Barnes, K. R., Hopkins, W. G., McGuigan, M. R., & Kilding, A. E. (2013). Effects of Different Uphill Interval-Training Programs on Running Economy and Performance. Int J Sports Physiol Perform.
An abstract can be found here on PubMed.
To summarize, 15 male and 5 female trained distance runners were evaluated for VO2max, RE, countermovement-jump peak force, and 5k time. The 20 subjects were then assigned to 1 of 5 groups. Each group completed a specific type of uphill interval training twice a week in addition to their normal training routine.
Below is a list of the five groups along with a brief description of the intervals they performed.
Group 1 (n = 3) -- 12-24 reps of 8-12s hill sprints, ~60s recovery; 18% gradient
Group 2 (n = 5) -- 8-16 reps of 40-45s, ~ 90s recovery; 15% gradient
Group 3 (n = 5) -- 5-9 reps of 2-2.5 min, ~ 4.5 min recovery; 10% gradient
Group 4 (n = 4) -- 4-7 reps of 4-5 min, ~6-7 min recovery; 7% gradient
Group 5 (n = 3) -- 1-2 reps of 10-25 min, equal recovery; 4% gradient
After completing 6 weeks of training with their respective intervals, the subjects returned to the lab to evaluate VO2max, RE, peak power, and 5k time.
Results: At best, one group improved their VO2max 4.1%. 5k time was improved ~2.0% (~20s) in all groups. The most interesting piece of information to me is that the group performing the 10s hill sprints showed the greatest improvement in countermovement-jump peak force (29%) and RE (2.4%). This is the first study to show that hill sprints improve RE. But this study also has some limitations...
First, there was no control group. So, one cannot say that uphill intervals are better than intervals on level-ground -- there is actually some evidence to suggest the opposite.
Second -- Statistical power. With only 3 subjects in a group, seeing a 2.4% improvement in RE, what is the likelihood that their result is not due to an error? I'm not sure, because we are not provided with p values, only percent changes from pre- and post-test measurements; and I am not a statistician. All this means is we cannot be certain that the results seen in this group can be attributed to the intervention (hill sprints); and we cannot definitively say that hill sprints improve running economy.
Third -- Treadmills. It appears that all interval training was conducted on treadmills. I understand the need to standardize speed, gradient, and intensity, but I think there are limitations to treadmill training. Hill sprints for example: hill sprints are balls-to-the-wall effort, not calculated at 120% vVO2max using an equation to extrapolate simulated velocity due to the incline of the treadmill. Just go outside and run up an 18% gradient for 8-10s as fast as possible -- Only then will you get a true maximal effort.
Fourth -- Male and female subjects? We are not told if any effort was made to accommodate for the women's menstrual cycles. Further, we do not know if the five female subjects were evenly distributed among the groups. And who knows? Women may not respond to uphill interval training in the same way men do.
But, despite the limitations of the study, it has shed some light on the topic of hill sprints and running economy and their relationship to 5k performance. In the academic world, we can now say that there is some evidence to suggest that short hill sprints do elicit similar improvements in peak power and RE as plyometrics and resistance training. But it seems coaches and athletes have known this for the last 50 years. It's only fitting that this study came out of New Zealand -- home to one of the pioneers of hill training, Arthur Lydiard.
Another interesting tid-bit from this study is that VO2max was improved the most between groups 3 and 4. Why did these groups see the largest improvements in VO2max? The authors theorize it was because these groups were training at intensities simulating VO2max.
I would like to see a comparison of muscle damage between runners doing level-ground interval training and uphill interval training. Just as coaches have been touting the benefits of uphill running for running economy, they also say uphill running is easier on the legs and may help prevent injuries...
In summary -- uphill interval training may be beneficial for runners. It can improve performance, RE, peak power output, increase stride rate and so on... but then again, so can level-ground interval training...