Running science nerd alert.

by Thomas Solomon PhD and Matt Laye PhD

September 2020.

Each month we compile a short-list of recently-published papers (full list here) in the world of running science and break them into bite-sized chunks so you can digest them as food for thought during your training sessions. Welcome to this month's installment of our "nerd alert". We hope you enjoy it.

Share this nerd alert:

Click the title of each article to "drop-down" the summary.

Full paper access: https://pubmed.ncbi.nlm.nih.gov/32887849/

What was the hypothesis or research question?
Humans naturally select an optimal stride frequency that minimizes metabolic cost. Even during fatiguing efforts of up to 1-h of high-intensity running, athletes self-select a near-optimal stride frequency. This study aimed to determine whether runners continue to do so during a 6-hour ultra-distance race.

What did they do to test the hypothesis or answer the research question?
The authors conducted a parallel-group repeated-measures study in 9 male runners who participated in a 6-hour ultramarathon race and 6 runners (5 male, 1 female) who did not compete and were used as controls. Both groups were experienced ultra-runners, had previously completed at least one ultramarathon, and had similar VO₂max values (67.3±11.5 and 66.2±10.2 mL/kg/min). All subjects completed counter-movement jump tests (to measure maximum jump height and vertical peak force), submaximal treadmill running tests (to determine subjects’ metabolic cost of running and preferred stride frequencies), and VO₂max tests. The “race” group repeated the counter-movement jump tests and submaximal treadmill running tests immediately after the 6-hour ultramarathon, which started at 8 am and was run on an 11.5 km loop (with an elevation change of 550 m) around the location of the lab. Subjects chose their own pace and were instructed to cover the longest distance possible, which was 44.3±2.8 km with 2276±487 m of elevation gain and loss.

What did they find?
— Maximum jump height and vertical peak force decreased (by 10.9% and a 7.4%) following the 6-hour ultra run in the experimental group, but were unchanged over the same period in the control group.
— Both groups showed the lowest energy cost of running at their preferred stride frequency.
— Increases or decreases in stride frequency away from their preferred frequency increased the energy cost of running in both groups.
— In the experimental group, the 6-hour ultra run did not affect subjects’ preferred stride frequency or the energy cost of running at subjects’ preferred stride frequency.

What were the strengths?
— High external validity due to the outdoor ultrarun with substantial elevation gain/loss.
— Reporting of effect sizes in addition to statistical significance.

What were the weaknesses?
— Small sample size, which was not justified using power calculations.
— The two groups of runners were not matched in a case-control sense, were imbalanced in number (9 experimental, 6 controls), and imbalanced in sex (9 males in the experimental group vs. 5 males and 1 female in the control group).

Are the findings useful in application to training/coaching practice?
Yes. The data adds to the evidence that humans, like all animals studied to date, optimise their stride frequency to minimise the energy cost of ambulation. Furthermore, this innate ability is maintained even following a fatiguing 6-hour ultramarathon.

Full paper access: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0238569

What was the hypothesis or research question?
As exercise intensity increases, motion artifact is the main source of error that decreases the accuracy of heart rate measured using photoplethysmography in wrist-worn wearables. Running outdoors can contribute to motion artefact. The authors aimed to determine heart rate validity of wearable monitors during a trail run and hypothesized that all devices would display acceptable validity with mean absolute percent error (MAPE) ≤5%, Lin’s Concordance Coefficient (r_C) ≥ 0.90, and intraclass correlation coefficient (ICC) ≥ 0.70.

What did they do to test the hypothesis or answer the research question?
Twenty-one young, healthy adult men (11) and women (10) volunteered to complete 3 self-paced 3.22 km trail runs while concurrently wearing photoplethysmography wearable devices and a criterion heart rate strap. Each run was an out-and-back format 1.61 km each way; the way out was net uphill and the way back was net downhill and each route varied by steepness, undulation, and elevation gain/loss (which ranged 48 to 104 metres). The criterion comparator device was the chest-worn ECG Polar H7 heart rate monitor. The experimental devices included the chest-worn ECG Suunto Spartan Sport watch plus several nonchest-worn photoplethysmography wearables: Garmin Fenix 5 (wrist), Jabra Elite Sport earbuds (ear), Motiv ring (finger), and the Scosche Rhythm+ (forearm).

What did they find?
— As expected, the chest-worn Suunto Spartan Sport watch performed excellently against the criterion Polar H7 heart rate monitor with mean absolute percent error (MAPE=1.9%), Lin’s Concordance Coefficient (r_C=0.955), and intraclass correlation coefficient (ICC=0.955) all within the acceptable ranges.
Note: MAPE indicates how much error there is in heart rate measurement between the test device and the comparator, while r_C and ICC demonstrate the size of the relationship between the test device-derived and comparator-derived measurements. MAPE, r_C, and ICC examine the data across the entire range of heart rate values. — Contrary to their hypothesis, the authors found that the photoplethysmography wearables (Garmin Fenix 5, Jabra Elite Sport earbuds, Motiv ring, and the Scosche Rhythm+) all performed very poorly.

Running science nerd alert Thomas Solomon and Matt Laye.

— The Bland-Altman plots shown in Figure 2 help us examine how the the difference in heart rate measurement between the test device and the comparator (y-axis/vertical axis) is related to the heart rate value being measured (x-axis/horizontal axis). This essentially allows you to determine the level of agreement between two devices, whereas a simple correlation only tells you if there is a relationship (i.e. if two devices have a strong correlation but one of the devices consistently measures 20 bpm higher, then their agreement would be poor). From the Bland-Altman plot you can calculate the 95% limit of agreement (i.e. the range within which 95% of the differences between the two measurement methods lie or the bias — the dashed lines in Fig 2, while the solid line represents the mean/average difference between measurements). Ideally, the solid line (the mean difference) would equal zero and the dashed lines (the bias) would be close to zero and all data points would exist within the dashed lines.

— The authors concluded that “regardless of device location (finger, wrist, ear, forearm), photoplethysmography-based devices do not provide acceptable heart rate validity during a trail run lasting longer than 20-min”.

What were the strengths?
— Sample size was justified using power calculations and the authors even used a more conservative effect size of 0.5 than the existing literature (0.91 to 0.95) to select a larger-than-needed sample size, which would prospectively increase the statistical power of their findings.
— The external validity is high since the trials were conducted on real outdoor trails using popular consumer devices.
— The consumer devices selected measure heart at different body sites: wirst, ear, finger, and forearm.

What were the weaknesses?
— Three trail routes were used but it is unclear whether all 21 subjects completed all three trail runs in a randomised, counterbalanced design or whether the 21 subjects were randomly assigned to (or even self-selected) one of the trail routes.
— Since no between-trial comparisons of the variance were conducted, I assume that subjects only completed one of the three routes, which increases the risk of within-trial bias since different within-trial running courses were used for different subjects.
— Since no pre-trial testing was conducted to ascertain the subject's VO₂max levels, we cannot know at what intensity the subjects ran their self-paced run. In the absence of such data, a simple self-assessed RPE value could have been recorded after each run to estimate the level of intensity exerted — but this too was missing.
— The results section of the paper does not make any attempt to describe the data and any interpretation of the Bland-Altman plots (Figure 2) is completely absent — this would have been very useful, especially since there are some very unusual data patterns. It would also have been more useful for the reader if the y-axes of the Bland-Altman plots in Figure 2 used the same range of values so that visual comparisons of the measurement error could have easily been compared between devices.

Are the findings useful in application to training/coaching practice?
Yes. This adds to the data that current photoplethysmography devices that attempt to measure heart rate during exercise are, at best, very poor. For folks who wish to measure heart rate during training and/or competition, it is best to measure it using an ECG-type device located where the organ is situated — in the chest.

Full paper access: https://pubmed.ncbi.nlm.nih.gov/32871363/

What was the hypothesis or research question?
The study aimed to investigate which pre-and post-training practices running coaches/running group leaders engage in with runners, which they believe are effective for injury risk reduction, how confident they feel in providing pre-and post-run injury prevention activities and advice, what they perceive causes injury in runners, and what barriers they perceive to exist which prevent prehab/injury prevention in running clubs/groups.

What did they do to test the hypothesis or answer the research question?
The authors invited coaches/running group leaders in the UK to take part in an online questionnaire. Recruitment (n=345) was via emails to club secretaries of all UK running clubs identified online and through running groups’ social media pages. A running coach was defined as “someone who offers training for a running or athletics club”, and a running group leader as “someone who supervises and motivates a running group in a less formal environment”. The questionnaire contained 23 questions including a mix of open, closed, and Likert-scale style. Data was analysed quantitatively for the Likert-scale questions while qualitative thematic analysis was used for the free-text questions and conducted by two independent analysts.

What did they find?
— 66 coaches and 34 running group leaders (100 in total) fully-responded to the questionnaire (mean ± SD age 51.1 ± 9.5 years; 58 female). The majority (62%) of coaches/running group leaders had 1 to 10 years’ experience , 29% had more than 10 years’ experience, and 9% had less than 1-year experience coaching or leading a running group.
— Most coaches/running group leaders performed a warm-up (97%), cool-down (94%), and gave advice on injury risk reduction (91%). A further 84% gave advice on recovery strategies, and 67% either included prehab exercises during training or directed runners to exercises out-with training.
— 95% said they would like to know more about pre- hab/injury prevention strategies.
— Median (interquartile range) length of warm-ups and cool-downs were 15 (10) and 10 (5) minutes respectively. Coaches/running group leaders thought prehab during a running session should ideally last 10 (5) minutes.
— Runners asked coaches/running group leaders for specific advice on specificity of injury management, timescales for return to running, diagnosis, and injury prevention.
— Coaches/running group leaders predominantly believed the cause of injuries to be multifactorial, including training errors, lack of conditioning, individuality in relation to shoes, runners not warming-up/cooling-up, and runners continuing to run when injured and not following advice.
— Coaches/running group leaders reported that main barriers to injury prehab were runners’ attitudes towards prehab, coaches/running group leaders’ lack of knowledge on why it was important and what to implement, lacking an appropriate environment, and insufficient time to add it to a running session.

What were the strengths?
— Provides real-life insight into the practices associated with injury risk reduction within running groups.
— Provides insight into general coach knowledge of the topic as well as their attitudes and their athletes’ attitudes to it.

What were the weaknesses?
— Since the questionnaire was sent to club/running group secretaries, the authors relied on them to actively distribute the questionnaire link to the coaches/running group leaders.
— Since most running club websites did not have any direct email addresses for their coaches/running group leaders, the questionnaire response rate was low (100 out of 345).
— Since the authors only targeted UK clubs, there may be regional biases that do not extrapolate to other countries.
— The wide range of experience of the coaches, which was not controlled for or sub-analysed, may increase bias.

Are the findings useful in application to training/coaching practice?
Coaches and running group leaders have a responsibility to provide safe and effective training programmes to their athletes that are appropriate given their athletes’ capabilities, experience, training history, and goals. Coaches/running group leaders also have a responsibility to work and advise within their remit and scope of their qualifications — most coaches/running group leaders are not qualified to diagnose injuries or prescribe therapeutic interventions to treat injuries, and few coaches/running group leaders are qualified in the strength and conditioning practices that can be used for prehab and injury prevention. The findings of this study perhaps do not inform coaching practice or athletic training directly but they do indicate areas that need strengthening within the British coaching structure to optimise the services coaches/running group leaders provide to their athletes.

Full paper access: https://pubmed.ncbi.nlm.nih.gov/32930647/

What was the hypothesis or research question?
Given that the only way to measure running economy is in a lab, the authors wanted to see whether running coaches could identify differences in running economy in athletes running around a track. This is important as running economy is one of the pillars of running performance and accounts for many of the differences in abilities. They hypothesized that those coaches with more experience would be better at differentiating poor and excellent running economy by watching someone run.

What did they do to test the hypothesis or answer the research question?
They recruited 33 recreational runners from the local community. They recruited 121 coaches who coached male runners across the country via digital means such as Letsrun.com. Running economy was measured on two consecutive days in 6-minute bouts at 3.57 meters/sec or 8 miles/hour. They then separated out the runners by running economy to create a spectrum of 5 different running economies from 5 different athletes with a difference of ~2mL/min/kg in running economy between each athlete. Those 5 athletes were then filmed on the treadmill from various angles and the films were attached to a survey so that each of the coaches could rank the 5 runners from least economical to most economical. They analyzed the data by different coaching characteristics including, years coaching, level of coaching, personal running experience, highest competition level personally, and education level.

What did they find?
— 35% of the coaches could not identify any of the runners in the correct order of running economy.
— 47% identified one, 12% identified two, and 6% identified three runners in the correct order of running economy.
— No coaching characteristics were predictive of the ability to identify varying running economies.
— Within the ranking, 59% of coaches were actually able to identify the least two economical runners.
— When coaches were asked what they were looking for, they often identified running traits that do align with traits that other studies have associated with running economy, such as stride rate, vertical displacement, and foot placement.

What were the strengths?
— They had a large number of coaches take their survey and they were careful to verify the running economies of the athletes by testing them twice on two consecutive days.

What were the weaknesses?
— The differences in running economy were small between the runners even though they are significant for running performance.
— The coaches only observed the running on the treadmill, which may be a less familiar location for them to observe running.
— Having 5 runners that needed to be ranked makes it extremely easy to get all of the runners wrong just by sequencing one runner incorrectly. So either more or less runners with differences in running economy would be useful.
— Despite the large sample size some of the statistics were underpowered and could have used more responses.
— Running economy was also quantified at only one speed which corresponded to 80% of VO2max.

Are the findings useful in application to training/coaching practice?
Coaches can often identify problems with running form that differ from the biomechanically “ideal”. However, we also know that each runner is an individual. For example, Paula Radcliffe is extremely efficient as a runner, but looking at her form you would not think that at all. Given the importance of running economy, it is recommended to focus on exercises that increase running economy such as strength training and plyometrics rather than trying to change someone's form based on what you see. If changes in someone's form are made then I would recommend actually testing their running economy on a metabolic cart to verify that the changes you are making with their form are indeed helping them improve.

Full paper access: http://www.ncbi.nlm.nih.gov/pubmed/32816636

What was the hypothesis or research question?
The authors wanted to identify factors that are associated with overtraining. The specific factor which they are interested in is muscle fiber type and they hypothesized that those with a higher proportion of type II fibers would display more overtraining symptoms than those with more type I fibers. They hypothesize this because of the increased fatigability of type II fibers relative to type I.

What did they do to test the hypothesis or answer the research question?
They recruited 24 elite middle-distance runners (n = 16, male with VO₂max = 73 mL/kg/min). The study lasted 7 weeks, with 3 weeks of normal training, 3 weeks of increasing training volume (10, 20, 30%), and one week of taper. Before and after each phase the runners underwent testing for time to exhaustion, ventilatory threshold, peak heart rate, and VO₂max using gas exchange. They also had blood taken and body composition and resting metabolic rate measured as potential markers related to overtraining. In addition, they had their muscle fiber types in their soleus and gastrocnemius muscles identified via an indirect measure which uses the amount of carnosine in the muscle as a marker of the relative amount of Type II or Type I fiber type (higher carnosine = type II). Total training load was calculated by time in one of three zones and measures of wellness were collected with surveys. Lastly, to determine overtraining they had to reduce their performance by half of the coefficient of variation from their time to exhaustion and to report a higher subjective fatigue level following the increased training volume. If subjects were more fatigued but increased their performance or it did not change then they were considered acutely fatigued.

What did they find?
— 12 of the runners were classified as overtrained as their time trial performance decreased by 3.15% or more. These runners were then analyzed separately as the FOR (functional overreach) group versus the AF (acute fatigue) group.
— The FOR group had lower HR peak and VO₂max following the high volume training.
— In looking directly at their hypothesis, those in the FOR group had lower carnosine and thus presumably higher Type II fiber proportion than those in the AF group.
— None of the blood markers or food intake differed between the groups.
— Training volume and intensity measured by distance and speed did not differ, but those in the FOR group reported more time at higher levels of RPE during week 3 of the higher volume compared to many of the other training weeks.
— Upper respiratory tract infections were more common in the FOR group as were sleep disturbances (both significant) and a moderate effect reported on fatigue levels.

What were the strengths?
— Captured a lot of data to see if there were relationships with functional overreaching.
Sample size was appropriate, although they may have got lucky with 12 and 12 being the FOR and AF groups.
— They also reported effect sizes which is always a plus.

What were the weaknesses?
— Most of the measures they did were looking at associations and nothing causal.
— Their training intensity only had 3 different levels in which the most intense one included a ventilatory threshold all the way to all-out sprinting, which have different physiological effects and recruit different fibers.
— Whether or not recovery from overreaching for more than a week leads to any differences would have been nice to know.
— The time to exhaustion test is not always the best way to measure performance as time trials seem to correlate better to actual performances.

Are the findings useful in application to training/coaching practice?
Most of us cannot measure fiber type, but can get an idea of whether our athletes are most type I or type II dominant athletes. Fiber type dominance might be important in deciding how we prescribe training, but it was not the only way to tell if someone was overreaching. Measures of RPE or general fatigue also indicated if an athlete was approaching overtraining. So while fiber type is cool to know, it may not be necessary. We should all be reminded that overreaching is good from time to time and that individual athletes will reach that point at different points and likely need different amounts of recoveries from those sessions.

Full paper access: https://pubmed.ncbi.nlm.nih.gov/32881840/

What was the hypothesis or research question?
The aim of the study was to compare the muscular effort of moderate to high intensity running to a similar effort on the stair stepper and elliptical machine. The rationale for this is that running injuries are common and cross training is a common strategy athletes and coaches use to maximize non-running fitness. However, which exercise modality provides the most similar running muscle activation pattern is not known. The authors hypothesized that running would require more muscle activation than the stair stepper and elliptical even at the same exercise intensity.

What did they do to test the hypothesis or answer the research question?
The subjects, 17 young moderately fit male and female runners (VO2max = 49.3 mL/min/kg), completed maximal exercise tests on the treadmill, stepper, and elliptical using a portable metabolic kit. Subjects then conducted a submaximal exercise test consisting of 3 minutes at 60%, 3 minutes at 70%, and 3 minutes of 80% of their machine specific VO2max. While conducting the submaximal test surface electromyographical (EMG) readings were performed on eight muscles; gastrocnemius medius, soleus, tibialis anterior, rectus femoris, vastus medialis, biceps femoris, semitendinosus, and gluteus medius. All EMG signals were normalized to their 80% submaximal treadmill values for each athlete due to person to person variations in EMG signals.

What did they find?
— They found that VO2 max did not differ between the different exercise modalities and it is possible to reach VO2max on each of the different pieces of equipment.
— The degree of muscle activation as measured by EMG was significantly higher at each of the submaximal exercise intensities for most muscle groups with the exception of rectus femoris and vastus medialis (two of the muscles in your quadriceps). For the muscles that significantly differed the effect sizes were large (Cohen d greater than 1.2).
— Some smaller differences between stepper and elliptical were also seen. The general trend was higher levels of muscle activation in the stepper compared to elliptical at some of the exercise intensities in some of the muscles. Moderate effect sizes were seen in the vastus medialis, rectus femoris, gastrocnemius, and soleus at one of more exercise intensities.

What were the strengths?
— They reported all effect sizes and seemed to account for differences in the normality of the data.
— A good representation of lower leg muscles were studied.

What were the weaknesses?
— No power analysis was done a priori.
— The EMG technique that they use has a number of different potential problems including having to normalize to each individual person and EMG signals could not be synchronized over gait cycles, which means the pattern most similar to running was not identified.
— The authors also report that because they do not have data on maximal force isometric production they cannot say whether the level of muscle activation is sufficient to elicit training adaptations.
— They also did not measure metabolic costs at those 3 submaximal exercise bouts.

Are the findings useful in application to training/coaching practice?
There are several applications to training here. First, for cardiovascular fitness either the stair stepper or elliptical is a suitable substitution for running. This is especially true if lower muscle activation and recruitment is needed due to an injury or a period of lower muscular strain is needed. It seems that the level of quadriceps activation is similar between the different modalities and so athletes and coaches should pay special attention to strength work focused on the hamstrings and lower legs if a significant amount of work is done on the elliptical or stair stepper. There were a few muscles that were slightly more activated in the stair stepper compared to elliptical so that might be the better of the two, but in general they showed similar peak muscle activations and given the likely differences in muscle activation pattern either is a good candidate for cross training.

That is all for this month's nerd alert. We hope to have succeeded in helping you learn a little more about the developments in the world of running science. Until next month, keep active, stay nerdy, and train smart.

Disclaimer: Any interpretations and recommendations we make are, and always will be, based on our own views and opinions shaped by the evidence available to us. Before making any changes to your training based on any information we provide, always ensure it is safe for you to do so and consult your doctor if you are unsure.

If you find value in these nerd-alerts, please help keep them alive and buy us a beer:

Buy me a beer.

If you are enjoying this free content, please like and follow @veohtu, @mjlaye and @thomaspjsolomon and share these articles on your social media pages.

Dr Thomas Solomon and Dr Matt Laye. Running science nerd alert.

About the authors:
Matt and Thomas are both passionate about making science accessible and helping folks meet their fitness and performance goals. They both have PhDs in exercise science, are widely published, have had their own athletic careers, and are both performance coaches alongside their day jobs. Originally from different sides of the Atlantic, their paths first crossed in Copenhagen in 2010 as research scientists at the Centre for Inflammation and Metabolism at Rigshospitalet (Copenhagen University Hospital). After discussing lots of science, spending many a mile pounding the trails, and frequent micro brew pub drinking sessions, they became firm friends. Thomas even got a "buy one get one free" deal out of the friendship, marrying one of Matt's best friends from home after a chance encounter during a training weekend for the CCC in Schwartzwald. Although they are once again separated by the Atlantic, Matt and Thomas meet up about once a year and have weekly video chats about science, running, and beer. This "nerd alert" was created as an outlet for some of the hundreds of scientific papers they read each month.

To read more about the authors, click the buttons:

Running science nerd alert.

Energetically optimal stride frequency is maintained with fatigue in trained ultramarathon runners.

Concurrent heart rate validity of wearable technology devices during trail running.

Running coaches and running group leaders' engagement with, and beliefs and perceived barriers to prehabilitation and injury prevention strategies for runners.

Visual Classification of Running Economy by Distance Running Coaches.

Muscle fiber typology is associated with the incidence of overreaching in response to overload training.

Muscular Activation Patterns During Exercise on the Treadmill, Stepper, and Elliptical Trainer.