Introduction and objectives Lower limb muscle strength measured by Quadriceps Maximal Voluntary Contraction (QMVC) provides valuable functional and prognostic information in people with COPD. Reference equations providing normal values for QMVC have been reported, some requiring measurement of muscle mass. It is unclear whether including muscle mass in the calculation significantly alters predicted values in COPD. We addressed this question by deriving reference equations for QMVC with and without the inclusion of whole body assessment of muscle mass in a cohort of healthy volunteers and subsequently comparing QMVC assessment using these reference equations in two separate cohorts of patients with COPD.

Methods Prediction equations were derived through multiple linear regression in a healthy control (HC) group. Age, gender, height and weight were inputted into the first model (FFM– model) and fat-free mass (FFM) added for the other (FFM+ model). The prediction equations were then applied to a Primary Care COPD (PCC) group and Complex Care COPD (CCC) group of patients where percentage predicted values were calculated and weakness determined using a threshold of the lower limit of normal.

Results 175 HC subjects (median (IQR) age: 54 (14) years, 31% male) were recruited. The PCC group comprised 87 patients (median (IQR) age: 68 (9) years, 71% male, FEV1 62 (20)% predicted) and the CCC group 189 patients (median (IQR) 66 (12) years, 57% male, FEV1: 29 (16)% predicted).

Prediction values for the HC and PCC were similar between the FFM– and FFM+ models as shown in the table. In the CCC percentage predicted values were lower and there were 11.9% more classed as weak by the FFM– model compared to the FFM+ model.

Conclusion The inclusion of fat-free mass did not significantly alter prediction of muscle weakness in the healthy cohort. In the COPD cohorts, including FFM in the model altered the proportion classified as having muscle weakness, most notably in the CCC cohort. This is likely to be due to a higher prevalence of muscle wasting in this population which resulted in an underestimate of predicted strength when muscle mass is included in the model.