Sato 2005a.

Methods	Randomised, double‐blind, placebo‐controlled trial using parallel group design (two intervention groups).
Participants	Country: Japan. Number of participants randomised: 96 hospitalised elderly women with post stroke hemiplegia mean age 74.1 years. Inclusion criteria: hospitalised elderly women with post stroke hemiplegia who had first‐ever cerebral infarction or haemorrhage more than two years before and were in a convalescent stage with post‐stroke hemiplegia. Exclusion criteria: dementia, total disability, or hospitalisation of less than two years' duration, receiving any drugs known to alter vitamin D metabolism, such as anticonvulsants, calcium, or vitamin D, during the 12 months preceding the trial.
Interventions	Participants were randomly assigned to receive: Intervention group 1: vitamin D2 (1000 IU) daily (n = 48); Intervention group 2 (Control group): matched placebo tablet daily (n = 48); for a two‐year period.
Outcomes	The primary outcome measure was number of falls. Secondary outcome measures were muscular strength and morphological changes of muscle.
Stated aim of study	"To evaluate the efficacy of vitamin D2 therapy in reducing the risk of falls in elderly women with stroke. Histochemical examination of skeletal muscles was performed to assess the effect of the therapy."
Notes	Additional information on mortality was received through personal communication with Dr Yoshiro Sato (05.02.2009).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Sequence generation was achieved using computer random number generation or a random number table.
Allocation concealment (selection bias)	Low risk	Allocation was controlled by a central and independent randomisation unit so that intervention allocations could not have been foreseen in advance of, or during, enrolment.
Blinding (performance bias and detection bias) All outcomes	Low risk	The trial was described as blinded, the parties that were blinded, and the method of blinding was described, so that knowledge of allocation was adequately prevented during the trial.
Incomplete outcome data (attrition bias) All outcomes	Low risk	The numbers and reasons for dropouts and withdrawals in all intervention groups were described.
Selective reporting (reporting bias)	Low risk	Pre‐defined, or clinically relevant and reasonably expected outcomes are reported on.
Industry bias	Unclear risk	The source of funding is not clear.
Other bias	Low risk	The trial appears to be free of other components that could put it at risk of bias.