Alexander Tolas

All Publications

• Fine-tuning LLMs in behavioral psychology for scalable health coaching. NPJ cardiovascular health Mantena, S., Johnson, A., Oppezzo, M., Schütz, N., Tolas, A., Doijad, R., Mattson, C. M., Lawrie, A., Ramirez-Posada, M., Schmiedmayer, P., Linos, E., King, A. C., Rodriguez, F., Kim, D. S., Ashley, E. A. 2025; 2 (1): 48

  Abstract

  Personalized, smartphone-based coaching improves physical activity but relies on static, human-crafted messages. We introduce My Heart Counts (MHC)-Coach, a large language model fine-tuned on the Transtheoretical Model of Change. MHC-Coach generates messages tailored to an individual's psychology (their "stage of change"), providing personalized support to foster long-term physical activity behavior change. To evaluate MHC-Coach's efficacy, 632 participants compared human-expert and MHC-Coach interventions encouraging physical activity. Among messages matched to an individual's stage of change, 68.0% (N = 430) preferred MHC-Coach-generated messages (P < 0.001). Blinded behavioral science experts (N = 2) rated MHC-Coach messages higher than human-expert messages for perceived effectiveness (4.4 vs. 2.8) and Transtheoretical Model alignment (4.1 vs. 3.5) on a 5-point Likert scale. This work demonstrates how language models can operationalize behavioral science frameworks for personalized health coaching, showing the potential for promoting long-term physical activity and reducing cardiovascular disease risk at scale.

  View details for DOI 10.1038/s44325-025-00083-5

  View details for PubMedID 40994879

  View details for PubMedCentralID PMC12454129

• Fine-tuning Large Language Models in Behavioral Psychology for Scalable Physical Activity Coaching. medRxiv : the preprint server for health sciences Mantena, S., Johnson, A., Oppezzo, M., Schuetz, N., Tolas, A., Doijad, R., Mattson, C. M., Lawrie, A., Ramirez-Posada, M., Linos, E., King, A. C., Rodriguez, F., Kim, D. S., Ashley, E. A. 2025

  Abstract

  Personalized, smartphone-based coaching improves physical activity but relies on static, human-crafted messages. We introduce My Heart Counts (MHC)-Coach, a large language model fine-tuned on the Transtheoretical Model of Change. MHC-Coach generates messages tailored to an individual's psychology (their "stage of change"), providing personalized support to foster long-term physical activity behavior change. To evaluate MHC-Coach's efficacy, 632 participants compared human-expert and MHC-Coach text-based interventions encouraging physical activity. Among messages matched to an individual's stage of change, 68.0% (N=430) preferred MHC-Coach-generated messages (P < 0.001). Blinded behavioral science experts (N=2) rated MHC-Coach messages higher than human-expert messages for perceived effectiveness (4.4 vs. 2.8) and Transtheoretical Model alignment (4.1 vs. 3.5) on a 5-point Likert scale. This work demonstrates how language models can operationalize behavioral science frameworks for personalized health coaching, promoting long-term physical activity and potentially reducing cardiovascular disease risk at scale.

  View details for DOI 10.1101/2025.02.19.25322559

  View details for PubMedID 40034753

• Unlocking insights: Clinical associations from the largest 6-minute walk test collection via the my Heart Counts Cardiovascular Health Study, a fully digital smartphone platform. Progress in cardiovascular diseases Kim, D. S., Schuetz, N., Johnson, A., Tolas, A., Mantena, S., O'Sullivan, J. W., Hershman, S. G., Myers, J. N., Christle, J. W., Oppezzo, M., Linos, E., Rodriguez, F., Mattsson, C. M., Wheeler, M. T., King, A. C., Taylor, H. A., Ashley, E. A. 2025

  Abstract

  The six-minute walk test (6MWT) is a prognostic sub-maximal exercise test used clinically as a measure of functional capacity. With the emergence of advanced sensors, 6MWTs are being performed remotely via smartphones and other devices. The My Heart Counts Cardiovascular Health Study is a smartphone application that serves as a digital platform for studies of human cardiovascular health, and has been used to perform 30,475 6MWTs on 8922 unique participants.As our 30,475 6MWTs represent the largest such collection of data available, we sought to identify associations with measured demographic and clinical variables with 6MWT distance at enrollment and separately determine if use of the My Heart Counts smartphone application led to changes in 6MWT distance.We present the public data release of our 30,475 6MWTs and the launch of a webpage-based data viewer of summary-level statistics, to compare the functional capacity of an individual by their age, gender, height, weight, and disease status (https://mhc-6mwts.streamlit.app). Using multivariable regression, we report associations of demographic and clinical variables with baseline 6MWT distance (N = 3606), validating prior associations with age, male gender, height, and baseline physical activity level with 6MWT distance. We also report associations of 6MWT baseline distance with employment status (+12.4 m ±4.9 m, P = 0.011) and feeling depressed (-3.65 m, ±0.79 m, P < 0.001). We separately found that cardiovascular disease status was significantly associated with decreased 6MWT distance for atrial fibrillation (-24.9 m ±7.8 m, P = 0.0013), peripheral artery disease (-41.7 m ±12.5 m, P < 0.001), and pulmonary arterial hypertension (-76.3 m ±24.8 m, P = 0.0022). Heart failure was associated with decreased 6MWT distance but was not statistically significant (-25.5 m ±14.5 m, P = 0.078). In a subset of participants who conducted repeat 6MWTs separated by at least 1 week but no greater than 3 months (N = 1129), we found that use of the My Heart Counts app was associated with a statistically significant increase in 6MWT distance (+17.5 m ±7.85 m, P < 0.001).We validate previously identified associations from clinic-performed 6MWTs, demonstrating the utility of a mobile method in collecting 6MWT data for clinicians and researchers. We also demonstrate that use of the My Heart Counts app is associated with small, but significant increases in 6MWT distance. Given the importance of 6MWTs in assessment of functional capacity, our publicly-available data will serve an important purpose as a health and disease-specific reference for investigators worldwide.

  View details for DOI 10.1016/j.pcad.2025.01.010

  View details for PubMedID 39884325

• StandUPTV: Preparation and optimization phases of a mHealth intervention to reduce sedentary screen time in adults. Contemporary clinical trials Keadle, S., Hasanaj, K., Leonard-Corzo, K., Tolas, A., Crosley-Lyons, R., Pfisterer, B., Legato, M., Fernandez, A., Lowell, E., Hollingshead, K., Yu, T. Y., Phelan, S., Phillips, S. M., Watson, N., Hagobian, T., Guastaferro, K., Buman, M. P. 2023: 107402

  Abstract

  Recreational sedentary screen time (rSST) is the most prevalent sedentary behavior for adults outside of work, school, and sleep, and is strongly linked to poor health. StandUPTV is a mHealth trial that uses the Multiphase Optimization Strategy (MOST) framework to develop and evaluate the efficacy of three theory-based strategies for reducing rSST among adults. This paper describes the preparation and optimization phases of StandUPTV within the MOST framework. We identified three candidate components based on previous literature: (a) rSST electronic lockout (LOCKOUT), which restricts rSST through electronic means; (b) adaptive prompts (TEXT), which provides adaptive prompts based on rSST behaviors; and (c) earning rSST through increased moderate-vigorous physical activity (MVPA) participation (EARN). We also describe the mHealth iterative design process and the selection of an optimization objective. Finally, we describe the protocol of the optimization randomized controlled trial using a 23 factorial experimental design. We will enroll 240 individuals aged 23-64 y who engage in >3 h/day of rSST. All participants will receive a target to reduce rSST by 50% and be randomized to one of 8 combinations representing all components and component levels: LOCKOUT (yes vs. no), TEXT (yes vs. no), and EARN (yes vs. no). Results will support the selection of the components for the intervention package that meet the optimization objective and are acceptable to participants. The optimized intervention will be tested in a future evaluation randomized trial to examine reductions in rSST on health outcomes among adults.

  View details for DOI 10.1016/j.cct.2023.107402

  View details for PubMedID 38000452

• Personalized digital behaviour interventions increase short-term physical activity: a randomized control crossover trial substudy of the MyHeart Counts Cardiovascular Health Study. European heart journal. Digital health Javed, A., Kim, D. S., Hershman, S. G., Shcherbina, A., Johnson, A., Tolas, A., O'Sullivan, J. W., McConnell, M. V., Lazzeroni, L., King, A. C., Christle, J. W., Oppezzo, M., Mattsson, C. M., Harrington, R. A., Wheeler, M. T., Ashley, E. A. 2023; 4 (5): 411-419

  Abstract

  Physical activity is associated with decreased incidence of the chronic diseases associated with aging. We previously demonstrated that digital interventions delivered through a smartphone app can increase short-term physical activity.We offered enrolment to community-living iPhone-using adults aged ≥18 years in the USA, UK, and Hong Kong who downloaded the MyHeart Counts app. After completion of a 1-week baseline period, e-consented participants were randomized to four 7-day interventions. Interventions consisted of: (i) daily personalized e-coaching based on the individual's baseline activity patterns, (ii) daily prompts to complete 10 000 steps, (iii) hourly prompts to stand following inactivity, and (iv) daily instructions to read guidelines from the American Heart Association (AHA) website. After completion of one 7-day intervention, participants subsequently randomized to the next intervention of the crossover trial. The trial was completed in a free-living setting, where neither the participants nor investigators were blinded to the intervention. The primary outcome was change in mean daily step count from baseline for each of the four interventions, assessed in a modified intention-to-treat analysis (modified in that participants had to complete 7 days of baseline monitoring and at least 1 day of an intervention to be included in analyses). This trial is registered with ClinicalTrials.gov, NCT03090321.Between 1 January 2017 and 1 April 2022, 4500 participants consented to enrol in the trial (a subset of the approximately 50 000 participants in the larger MyHeart Counts study), of whom 2458 completed 7 days of baseline monitoring (mean daily steps 4232 ± 73) and at least 1 day of one of the four interventions. Personalized e-coaching prompts, tailored to an individual based on their baseline activity, increased step count significantly (+402 ± 71 steps from baseline, P = 7.1⨯10-8). Hourly stand prompts (+292 steps from baseline, P = 0.00029) and a daily prompt to read AHA guidelines (+215 steps from baseline, P = 0.021) were significantly associated with increased mean daily step count, while a daily reminder to complete 10 000 steps was not (+170 steps from baseline, P = 0.11). Digital studies have a significant advantage over traditional clinical trials in that they can continuously recruit participants in a cost-effective manner, allowing for new insights provided by increased statistical power and refinement of prior signals. Here, we present a novel finding that digital interventions tailored to an individual are effective in increasing short-term physical activity in a free-living cohort. These data suggest that participants are more likely to react positively and increase their physical activity when prompts are personalized. Further studies are needed to determine the effects of digital interventions on long-term outcomes.

  View details for DOI 10.1093/ehjdh/ztad047

  View details for PubMedID 37794870

  View details for PubMedCentralID PMC10545510

• Evaluation of Within- and Between-Site Agreement for Direct Observation of Physical Behavior Across Four Research Groups JOURNAL FOR THE MEASUREMENT OF PHYSICAL BEHAVIOUR Keadle, S., Martinez, J., Strath, S. J., Sirard, J., John, D., Intille, S., Arguello, D., Amalbert-Birriel, M., Barnett, R., Thapa-Chhetry, B., Cox, M., Chase, J., Dooley, E., Marcotte, R., Tolas, A., Staudemayer, J. W. 2023; 6 (3): 176-184

  View details for DOI 10.1123/jmpb.2022-0048

  View details for Web of Science ID 001289031600002