PROJECT SUMMARY/ABSTRACT People seek out upper limb rehabilitation services to improve performance of activities in daily life, with performance defined by the World Health Organization ICF model as what someone actually does in the unstructured, free-living environment. Wearable, motion sensors now allow for direct measurement of upper limb performance in daily life, but performance is rarely measured or considered in research and care. Data from our 2nd cycle make it clear that improvements in impairments and/or functional capacity assessed in the structured environment of the clinic or laboratory most often do not translate to improvements in performance of activities in daily life. These data point to the clear need to move assessment of performance from the specialized research realm into more universal use in upper limb research studies and clinical care. A major barrier to more wide-spread use of sensors for routine upper limb performance assessment is lack of clinical validation. An elegant solution to the many clinical validation challenges is to move from single variables in individual populations to multivariate categories of upper limb performance in daily life that span patient populations. Our central hypothesis is that, despite the immense variability in biological conditions and their resulting upper limb impairments and functional capacity profiles, upper limb performance in daily life can be organized into a relatively small number of categories. Defined, validated categories could be used across populations to push upper limb rehabilitation research forward and personalize clinical decision-making. We now propose to expand and validate preliminary categories from stroke into a variety of conditions that send people to upper limb rehabilitation services and result in upper limb disability. Wearable sensors will be used to assess upper limb performance in a heterogeneous, longitudinal, observational cohort (N=340) of persons with (stroke, multiple sclerosis, UL fracture, adhesive capsulitis, breast cancer) and without UL disability (controls), and a retrospective, pediatric cohort (n ≥ 400) to address three aims. Aim 1 will identify and validate categories of upper limb performance in daily life that can be applied across conditions that cause upper limb disability in adults. Aim 2 will evaluate responsiveness to change across categories. And Aim 3 will explore how categories identified in adults map to upper limb performance in childhood. Expected outcomes at the end of the proposed cycle are validated categories of upper limb performance that are responsive to change and span patient populations. These categories (and the software algorithms to derive them) will immediately be available for intervention studies to measure outcomes that are most relevant to patients. With anticipated advances in hardware and software over the next few years, future work will implement validated categories readily into clinical rehabilitation car...