Abstract

Background: Cross sectional studies have shown weak correlation between muscle function and physical activity (PA) in COPD. However, there is a lack of longitudinal studies to establish directionality of the association.

Objectives: To study the relationship between PA and muscle mass and function by studying baseline and changes over time in patients with COPD. Methods: 236 patients with COPD (159 males, 67 [62.5-74] years, BMI 26.5 [23.3-29.5]kg.m−2, FEV1 57 [42-71] %pred, 3660 [2382-5841] steps/day) from 6 European countries were assessed at baseline and after 12 months.

Assessments included: fat free mass (FFM) and fat free mass index (FFMI) by bioimpedance, dyspnoea (mMRC), six-minute walk test (6MWT), quadriceps maximal voluntary contraction (QMVC) and levels of PA (Actigraph GT3x).

Results: There were weak correlations between PA levels (steps/day, VMU and walking time) and QMVC at baseline (0.19 ≤ r ≤ 0.33 p<0.001 for all). No correlations were found between changes in PA and QMVC (-509 [-1295-362] vs -0.4 [-3.5-2.6] - 12 months minus baseline - respectively) and future FFM, FFMI (p>0.05). Baseline steps/day, VMU and walking time are related to future QMVC (0.30 ≤ r ≤ 0.41, p<.0001) but not with FFM or FFMI. Baseline PA did not relate to changes over time in FFM, FFMI or QMVC. Regression analysis (dependent variable at one year) showed that walking time is weakly influenced by age, MRC and 6MWT (r2=0.22).

Conclusions: PA levels are not related to one-year changes in FFM, FFMI and QMVC. However, higher PA levels at baseline are related to have higher muscle strength at one-year. Age, mMRC and 6MWT are weak prospective predictors of PA in COPD.