The recovery of walking ability is crucial to promote independence in daily living and is one of the major goal of neuromotor rehabilitation.

Currently, standard rehabilitative programs are usually based on straight-walking training (SWT) and the assessment of their effects is performed through functional scales based on straight-walking trajectories, e.g. Timed Up and Go (TUG), 10 meters walking test (10mWT).

Curved-walking training (CWT) may be interesting to provide an ecological and challenging context during rehabilitation. Indeed, CWT is based on demanding neural processes that drive an asymmetrical contribution at lower limb level, challenging balance ability and complex adaptation such as body weight shifting in response to centrifugal force and production of different step lengths.

Up to now, literature has investigated CWT in healthy adults in terms of muscular activation, kinematics and kinetics of the movement. Results showed that CWT needs a different biomechanical strategy with respect to SWT. Nevertheless CWT has not been investigated in pathological adults.

The present study aims at assessing the effectiveness of a rehabilitative physical therapy based on CWT with respect to traditional SWT for the recovery of locomotor abilities in neurological and orthopaedic patients.

The hypothesis is that a training based on curved-walking is ecologically meaningful and may be superior with respect to standard training in improving balance, walking abilities, and independence in activity of daily live of patients.

A secondary aim of the project is to propose an innovative functional scale based on the timed up and go on curved trajectory (CTUG), and to determine its reliability and responsiveness, establishing the minimum Detectable Change (MDC) and the Minimal Clinically Important Difference (MCID).

A single-blind randomized controlled study is being carried out on three different populations:

• Post-acute stroke patients
• Idiopathic Parkinson Disease
• Femoral fracture

A healthy group is also being recruited to provide reference values of CTUG. For each of the three populations, subjects are randomized into two groups. The experimental one performs a novel rehabilitative program composed by a 30-minute training on curved trajectory ("S" trajectory composed by two semicircle with a radius of 1.2 m) in addition to usual care. The control group performs an equal dose of traditional treatment on straight trajectories.

Both groups undergo 20 90-minutes sessions of training (three times a week for seven weeks).

Participants are evaluated at baseline (T0), after training (T1), and at a three-months follow-up visit (T2).

The primary outcome measure is the 10mWT (minimal clinically important difference of 0.16 m/s identified by Tilson and colleagues). On the basis of this measure, a sample size of 70 subjects for each population was computed.

Inclusion Criteria:

• Post-acute stroke patients experiencing a stroke less than 6 months before recruitment
• Idiopathic Parkinson Disease with disability level from moderate to severe (modified Hoehn & Yahr scale 2.5-4)
• Femoral fracture, less than 1 month form surgery

Exclusion Criteria:

• Cognitive deficits (Mini Mental Scale Evaluation < 24)
• Hemineglect
• Modified Ashworth Scale of lower limb >2
• Unstable pharmacological treatment for Parkinson's Disease during the 15 days before the recruitment
• cardiopathic conditions
• metabolic conditions (e.g. dialysis) that prevent patients from aerobic training
• Previous history of major neurological, vascular, musculoskeletal disorders
• Body Mass Index > 30 Kg/m2
• Invasive pharmacological treatment or surgery for Parkinson's disease
• lower limb pain (VAS >3)