Intervenções para mudança do comportamento de movimento e o desenvolvimento de uma plataforma para o monitoramento de atividade em reabilitação (MARe) para indivíduos pós-acidente vascular cerebral

Abstract

After a stroke, it has been identified that most individuals adopt a lifestyle with sedentary behaviors (SB) and low levels of physical activity (PA), which can cause damage to the individual, accentuating disabilities, limiting the performance of daily activities, restricting participation and society, and increasing the risk of new cerebrovascular events. Movement behavior (SB and PA) is an area of great interest in Neurofunctional Physiotherapy. Thus, interventions aiming to reduce SB and increase PA after stroke have been studied in the last decades and shown promising results. However, most studies end up using measurements of PA and SB self-reported by patients, thus not providing sufficiently reliable objective data to reflect the person's daily context. Thus, accelerometers and other motion sensors integrated into inertial measurement units (IMUs) bring important technological advances to rehabilitation, with systems capable of continuously detecting aspects of real-world mobility. However, due to asymmetry and decreased gait speed in this population, many of these approaches do not provide valid and reliable metric properties for measuring movement behavior in individuals with stroke. Given this, this thesis aimed to systematically synthesize the literature on the effectiveness of interventions, which used outcomes based on the use of activity monitors, to reduce SB and increase PA levels. In addition, it aimed to evaluate the measurement properties of a platform for activity monitoring in rehabilitation (AMoR) for post-stroke individuals. The first study (Manuscript 1) was a systematic review with meta-analysis, in which searches of the main databases of the area were performed. Data extraction was performed and analyses of the methodological quality, risk of bias and quality of evidence of the included studies were evaluated. Seventeen studies involving 1,031 patients were included in this systematic review. The meta-analysis showed that in post-stroke individuals, there is evidence of high quality and statistical significance in favor of exercise interventions combined with behavioral change techniques in the acute/subacute phase and in favor of exercise-only training in the chronic phase for increasing the number of daily steps. Overall, the risk of bias in the included studies was low. The second study (Manuscript 2), was a cross-sectional observational study, with 21 individuals in the chronic phase of post-stroke, who used the AMoR platform, during activities of daily living (ADLs), while being monitored by a video camera. These activities were filmed in a simulated environment. The validity and reliability of measurements between the platform and cameras were analyzed by Spearman's correlation coefficient and mean absolute percentage error (MAPE), and intraclass correlation coefficient (ICC) and Bland-Altman plot analyses, respectively. The following variables were considered: number of steps, time spent sitting/lying, and postural change from sitting to standing (Sit-Std) and standing to sitting (Std-Sit). The AMoR platform showed very high correlation values for number of steps and time spent sitting/lying and high correlation for postural change from Sit-Std and Std-Sit. An error percentage above 5% was observed only for the Sit-Std postural change (7.13%). The ICC values show excellent agreement of the AMoR platform with the video for number of steps and time spent sitting/lying and good agreement for Sit-Std and Std-Sit postural change. Difference values of the AMoR platform and video for step counting, sitting/lying time and postural change were within the limit of agreement in the Bland-Altman plot analyses. In conclusion, this thesis showed that: 1) by the literature analysis, the combination of exercise with behavioral interventions is associated with a significant increase in PA level in the acute and subacute phase after stroke. Furthermore, exercise-based interventions are associated with a significant increase in daily steps in chronic post-stroke participants; 2) the AMoR platform was valid and reliable for counting number of steps, time spent sitting/lying and identification of Sit-Std and Std-Sit postural changes during tests in a simulated environment in post-stroke individuals.