Understanding mediators and moderators is critical for workload-injury investigations and exploring which athletes are at higher risk for injury.
The relationship between training load spikes and injury has received much press in sports science in recent times. Yet, attributing injuries solely to workload spikes is an oversimplification of the process. Instead, we must delve deeper into how and why (and even if) workload spikes lead to injuries, as well as identify which athletes are more susceptible or resilient to such risks. This is where the concepts of moderators and mediators come into play, shedding light on the complex interplay between workload and injury.
Domino or Dimmer? Differentiating 'Mediators' and 'Moderators'
Mediators and moderators in the workload-injury relationship have been best described in sports science by a 2017 British Journal of Sports Medicine editorial by Johann Windt and colleagues. You can download this paper for free on ResearchGate here.
Mediators act as the intermediary steps that explain the association between an observed variable and an outcome. This is likened to dominoes being knocked over. In our case, they may help to answer the question: "Why do changes in workloads cause injuries?"
For instance, consider neuromuscular fatigue as a mediator. One potential explanation for an increased injury risk with a workload spike is increased workloads cause higher levels of neuromuscular fatigue, which could subsequently increase injury risk, such as while performing a cut. As such, neuromuscular fatigue may mediate this relationship.
On the other hand, moderators can be likened to dimmer switches, modifying the effect of workload on injury risk. These variables alter the relationship between workload and injury outcome, and may also be referred to as interactions or effect modifiers. Here, they answer the question: “What characteristics make certain athletes more robust or more susceptible to injury at given workloads?”.
Various physical capacities may act as moderators, potentially reducing – or ‘dimming’ – the risk of injury. For example, Gaelic footballers with lower fitness (assessed via 1 km time trial time) had a significantly higher odds ratio for injury with a workload spike (Malone et al., 2016).
Similarly, another study led by Shane Malone, this time in amateur hurling players, found strength and repeated sprint qualities may also moderate injury risk at given workloads (Malone et al., 2018). Stronger athletes also had a reduced odds ratio for injury compared to weaker athletes with large week-to-week changes in training loads.
These studies have their limitations of course, but add weight to the notion that physical capacities, among many other factors, may moderate the relationship between training load and injury.
Dominoes and Dimmer Switches in a Web of Determinants
While the domino and dimmer switch analogy provides a simplified view, the reality of injury causation is more complex. According to Johann Windt’s editorial, the relationship between workload and injury may be more appropriately viewed as moderated mediation.
This means that the effect of the mediator is moderated by another variables. In this case, while workload may lead to increased neuromuscular fatigue, the strength of this relationship may be moderated by other factors such as aerobic fitness. It is the relationship between all of these variables that we should not underplay.
Indeed, injuries often result from a dynamic interplay of multiple factors, making them context-dependent and multifactorial. This is best illustrated in Dr Natalia Bittencourt and colleagues’ complex model for sports injury (2016).
The Importance of Information Architecture
Ultimately, considering the workload-injury relationship via ‘moderators and mediators’ and the complex system approach reinforces the need for robust information architecture in sports science.
Understanding the characteristics that act as moderators to ‘dim’, or conversely intensify workload-related injury risks, is important to appropriately manage training load. Therefore, physical capacity information should be accessible alongside training load monitoring data. Likewise, connecting readiness information, which may indicate the presence of neuromuscular fatigue, to training load data allows practitioners to analyse the dose-response relationship more comprehensively.
Athlete management systems should seamlessly connect these various data streams, including workload metrics, readiness information, physical assessments and injury data. By integrating these datasets, practitioners can assess the potential influence of moderators and mediators on the training process.
As analytics and artificial intelligence evolves, we may have greater ability to understand such relationships. I previously discussed how data streams beyond training load and injury alone can potentially strengthen model performance in sports science. That article was in response to Zone7’s white paper: ‘The Importance of Workload, Strength, Recovery and Environmental Context Data in an Injury Risk Forecasting Model’, which demonstrated their findings in incorporating such moderators into injury risk forecasting models.
Final Thoughts
In conclusion, moderators and mediators play pivotal roles in shaping the relationship between workload and injury in sports training. Put simply, mediators may explain a relationship, reflecting the domino effect. Whereas, moderators may act as a dimmer switch, altering the relationship between a factor and an outcome.
In reality of course, these factors exist in a more complex relationship, dynamically interacting within a web of determinants. Yet, these simple analogies remind us that there is more to workload and injury risk than isolated GPS metrics. To be able to tailor training programmes and make suitable interventions, we must have efficient information architecture that connects these various data streams together.