization for viral wheezing episodes at young age and decreasing the statistical power for asthma prediction in later years in the within-cohort analyses. Slightly more than half (51.0%) of the study group had exposure to tobacco smoke during infancy, and 48% of these individuals were current smokers at the time of last follow-up visit. The prevalence of smoking in our study group far exceeded that of the current prevalence of smoking (mean: 11%) among young Finnish adults.24
In algorithms aimed at predicting the risk of asthma in school-aged children, a high eosinophil count (i.e. an absolute count of ≥0.45 × 109/L) at preschool age has been linked to an elevated risk of asthma at later life.25 It should be noted that eosinophil sampling during an infection, as in the present study, is not optimal, as infections often trigger a decrease in eosinophils.26 Thus, the finding of a normal eosinophil count in such cases would actually be abnormal finding.26
The association of asthma with respiratory allergies is well known and is termed united airway disease.27 Previous Swedish and Finnish cohort studies investigated risk factors for asthma in 46 patients with a diagnosis of RSV-related bronchiolitis in childhood who were followed up at a mean age of 18 years8 and 82 patients with a diagnosis of bronchiolitis in childhood who were followed up at a mean age of 27 years, resepectively.9Both studies reported that current allergy, especially allergic rhinoconjunctivitis, was a significant risk factor for asthma in adulthood.8,9 In the present cohort, more than 75% of those with current asthma reported allergic rhinitis and/or were sensitized to airborne allergens.
RSV and especially rhinovirus infections, when associated with wheezing in early childhood, is known to be associated with an increased risk of asthma in later life.3,11,28 On the other hand, a Danish cohort study on 313 children found that the number of respiratory infections in the first years of life, independently from the causative viral agents, was associated with asthma at the age of 7 years.29 As described earlier, hospitalization in early childhood because of viral wheezing episodes increased the risk of asthma in adulthood compared to controls in this cohort, and rhinovirus increased the risk more than other viruses.17
The prevalence of asthma in our cohort was much higher (53%) than that found in other prospective post-bronchiolitis follow-ups until adulthood conducted in Finland and Sweden, which reported asthma in 18% and 39%, respectively, of cohort subjects aged 18–29 years.7-9There are no other, similar published prospective, longitudinal studies on patients hospitalized for viral wheezing episodes and followed up until adulthood. An increasing trend of asthma prevalence by time was clearly seen in these Finnish and Swedish studies; new studies reported a higher asthma prevalence than old studies. Thus, the prevalence of asthma in our study (53%) is in line with the current trend, although higher than expected since the allergy and asthma epidemics have decreased also in Nordic countries.30 In non-selected birth cohorts, the prevalence of asthma varied from 12% to 15% in longitudinal studies on adolescents and young adults aged 11–24 years.31 The prevalence of asthma among Finnish adults was 10.9% in a population-based survey based on data collected using postal questionnaires.30 The higher prevalence of asthma among this cohort population might be due to participation bias - individuals with asthma-related symptoms more likely to volunteer for follow-up visits than those without symptoms.
Our cohort study had some limitations. The most important limitation was the small number of participants, which resulted in underpowered analyses and a risk of type-2 errors. On the other hand, all the factors, such as previous asthma diagnoses at different ages through childhood, that showed statistically significant associations in the multivariate analyses in this small material, confirmed a real association with high asthma risk after early life viral wheezing episodes. Moreover, this cohort has been followed for more than 18 years, and therefore, dropouts are understandable. In terms of the strengths of the study, the cohort offers unique longitudinal data on the associations of viral wheezing episodes requiring hospitalization, early childhood wheezing and asthma at preschool age with asthma outcomes in adulthood.
In conclusion, this cohort study confirmed the results of previous studies that an asthma diagnosis at any age during childhood and an increased blood eosinophil count in early childhood are independent predictive factors for asthma in early adulthood.