Your Step Count Can Now Predict Your Biological Age With Surprising Accuracy

Biological age clocks have traditionally required expensive laboratory testing — blood panels, epigenetic methylation assays, or metabolic profiling — making widespread adoption difficult. MoveIt! Age addresses this gap by predicting biological age using only step count data from consumer wearables, making it potentially accessible to anyone with a fitness tracker or smartphone. The clock was trained on the United States National Health and Nutrition Examination Survey (NHANES) dataset, one of the largest population health databases available, and validated in two independent international cohorts across very different populations.

The model was constructed using three inputs: chronological age, maximum daily step count, and step count variability across the measurement period. These features were used to predict PhenoAge, a well-established blood biochemistry biological age score derived from nine clinical biomarkers. In the NHANES testing dataset, MoveIt! Age achieved a correlation of r=0.97 with chronological age and a root mean square error (RMSE) of 5.4 years — strong predictive performance for a model using only movement data. Importantly, delta MoveIt! Age (predicted biological age minus chronological age) was more significantly associated with all-cause mortality than PhenoAge in the same dataset, suggesting the step-based clock captures health-relevant variance beyond simple aging patterns.

The MitoHealth cohort (N=55, healthy young and older Dutch adults) provided a critical molecular validation. Delta MoveIt! Age correlated significantly with muscle NAD+ levels (r=−0.37, p=0.023), while chronological age showed no significant association (p=0.416). NAD+ is a key coenzyme in cellular energy metabolism whose decline with age is implicated in mitochondrial dysfunction and is a major target of longevity interventions including NMN and NR supplementation. This finding links step count patterns to cellular-level aging biology in a way that chronological age alone cannot capture.

The RESORT cohort (N=145, geriatric rehabilitation inpatients from Australia) enabled direct comparison of MoveIt! Age against two established hematological aging clocks: SenoClock-BloodAge and PhenoAge. Delta MoveIt! Age demonstrated stronger associations with physical function outcomes including frailty status, handgrip strength, and performance on functional assessments. This is clinically significant: in a geriatric rehabilitation setting, predicting functional outcomes matters more than laboratory aging scores. The fact that a wearable-only tool matched or exceeded blood-based clocks in this domain is a notable finding.

The study has important implications for scaling biological age assessment. Blood-based clocks require phlebotomy, laboratory processing, and clinical infrastructure. MoveIt! Age requires only a wearable device and a brief monitoring period. This could enable biological age tracking in low-resource settings, remote monitoring programs, rehabilitation centers, and large-scale population health studies. The authors envision integration into consumer health platforms and clinical monitoring tools. Key caveats include the relatively small validation cohort sizes, the cross-sectional design limiting causal inference, and the fact that step count is itself influenced by environmental and motivational factors beyond pure physical capacity. Replication in larger and more diverse populations is warranted before broad clinical deployment.