NOTE: End date changed to 6/30/2016 per NSBRI (Ed., 6/29/16)

NOTE: Extended to 7/31/2016 per NSBRI (Ed., 6/16/16)

Key Findings since last report: In Year 3, we continued work on Aims 2 and 3. Aim 2 was a short-term study (6 days) to determine the most effective partners to enhance exercise intensity.

In Aim 2, we tested an SG partner in one of three modes against an individual control condition: (a) coacting mode, where the subject cycled with the SG partner but whose performance was independent of the partner (i.e., they were not teammates), (b) conjunctive-teammate mode, where scores were based on the slower performer, and (c) choice mode where subject could choose coacting mode or conjunctive teammate mode for all trials. The experiment used activity routines developed by Ploutz-Snyder that consist of (a) 30 min. of continuous aerobic exercise on a cycle ergometer at or above 75% of maximum heart rate (HR max) and (b) high-intensity interval training involving 4 repetitions of 4 min. at or above 90% HR max. In Aim 2 (Year 3), we tested an additional 38 chronic exercisers (22 female). Subjects (N = 82, 47 female; M age = 44.89, 9.45) were allowed to adjust the power output (watts) during workouts. The main outcome variable was average watts cycled above target prescribed watts. Results showed no condition main effects from baseline; however, during the 4 min. interval sessions conjunctive (Mdiff = 5.23 watts) trended toward greater exercise effort than control (Mdiff = 4.45 watts) and coactive conditions (3.46 watts). The 4 min. intervals probably represent the most motivationally demanding workout at 90% HR max. Increases in effort above target watts positively correlated with Enjoyment (r = .44), Self-efficacy (r = .51), and Team perceptions (r = .21) during the 4 min. intervals. No relationships were found with less intense continuous exercise. Although this pilot study did not find significant between group increases in performance effort, participants significantly increased their effort with our exercise video game (i.e., simulated bike paths).

In Aim 3, long-term study, we tested an SG exercise partner in one of two modes: (a) conjunctive-teammate mode and (b) conjunctive teammate who is not always superior (NAS) and is sometimes surpassed by the subject (~15% of the time), plus (c) individual control condition. We finished both cohorts of subjects (Cohort 1 = 23; 11 female; Cohort 2 = 18; 7 female). Subjects (N = 41; M age = 45.53 8.12) exercised 6 days/week for 24 weeks, using the following aerobic routines developed by Ploutz-Snyder: (a) 30 min. of continuous aerobic exercise on a stationary cycle at or above 75% HR max, (b) 4x4 min. intervals at or above 90% HR max with 3 min. active rest, (c) 6x2 min. intervals at varying intensities with 2 min. active rest, and (d) 30 sec. sprint intervals at maximal effort with 20 sec. active rest. We first tested whether the Conjunctive or NAS groups adhered more to the protocol than Controls. Conjunctive (M = 122.67) and NAS (M = 123.69) conditions averaged 15 more days completing the protocol than the Control (M = 107.00). However, group differences were not statistically different. In terms of our primary dependent measure, effort (watt increases above one's target), analyses are based on the continuous and 4-min. interval sessions. Subjects were not allowed to increase their intensity on the 2-min. intervals. Similarly, the dependent measure for the 30 sec. sprints was number of intervals completed, which all subjects completed. During Week 20 (last week with >80% of subjects remaining), NAS subjects increased their effort more (M = 8.9 watts) compared to Controls (M = 1.5 watts; d = 0.37) and Conjunctive subjects (M = 3.9 watts) on the 4-min. interval workout. Throughout the duration of the study, NAS subjects always outperformed Controls on the 4-min. intervals. All groups had large significant increases in VO2max from baseline (M = 34) to midpoint (M = 39), then values leveled off from midpoint to final (M = 38). Social connectedness rose significantly from midpoint (M = 2.96) to final (M = 3.48), and those with an SG partner increased their teams' perceptions from midpoint (M = 3.39) to final (M = 4.22). Subjects in NAS (M = 7.85) and Conjunctive conditions (M = 7.26) had higher self-efficacy beliefs than Controls (M = 6.86 2.83) after 1 week with the SG partner. Enjoyment remained stable, above response scale midpoint, across 24 weeks.

Impact of Key Findings: Aim 3: The findings in Aim 3 suggest that having an SG partner, where the subject can intermittently be superior (~15% of time in NAS condition) is most motivating in the most demanding 4-min interval workouts.

Proposed Research Plan for Coming Year: We will prepare manuscripts to submit to journals and present results at professional conferences.

We began our long-term study (Aim 3) before finishing our short-term, 6-day study (Aim 2) in order to stay on schedule.

In Year 3, also we finished Study 1 (Aim 2) to provide a larger sample comparing the 30 min. continuous cycling (at 75% VO2 max) and high-intensity interval cycling involving 4 repetitions of 4 min. (at 90% VO2 max) under SG partner/no partner conditions. We tested an additional 38 chronic exercisers (22 female). Total subjects were 86 (47 female; M age = 44.89, 9.45). We have presented our preliminary findings for both studies at regional, national, and international conferences. We have begun drafting manuscripts to submit to various research journals and conducting more fine-grained analyses.

NOTE: End date changed to 6/30/2016 per NSBRI (Ed., 6/29/16)

NOTE: Extended to 7/31/2016 per NSBRI (Ed., 6/16/16)

Key Findings since last report: In Year 2, we worked on Aims 2 and 3, first testing various design features of the SG partner in an exercise video game: Train Like an Astronaut.

Aim 2 was a short-term study (6 days) to determine the most effective partners to enhance exercise intensity. In Aim 2, we tested an SG exercise partner in one of three modes: (a) a coacting mode, where the subject cycled with the SG partner but whose performance was independent of the SG partner (i.e., they are not teammates), (b) a conjunctive-teammate mode, where scores were based on the slower/weaker performer, and (c) a choice mode that had the option for the participant to choose coacting mode or conjunctive teammate mode for all trials. We tested these modes against an individual control condition. The experiment used the activity routines developed by Ploutz-Snyder that consist of (a) 30 min. of continuous aerobic exercise on a stationary bike at or above 75% of maximum heart rate and (b) high-intensity interval training involving several repetitions of 4 min. at or above 90% maximum heart rate. Thus far in Aim 2, we have tested 19 male and 25 female chronic exercisers ages 30-62 yrs (M = 49.31; ± = 7.77). Subjects were allowed to change intensity (in watts) during each workout, and effort was measured by average watts cycled. Subjects in all conditions increased watts significantly from baseline, but the conjunctive condition (M = 7.47 ± 8.40) trended toward a greater increase compared to the control (M = 5.66 ± 9.11). Thus, with the small sample tested so far, exercising with a conjunctive partner showed a small effect (d = .20) over exercising alone. By end of 6 days, increases in effort above target watts were positively correlated with Enjoyment (r = .47), Self-efficacy (r = .49), and Feelings toward partner (r = .36). Although testing is not finished in the short-term study, there was enough evidence to design the long-term study (Aim 3). Feedback from subjects revealed that they wanted more varied scenery, that an interactive dialogue with the partner during the introduction phase would help in making a social connection, that subjects should match the virtual partner longer during exercise sessions, and that the conjunctive task demand should be more obvious.

In Aim 3, we added more social interactions throughout the sessions, more varied scenery, a longer time where the subject could lessen the gap between him/her and the SG partner, and we added a catch-up and surpass condition. We also added features that made the conjunctive task demand more obvious to the subject. Thus, in the Aim 3 long-term study, we are testing an SG exercise partner in one of two modes: (a) a conjunctive-teammate mode, where scores were based on the slower/weaker performer, and (b) a conjunctive teammate who is sometimes surpassed by the subject, plus an individual control condition. The first cohort of subjects (N = 23; 11 women, 12 men) started in January, 2015 (M age = 46.74; ± = 6.98). Baseline fitness measures of aerobic capacity, body fat, and thigh strength were obtained. Subjects exercise 6 days per week, using the complete aerobic routines developed by Ploutz-Snyder that consist of (a) 30 min. of continuous aerobic exercise on a stationary cycle at or above 75% of maximum heart rate, (b) 4 x 4 min. intervals at or above 90% maximum heart rate with 3 minutes active rest, (c) 6 x 2 minute intervals at varying intensities with 2 minutes active rest, and (d) 30 sec. sprint intervals at maximal effort with 20 seconds active rest. Subjects are in Week 15 of the 24-week testing period. Thus far, one participant in the control condition (female), dropped out at Week 3. We do not have performance or physiological data for Aim 3, as yet. However, initial data on team perceptions and feelings toward one's SG partner indicate that when subjects feel a connection to the partner, they identify more as a team (r = .56) and have more confidence in their partners to help them achieve their own goals (r = .32).

Impact of Key Findings on Hypotheses, Technology Requirements, Objectives, and Specific Aims: Aim 2: The preliminary findings in Aim 2 provided the basis to continue building on the conjunctive task condition, enhance the design effects of the training game, and make the dialogue with the SG partner more interactive.

Proposed Research Plan for the Coming Year: In Year 3, we will continue to pursue Aim 3 with a second cohort, using the same research and training design. Because of summer vacations, we most likely will need to stagger the start dates for subjects, and thus, will not try to change any design aspects of the study.

We have collected and analyzed data on 44 subjects (female = 25) in Study 1. We surpassed the number of women we hoped to recruit to the study, and subjects were similar in age (M = 49.31; SD = 7.77) and aerobic fitness to experienced astronauts. We were not able to recruit as many subjects as we wanted before our deadline for starting Study 2 (Aim 3), but had enough data and feedback from subjects to design the long-term study (Study 2).

For Study 2, we varied the scenery of the bike paths, added interactive dialogue with the partner in the introduction phase, allowed the subject to match the partner's exercise intensity for longer periods of time during sessions, and made the connection between subject and partner more obvious. Study 2 involved two additional high intensity training routines in our exergame: (a) 8 repetitions of 30 s sprints and (b) 6 repetitions of 2-min. each in a ladder fashion of 70, 80, 90, 100, 90, 80% VO2max intervals. High intensity sessions are alternated with 30 min. of continuous aerobic sessions for 6 days per week of training for the 24-week study. Study 2 data collection began in January, 2015, with a cohort of 23 subjects (female = 11, including one dropout). Subjects are similar in age (M = 46.74; ± = 6.95) and aerobic fitness to experienced astronauts. Baseline and midway fitness data have been collected as well as measures of enjoyment, self-efficacy, perceived autonomy, and social connectedness. We are a little more than halfway through data collection for this first cohort and expect to have the second cohort going by July.

Key Findings: We have completed Aim 1 (developing SG exercise partners, interfaced with stationary bikes) to be used in Aim 2. We first demonstrated the software to a small group of astronauts and physical trainers at the Johnson Space Center to help refine the look of the SG partner and then conducted focus groups of highly active male and female athletes/exercisers over 35 years of age to further refine the appearance of the SG partners, the appearance of the exergame interface, and the nature and quality of interactions between participants and their SG partners (e.g., detail of introductions, greetings). Results showed that focus group participants desired a virtual partner who appeared to be fit but not overly muscular and were dressed age-appropriately. Females preferred partners who had muscle tone yet remained feminine while males indicated they wanted a partner who was both a peer and in appropriate shape for the task. Both males and females wanted the option of choosing their SG partner. Males focused on choosing a partner based on the specific workout while females focused on gender, ethnicity, and age. Lastly, in terms of the game interface, both genders preferred realistic scenery that matched the task to be completed (i.e., for a long, continuous bike ride participants preferred outdoor scenery and sounds vs. a track or velodrome simulation for interval training cycling). After conducting focus groups, the games were pilot tested on a convenience sample of eight highly active university students who played the game (i.e., rode a stationary bike on a simulated bike path) for 30 min. each day for 6 days. The first day was without the game and Days 2 – 6 were with the game. Watts (units for cycle power output) was the dependent measure of intensity. Preliminary results showed that participants put forth more effort when they cycled with an exercise video game on Days 2-6 than they did cycling alone without a game. Further, the Watts in the game environment stayed consistently higher.

Impact of Key Findings on Hypotheses, Technology Requirements, Objectives, and Specific Aims: We had no hypotheses for Aim 1. However, our focus group findings suggest that participants want SG partners who look more like themselves. Our pilot work suggests that riding a stationary bike on simulated bike path results in more effort than riding without the simulation.

Proposed Research Plan for the Coming Year: In Year 2, we will pursue Aims 2 and 3, first testing various design features of the SG partner in a short-term study (6 days) to determine the most effective partners to enhance exercise intensity. We will present an SG exercise partner in one of three modes: (a) a coacting mode, where the subject cycles with the SG partner but whose performance is independent of the SG partner (i.e., they are not teammates); (b) a conjunctive-teammate mode, where scores are based on the slower/weaker performer; and (c) a choice mode that has the option for the participant to choose coacting mode or conjunctive teammate mode for all trials. The experiment will use the aerobic routines developed by Ploutz-Snyder that consist of (a) 30 min. of continuous aerobic exercise on a stationary cycle at 75% heart rate max and (b) high-intensity interval training involving several repetitions of 4 min. at 90% heart rate max. Based on findings from the short-term study, we will design the long-term study to test how well the motivation gains we expect to find will persist across a 24-week period using a combination of the 30 min. continuous and various interval routines.

The specific aims of the proposed project are to 1) Develop the software to create SG exercise partners to interface with the exercise equipment; 2) Test various design features of the SG partner within designed exercise games to determine the most effective features for enhancing motivation to exercise, enjoyment, confidence, and social connectedness; and 3) Test whether exercising over an extended time period with an SG partner, compared to exercising alone, results in better aerobic capacity and muscle strength, adherence to the exercise regimen, and enhanced enjoyment in the activity, maintenance efficacy beliefs, and sense of social connectedness.