We modeled a hypothetical set of 100,000 capillaries, applying the proportional risk of obstruction (by branch order) and global risk (30%) of subsequent pruning every 2 hr. (A) Plot showing the relative risk of capillary obstruction decreased over time as at-risk capillaries were lost, leaving a higher proportion of less frequently obstructed capillaries (usually of higher branch orders). Note that this ignores possible changes in branch order due to pruning as capillaries were modeled independently with fixed branch orders. (B) Relative fraction of each capillary branch order over time. Over 10,000 hr the fraction of frequently obstructed capillaries were reduced while the number of rarely obstructed capillaries stayed constant and thus comprised a greater fraction over time. (C) Comparing our predicted and experimentally derived estimate of capillary loss to previous aging studies in rodents. Note that some previous studies used histochemical markers to label vasculature such as alkaline-phosphatase, which might decrease in aged animals’ independent of actual vessel less, leading to over-estimates of capillary loss. Data shown was calculated from published numbers in each study, ±S.D. when available.