After another great conference I have just been to I thought I should point out this great and useful piece of evidence about variability in running and prevention of injury. Read for yourself below, but essentially it shows that runners with more variability in their time on the ground and stride length had less injury. 

This implies that those with less variability in their gait are consistently loading the same area of soft tissue and therefore making this area more prone to overload and injury. So if your pattern of running is consistently poor and you keep loading the same way, you could be setting yourself up for injury with every step you take.

The ability to have the strength to vary your gait, control yourself over different terrains and conditions underfoot and be adaptable must be key to keep us all injury free.

FLUCTUATIONS IN STRIKE INDEX AND SPATIOTEMPORAL PARAMETERS IN PREVIOUSLY INJURED VS UNINJURED RUNNERS
R Mann, K Meijer, L Malisoux, R Brunner, A Urhausen, C Nuehrenboerger, D Theisen
Br J Sports Med 2014;48:632-633 doi:10.1136/bjsports-2014-093494.196
Abstracts from the IOC World Conference on Prevention of Injury & Illness in Sport, Monaco 2014
Background: Long-range correlations in running stride interval have been proposed as an indicator of running-related injury (RRI) and found to change with running speed. They have not yet been used to distinguish running style between runners having previously sustained a RRI or not.
Objectives: To compare long-range correlations and stride-to-stride variability of strike index (SI) and spatiotemporal parameters in previously injured and uninjured runners over various running speeds. We hypothesized significant effects of groups and running speed for these parameters.
Design: Retrospective, case-control study.
Setting: Recreational runners.
Participants: 69 regular male and female runners (currently asymptomatic) either having sustained a RRI during the previous year (n=26) or not (n=43).
Risk factor assessment: Five 2-minute incremental running intervals were performed at 80, 90, 100, 110 and 120% of the individual typical running speed on a treadmill. SI, contact time (CT), flight time (FT), stride time (ST), duty factor (CT/ST), stride length (SL) and frequency were measured using an insole-based system. Coefficient of variance (CV) and the scaling exponent α of the detrended fluctuation analysis were calculated, and effects were analysed using repeated measures analyses of variance.
Main outcome measurements: RRI, defined as a pain or complaint to the lower limbs, progressive in nature and influencing running activity for >1 week.
Results: The RRI group displayed lower CV than the controls for CT (P=.01) and FT (P=.027). Significant differences in α between normalised speeds were found for all variables (P<.048) except SL. CV differed significantly between normalised speeds for ST and SL. No significant interactions were found.
Conclusions: Previously uninjured runners display significantly greater CV in CT and FT. Long-range correlations of SI and spatiotemporal parameters change according to running speed, but are not significantly different between previously injured and uninjured runners. Further prospective study to confirm these findings is warranted.