Objective measurement of physical activity in a multiple sclerosis population

Multiple Sclerosis (MS) is a demyelinating neurodegenerative disease of the central nervous system. Gait impairment is among the most common symptoms. As a result, people with MS are more at risk of developing hypokinetic diseases. The promotion and maintenance of regular physical activity (PA) in persons with MS (PwMS) is of key importance in preventing secondary cardiovascular conditions and maintaining functional independence. A key issue in the research around PA and PwMS relates to the need for a objective measurement tool with strong psychometric properties which is capable of measuring both PA and energy expenditure. Following a review of the literature, a number of gaps in the research were identified. An initial review of the literature identified the conceptual and measurement difficulties associated with PA and the current level of knowledge in relation to PA and MS. This thesis aimed to address these gaps by establishing the criterion validity of two activity monitors - the ActivPAL (AP) and the Sensewear Armband (SWA) - as a measure of PA and energy expenditure in thirty PwMS. Finally, this thesis aimed to compare the energy expenditure between healthy individuals and PwMS during activities of daily living (ADL). The literature review found that PA is a multidimensional and complex concept. Although PA itself is not a quantifiable scientific unit, PA consists of numerous measureable parameters which can be utilized as proxy measures of PA. Defining an appropriate research question is of key importance in choosing the optimal PA measurement tool. The measurement of PA in PwMS is effected by alterations in movement patterns and varying energy cost of movement. To date only two accelerometers-the TriTrac RT3 and the ActiGraph- have been examined in terms of reliability and convergent validity in PwMS, however, the optimal measurement tool remains unknown. There is a need to establish the criterion validity of activity monitors in PwMS to determine the optimal measurement tool. This study found that the SWA was a valid measure of PA- step count and energy expenditure- in PwMS as indicated by Intraclass Correlation Coefficient (ICC) values and Bland and Altman analysis. The validity of the SWA was affected by level of gait impairment for step count. The SWA demonstrated an ICC of 0.837 for PwMS with minimal gait impairment (MSA) and 0.441 for PwMS who used a bilateral walking aid (MSB). The ActivPAL was found to be a valid measure of steps during walking tasks for the total group (ICC for walking: 0.883) but not during mixed upper and lower limb ADL (ICC for folding laundry: 0.275). The SWA provided a valid measure of energy expenditure during ADL with an ICC value of 0.799 for MSA subgroup (mean difference: 16.51 kilocalories) and an ICC of 0.708 for MSB subgroup (mean difference: 5.79 kilocalories). However it overestimated energy expenditure during high intensity ADL and underestimated energy expenditure during low intensity ADL. PwMS who used a bilateral walking aid had greater energy expenditure per step during ADL than healthy individuals (p<0.001). In conclusion, future research should utilize the SWA as a measure of PA, if measuring both steps and energy expenditure are the parameters of interest. Future research should investigate the causative factors for the increased energy cost for PwMS during ADL, the long-term activity patterns of PwMS and the efficacy of exercise interventions for modifying long-term PA patterns in PwMS