Table 1– Studies on physical inactivity and lung function decline or COPD incidence
First author [ref.]Design and settingSubjects and follow-upSubjects’ characteristics#]Physical activity assessmentOutcome(s)Main Results
Jakes [26]Population-based cohort
UK		n = 12 283
Mean 3.7 years follow-up		45% male
Age: 59±9 years
FEV1: 3.0±0.7 L in males, and 2.2±0.5 L in females		Standardised questionnaireFEV1 decline mL·year−1Climbing more stairs is associated with slower FEV1 decline in females (-0.31, -0.11, -0.03 and 0.04% change in FEV1 in none, 1–5, 6–10 and >10 flights of stairs per day; p-value for trend <0.004)
Participation in vigorous leisure time activities is associated with slower FEV1 decline in females (-0.22, -0.34, 0.27 and 0.03% change in FEV1 in none, <0.25, 0.25–1, and >1 h per week; p-value for trend<0.004)
Based on a linear regression model adjusted for age, smoking status, and per cent change in weight
No association found in males
Pelkonen [27]Cohort of rural men with a high physical activity level
Finland		n = 186
25 years follow-up		100% male
Age: 54±5 years
FEV0.75: 2.9±0.6 L		Validated questionnaireFEV0.75 decline mL·year−1Decline in FEV0.75 over 25 years was -44.4, -40.5 and -36.5 mL·year−1 for low, medium and high physical activity; p-value for trend = 0.035
Based on a linear regression model adjusted for age, height and smoking habits
Cheng [28]Convenience sample from a clinic
USA		n = 5707
Mean 1.6 years follow-up		87% male
Age range: 25–55 years		Self-reportFEV1 and FVC decline mL·year−1Males who remained in the active category during the follow-up increased FEV1 by 50 mL and FVC by 70 mL while subjects who remained in the sedentary group reduced FEV1 and FVC by 30 and 20 mL, respectively
Based on generalised least squared regressions adjusted for smoking and drinking habit change, age, baseline height, baseline lung function levels, and follow-up time
No association found in females
Garcia-Aymerich [29]Population-based cohort
Denmark		n = 6790
10 years follow-up		43% male
Age: 52±12 years
FEV1: 2.7±0.9 L
FVC: 3.3±1.0 L		Validated questionnaireCOPD incidence (FEV1/FVC ≤70%)Active smokers with moderate-to-high physical activity show a reduced risk of developing COPD compared with the low physical activity group (OR = 0.77; p = 0.027)
Based on logistic regression models adjusted for sex, age, education, BMI, weight change during follow-up, asthma, dyspnoea, sputum, smoking status and smoking duration
No association found in former or never-smokers
FEV1 and FVC decline mL·year−1Active smokers with moderate and high physical activity show a reduced FEV1 decline compared with those with low physical activity (relative change in FEV1 of +2.6 and +4.8 mL·year−1, respectively; p-value for trend = 0.006)
Active smokers with moderate and high physical activity show a reduced FVC decline compared with those with low physical activity (relative change in FVC of +2.6 and +7.7 mL·year−1, respectively; p-value for trend<0.0001)
Based on linear regression models adjusted for sex, age, education, BMI, weight change during follow-up, ischaemic heart disease, dyspnoea, sputum, smoking status, smoking duration, alcohol consumption and baseline lung function levels
No association found in former or never-smokers
Garcia-Aymerich [30]Population-based cohort
Denmark		n = 6568
Mean 16 years follow-up		41% male
Age: 49±11 years
FEV1: 2.7±0.8 L
FVC: 3.4±1.0 L		Validated questionnaireCOPD incidence (FEV1/FVC ≤70%)Subjects with moderate-to-high physical activity show a reduced risk of developing COPD compared with the low physical activity group (OR = 0.79; p = 0.025)
Based on a logistic regression model adjusted for sex, age, education, BMI, sputum, asthma and smoking, and weighted using marginal structural models to allow repeated measures of physical activity, lung function and covariates
FEV1 and FVC decline mL·year−1Subjects with moderate and high physical activity show a reduced FEV1 decline compared with those with low physical activity (relative change of +7.4 and +10.3 mL·year−1 for FEV1; p-value for trend<0.001)
Subjects with moderate and high physical activity show a reduced FVC decline compared with those with low physical activity (relative change of +6.9 and +10.0 mL·year−1 for FVC; p-value for trend<0.001)
Based on linear regression models adjusted for sex, age, education, BMI, sputum, smoking, alcohol and baseline lung function levels, and weighted using marginal structural models to allow repeated measures of physical activity, lung function and covariates

  • ↵COPD: chronic obstructive pulmonary disease; FEV1: forced expiratory volume in 1 s; FEV0.75: forced expiratory volume in 0.75 s; FVC: forced vital capacity; BMI: body mass index. #: subjects’ characteristics are presented as mean±sd, unless otherwise stated.