Children with cerebral palsy have greater stochastic features present in the variability of their gait kinematics.

Researchers filmed the kids with spastic diplegic cerebral palsy and 12 typically-developing peers walking. They used a new math tool called Langevin analysis to split each child's stride-to-stride changes into two buckets: predictable patterns (deterministic) and random noise (stochastic).

The goal was to see if the extra gait wobble in CP comes from hidden patterns the brain could fix, or from plain neuromuscular noise that is harder to train away.

Kids with CP had almost three times more stride-to-stride variability than their TD peers. Almost all of that extra scatter was random noise, not a hidden pattern.

In plain words: the limp is less about 'wrong rhythm' and more about 'static on the line' between brain and muscles.

Antunes et al. (2016) saw immediate gait gains after one hippotherapy session. Their short-term improvement shows some features can be tweaked, but Matson et al. (2013) warn the underlying noise remains—so hippotherapy benefits may fade quickly.

Sacco et al. (2010) used fancy stats to split autism behaviors into four clean parts. Matson et al. (2013) pull the same trick on gait, giving CP therapists a clearer picture of what is trainable versus what is background noise.

Dagnan et al. (2005) built switch-access tools for clients with minimal movement. Both papers push the field toward assessments that work even when the body gives noisy data.

When you see a child with CP walk unevenly, blame neuromuscular noise, not poor timing. This reframes goals: reduce noise with strength or orthotics first, then add rhythm drills. It also justifies using objective noise metrics to show small but real progress that step-count alone can miss.