Trial Outcomes & Findings for Companion: Sensor-based Two-way Communication for Physical Activity in Older Adults (NCT NCT04827212)

Results Overview

Total time spent being active will be measured using a body worn motion sensor (accelerometer) during waking hours over a period of one week. Measures for each day (hours) will be averaged to obtain a representative estimate of daily habitual behavior.

Recruitment status

COMPLETED

Study phase

EARLY_PHASE1

Target enrollment

51 participants

Primary outcome timeframe

Baseline

Results posted on

2024-07-03

Participant Flow

5 participants were excluded from the study before assignment to groups. 3 chose to not continue participation and 2 were recommended to no continue participation by their personal physician. 46 participants were randomized to th control and treatment groups

Participant milestones

Participant milestones
Measure	Control Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Baseline to 16 Weeks STARTED	23	23
Baseline to 16 Weeks COMPLETED	21	21
Baseline to 16 Weeks NOT COMPLETED	2	2
16 to 24 Weeks STARTED	21	21
16 to 24 Weeks COMPLETED	18	17
16 to 24 Weeks NOT COMPLETED	3	4

Reasons for withdrawal

Reasons for withdrawal
Measure	Control Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Baseline to 16 Weeks Lost to Follow-up	2	2
16 to 24 Weeks Lost to Follow-up	3	4

Baseline Characteristics

Companion: Sensor-based Two-way Communication for Physical Activity in Older Adults

Baseline characteristics by cohort

Baseline characteristics by cohort
Measure	Control n=23 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=23 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.	Total n=46 Participants Total of all reporting groups
Age, Continuous	72.7 years STANDARD_DEVIATION 6.1 • n=5 Participants	68.9 years STANDARD_DEVIATION 6.5 • n=7 Participants	71 years STANDARD_DEVIATION 6.5 • n=5 Participants
Sex: Female, Male Female	13 Participants n=5 Participants	18 Participants n=7 Participants	31 Participants n=5 Participants
Sex: Female, Male Male	10 Participants n=5 Participants	5 Participants n=7 Participants	15 Participants n=5 Participants
Ethnicity (NIH/OMB) Hispanic or Latino	1 Participants n=5 Participants	4 Participants n=7 Participants	5 Participants n=5 Participants
Ethnicity (NIH/OMB) Not Hispanic or Latino	22 Participants n=5 Participants	19 Participants n=7 Participants	41 Participants n=5 Participants
Ethnicity (NIH/OMB) Unknown or Not Reported	0 Participants n=5 Participants	0 Participants n=7 Participants	0 Participants n=5 Participants
Race (NIH/OMB) American Indian or Alaska Native	0 Participants n=5 Participants	0 Participants n=7 Participants	0 Participants n=5 Participants
Race (NIH/OMB) Asian	2 Participants n=5 Participants	0 Participants n=7 Participants	2 Participants n=5 Participants
Race (NIH/OMB) Native Hawaiian or Other Pacific Islander	0 Participants n=5 Participants	0 Participants n=7 Participants	0 Participants n=5 Participants
Race (NIH/OMB) Black or African American	4 Participants n=5 Participants	3 Participants n=7 Participants	7 Participants n=5 Participants
Race (NIH/OMB) White	17 Participants n=5 Participants	19 Participants n=7 Participants	36 Participants n=5 Participants
Race (NIH/OMB) More than one race	0 Participants n=5 Participants	1 Participants n=7 Participants	1 Participants n=5 Participants
Race (NIH/OMB) Unknown or Not Reported	0 Participants n=5 Participants	0 Participants n=7 Participants	0 Participants n=5 Participants
BMI	30.5 kg/^2 STANDARD_DEVIATION 3.8 • n=5 Participants	30.4 kg/^2 STANDARD_DEVIATION 4.9 • n=7 Participants	30.4 kg/^2 STANDARD_DEVIATION 4.3 • n=5 Participants

PRIMARY outcome

Timeframe: Baseline

Population: Accelerometer data were missing for 1 participant in the control group.This results in a discrepancy with overall participant flow table which represents the number of participants with any available data.

Total time spent being active will be measured using a body worn motion sensor (accelerometer) during waking hours over a period of one week. Measures for each day (hours) will be averaged to obtain a representative estimate of daily habitual behavior.

Outcome measures

Outcome measures
Measure	Control n=22 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=23 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Physical Activity Hours Per Day Measured Using a Body-worn Accelerometer at Baseline	1.3 h/day Standard Deviation 0.5	1.6 h/day Standard Deviation 0.5

PRIMARY outcome

Timeframe: 4 months

Population: There were 2 participants lost to followup in both groups and 1 participant was missing accelerometer data in the control group. This results in a discrepancy with overall participant flow table which represents the number of participants with any available data.

Total time spent being active will be measured using a body worn motion sensor (accelerometer) during waking hours over a period of one week. Measures for each day (hours) will be averaged to obtain a representative estimate of daily habitual behavior.

Outcome measures

Outcome measures
Measure	Control n=20 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=21 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Physical Activity Hours Per Day Measured Using a Body-worn Accelerometer After 4 Months	1.5 h/day Standard Deviation 0.2	1.5 h/day Standard Deviation 0.2

PRIMARY outcome

Timeframe: 6 months

Population: 5 and 6 participants were lost to followup and additionally, accelerometer data were missing for 1 participant each in the control and treatment groups respectively. This results in a discrepancy with overall participant flow table which represents the number of participants with any available data.

Total time spent being active will be measured using a body worn motion sensor (accelerometer) during waking hours over a period of one week. Measures for each day (hours) will be averaged to obtain a representative estimate of daily habitual behavior.

Outcome measures

Outcome measures
Measure	Control n=17 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=16 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Physical Activity Hours Per Day Measured Using a Body-worn Accelerometer After 6 Months	1.5 h/day Standard Deviation 0.24	1.5 h/day Standard Deviation 0.24

PRIMARY outcome

Timeframe: Baseline

Population: Accelerometer data were missing for 1 participant in the control group.This results in a discrepancy with overall participant flow table which represents the number of participants with any available data.

Total time spent sedentary will be measured using a body worn motion sensor (accelerometer) during waking hours over a period of one week. Measures for each day (hours) will be averaged to obtain a representative estimate of daily habitual behavior.

Outcome measures

Outcome measures
Measure	Control n=22 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=23 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Sedentary Behavior Hours Per Day Measured Using a Body-worn Accelerometer at Baseline	10.6 h/day Standard Deviation 2.0	10.2 h/day Standard Deviation 2.0

PRIMARY outcome

Timeframe: 4 months

Population: There were 2 participants lost to followup in both groups and 1 participant was missing accelerometer data in the control group. This results in a discrepancy with overall participant flow table which represents the number of participants with any available data.

Total time spent sedentary will be measured using a body worn motion sensor (accelerometer) during waking hours over a period of one week. Measures for each day (hours) will be averaged to obtain a representative estimate of daily habitual behavior.

Outcome measures

Outcome measures
Measure	Control n=20 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=21 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Sedentary Behavior Hours Per Day Measured Using a Body-worn Accelerometer After 4 Months	10.4 h/day Standard Deviation 1.9	9.6 h/day Standard Deviation 1.9

PRIMARY outcome

Timeframe: 6 months

Population: 5 and 6 participants were lost to followup and additionally, accelerometer data were missing for 1 participant each in the control and treatment groups respectively. This results in a discrepancy with overall participant flow table which represents the number of participants with any available data.

Total time spent sedentary will be measured using a body worn motion sensor (accelerometer) during waking hours over a period of one week. Measures for each day (hours) will be averaged to obtain a representative estimate of daily habitual behavior.

Outcome measures

Outcome measures
Measure	Control n=17 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=16 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Sedentary Behavior Hours Per Day Measured Using a Body-worn Accelerometer After 6 Months	10.3 h/day Standard Deviation 1.8	10.2 h/day Standard Deviation 1.7

OTHER_PRE_SPECIFIED outcome

Timeframe: Baseline

This is an exploratory outcome. The amount of fat in the body will be computed as a percentage

Outcome measures

Outcome measures
Measure	Control n=23 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=23 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Percent Body Fat Measured Via Bioelectrical Impedance at Baseline	38.9 % body fat Standard Deviation 5.8	41.1 % body fat Standard Deviation 5.8

OTHER_PRE_SPECIFIED outcome

Timeframe: 4 months

Population: Missing data from 2 participants who were lost to follow-up in each group

This is an exploratory outcome. The amount of fat in the body will be computed as a percentage

Outcome measures

Outcome measures
Measure	Control n=21 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=21 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Percent Body Fat Measured Via Bioelectrical Impedance After 4 Months	38.2 % body fat Standard Deviation 5.6	40.5 % body fat Standard Deviation 5.6

OTHER_PRE_SPECIFIED outcome

Timeframe: 6 months

Population: 5 and 6 participants from the control and treatment group were lost to followup, respectively.

This is an exploratory outcome. The amount of fat in the body will be computed as a percentage

Outcome measures

Outcome measures
Measure	Control n=18 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=17 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Percent Body Fat Measured Via Bioelectrical Impedance After 6 Months	37.7 % body fat Standard Deviation 5.2	40.5 % body fat Standard Deviation 5.1

OTHER_PRE_SPECIFIED outcome

Timeframe: Baseline

This is an exploratory outcome - Average age-adjusted national percentile of NIH Toolbox Flanker Inhibitory Control and Attention Test score measured at Baseline. Each participant's raw score on the NIH Toolbox Flanker Inhibitory Control and Attention Test is standardized against a nationally representative sample to get the participant's national percentile. Percentiles are then adjusted for age and averaged across participants. Percentiles can range from 0 to 100, with a higher percentile indicating better performance.

Outcome measures

Outcome measures
Measure	Control n=23 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=23 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Attention and Executive Function Measured Using the NIH Tool Box Measured at Baseline	36.2 average age-adjusted percentile Standard Deviation 25.3	42.4 average age-adjusted percentile Standard Deviation 25.3

OTHER_PRE_SPECIFIED outcome

Timeframe: 4 months

Population: Two participants each were lost to follow-up in the control and treatment groups

This is an exploratory outcome - Average age-adjusted national percentile of NIH Toolbox Flanker Inhibitory Control and Attention Test score measured at Baseline. Each participant's raw score on the NIH Toolbox Flanker Inhibitory Control and Attention Test is standardized against a nationally representative sample to get the participant's national percentile. Percentiles are then adjusted for age and averaged across participants. Percentiles can range from 0 to 100, with a higher percentile indicating better performance.

Outcome measures

Outcome measures
Measure	Control n=21 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=21 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Attention and Executive Function Measured Using the NIH Tool Box After 4 Months	46.0 average age-adjusted percentile Standard Deviation 25.1	52.2 average age-adjusted percentile Standard Deviation 25.5

OTHER_PRE_SPECIFIED outcome

Timeframe: 6 months

Population: 5 and 6 participants from the control and treatment group were lost to followup, respectively.

This is an exploratory outcome - Average age-adjusted national percentile of NIH Toolbox Flanker Inhibitory Control and Attention Test score measured at Baseline. Each participant's raw score on the NIH Toolbox Flanker Inhibitory Control and Attention Test is standardized against a nationally representative sample to get the participant's national percentile. Percentiles are then adjusted for age and averaged across participants. Percentiles can range from 0 to 100, with a higher percentile indicating better performance.

Outcome measures

Outcome measures
Measure	Control n=18 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=17 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Attention and Executive Function Measured Using the NIH Tool Box	50.3 average age-adjusted score Standard Deviation 23.2	55.3 average age-adjusted score Standard Deviation 23.8

OTHER_PRE_SPECIFIED outcome

Timeframe: Baseline

This is an exploratory outcome - Average age-adjusted national percentile of NIH Toolbox Picture Sequence Memory Test score measured at Baseline. Each participant's raw score on the NIH Toolbox Picture Sequence Memory Test is standardized against a nationally representative sample to get the participant's national percentile. Percentiles are then adjusted for age and averaged accross participants. Percentiles can range from 0 to 100, with a higher percentile indicating better performance.

Outcome measures

Outcome measures
Measure	Control n=23 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=23 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Working Memory Measured Using the NIH Tool Box at Baseline	59.2 average age-adjusted percentile Standard Deviation 28.8	59.0 average age-adjusted percentile Standard Deviation 29.0

OTHER_PRE_SPECIFIED outcome

Timeframe: 4 months

Population: Two participants each from the control and treatment group were lost to followup.

This is an exploratory outcome - Average age-adjusted national percentile of NIH Toolbox Picture Sequence Memory Test score measured at Baseline. Each participant's raw score on the NIH Toolbox Picture Sequence Memory Test is standardized against a nationally representative sample to get the participant's national percentile. Percentiles are then adjusted for age and averaged across participants. Percentiles can range from 0 to 100, with a higher percentile indicating better performance.

Outcome measures

Outcome measures
Measure	Control n=21 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=21 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Working Memory Measured Using the NIH Tool Box After 4 Months	62.1 average age-adjusted-percentile Standard Deviation 28.2	61.7 average age-adjusted-percentile Standard Deviation 28.6

OTHER_PRE_SPECIFIED outcome

Timeframe: 6 months

Population: 5 and 6 participants from the control and treatment group were lost to followup, respectively.

This is an exploratory outcome - Average age-adjusted national percentile of NIH Toolbox Picture Sequence Memory Test score measured at Baseline. Each participant's raw score on the NIH Toolbox Picture Sequence Memory Test is standardized against a nationally representative sample to get the participant's national percentile. Percentiles are then adjusted for age and averaged across participants. Percentiles can range from 0 to 100, with a higher percentile indicating better performance.

Outcome measures

Outcome measures
Measure	Control n=18 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=17 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Working Memory Measured Using the NIH Tool Box After 6 Months	62.2 average age-adjusted percentile Standard Deviation 26.1	63.3 average age-adjusted percentile Standard Deviation 26.4

OTHER_PRE_SPECIFIED outcome

Timeframe: Baseline

Average age-adjusted national percentile of NIH Toolbox Pattern Comparison Processing Speed Test score measured at Baseline. Each participant's raw score on the NIH Toolbox Pattern Comparison Processing Speed Test is standardized against a nationally representative sample to get the participant's national percentile. Percentiles are then adjusted for age and averaged across participants. Percentiles can range from 0 to 100, with a higher percentile indicating better performance.

Outcome measures

Outcome measures
Measure	Control n=23 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=23 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Relative Processing Speed Measured Using the NIH Tool Box at Baseline	39.7 Average age-adjusted percentile Standard Deviation 38.2	52.7 Average age-adjusted percentile Standard Deviation 38.2

OTHER_PRE_SPECIFIED outcome

Timeframe: 4 months

Population: Two participants each from the control and treatment group were lost to followup.

This is an exploratory outcome - Average age-adjusted national percentile of NIH Toolbox Pattern Comparison Processing Speed Test score measured at Baseline. Each participant's raw score on the NIH Toolbox Pattern Comparison Processing Speed Test is standardized against a nationally representative sample to get the participant's national percentile. Percentiles are then adjusted for age and averaged accross participants. Percentiles can range from 0 to 100, with a higher percentile indicating better performance.

Outcome measures

Outcome measures
Measure	Control n=21 Participants Control group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks	Treatment n=21 Participants Treatment group will engage in a 16-week supervised exercise program and then be followed-up after 24 weeks. In addition, the Companion will be deployed in the treatment group only. Companion: Companion integrates wearable sensing with the smartphone to develop a behavior-aware, virtual system that uses a "human-in-the-loop" approach to enable meaningful two-way communication. Companion builds rich models of typical behavior using sensors and context sensitive ecological momentary assessment to deliver intervention components and behavior change strategies using socially engaging, contextually salient, and tailored text-message conversations in near-real-time.
Relative Processing Speed Measured Using the NIH Tool Box Measured After 4 Months	43.5 Average age-adjusted percentile Standard Deviation 37.1	52.0 Average age-adjusted percentile Standard Deviation 37.4

OTHER_PRE_SPECIFIED outcome

Timeframe: 6 months

This is an exploratory outcome - Average age-adjusted national percentile of NIH Toolbox Pattern Comparison Processing Speed Test score measured at Baseline. Each participant's raw score on the NIH Toolbox Pattern Comparison Processing Speed Test is standardized against a nationally representative sample to get the participant's national percentile. Percentiles are then adjusted for age and averaged across participants. Percentiles can range from 0 to 100, with a higher percentile indicating better performance.