Franco M. Impellizzeri works at the Human Performance Research Centre, Faculty of Health, University of Technology, Sydney. He has authored about 150 publications in peer-reviewed and indexed journals in the area of sport science and orthopedics. He started his career as a coach before becoming head of research at the MAPEI Sport Research Centre (Italy), where he took care of the training and testing of elite and top professional-level athletes. Professor Impellizzeri has also worked in the clinical setting (head of the Lower Limb Clinical Outcome Unit) as a senior research fellow at the Schulthess Clinic (Zurich), developing his research in the area of clinimetrics (patient-reported outcomes).
Freelap USA: You have been outspoken about the limitations of the ACWR (acute:chronic workload ratio) for some time now, and it seems that some want to keep the metric alive because they don’t know an alternative. Is it safe to say that we need multiple data points to make smart decisions? It looks like the convenience of a daily number is removing the thinking process of monitoring.
Franco Impellizzeri: I have tried to advise others about the flaws of this metric for three main reasons: 1) the scientific process used to arrive at this “model” and this new metric is methodologically and conceptually wrong; 2) it is an oversimplification of a complex problem (injury) that cannot be addressed and reduced to a few metrics; and 3) the “practical applications” are unreasonable unless you interpret them in such a liberal way that the recommendations are no longer based on the study’s results but common sense (which means the studies are worthless). I am going to write more formally about these issues soon (and other researchers I know are doing the same), so I don’t want to use the space here to discuss this.
But I believe that the main reason why practitioners jumped on this metric, and more generically on this approach (metric-based), is that we don’t like to live with too many uncertainties and face this “interior conflict.” We have created the illusion we can in some way control “injuries,” which is now also the main reason people are sacked (or recruited) in professional teams. It is an illusion because, scientifically speaking, we still don’t have prognostic factors strongly associated with the risk of injuries, and even more so in terms of etiology and causal relations. There are studies and opinion pieces reminding us that association is not prediction. But we should also remember that association is not causation.
Although we all apparently know this, the majority of the studies on training load and injuries suggest manipulating training load to decrease the injury risk—that is assuming a causal association never supported by the studies. I am not saying the training load has no absolute effect on injury risk. It may not, but we don’t know how at this current moment. Studies showing that high ACWR (in total distance) was related to higher injury risk have also shown (in the same study) that increasing more than twice or even triple the acute load (high-speed running) has no relation, or was even associated, to a decrease in the injury risk! But these results are never discussed and are often ignored.
We just build conclusions based on the findings that make more sense to us, such as that it is better to slowly increase the load. This conclusion does not come from an analysis of the studies, but from well-known training principles such as load progression. So, we basically select among the inconsistent results presented in the literature—those that fit our beliefs—and these beliefs are often based on our experience as coaches and the traditional training principles we grew up with.
My opinion is that we should use training load monitoring mainly to see whether the program we planned has really been done by our athletes (especially in team sports with spontaneous activities such as small-sided games and tactical training) and to see how they cope with these demands. We usually modify the training based on this latter feedback and not the metric. Most of the instruments used are a way to quantify how the athletes are coping with/tolerating the training.
The plan moving forward is ultimately a decision of the coach, whereas the support staff just provides information so that the coach can make the best decision. Injury prevention is somewhat related to the “plan,” for instance by targeting training specific outcomes that we believe are important, or if there is some evidence it may be a risk factor. If I think my athletes need sprinting for reducing the injury risk, I do it and I don’t care if there are studies saying that sprinting more than 9 meters per session dramatically increases the injury risk. This is mainly because these studies are weak and not really worth considering.
When we’ve done all we believe can help prevent injuries, we should be realistic and acknowledge that injury risk is mostly outside our control, and the risk is always there. Share on XThe problem is that practitioners are often unable to judge the quality of peer-reviewed studies, and sport scientists sometimes cannot do this properly either (or they don’t explain the limitations and level of evidence). And when we have done all we believe can help prevent injuries, we should be realistic and acknowledge that injury risk is mostly outside our control and the risk is always there. Instead, I have recently seen coaches and especially managers being scared about injuries and training “too much,” selling the idea that someone particularly gifted can control the injury risk. Instead, we should educate all the stakeholders that injuries cannot be controlled, and whatever we do is an attempt, but nothing more than that. We need to accept and educate to live with this uncertainty.
Freelap USA: The research you performed on general running versus small-sided games demonstrated that general fitness from running and games are nearly identical. Some coaches elect to reduce specific practice volume and do tempo runs to help with muscle injury reduction with success. What are your thoughts about nonspecific conditioning?
Franco Impellizzeri: What we have shown years ago is that as soon as the physiological stimulus is comparable, the outcomes are similar (in terms of aerobic conditioning). And this also fits with our internal-external training load framework. The equivalence in terms of outcomes suggests that the coach has various options to reach similar results, and it is a coach’s decision what, and how much, to use of the two strategies (generic vs. specific).
The idea that small-sided games-based training increases the risk of injuries is not really supported, even if a high risk of contact injuries is reasonable. The problem with these forms of specific training is that you cannot control and plan the load accurately because the activity is influenced by contextual factors and is spontaneous. So, the training stimulus may be not consistent within and/or between players. This is the reason why monitoring is very important.
On the other hand, tempo runs are much more controllable, and you can plan in advance the stimulus with less inter- and intra-individual variability. So personally, I think it is better to use both, also based on how the players “accept” one form or the other, but I favor the non-specific and “more controllable” conditioning when possible. Clearly, this is a generic opinion, but the choice also depends on other contextual factors. The ability to “adapt” to the context is a characteristic that a coach and trainer must possess.
Freelap USA: Sand training is becoming popular again because of the summer here in the U.S. What are the pros and cons of the modality we should know about besides what you have published? Any nuances that you wish to share that you didn’t have a chance to within your study?
Franco Impellizzeri: I designed a study on sprint and plyometric training on sand because, some years ago, it was becoming quite popular in soccer and professional teams, and I usually try to address topics of interest among practitioners. For a while the interest went down, but now I see a new rise in the popularity of this kind of training. I was concerned about the effects, given that the adaptations are also specific to the nature of training. So, we designed a study to see whether our hypothesis was supported, and it was.
I think it is a good option as a complementary form of training, but it cannot completely replace training on other surfaces, especially if the performance is required on these other surfaces. It is also used in the rehabilitation phase for reducing the stress on the structure, and I think that makes sense. The problem, as sometimes happens, is when a form of training becomes exclusive and replaces other approaches. I am not saying that we should not change, but we need to consider when training can really be a substitution and when it can be a good integration.
We need to consider when a form of training can really be a substitution and when it can be a good integration, says @francoimpell. Share on XIn my study, I examined a “substitution,” but purely for research requisites: I had to isolate the effect of the independent variable that was the surface. But it would be more ecologically valid to also have a group combining training on grass and sand. This (internal vs. external validity) is an issue people should always consider to avoid interpreting study results in the wrong way when extrapolating practical applications. For research design necessity, we often study a training strategy in isolation, but in real life we use several approaches and exercises of a similar nature concurrently. The study results should often be interpreted as a sort of “proof of principle,” and we shouldn’t use the study to take out the “protocol.”
Freelap USA: You have great knowledge of blood analysis in sport and did a wonderful job investigating adaptations and changes with your research. What should team sports do to help connect biomarkers and player tracking to get more out of the data besides screening for anemia?
Franco Impellizzeri: I am actually not an expert on blood analysis. I participated in studies examining the changes in blood parameters (e.g., during cycling stage races), and I worked and collaborated with professional teams of various sports (mainly endurance), attempting to use these biomarkers as additional parameters to understand how an athlete tolerated the training load. My personal opinion, based on my experience, is that biochemical blood parameters are not easy to interpret to understand how the athlete is coping with the training stimulus.
The main drawback is that when important alterations are found, it is too late—for example, in terms of overtraining or overreaching. Working with hundreds of professional cyclists, we found very few “unexplained” declines in performance that could be classified as overtraining. By having completed baseline screening, for example, we could link some form of overtraining to post-viral syndrome and ongoing infections. This would not have been possible without baseline data.
But when the hormonal responses after a maximal test were altered, the athletes were clearly already overtrained or overreached because the performance and symptoms were present for weeks or months already (i.e., too late). To use these parameters to manipulate the training load, we should be able to differentiate normal from abnormal variations and acute from chronic changes, which is quite difficult: It is really challenging to understand when a variation is really a warning signal. We had athletes with very high cortisol levels and quite low testosterone levels before winning important competitions, just as an anecdote on two commonly monitored hormones.
The interpretation is very subjective, and it is just additional information at your disposal. Support staff should include sport scientists with adequate knowledge of physiology able to interact with medical doctors in order to try to interpret the results (from a physiological and medical point of view).
Freelap USA: Heart rate monitoring seems like a lost art now. Can you explain how a team can manage to get more out of their internal load and response data? Your study on pre-season training and polarized approaches is very thought-provoking for those wanting to gain fitness but reduce unnecessary risk.
Franco Impellizzeri: Yes, this is not so “fancy” anymore, but still I believe HR can provide important information. The combination of external and internal load can give more insights on how the athlete is coping with the training and quantify the actual psychophysiological stimulus. We need both sources of information, of course, because the external load gives us more indications about the nature of the training stimulus, while the internal load gives more on what this external stimulus is inducing on our athletes.
In the study where we found that 7–8% of training in the pre-season was spent at high intensity (based on HR and physiological thresholds), we also found a relation between the time spent at high intensities and improvements in specific and generic aerobic outcomes. This reminds us, once again, that it is the physiological stimulus determining adaptations and changes. So, we should concentrate and think about what an exercise will induce in our athletes and not only what the exercise looks like.
The “polarized” distribution is quite common and expected. First, because the time at disposal for training is fixed, and when you do something, you cut other forms and quantities of training. So, if you dedicate time to tactical and technical training, for example, you have less time to give to something else. If you add gym and recovery sessions, at the end, the time at your disposal for conditioning training is reduced. So, it is not quite thought-provoking, it is just the practical consequence of training scheduling and team choices.
From a practical point of view, I found it very useful in both team and endurance sports to know the amount of high-intensity training in terms of internal load (e.g., heart rate). This gave me the ability to understand how much conditioning training the athletes completed at high intensity (physiologically) every week and try to associate that amount to the athlete condition (objective or subjective). Knowing both the internal and external load can help in identifying dissociations between various indicators; for example, by examining and comparing the amount of high-intensity distance versus accumulated HR response versus perceived exertion.
Forty years ago, Banister and Calvert tried to quantify training load using a single number. Since then, we still struggle to find a way, but no methods that are established or free of issues exist. They are all proxy measures of training load, each with strengths and limitations. These limitations can be only partially addressed by the use of various measures.
What I always try to explain to the coach is that they are like a doctor deciding for a patient. They (may) use the available evidence to define the “intervention,” combined with their experience and the athlete necessities (EBP). But to define which intervention (or how to modify the intervention), the doctor makes an informed decision based on the various pieces of information they have at their disposal.
The surgeons with whom I worked for 10 years collected information on the patients (anamnesis), they talked to them, they used the patient-reported measures (validated and valid!) to quantify symptoms, they used image diagnostic, clinical tests and biochemical markers—and based on this information, they made a decision. And do you know what? With the same information, different doctors can make different decisions!
Like surgeons, coaches collect information from qualitative sources, examine it in quantitative ways, and often arrive at a different decision than other coaches, says @francoimpell. Share on XSimilarly, as coaches, we collect information with the help of the supporting staff and we make a decision, and this decision can be different than that of another coach even with the same information—and there is nothing wrong with this! As the surgeon relies on several quantitative sources of information examined in a qualitative way without trying to create fancy metrics and indexes, so too does the coach. It is the job of a coach to make a decision and take responsibility for that decision.
The supporting staff just provides information to support this process and has to take the responsibility if the information provided is not good or scientifically valid. I would not sack medical staff or sport scientists for the increase in injury rate from one season to another. However, I would fire all if I found out that the information “sold” to the coach or management as scientific or evidence-based was actually based on unsupported scientific claims and weak evidence.
Since you’re here…
…we have a small favor to ask. More people are reading SimpliFaster than ever, and each week we bring you compelling content from coaches, sport scientists, and physiotherapists who are devoted to building better athletes. Please take a moment to share the articles on social media, engage the authors with questions and comments below, and link to articles when appropriate if you have a blog or participate on forums of related topics. — SF
Really good thank you so much for your amazing knowledge.