Longitudinal Analysis of Gait Variability to Predict Falls in Parkinson's Disease

The broad goal of this research project is to improve the prediction of falls in patients with Parkinson's disease (PD) through a comprehensive multidisciplinary approach that includes longitudinal measurements of walking, cognitive and functional performances. PD is a life-changing disorders affecting one million Americans, with more than 60,000 new cases reported every year. Within 3 years of diagnosis, more than 85% of people with clinically probable PD develop gait problems, which in turn lead to falls resulting in serious injury and reduced quality of life. There is a pressing need to identify fall risk factors before the occurrence of the first fall, and to better understand behavioral and cognitive changes leading to falls in PD patients.

Participants will take part in a baseline study session, followed by a year of weekly reports of their number of near falls and falls. The same participants will repeat the same study session six months and twelve months after the baseline session. Each of the three study sessions will have a similar design. Each participant will make a visit of approximately 2 hours to the laboratory. At baseline, all participants will complete a medical history questionnaire which will confirm that all inclusion criteria are met. At each measurement interval (i.e., baseline, six-month and twelve-month), participants will be evaluated through a series of examinations and questionnaires prior to data collection. PD patients will be tested at their self-reported peak dose of medication ('on'-state). Upon completion of the tests and questionnaires, participants will be prepared for data collection (footswitches and strap-based IMUs). The effect of dual tasking will be examined through three conditions performed in the following order: single-cognitive testing, which will consist in a phoneme monitoring task (seated in a quiet room with headphones); single-walking, which will consist of simply walking normally (silent headphones); and dual-task walking, which will consist in walking while performing the phoneme monitoring task. The phoneme monitoring task will consist in listening to a text through headphones, and participants will be asked to i) answer ten questions regarding the content of the text after the task, and ii) count how many times two predefined words appeared in the text. During single and dual-task walking, participants will be asked to walk around the track at a speed that they feel would be comfortable while walking through the grocery store or walking to their car, with no instructions regarding the priority to give to each task (walking vs. cognitive task). The texts presented during single-task cognitive testing and dual-task walking will be different. Each trial will last approximately ten minutes, which is required to collect at least 500 stride time intervals for a reliable estimation of the temporal ordering of stride-to-stride fluctuations. Participants will rest between trials for at least five minutes and will be able to rest more if they so desire. Two investigators will always be present during data collection and will walk behind the participant during the walking trials for safety purposes. All participants will be followed for twelve months and will report their falls and near falls experiences. After the baseline session, participants will be provided with a fall diary, in which they will write the date, time, cause, circumstances under which falls or near falls occurred, and the occurrence of any fall-related injury. Participants will be instructed to daily fill the diary and return it every two weeks (with self-addressed stamped return envelope), or to communicate the information via email. In addition, each month participating personnel will call participants to ensure all falls and near falls were reported in case participants fail to fill in the fall diary. Participants will also be called if the investigators do not receive their report a week after the expected date of reception.

Participants will take part in a baseline study session, followed by a year of weekly reports of their number of near falls and falls. The same participants will repeat the same study session six months and twelve months after the baseline session.

• The task will consist in walking while performing the phoneme monitoring task. The phoneme monitoring task will consist in listening to a text through headphones, and participants will be asked to i) answer ten questions regarding the content of the text after the task, and ii) count how many times two predefined words appeared in the text. During single and dual-task walking, participants will be asked to walk around the track at a speed that they feel would be comfortable while walking through the grocery store or walking to their car, with no instructions regarding the priority to give to each task (walking vs. cognitive task).

Using the detrended fluctuation analysis (DFA). Briefly, DFA estimate the relationship between the size of fluctuations and the size of the 'box' from which these fluctuations are observed. Stride time fluctuations typically scale with the size of the box in a power-law relationship. The slope of this relationship in bi-logarithmic coordinate thus provide the scaling exponent which reflect the degree of scale invariance.

Sub-phases of stride cycles, in particular the stance phase (i.e., when the foot is in contact with the ground) and the swing phase (i.e., when the foot is in the air), which will provide information about balance control mechanisms; medial-lateral center of mass movements and local stability, estimated with the Lyapunov exponent, using data from the pelvis sensor, which will provide information about gait stability; coordination between arm and leg movements, which will inform about the degree of rigidity and central synchronization between the arms during walking.

Inclusion Criteria: Able to provide written informed consent 60 years or older Able to walk 10 min uninterrupted without assistance for PD group only: must have a confirmed diagnosis by a trained clinician of idiopathic PD Exclusion Criteria: other known neurological, orthopedic or cardiothoracic conditions that could interfere with gait and cognitive assessment. surgery affecting gait and balance within 6 months prior to baseline session.