ABSTRACT
This study examines the validity of the Marshall Sitting Questionnaire (MSQ) in people with multiple sclerosis (MS) and presents a prediction model to improve the accuracy of the MSQ estimates. Participants (n = 63 [15M and 48F], age: 56.6 ± 7.3 years) wore an accelerometer on the hip for a 7-day period and completed the MSQ. Sedentary behaviour (SB) estimates were computed as min/day for both the MSQ and accelerometer. Associations between the two methods were examined using the Pearson correlation and agreement was assessed using a Bland-Altman plot. A linear prediction model was developed to improve the accuracy of the MSQ estimates using a selection of predictor variables routinely collected in MS research. There was a moderate correlation between MSQ and accelerometer SB estimates (r = 0.34, p < 0.01). The Bland-Altman plot indicated that the MSQ overestimated SB (mean bias: 80.54 min/day, 95% limits of agreement: −410.5 to 571.5 min/day). The prediction model improved the MSQ estimates by 39% and virtually eliminated measurement bias (mean bias:-0.21 min/day; 95% limits of agreement:-109.8 to 109.4 min/day). The results indicate preliminary evidence for the validity of the MSQ in people with MS, and support the application of an alternative prediction model to improve the accuracy of the MSQ estimates.
Abbreviations: SB = sedentary behaviour; MS = multiple sclerosis; MSQ = Marshall Sitting Questionnaire; PA = physical activity; EDSS = Expanded Disability Status Scale; LPA = light physical activity; MVPA = moderate-to-vigorous physical activity; GLTEQ = Godin Leisure Time Exercise Questionnaire; PDDS = Patient-Determined Disease Steps; RMSE = root mean square error.
Acknowledgments
This research was funded by a grant from the National Multiple Sclerosis Society (IL0009). Jeffer E Sasaki was supported by a mentor-based fellowship from National Multiple Sclerosis Society (MB0011).
Disclosure statement
No potential conflict of interest was reported by the authors.