Physical activity (PA) is a critical, modifiable determinant of health. The relationship between PA behavior and health outcomes has been increasingly examined using objectively measured accelerometer data. Accelerometer□derived PA features have been consistently associated with a wide range of health outcomes and have shown potential in predictive modeling. However, traditional summary□statistic features are limited in their ability to capture the high□resolution and dynamic accelerometer data, and the predictive utility of accelerometer data across diverse health outcomes has not been comprehensively evaluated. Here, we present PABformer, a foundation model for accelerometer data pretrained on the UK Biobank to learn behavior□level representations of PA. PABformer employs a channel□separation strategy to disentangle heterogeneities inherent in accelerometer data and leverages a multi□channel Transformer encoder to extract channel□specific representations. We finetuned PABformer on benchmark tasks, including demographic attribute inference and Parkinson’s disease classification, where it consistently outperformed baseline models across multiple evaluation metrics. Extending beyond benchmarks, we applied the pretrained model to 157 chronic diseases spanning seven organ systems, as well as all□cause mortality. PABformer outperformed traditional covariate□based models in 48% of prevalent diseases, 54% of incident diseases, and mortality, with particularly marked improvements in three prevalent and nine incident diseases. These findings establish PABformer as a generalizable and scalable foundation model for accelerometer data, enhancing the utility of accelerometer measurements for health outcome assessment and offering the potential to advance disease risk prediction and enable precision health applications.
Competing Interest StatementThe authors have declared no competing interest.
Funding StatementWe sincerely thank the participants and staff of the UK Biobank for the invaluable contributions. K. H. and M.W. were supported by M.W.s startup fund from the Institute for Heart and Brain Health. This research was supported in part through computational resources and services provided by Advanced Research Computing at the University of Michigan, Ann Arbor.
Author DeclarationsI confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.
Yes
The details of the IRB/oversight body that provided approval or exemption for the research described are given below:
Ethics committee/IRB of UK Biobank gave ethical approval for the use of the UK Biobank data. This research is conducted under the approved application code 197947, with the data released to the University of Michigan.
I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.
Yes
I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).
Yes
I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable.
Yes
Data availabilityAll accelerometer, phenotype, and health outcome data used in this study were obtained from the UK Biobank. Access to these data is available through the UK Biobank Research Analysis Platform (https://www.ukbiobank.ac.uk/) upon application and approval.
Comments (0)