Quality of Life and Fitness in Cystic Fibrosis
In this study we found that aerobic fitness and reported physical activity were directly and moderately associated with HRQoL scales cross-sectionally and to a lesser extent also longitudinally in a sample of 76 patients with CF and mild to moderate disease. Established determinants of HRQoL such as FEV1 and BMI correlated with fewer scales. Based on scarce data in the literature looking at the relation between HRQoL and aerobic fitness and physical activity in CF, these findings are most important and relevant from a clinical perspective. The novel aspect of including longitudinal associations strengthens the findings.
HRQoL in our study was similar to that reported for patients with mild to moderate disease severity. In contrast to the German and American validation studies of the CFQ-R, data of patients with an FEV1 below 35% predicted were not included in our analyses which reduced the range of FEV1 in our project (35–107% predicted) compared to the respective validation studies (Germany: 12–106% predicted; United States: 17–130% predicted). Possibly as a consequence of the smaller range in FEV1, surprisingly few significant associations were observed between HRQoL-scales and FEV1 in the current project. Nevertheless, the power of our study was sufficient to detect clear and consistent positive associations between physical fitness and multiple HRQoL-scales.
In contrast to our hypothesis, but in accordance to another comparable study in children with CF no significant associations were observed between objectively measured physical activity and any HRQoL-scale. For reported physical activity, only one significant direct correlation was detected with the scale 'Role limitations' cross-sectionally (Table 4A). In a previous cross-sectional study, physical activity was found to correlate quite strongly with Quality of Well-being reported by patients with CF. The discrepancy to our findings might reflect differences in the approach to determine HRQoL between the generic Quality of Well-being Scale and the multi-dimensional disease-specific CFQ-R. The Quality of Well-being Scale asks specifically whether or not the patient performs certain activities while the CFQ-R assesses the patient's feelings and difficulties while performing activities.
Changes in reported physical activity, however, were positively associated with changes in three scales of the CFQ-R in the current study, namely 'Role limitations', 'Feelings of embarrassment', and 'Weight problems'. This is important, as it shows that an increase in physical activity could increase HRQoL even after just 6 months. Other studies have also shown significant positive relationships between self-reported physical activity and HRQoL in healthy adults or patients with chronic obstructive pulmonary disease.
While reported physical activity showed some direct association with HRQoL, objectively measured activity did not. This finding is difficult to interpret. Possibly, specific aspects of our study design have obscured a significant relationship: Accelerometry was performed within a given period of three weeks before or after assessment of HRQoL, while anthropometry, pulmonary function assessment, exercise testing, and completion of activity and quality of life questionnaires were all done on the same day. It is possible that physical activity assessed by accelerometry over a few days did not mirror physical activity a few weeks earlier or later which may have been confounded by short-term alterations in disease state or weather conditions. Furthermore, the approach chosen to analyze accelerometer data including the selection of epoch times, definition of non-wear time or cutoffs for MVPA has been shown to influence the results and may also have contributed to the attenuated relation between physical activity and HRQoL.
Furthermore, due to incomplete accelerometry data, sample size for the respective analyses and thus statistical power was lower than for other variables. Further studies are required to explore the associations between objectively measured physical activity and HRQoL in CF.
The main finding of the present study was that physical fitness, mainly aerobic fitness, was positively associated with HRQoL cross-sectionally and longitudinally. This observation is in line with the positive association between changes in aerobic fitness and changes in Quality of Well-being reported by others and the direct associations between VO2peak and the HRQoL-scales 'physical functioning', 'emotional functioning', 'social functioning', and 'treatment burden' observed by Groenevelt et al.. Furthermore, the significant and positive correlations between indicators of aerobic fitness and the disease-specific HRQoL-scales were as strong and sometimes stronger than the associations of these scales with more traditional measures of disease severity such as age, FEV1 or body composition. This finding underlines the importance of a good physical fitness for HRQoL in CF especially in disease-specific scales, and therefore the importance of physical fitness assessments in patients with CF. In addition, our findings strengthen the construct validity of the CFQ-R.
After adjusting for age and gender, the only other cross-sectional study testing the relation between physical activity, physical fitness, and HRQoL scales did not find any significant correlation. This may be due to the much smaller sample size in their study group and the large age ranges as important confounders. Furthermore, the expression of the aerobic fitness per kg bodyweight rather than per predicted value may have "penalized" those who maintained body weight (leading to lower VO2peak per kg values) and "favorized" those who lost bodyweight (leading to higher VO2peak per kg values) – the opposite scenario from what would be expected from a clinical perspective. Indeed, when analyzing our data with a multiple regression model and adjusting for age, gender and nationality, direct relationships between VO2peak and 9 HRQoL-scales were observed (see Additional file 1).
At first sight it seems surprising that the HRQoL-scales 'Physical functioning' and 'Respiratory symptoms' were not responsive to changes in physical fitness. It is notable, however, that the actual data were taken from a training study and not from a longitudinal set of observational data. Thus, the patients included in this project were willing and felt generally capable of participating in regular exercise. This approach might have affected HRQoL perception related to changes in physical fitness. Moreover, several training studies have shown that physical conditioning may improve physical fitness while FEV1 remains relatively unaffected. Thus, although there was a significant direct correlation between peak oxygen uptake and the 'Respiratory symptom' scale at baseline, the subjective and objective respiratory situation in many patients might not have changed enough to be perceived positively despite improvements in aerobic fitness.
One limitation of the current project is the relatively small number of patients in the five treatment conditions that limits statistical power in detecting effects of the different treatment regimens on HRQoL. Nevertheless, supervised strength training three times per week over six months was associated with significant decreases in the HRQoL-scale 'Vitality', i.e. less vitality. It is possible that the supervised strength training might have been too strenuous to maintain or even improve 'Vitality' despite improvements in aerobic fitness and pulmonary function. A decrease in feelings of energy and an increase in perceived fatigue were observed in competitive swimmers progressively increasing their training volume, which supports our "overload" hypothesis. In contrast to our findings, others did not observe a negative effect of a presumably less strenuous partially supervised strength training program on Quality of Well-being which presumes that the negative effects on some HRQoL-scales in our study were rather a consequence of the intensity than of the strength training per se.
The discrepancies in time spent active between the subjective and objective assessments might be perceived as a further limitation of the study. However, this discrepancy is possibly related to the already mentioned different time frame that the two measurement tools covered and the way accelerometer data were analyzed that may at least in part have attenuated associations with HRQoL. In addition, reporting bias based on social desirability and social approval may have occurred.
Furthermore, as this study was a training study, it cannot be completely excluded that the interaction of the training supervisors with the participants may have affected some of the associations between the fitness, activity and HRQoL variables. Nevertheless, the fact that there was merely no difference in changes of HRQoL among groups suggests that this influence was minor. Finally, there is a possibility of shared variance amongst a number of the predictor variables in the regression model explaining HR-QoL that could be responsible for a reduced association between variables. Yet, this potential multicollinearity does not reduce the predictive power or reliability of the model as a whole but rather affects calculations regarding individual predictors.
Since physical conditioning programs in CF have been shown to improve pulmonary function, physical fitness and HRQoL in concert, patients should be encouraged to engage in regular at least moderate physical activities efficient to improve physical fitness even though we could only show beneficial effects of increased physical activity on some HRQoL-scales. To maintain or even improve 'Vitality', too strenuous training sessions without sufficient recovery periods such as unfamiliar strength training exercises three times per week need to be avoided in order to not compromise HRQoL.
Discussion
In this study we found that aerobic fitness and reported physical activity were directly and moderately associated with HRQoL scales cross-sectionally and to a lesser extent also longitudinally in a sample of 76 patients with CF and mild to moderate disease. Established determinants of HRQoL such as FEV1 and BMI correlated with fewer scales. Based on scarce data in the literature looking at the relation between HRQoL and aerobic fitness and physical activity in CF, these findings are most important and relevant from a clinical perspective. The novel aspect of including longitudinal associations strengthens the findings.
HRQoL in our study was similar to that reported for patients with mild to moderate disease severity. In contrast to the German and American validation studies of the CFQ-R, data of patients with an FEV1 below 35% predicted were not included in our analyses which reduced the range of FEV1 in our project (35–107% predicted) compared to the respective validation studies (Germany: 12–106% predicted; United States: 17–130% predicted). Possibly as a consequence of the smaller range in FEV1, surprisingly few significant associations were observed between HRQoL-scales and FEV1 in the current project. Nevertheless, the power of our study was sufficient to detect clear and consistent positive associations between physical fitness and multiple HRQoL-scales.
Relationship Between HRQoL and Physical Activity
In contrast to our hypothesis, but in accordance to another comparable study in children with CF no significant associations were observed between objectively measured physical activity and any HRQoL-scale. For reported physical activity, only one significant direct correlation was detected with the scale 'Role limitations' cross-sectionally (Table 4A). In a previous cross-sectional study, physical activity was found to correlate quite strongly with Quality of Well-being reported by patients with CF. The discrepancy to our findings might reflect differences in the approach to determine HRQoL between the generic Quality of Well-being Scale and the multi-dimensional disease-specific CFQ-R. The Quality of Well-being Scale asks specifically whether or not the patient performs certain activities while the CFQ-R assesses the patient's feelings and difficulties while performing activities.
Changes in reported physical activity, however, were positively associated with changes in three scales of the CFQ-R in the current study, namely 'Role limitations', 'Feelings of embarrassment', and 'Weight problems'. This is important, as it shows that an increase in physical activity could increase HRQoL even after just 6 months. Other studies have also shown significant positive relationships between self-reported physical activity and HRQoL in healthy adults or patients with chronic obstructive pulmonary disease.
While reported physical activity showed some direct association with HRQoL, objectively measured activity did not. This finding is difficult to interpret. Possibly, specific aspects of our study design have obscured a significant relationship: Accelerometry was performed within a given period of three weeks before or after assessment of HRQoL, while anthropometry, pulmonary function assessment, exercise testing, and completion of activity and quality of life questionnaires were all done on the same day. It is possible that physical activity assessed by accelerometry over a few days did not mirror physical activity a few weeks earlier or later which may have been confounded by short-term alterations in disease state or weather conditions. Furthermore, the approach chosen to analyze accelerometer data including the selection of epoch times, definition of non-wear time or cutoffs for MVPA has been shown to influence the results and may also have contributed to the attenuated relation between physical activity and HRQoL.
Furthermore, due to incomplete accelerometry data, sample size for the respective analyses and thus statistical power was lower than for other variables. Further studies are required to explore the associations between objectively measured physical activity and HRQoL in CF.
Relationship Between HRQoL and Physical Fitness
The main finding of the present study was that physical fitness, mainly aerobic fitness, was positively associated with HRQoL cross-sectionally and longitudinally. This observation is in line with the positive association between changes in aerobic fitness and changes in Quality of Well-being reported by others and the direct associations between VO2peak and the HRQoL-scales 'physical functioning', 'emotional functioning', 'social functioning', and 'treatment burden' observed by Groenevelt et al.. Furthermore, the significant and positive correlations between indicators of aerobic fitness and the disease-specific HRQoL-scales were as strong and sometimes stronger than the associations of these scales with more traditional measures of disease severity such as age, FEV1 or body composition. This finding underlines the importance of a good physical fitness for HRQoL in CF especially in disease-specific scales, and therefore the importance of physical fitness assessments in patients with CF. In addition, our findings strengthen the construct validity of the CFQ-R.
After adjusting for age and gender, the only other cross-sectional study testing the relation between physical activity, physical fitness, and HRQoL scales did not find any significant correlation. This may be due to the much smaller sample size in their study group and the large age ranges as important confounders. Furthermore, the expression of the aerobic fitness per kg bodyweight rather than per predicted value may have "penalized" those who maintained body weight (leading to lower VO2peak per kg values) and "favorized" those who lost bodyweight (leading to higher VO2peak per kg values) – the opposite scenario from what would be expected from a clinical perspective. Indeed, when analyzing our data with a multiple regression model and adjusting for age, gender and nationality, direct relationships between VO2peak and 9 HRQoL-scales were observed (see Additional file 1).
At first sight it seems surprising that the HRQoL-scales 'Physical functioning' and 'Respiratory symptoms' were not responsive to changes in physical fitness. It is notable, however, that the actual data were taken from a training study and not from a longitudinal set of observational data. Thus, the patients included in this project were willing and felt generally capable of participating in regular exercise. This approach might have affected HRQoL perception related to changes in physical fitness. Moreover, several training studies have shown that physical conditioning may improve physical fitness while FEV1 remains relatively unaffected. Thus, although there was a significant direct correlation between peak oxygen uptake and the 'Respiratory symptom' scale at baseline, the subjective and objective respiratory situation in many patients might not have changed enough to be perceived positively despite improvements in aerobic fitness.
One limitation of the current project is the relatively small number of patients in the five treatment conditions that limits statistical power in detecting effects of the different treatment regimens on HRQoL. Nevertheless, supervised strength training three times per week over six months was associated with significant decreases in the HRQoL-scale 'Vitality', i.e. less vitality. It is possible that the supervised strength training might have been too strenuous to maintain or even improve 'Vitality' despite improvements in aerobic fitness and pulmonary function. A decrease in feelings of energy and an increase in perceived fatigue were observed in competitive swimmers progressively increasing their training volume, which supports our "overload" hypothesis. In contrast to our findings, others did not observe a negative effect of a presumably less strenuous partially supervised strength training program on Quality of Well-being which presumes that the negative effects on some HRQoL-scales in our study were rather a consequence of the intensity than of the strength training per se.
The discrepancies in time spent active between the subjective and objective assessments might be perceived as a further limitation of the study. However, this discrepancy is possibly related to the already mentioned different time frame that the two measurement tools covered and the way accelerometer data were analyzed that may at least in part have attenuated associations with HRQoL. In addition, reporting bias based on social desirability and social approval may have occurred.
Furthermore, as this study was a training study, it cannot be completely excluded that the interaction of the training supervisors with the participants may have affected some of the associations between the fitness, activity and HRQoL variables. Nevertheless, the fact that there was merely no difference in changes of HRQoL among groups suggests that this influence was minor. Finally, there is a possibility of shared variance amongst a number of the predictor variables in the regression model explaining HR-QoL that could be responsible for a reduced association between variables. Yet, this potential multicollinearity does not reduce the predictive power or reliability of the model as a whole but rather affects calculations regarding individual predictors.
Since physical conditioning programs in CF have been shown to improve pulmonary function, physical fitness and HRQoL in concert, patients should be encouraged to engage in regular at least moderate physical activities efficient to improve physical fitness even though we could only show beneficial effects of increased physical activity on some HRQoL-scales. To maintain or even improve 'Vitality', too strenuous training sessions without sufficient recovery periods such as unfamiliar strength training exercises three times per week need to be avoided in order to not compromise HRQoL.
Source...