Six studies (n=1,857) were included in the review. Sample sizes ranged from 110 to 702 participants. One study fulfilled all or most criteria, four fulfilled some criteria and one fulfilled few or no criteria. Only two studies used blinded outcome assessors. Drop-out rates ranged from 3% to 29%.
Three studies reported a statistically significant reduction in risk of falling for the Tai Chi group compared to control/other treatments, in relatively healthy populations living in the community. Three studies of more frail dependent patient populations reported no significant differences, although there were some differences for subpopulations including an increased risk of falling with Tai Chi in frail patients (one study).