Fig. 3

Forest plot and meta-analysis of restricted activity and mortality, with a long follow up (≥ 25). Analysis is sub-grouped by the type of restricted activity (bed rest, restricted movement, and activities of daily living (ADL) dependency). Square markers indicate the point estimate of the effect size within the different studies, with the whiskers indicating the confidence intervals. The size of the box correlates to the inverse variance of the effect estimate, which indicates the weight given to the study in the pooled analyses. The diamond markers indicate the pooled effect estimate across sub-groups. Superscript denotes the control for each study. * = Not adjusted. ** = Adjusted.

¹ No cut-down in usual activity. ²Independent for ADL, or walking. ³Spending day moving around. ⁴ Spending day sitting or moving around. ⁵Less than the defined cut off (16 h) hours in bed. ⁶Less than the defined cut off (6 days) bed rest