Generic Respiratory Symptoms and Branded Lung Diseases. Same Difference?

From time immemorial, the learned have known that most things in the universe exist along a continuum. Lesser mortals and clinicians must contend with deviations from the normal arc, and consensus thresholds for normality and the diagnosis of disease. In this issue of the Journal, Kalhan and colleagues (pp. 1616–1624) blur the boundaries of normal health and lung disease (1). There is a growing awareness that lung function later in life is substantially influenced by early life exposures, starting almost from the time of conception (2). Airway branching is complete in late pregnancy, and the branching pattern can significantly influence the likelihood of lung disease later in life (3). Airway caliber and function are influenced by early life exposures, and in utero exposures to environmental pollution, cigarette smoke, toxins, and stress have been well documented to result in airway hyperresponsiveness, immunological impairments, and structural changes in the airways. Recent studies suggest that alveolar growth continues well into adulthood, and lung function appears to peak in most individuals at approximately 25 years of age. A combination of the peak achieved and the decline from this peak determines the overall deviation from the arc of optimal lung health.

Multiple cohort studies have documented several early-life risk factors with substantial effect sizes for lung function impairment and chronic obstructive pulmonary disease. These include in utero exposure to maternal smoking (3.4-fold for one pack of cigarettes a day) (4), parental asthma (2.5-fold), and childhood asthma (6- to 32-fold) (5, 6). Once peak lung function (optimal or suboptimal) is attained, established risk factors for a decline in lung function over and above the age-associated decline are few, and include cigarette smoking and environmental pollution. What drives the subclinical loss of lung function in some individuals in the absence of overt exposures is of great interest for investigators seeking to develop preventive strategies aimed at preserving lung health. Kalhan and colleagues hypothesized that persistent respiratory symptoms early in life, even in the setting of normal lung function, are associated with an accelerated decline in lung function (1). To test this hypothesis, they analyzed data from the CARDIA (Coronary Artery Risk Development in Young Adults) study, in which participants aged 18–30 years at enrollment underwent spirometry testing at baseline and every 5 years, and respiratory symptoms were ascertained annually for approximately 30 years. Computed tomography (CT) was performed at Year 25. They defined persistent respiratory symptoms by their presence at both baseline and Year 2. They found that the persistent presence of any respiratory symptom was associated with an additional 2.7 ml/year decline in FEV₁ and a 2.2 ml/year decline in FVC, as well as 1.6-fold greater odds of developing airflow obstruction, and 1.4-fold higher odds of showing restrictive spirometry at Year 30. They also reported a differential risk, in that cough-related symptoms were associated with greater risk of airflow obstruction and CT emphysema, and that dyspnea was associated with spirometry suggestive of restriction. Although the greater decline in lung function in symptomatic individuals compared with those without symptoms has previously been documented (7, 8), the current study reports novel findings of accelerated lung function decline that are different for FEV₁ and FVC in cough- and dyspnea-predominant phenotypes, respectively. In contrast to multiple other studies, this study also assessed changes in airflow obstruction by estimating the decline in FEV₁/FVC.

Kalhan and colleagues raise important questions as to the significance of respiratory symptoms in the absence of a known chronic disease. Presumably, these symptoms did not represent major respiratory illnesses (e.g., pneumonia) that could result in significant structural damage. What, then, could be the mechanisms for the accelerated decline in the absence of overt persistent exposure? Episodes of bronchitis or mild parenchymal infections can result in structural damage with consequent mechanical effects (9). Airway inflammation can result in bronchiectasis or more subtle mucus clearance and hypersecretion issues (10). Mild respiratory infections can result in alterations in the lung microbiome, and the reduced complexity of the microbiome can result in persistent inflammation (11). Could atopy explain some of these persistent symptoms, given that atopy is associated with an accelerated decline in lung function (12)? Forty percent of the participants had respiratory symptoms at baseline and at Year 30. Misclassification bias, resulting in some cases of asthma being missed, might explain some of the findings. In individual patients, however, it is equally likely that mild respiratory infections result in a decrease in lung function that does not return to baseline, and these individuals then suffer an age-related decline. Exacerbation-like events in the setting of “normal” lung function are associated with an accelerated rate of lung function decline (13). Exacerbations of established disease are associated with a drop in lung function that does not return to baseline in a substantial proportion of patients even 3 months after an exacerbation, and in some patients, it is possible that it may not return to baseline at all (14). Further, exacerbations with prolonged recovery are associated with a faster lung function decline (14). It is also possible that these persistent symptoms are a manifestation of the same early-life exposures that are risk factors for adult-onset chronic lung disease (15).

Several limitations of the study preclude definitive answers. For example, it is not known how chronic these persistent symptoms were, as they were ascertained at two time points with no information on symptom burden in the interim. In addition, the primary outcome of FEV₁ changes was treated as an annualized change between Years 5 and 30. This raises concerns as to whether the participants actually experienced an accelerated decline or simply a stepwise drop in lung function with acute respiratory events followed by a period of age-related decline. The authors tested generalized estimating equation models to account for repeated measures and showed that at a population level, persistent symptoms were associated with decline beyond that anticipated for aging, but the spirometry intervals were at least 5 years apart. The absolute decline in FEV₁, although statistically significant, is of unclear clinical significance. Over 25 years, an excess decline of 2.7 ml/year translates into an absolute decrease of 68 ml. The authors address this partially by demonstrating a greater risk for obstructive and restrictive spirometry, as well as parenchymal disease on CT, but the magnitude of decline is not commensurate with the occurrence of disease. Smoking status was ascertained at baseline, but the cumulative pack-years of smoking burden were not accounted for in the assessment of the odds of incident disease. Indeed, a substantial proportion of the subjects who reported symptoms at baseline were active smokers. Approximately 29% of the participants were lost to follow-up at Year 30, and those lost to follow-up had greater respiratory symptoms at baseline.

Despite these limitations, this study is a strong reminder that health is more than the mere absence of disease, and that we should not ignore persistent symptoms in early life even in the presence of “normal” lung function. Unfortunately, clinicians typically pay scant attention to early life exposures (16). As previously noted by Keyes, health and illness are not opposite ends of the same continuum, and measures of health correlate modestly with measures of disease (17). The absence of disease does not always equate to healthy lungs, and a deviation from normal does not always equate to lung disease. Perhaps health and disease exist on parallel continuums. Figure 1 shows a two-continuum model of lung health and lung disease. At times, we may be ill-served by thresholds and labels, and should change our perspective to evaluating patients on a continuum to enable better lung health practices. This is akin to the continuum of coronary atherosclerosis that has not been labeled as a significant coronary artery disease but is still important and modifiable. Preserving lung health involves a combination of a number of existing interventions, such as avoiding certain environmental exposures and cigarette smoke, early and adequate treatment of atopy and asthma, and vaccination against seasonal influenza, as well as the development of novel interventions. It is easy to ascertain respiratory symptoms in clinical practice, and such symptoms should be considered risk factors for the loss of lung health and for future lung disease.

Figure 1.

Two-continuum schema for lung health and lung disease. Individuals can be on different points along the spectrum of lung disease and lung health; the two are not on a single continuum. At any level of lung health that is not optimal, individuals do not function at optimal capacity and may have symptoms or signs of lung disease even if they do not fit a predefined label.