The Brief Symptom Inventory (BSI), anxiety index, consists of 6 items rated on a 5-point Likert scale (0 = not at all; 4 = extremely). The total score ranges from 0 to 24, with a higher score reflecting greater anxiousness.

The Vulnerable Elders Survey 13 (VES-13) includes four dimensions: age, self-assessment of health, physical function, and living function. Each domain is scored differently. Age is scored as follows: 75-84 years = 1 point, ≥ 85 years = 3 points. Self-rated health is categorized into "fair and poor" or "good", "very good", or "excellent", with only the "fair and poor" category scoring 1 point. The total score ranges from 0 to 15, with higher scores indicating greater vulnerability.

Short-term Heart Rate Variability (HRV) will be collected using the Polar H9 heart rate monitor (Polar Electro OY, Kempele, Finland). The investigator will send a reminder about the precautions to be taken (e.g., avoiding caffeine, alcohol, smoking, and intense physical activity, as these factors could affect the HRV measurement for successful HRV measurement the day before the scheduled test. The investigator will again ask if the instructions have been followed when they arrive for data collection.

Study patients will be instructed to lay supine for 10 minutes without speaking and to remain as still as possible. The investigator will place the Polar H9 heart rate monitor on the center of the sternum to measure HRV. During HRV data collection, the study patients will also wear the ActiGraph on the waist because the Polar H9 communicates with the ActiGraph.

Muscle strengths will be measured using the handheld dynamometry (HHD, Lafayette dynamometer, model01165APP; Lafayette Instrument Company, Lafayette, Ind., USA). This device measures the peak force exerted in kilograms for five seconds during muscle contraction on hips and quadriceps.

Balance will be measured using an Inertial Measurement Unit device (DOTIMU; Xsens Technologies, Enschede., the Netherlands). The investigators attach the IMU sensor (a small device the size of a large watch face) to the participant's lower back at the level of L3 vertebra. Patients will be asked to take off their shoes and stand up on the mat with their hands on their hips for 30 seconds. The investigators will measure balance 3 times, with 30 seconds of rest between assessments. The total estimated time for the balance test is 3-5 minutes. All signal processing, subsequent feature extraction, and analysis will be performed using MATLAB version R2018B (The Mathworks Inc., Natick, MA, USA). The root-mean-square (RMS) value of the acceleration signal will be used to quantify the magnitude and trajectory of postural sway in each direction. The RMS of the magnitude vector of the 3-axis acceleration signal will be used as the summary score.

Physical activity (PA) will be measured using actigraphy, the gold-standard for measuring physical activity. The investigator will use the ActiGraph (GT3X Link model), a triaxial instrument, which is the most widely used wearable accelerometer in clinical research. All patients will be instructed to wear the device for nine consecutive days, including at least one weekend day. Both baseline and 3-month values were calculated in a similar way.

Muscle strengths will be measured using the handheld dynamometry (HHD, Lafayette dynamometer, model01165APP; Lafayette Instrument Company, Lafayette, Ind., USA). This device measures the peak force exerted in kilograms for five seconds during muscle contraction on hips and quadriceps.

Total of all reporting groups

The completion rate of the study will be calculated by dividing the total number of patients by the number of patients who will complete the study and then multiplying it by 100.

The intervention adherence rate will be calculated by dividing the total number of sessions attended by the total number of intervention sessions (i.e., 24) and multiplying that number by 100.

This measure was only taken by the intervention group

The investigators will use modified versions of the Treatment Acceptability Adherence Scale (TAAS) to assess acceptability of the yoga intervention at 3-months follow-up. The TAAS is a 10-item self-report questionnaire designed to evaluate treatment acceptability and adherence. Items are rated on a 7-point Likert-type scale (1 = disagree strongly; 7 = agree strongly). The total score is a sum of ten items, ranging from 10 to 70 where a higher score means greater adherence.

Only participants in the intervention arm completed this satisfaction survey. One out of 16 participants did not respond to this survey and was excluded, leaving 15 participants for analysis.

The investigators will use modified versions of the Client Satisfaction Questionnaire (CSQ) to assess the satisfaction of the yoga intervention at 3-month follow-up. The CSQ is an 8-item self-report questionnaire to evaluate participant satisfaction with a specific intervention. Items are rated on a 4-point Likert-type scale (1 = very satisfied; 4= quite dissatisfied). The total score is the sum of eight items, ranging from 8 to 32, where a higher score indicates greater satisfaction with the intervention. This instrument was validated for patients and has internal consistency reliability.

Only participants in the intervention arm completed this satisfaction survey.

The Patient Health Questionnaire-9 (PHQ-9) was developed to correspond to symptoms used in the DSMV-IV to diagnose clinical depression. Nine symptoms are rated on a 4-point Likert scale (0 = 'Not at all; 3 = 'Nearly every day'), with possible total scores ranging from 0 to 27 and higher scores indicating higher levels of depressive symptoms.

The Brief Symptom Inventory (BSI), anxiety index, consists of 6 items rated on a 5-point Likert scale (0 = not at all; 4 = extremely). The total score ranges from 0 to 24, with a higher score reflecting greater anxiousness.

The Vulnerable Elders Survey 13 (VES-13) includes four dimensions: age, self-assessment of health, physical function, and living function. Each domain is scored differently. Age is scored as follows: 75-84 years = 1 point, ≥ 85 years = 3 points. Self-rated health is categorized into "fair and poor" or "good", "very good", or "excellent", with only the "fair and poor" category scoring 1 point. The total score ranges from 0 to 15, with higher scores indicating greater vulnerability.

Although baseline data were collected, it was pre-specified to report a change value at month 3. No data were collected at month 3 for participants because the use of the survey was abandoned and therefore data were not collected for the Outcome Measure

Short-term Heart Rate Variability (HRV) will be collected using the Polar H9 heart rate monitor (Polar Electro OY, Kempele, Finland). The investigator will send a reminder about the precautions to be taken (e.g., avoiding caffeine, alcohol, smoking, and intense physical activity, as these factors could affect the HRV measurement for successful HRV measurement the day before the scheduled test. The investigator will again ask if the instructions have been followed when they arrive for data collection.

Study patients will be instructed to lay supine for 10 minutes without speaking and to remain as still as possible. The investigator will place the Polar H9 heart rate monitor on the center of the sternum to measure HRV. During HRV data collection, the study patients will also wear the ActiGraph on the waist because the Polar H9 communicates with the ActiGraph.

Although some baseline data were collected, it was pre-specified to report a change value at month 3. No participants were assessed at month 3 and therefore the pre-specified change value cannot be calculated. Data not collected.

Muscle strengths will be measured using the handheld dynamometry (HHD, Lafayette dynamometer, model01165APP; Lafayette Instrument Company, Lafayette, Ind., USA). This device measures the peak force exerted in kilograms for five seconds during muscle contraction on hips and quadriceps.

Balance will be measured using an Inertial Measurement Unit device (DOTIMU; Xsens Technologies, Enschede., the Netherlands). The investigators attach the IMU sensor (a small device the size of a large watch face) to the participant's lower back at the level of L3 vertebra. Patients will be asked to take off their shoes and stand up on the mat with their hands on their hips for 30 seconds. The investigators will measure balance 3 times, with 30 seconds of rest between assessments. The total estimated time for the balance test is 3-5 minutes. All signal processing, subsequent feature extraction, and analysis will be performed using MATLAB version R2018B (The Mathworks Inc., Natick, MA, USA). The root-mean-square (RMS) value of the acceleration signal will be used to quantify the magnitude and trajectory of postural sway in each direction. The RMS of the magnitude vector of the 3-axis acceleration signal will be used as the summary score.

Physical activity (PA) will be measured using actigraphy, the gold-standard for measuring physical activity. The investigator will use the ActiGraph (GT3X Link model), a triaxial instrument, which is the most widely used wearable accelerometer in clinical research. All patients will be instructed to wear the device for nine consecutive days, including at least one weekend day. Both baseline and 3-month values were calculated in a similar way.

Muscle strengths will be measured using the handheld dynamometry (HHD, Lafayette dynamometer, model01165APP; Lafayette Instrument Company, Lafayette, Ind., USA). This device measures the peak force exerted in kilograms for five seconds during muscle contraction on hips and quadriceps.

Participants in the intervention group will attend two virtual 60-minute gentle yoga sessions per week for 12 weeks.

Gentle Yoga: Participants will attend two virtual 60-minute gentle yoga sessions per week for 12 weeks. Each session will be conducted via videoconference using the Zoom platform, with participants joining from their homes.