Longitudinal relationship between physical activity and lung health in patients with cystic fibrosis

Introduction

Advances in the treatment of cystic fibrosis (CF) have resulted in an increase in the median survival age for CF patients; however, the disease continues to be life limiting [1]. Nutritional status, pulmonary function, genotype, age at diagnosis and infection with Pseudomonas aeruginosa have been identified as predictors of mortality in patients with CF [2, 3]. A growing body of research has demonstrated that exercise training and physical activity contribute positively to outcome in patients with CF, with improvements in aerobic capacity, activity level, quality of life, weight gain, lung function and leg strength [4–6] and an association with a reduced rate of decline in lung function [7, 8].

While physical activity has well-documented benefits for healthy children [9], additional advantages for patients with CF include enhanced airway clearance [10] and improved ion channel function, possibly leading to better mucus hydration and enhanced mucus clearance [11]. Moreover, in patients with CF with severe lung disease, physical activity has been related to aerobic capacity [12], suggesting a direct relationship between aerobic capacity and survival.

Habitual physical activity (HPA) refers to activity that is incorporated into daily life, is less structured than traditional exercise training and can encompass a wide range of intensity levels. Thus far, reduced HPA has been reported to be related to lung function decline in females with CF over a short observation period [8] and adults with CF have been shown to accumulate less HPA than their non-CF peers [13]. While exercise training has been recommended for inclusion into CF routine therapy [14], issues such as the burden of disease [9] and inadequate adherence [15] have made it a challenge to incorporate into a treatment programme. Knowing that focusing patients and parents on the importance of regular activity could potentially change habitual levels over time, we sought to study prospectively the long-term relationship between changes in HPA levels and lung function in patients with CF.

Results

212 patients with CF (aged 7–17 years at baseline, 108 (50.9%) females) were recruited over a 9-year period. Baseline characteristics for the study population are shown in table 1. Males had significantly higher pulmonary function, activity levels and V′_O2peak values. Mean±sd HPA for the overall group at baseline was 5.47±2.78 h·day⁻¹, with males engaging in greater activity than females (p=0.028).

Length of follow-up is presented in table 2. The median (range) length of follow-up was 5.21 (0–9.7) years. A median (range) of nine (1–27) and 10 (1–29) measures were collected for HPA and FEV₁, respectively, over the study period.

Of 40 patients from Montreal Children’s Hospital, 16 (40%) moved to adult care during the study period; these patients were not followed in the study after the age of 18 years. Of 172 patients from the Hospital for Sick Children, 86 (50%) reached 18 years of age before the end of the study, of which 62 (72%) were followed after the age of 18 years.

As the majority of patients (n=166) were recruited in the first year, we compared the baseline characteristics of recruited patients to the overall eligible clinic population in the major recruitment centre (the Hospital for Sick Children). Patients participating in the study were younger (11.8±2.9 years versus 13.4±3.3 years, p<0.0005) and had higher lung function (FEV₁ 85.6±17.4% pred versus 67.2±23.0% pred, p<0.0001) than patients not participating in the study (n=67). There was no significant difference in baseline FEV₁ (p=0.5048) or V′_O2max (p=0.7126) between patients from the two paediatric centres (data not shown).

Adjusting for the potential confounders of sex, baseline age and FEV₁, mucoid P. aeruginosa and CFRD, overall FEV₁ decreased at a mean±sd rate of 1.63±0.08% per year (p<0.0001) and HPA increased at a mean±sd rate of 0.28±0.03 h·day⁻¹ per year (p<0.0001) over the study period. There was a significant positive correlation between rates of change of activity level and change in FEV₁ decline, indicating that an increase in activity was associated with a slower rate of decline in lung function over the study period (r=0.19, p<0.007).

Participants were divided into high (above the mean rate of change of activity) and low (below the mean rate of change of activity) groups. table 3 indicates that all evaluated potential confounder variables were evenly distributed between the two groups. The high group had a rate of increase in HPA of 0.59 h·day⁻¹ per year, while the low group had a rate of decline of activity of 0.15 h·day⁻¹ per year over the study period. Mixed model analysis results are presented in table 4 and indicate that FEV₁ was significantly associated with baseline FEV₁ (p=0.0001), CFRD (p=0.0452) and change in activity level over time, such that the rate of decline of FEV₁ was less steep for the high group (-1.39% pred per year) compared to the low group (-1.90% pred per year) (p=0.0001).

189 patients performed Stage I exercise testing, with a total of 493 tests included for analysis (range 1–7 tests per patient). Adjusting for baseline FEV₁, there was a positive relationship between FEV₁ (% pred) and V′_O2peak (p=0.0194) over the study period. Similarly, adjusting for baseline FEV₁, there was a positive relationship between FEV₁ (% pred) and WR_peak (% pred) (p=0.0004) over the study period. There was no significant association between V′_O2peak and HPA (p=0.7457).

To confirm whether an increase in HPA was associated with a slower decline in FEV₁, we performed the following additional analysis. Between T1 (baseline) and T2 (2.5 years) we classified the patients into either increasing or reducing HPA over time. We then used the data between T2 and T3 (6.6 yrs) to look at the rate of FEV₁ decline for the two groups. The results showed that the rate of decline in FEV₁ between T2 and T3 was significantly less steep for patients who incresed their HPA (-0.58% pred per year) compared to those with reduced HPA (-2.15% pred per year) (p=0.0231). Using the same method, we found no differences in change of V′_O2peak between the two groups (p=0.7590).

Defining pulmonary exacerbation as hospitalisation for respiratory symptoms requiring antibiotics [24], 57 (32.2%) of our study patients had at least one exacerbation with a median (range) of three (1–16) over the study period. Hospital admissions were infrequent and there was no significant association between V′_O2peak and hospital admissions per year (r= -0.07, p=0.35).

Discussion

In this prospective longitudinal study, after accounting for baseline characteristics known to affect clinical course, patients with CF with increasing activity levels had a reduced rate of decline of their FEV₁ compared to those that did not become more active over the study period. If the goal in the treatment of CF is to preserve lung function for as long as possible thus potentially extending survival [25], these results would suggest that enhancing physical activity should be an integral part of the management of the disease.

We found a significant overall increase in physical activity of 0.28 h·day⁻¹ per year over the study period, which translates into an increase of ∼17 min·day⁻¹. This contrasts with earlier reports of declining levels of HPA for the majority of youth over a 5-year period of growth from childhood to adolescence, regardless of sex or weight status [26] and the trend of decreasing activity levels of children and youth in general [27]. It is possible that monitoring a patient’s activity level while participating in a study may be a sufficient stimulus to effectively maintain what would otherwise be a predictable pattern of decreasing activity level with age [28]. This highlights a valuable opportunity to have a positive impact on lung function decline, by promoting a physically active lifestyle throughout childhood, to encourage the maintenance of activity and enhance the carry-over into adulthood [29].

Previous reports of activity employing the HAES have indicated that children with CF engage in similar total amounts of HPA to their healthy peers [30]. Current values for HPA are similar to those reported in a small number of youths with CF [22] and somewhat lower than those reported by Ruf et al. [31] of 7.4 h·day⁻¹ (females) and 6.0 h·day⁻¹ (males) for patients with CF (aged 12–41 years). The latter may be explained by their inclusion of weekend activity, which was excluded in our analysis due to greater variability in preliminary analyses [8].

We investigated the relationship between lung function and HPA, as we were interested in documenting a “lifestyle” variable in patients with CF, one that did not require an “exercise prescription”, which is generally difficult to maintain once the research study resources have ended [4, 32]. HPA refers to the level of activity that is inherent in daily life, such as free play periods, climbing stairs, sporting activities, activities of daily living and activities of lower intensity such as walking. Therefore, changes in HPA represent a lifestyle modification that should result in longer lasting benefits. Enhancing HPA presents an interesting, widely feasible possibility for CF patients to improve their health and quality of life.

While multiple studies have highlighted factors that negatively affect the rate of decline of lung function, there are few that illustrate positive factors [25, 33]. Likewise, most interventions in CF have been assessed regarding their short-term efficacy in improving lung function rather than their long-term effect on lung function decline, as the latter requires a much longer observation period to demonstrate an effect. This is the first prospective, longitudinal study to illustrate that increasing habitual physical activity has a positive effect in attenuating the rate of decline of lung function in patients with CF, possibly affecting survival.

There is evidence to suggest that physical activity affects lung function even in the absence of lung disease. Berntsen et al. [34] reported that among a cross-sectional group of 2537 healthy children (aged 9–10 years), FEV₁ was highest among those who were physically active four or more times per week, once adjusted for potential confounders (p=0.02). The results of this study would suggest that any intervention assessing pulmonary function decline as an outcome should adjust for activity level, as it is a significant factor in addition to the traditional potential confounders.

Our findings of a positive longitudinal relationship between the rate of decline of FEV₁ and both V′_O2peak and WR_peak agree with results of a 2-year follow-up of children with CF [35] and support earlier reports of a relationship between V′_O2peak and survival [36, 37]. Thus, it would suggest that both HPA and maximal aerobic fitness positively influence the outcome of patients with CF, reinforcing the association between physical activity and V′_O2peak in patients with CF [38].

There are some inherent limitations to this study. Participants enrolled in the study were younger and had a higher FEV₁ than the remaining eligible clinic patients not recruited that year. Self-selection by those patients with mild disease is not uncommon in CF research studies [15], and in this case they would be healthier and less likely to have a chest exacerbation, and therefore eligible to participate in the activity study. Conversely, the included population represented the majority of patients followed in our CF centre and demonstrating a positive effect of physical activity on lung function decline in patients having less severe lung function abnormalities highlights the opportunity to utilise enhancing physical activity as an early intervention strategy.

Given that HPA is an important adjunct to treatment for patients with CF, future research is needed to determine the most efficacious strategies to help patients increase their HPA and build a lifestyle conducive to lifelong involvement in physical activity.

In summary, we demonstrate that patients with CF with increasing activity levels have a reduced FEV₁ decline compared with those who did not become more active. This supports the hypothesis that even low-intensity activity, such as that reported by our adolescents, is able to preserve lung function. This study highlights the importance of incorporating, facilitating and encouraging long-term habitual physical activity in the clinical management of CF.