Predicting persistence of wheezing: one algorithm does not fit all

From the authors:

Scott et al. 1 have commented, in the light of outcomes of their “Isle of Wight Birth Cohort study” (IoWBC) 2, 3, on our findings on the risk factors for asthma in the Multicentre Allergy Study (MAS) birth cohort study 4. We agree that the two studies lead to similar conclusions in regard to early origins of persistent wheezing, heterogeneity of childhood wheezing and the relevance of an atopic background. In addition, we followed their suggestion to build a prognostic score in accordance with the IoWBC 3 and to test its performance in the MAS database. Here we report the results of our analysis.

The IWBC prognostic algorithm comprises four risk factors: a family history of asthma; recurrent chest infections in infancy; absence of nasal symptoms at age 1 yr; and atopic sensitisation at age 4 yrs 3. To make the data of the two cohorts comparable, we reduced our longitudinal study with follow-ups every year to the four time points of the Isle of Wight Study: age 1, 2, 4 and 10 yrs. Due to missing values, this reduction resulted in a fairly smaller population sample (n = 765); therefore we increased the older age group by 8 and 9 yr olds and obtained a proportion of examined children similar to the one obtained in the IoWBC. The parameters we used were quite similar, although not identical, to those used in the IoWBC study (table 1⇓). The frequency distribution of the risk score in the MAS cohort was similar to the one observed in the IoWBC; however, only three children with early wheezing had a risk score of 4 (table 2⇓). A cut-off to predict persistent wheeze in the MAS cohort was therefore placed at a risk score of ≥3. The positive predictive value (PPV) we obtained with this cut-off was 0.65 (95% CI 0.44–0.82) (table 3⇓). Interestingly, when adopting a cut-off of 3, the PPV in the IoWBC study was 0.68, i.e. almost the same as the one obtained in the MAS cohort (table 3⇓) 3.

On this basis we may conclude, as proposed by Scott et al. 1, that the IoWBC score can be a starting point in developing a clinically relevant predictive tool. Preventive measures should not be advised to children whose wheezing is likely to undergo spontaneous remission; therefore, a high PPV is desirable. On this basis, we feel that an algorithm based on early wheezing, sensitisation to mites and elevated allergen exposure, might be also clinically useful because it resulted in a better prediction in the MAS cohort (PPV 0.83). On the other hand, a third algorithm, based on the presence of atopic eczema, immunoglobulin (Ig)E sensitisation to food allergens and specific polymorphisms of the filaggrin gene, was able to predict another subset of asthmatic children of the MAS cohort with a PPV and equal to 1.00 (95% CI 65–100) 5. Given the great heterogeneity of mechanisms and risk factors for wheezing disorders, multiple algorithms are likely necessary to predict, with enough confidence, persistence of wheezing in the children with early wheezing.

A second point raised by Scott et al. 1 concerns the role of tobacco smoke and the genetic background in the determination of wheezing persisting up to puberty. The role of smoke and genetic predisposition in the MAS cohort has been recently published 6. This study demonstrated that among children with two allergic parents, a mother who smoked regularly significantly increased the odds for allergic sensitisation (adjusted OR 4.8, 95% CI 1.3–18.2) and wheezing (adjusted OR 5.7, 95% CI 1.7–19.0) in her child compared with children who were never exposed. By contrast, maternal smoking had no effects in children without allergic parents. The results of the MAS cohort study therefore agree also in this respect with those obtained by IoWBC study and support the existence of gene-by-environment interaction (in our example “atopic heredity-by-maternal smoke” interaction) in the inception of asthma.

Similarities in the IoWBC and the MAS studies show that our research toward understanding the origins of asthma, by following the footsteps left by children developing persistent wheezing, is on the right track. We are probably starting to make good use of another Sherlock Holmes' famous aphorism: “In solving a problem of this sort, the grand thing is to be able to reason backward” 7.

Statement of interest

None declared