Twelve studies (n=763, range five to 127) were included in the review: seven RCTs; four repeated measures with randomised cross-over (RXT); and one non-RCT. Studies scored between 2 and 4 on the quality rating scale; six were rated as strong, five as moderate and one as weak. There were some discrepancies between details reported in the text and those reported in the tables (we have mainly used information provided in the text).
Six of the 12 studies showed that interventions were statistically significantly more effective than attention controls or other treatments.
Activity/recreation (one RCT, one cross-over trial): The RCT found that contentment and interest was greater in the three treatment arms (activities of daily living care and psychosocial activities, activities of daily living care alone and psychosocial activities alone) compared to attention control (personal interaction) (p<0.05), but no intervention was superior over the other. The cross-over trial found a moderate effect (0.54) in favour of activity/recreation suited to interests and skills compared to activity/recreation suited to interest or skills.
Exercise, movement and relaxation (two RCTs, one cross-over trial): One RCT found that patients who received an intervention that comprised strength, balance and flexibility exercises was statistically significantly more effective than a walking group and a social conversation control group (p<0.05). One RCT and one cross-over trial reported no statistically significant differences between groups.
Music and sensory enrichment interventions (one RCT, two cross-over trials): It appeared that one cross-over trial found a moderate effect (0.75) on irritability, depression and fear in favour of patients who received music while eating their dinner compared to patients who did not receive music (although this was not supported by the data provided in the table). One RCT and one cross-over trail reported no statistically significant differences between treatment groups.
Reminiscence and validation therapy (two RCTs): One RCT showed a moderate effect (0.5) on depression (but not irritability and social withdrawal) in favour of validation therapy compared to attention control (social group). The second RCT showed a statistically significant treatment by time interaction (p<0.05) in favour of reminiscence therapy compared to attention control (social group)
No statistically significant differences were found between dementia carer education interventions and attention controls (one RCT, 1 non-RCT).