Base of support (cm), the distance between heels, quantified during gait analysis using the GAITRite electronic walkway system.

Step width variability (%) quantified during gait analysis using the GAITRite electronic walkway system. Variability calculated by the coefficient of varation=(standard deviation/mean)x100.

Single support time (s), an indirect measure of dynamic balance and gait certainty, quantified during gait analysis using the GAITRite electronic walkway system

Double support time (s), an indirect measure of dynamic balance and gait certainty, quantified during gait analysis using the GAITRite electronic walkway system

Stride time (s), also referred to as gait cycle time, quantified during gait analysis using the GAITRite electronic walkway system

Stride time variability (percent) quantified during gait analysis using the GAITRite electronic walkway system. Variability calculated by the coefficient of varation=(standard deviation/mean)x100.

Swing time (s), the time one leg is in the air during one gait cycle, quantified during gait analysis using the GAITRite electronic walkway system.

Swing time variability (percent) quantified during gait analysis using the GAITRite electronic walkway system. Variability calculated by the coefficient of varation=(standard deviation/mean)x100.

Standardized assessment of general mobility in older adults, time measured in seconds using a stopwatch

After performing the Timed Up and Go Test, the participant remains seated and imagines performing the Timed Up and Go Test. Test time is measured in seconds using a stopwatch. Time for imagined test performance that is twice as fast or faster than real test performance time reflect possible cognitive difficulties as well as walking difficulties, particularly under dual or multi-task conditions.

Standardized assessment of physical performance with measurements of muscle strength, balance and walking. Results range from 0 to 12 points, higher point score represents better physical performance.

Strandardized assessment of physical functional performance of 10 tasks of daily living. Task results measured as time (s), weight (kg) and/or distance (cm or m, depending on task) are transformed by the assessment software into scaled scores ranging from 0 to 100 points, with higher point scores reflecting better performance and greater liklihood of independent living. Total score (0-100 points) and well as subscores (0-100 points) of five domains of physical functional perfomance are provided: upper body strength, upper body flexibility, lower body strength, balance and coordination, and endurance

Standardized, self-report questionnaire to measure physical activity levels in older people as well as to assess the effectiveness of exercise interventions. From a list of 12 types of activities and referring to the previous week, participants report how many days per week and for how many hours per day they performed each activity. Scores are calculated by multiplying the weight (activity weights are provided in a table for PASE scoring) by the frequency values for each of 12 types of activities. Scores may range from zero to 400 or more. Higher scores reflect higher levels of physical activity.

Grip strength, as a surrogate marker of overall muscle strength, measured using a Martin vigorimater (kPa)

General functional measure of lower body strength, measured in seconds using a stop watch, performed as part of the Short Physical Performance Battery.

Assesses balance under four test conditions. Time to perform each task is measured in seconds with a stopwatch, results range from not being able to perform a task to maximal task time of 30 seconds. Postural sway will be measured during each task with the APDM system (measured in angle, m/s2, Hz or m2/s4, depending on paramter)

Fall incidence will be documented from the pre-intervention assessment until the end of the follow-up period via monthly diaries completed by study participants as well as via monthly phone calls with the study team.

Self-reported using the Falls Efficacy Scale-lnternational version (FES-I) questionnaire (scores range from 1 to 64 points, higher point scores reflect greater subjective fear of falling), as well as binary responses (yes/no) of participants at each clinical visit

Standardized test of general cognition, administered by trained site staff. Scores range from 0 to 30 points, higher point scores reflect better cognition

This self-report questionnaire is a reliable and valid tool for measuring health-related quality of life. It consists of 36 questions (acute recall = previous week) reflecting eight domains of health and provides eight scaled scores, which are the weighted sums of the questions in each respective domain. Each scale is directly transformed into a score of 0 to 100 points on the assumption that each question carries equal weight. Lower scores are associated with greater disability and lower quality of life.

Number of intervention courses from maximum 24 that each participant attends, as a marker of intervention adherence, will be documented by course instructors at each class