Assessing the exacerbations risk of influenza-associated chronic occupational asthma

Abstract

The purpose of this article was to conduct a risk-based study based on a linkage of experimental human influenza infections and fluctuation analysis of airway function to assess whether influenza viral infection was risk factor for exacerbations of chronic occupational asthma. Here we provided a comprehensive probabilistic analysis aimed at quantifying influenza-associated exacerbations risk for occupational asthmatics, based on a combination of published distributions of viral shedding and symptoms scores and lung respiratory system properties characterized by long-range peak expiratory flow (PEF) dynamics. Using a coupled detrended fluctuation analysis-experimental human influenza approach, we estimated the conditional probability of moderate or severe lung airway obstruction and hence the exacerbations risk of influenza-associated occupational asthma in individuals. The long-range correlation exponent (alpha) was used as a predictor of future exacerbations risk of influenza-associated asthma. For our illustrative distribution of PEF fluctuations and influenza-induced asthma exacerbations risk relations, we found that the probability of exacerbations risk can be limited to below 50% by keeping alpha to below 0.53. This study also found that limiting wheeze scores to 0.56 yields a 75% probability of influenza-associated asthma exacerbations risk and a limit of 0.34 yields a 50% probability that may give a representative estimate of the distribution of chronic respiratory system properties. This study implicates that influenza viral infection is an important risk factor for exacerbations of chronic occupational asthma.

Figure 1

Schematic representation of computational algorithm used in this study (see main text for detailed symbol meanings).

Figure 2

(A) Influenza viral shedding and normalized respiratory symptoms scores dynamics. (B) Reconstructed relationships between respiratory symptoms scores and viral shedding fitted by a three‐parameter Hill equation model with 95% confidence intervals.

Figure 3

(A) Time course of influenza viral shedding and decrease in% PEF for influenza‐induced asthma individuals. (B) Reconstructed models fitted by a three‐parameter Hill equation model with 95% confidence intervals describing the relationships between decreasing in% PEF and viral shedding.

Figure 4

(A) Conditional probability of P(RSS|PEF) that is optimal fitted by a Hill equation where RSS denotes respiratory symptoms scores. (B) The exceedence risk of respiratory symptoms scores in asthmatic and nonasthmatic individuals based on the representative probability density functions of PEF shown in (C) and (D), respectively. Two vertical lines represent 60% and 80% of predicted PEF values (the normal value is estimated to be 628.5 L min⁻¹) to delineate the severe and moderate regions based on GINA.^{( 2 )}

Figure 5

(A) The reproduced time series of PEF measurements in three representative scenarios I, II, and III representing different levels of lung function based on the published measurement data of six‐daily basis PEFs (B–D). (E) Plot of log F(n) vs. log n for three scenarios I, II, and III of PEF time series fitted by power law function. (F) Best fitted linear regression model describing the relationship between predicted PEF and correlation exponents (α).

Figure 6

(A, B) Conditional probability of P(exacerbation risk|α) fitted by a Hill model by which the influenza‐induced asthma exacerbations risk for scenarios I, II, and III can be estimated based on correlation exponent (α) in which the box and whisker represent the uncertainty in three scenarios. (C, D) Conditional probability of P(exacerbation risk|wheeze) by which the influenza‐induced asthma exacerbations risk can be predicted from normalized wheeze symptoms scores based on the relationships between wheeze symptoms scores and PEF (% predict).