The interstitial lung diseases (ILDs) are a disabling and diverse group of chronic lung conditions that have been broadly classified into four groups: ILD of known cause such as occupational or environmental exposures and/or collagen vascular disease; granulomatous ILD such as sarcoidosis; idiopathic interstitial pneumonias including idiopathic pulmonary fibrosis (IPF) and nonspecific interstitial pneumonia (NSIP); and other rare forms of ILD including lymphangioleiomyomatosis, pulmonary Langerhans’ cell histiocytosis/histiocytosis X, and eosinophilic pneumonia . Many ILDs are characterised by the development of irreversible and progressive interstitial fibrosis of the lung parenchyma  resulting in altered respiratory mechanics, impaired gas exchange, reduced exercise capacity and dyspnoea on exertion [3–5]. Skeletal muscle dysfunction and weakness may occur, leading to worsening exercise capacity and increasing symptoms [2, 6, 7]. Health-related quality of life (HRQoL) is frequently markedly reduced and those with the greatest exercise limitation report the worst quality of life . As disease progresses, severe hypoxemia and pulmonary hypertension may develop [9, 10], with patients often becoming dependent on supplemental oxygen.
The classification of the ILDs has been the subject of criticism, due to its failure to reflect the marked heterogeneity in clinical course within disease subgroups. Idiopathic pulmonary fibrosis (IPF), the best characterized of the ILDs, is largely a fatally progressive disease with a median survival of 3–5 years . The prognosis of NSIP is more variable, and, although a minority of patients may have an accelerated decline similar to IPF , survival is generally significantly longer than in IPF. Dust and connective tissue disease- related ILD may be associated with a better overall survival rate but can result in significant and progressive morbidity over many years [12, 13]. Prognosis of sarcoidosis is again variable and difficult to predict with stabilisation or improvement in some patients and the development of progressive pulmonary fibrosis in others [10, 14].
Few treatments have demonstrated improvements in either HRQoL or community functioning for any of the ILDs [15, 16]. In IPF, the most common and most lethal ILD, the options for pharmacological treatment are very limited . Therapies that can improve dyspnoea, fatigue, exercise capacity and quality of life are highly sought after in ILD . Exercise is one of the few treatments to show positive changes in functional capacity and symptoms. We have previously shown that exercise training could significantly improve exercise capacity and reduce dyspnoea and fatigue symptoms in patients with ILD of varying aetiology . Nishiyama et al. found similar positive effects from exercise training in patients with IPF only . Additionally, several observational studies evaluating the benefits of pulmonary rehabilitation, of which exercise training is an essential component, demonstrated statistically and clinically significant improvements in functional capacity, dyspnoea and HRQoL in patients with ILD of varying aetiology [19–23]. Despite these promising outcomes exercise training is not yet widely recommended for people with ILD. Only weak recommendations regarding exercise training are provided in the most recent clinical guidelines for the diagnosis and management of IPF  and ILD .
Uncertainty remains regarding the clinical relevance of exercise training across the entire range of ILDs. Patients with IPF appear to have smaller gains in functional capacity than those with ILD of other aetiology . This raises the possibility that some forms of ILD may respond to exercise training better than others. Common manifestations of ILD, such as exercise induced hypoxia and pulmonary hypertension [2, 9], may also affect the improvements that may be achieved. Hypoxaemia impairs maximal exercise performance  and pulmonary hypertension in ILD is associated with considerably reduced exercise capacity and greater exercise limitation [25–27]. In an uncontrolled study evaluating the relationship between response to exercise and disease aetiology and severity in forty-four subjects with ILD of varying aetiology, less severe lung function, less oxyhaemoglobin desaturation and less pulmonary hypertension were associated with greater improvement in functional capacity in patients with IPF . This relationship persisted at six months, suggesting that those with less advanced IPF may be able to achieve sustained benefits from exercise training. This relationship was not seen in subjects with other ILDs. It is therefore possible that the timing of exercise training may be important for patients with IPF, whereas patients with other forms of ILD may benefit regardless of disease severity.
In order for exercise training to be widely adopted in clinical practice, clinicians require more information regarding its role across the disease spectrum. The aims of this study are 1) to establish the impact of exercise training on ILDs of different aetiology and severity and 2) to identify whether there is an optimal timing for exercise training to achieve maximal benefit. We hypothesise that exercise training will be effective regardless of disease severity in patients with non-IPF related ILD, whereas in patients with IPF, the response to exercise training will be greatest in those with less severe disease.