The ACSM defines FDA approved Drug Library physical activity as body movement that is produced by the contraction of skeletal muscles and that increases energy expenditure ( Garber et al 2011) and goes on to affirm that physical activity broadly encompasses exercise, sports, and physical activities. We acknowledge that most trials included in this review centred on investigating the effectiveness of structured exercise, and that sub-grouping trials according to the type of exercise might yield different results, however this was outside the scope of our review. We also acknowledge the diversity of exercise programs assessed by the included trials would potentially introduce

unwanted heterogeneity in our pooled analyses. However, a statistically significant level of heterogeneity (p = 0.006) was only observed in the pooled analysis of endurance. We recommend caution when interpreting these results. We have based our conclusions about the size buy PLX3397 of effects of interventions on the widely used cut offs for clinical significance proposed by Cohen (1988), suggesting that standardised effect sizes of 0.2 should be considered small, those of 0.5 considered moderate, and those of 1.0 considered large (Cohen 1988). However, variations exist (Norman et al 2003), and by using different cut-offs

we could have concluded differently. These benchmarks have been derived mainly from social science research; interpretations mainly reflect the opinions of researchers, rather than consumers (Ferreira et al 2012). Many of the included trials were small and conducted in a research setting. The strength of a meta-analysis is that it can combine small trials that would not be individually powered to detect statistically significant effects of interventions.

However the small size and research setting of many included trials means that it is difficult to draw conclusions about the feasibility of widespread implementation of these interventions in community settings. The majority of the included trials did not appear to use blinded outcome assessment or concealed random allocation to groups. It is possible that this would increase the size of the effects through seen. However, even if the true effect of physical activity intervention in this population is smaller than seen in the review we suggest that it still likely to be large enough to be useful. No trials of the effectiveness of physical activity programs on short-term falls in middle-aged people were found. Although people in this age group do experience falls, which may be indicative of early problems with balance and strength, the overall incidence of falls is lower than in people aged 65 and older. Therefore very large sample sizes would be required to assess effects of physical activity on falls in this population.