Using submaximal fitness tests to understand fitness and fatigue

Practitioners working with team sport athletes share similar goals: optimising our athletes’ performance, well being, and recovery while minimising the risk of injury. Exercise assessments are an integral ingredient for achieving these goals by letting us objectively monitor athletic status. However, the demanding schedules and extended competition periods of today’s team sports landscape present challenges to routinely performing these tests ^[1].

Submaximal fitness tests (SMFT) are short bouts of running or cycling at a standardized, non-exhaustive intensity ^[2]. SMFT have become a pragmatic alternative within team sports, offering a short, time efficient, and minimally fatiguing approach to assessing athletes’ responses to a standardised physical stimulus ^[2].

The stimulus determines the protocol. Surveying team sport practitioners revealed numerous SMFT protocols. However, they mostly fall into five categories, based on the exercise regimen – continuous or intermittent – and the intensity changes throughout the activity: fixed, incremental, or variable (Figure 1) ^[2].

The response measures from each test serve as proxy indicators of athletes’ physiological state, encompassing the overall physical performance and/or training adaptations ^[3].

Cardiorespiratory/metabolic	Subjective	Mechanical
HR-derived indices – Exercise heart rate (HRex) – Heart rate recovery (HHR) – Heart rate variability (HRV)	Rating of perceived exertion – Global RPE – Differential RPE	Locomotor outputs – Total distance covered – Distance/time/count in zones – Accelerations, changes of direction
Blooad markers – End-bout lactate concentration		IMU derived data – Overall accelerometry load – Uniaxial accelerometry load – Ground contact/flight time – Stride length
Oxygen kinetics – Oxygen consumption
HR: heart rate; IMU: inertial measurement units; RPE: rating of perceived exertion

Submaximal fitness test protocols: One size does not fit all

Practitioners often seek one “magic protocol.” But different protocols may have their place over the course of a season, and practitioners can adapt them to suit their context, e.g., changing the exercise regimen, or increasing / decreasing the running speed. Rather than relying on rigid rules, consider some key principles to guide your protocol.

The first questions you should ask yourself are, “What is my primary outcome measure, and what physiological system does it represent?” These two questions will probably set the groundwork for the majority of your decisions.

Other questions – and decisions – can follow accordingly. “Does my protocol align effectively with my training plan?” “Is this a protocol we can administer routinely?”

One of the primary advantages of SMFT is that they let us test a group of athletes simultaneously, right on the training field. The only way to establish a sustainable test protocol under those conditions is by standardising the external load parameters (e.g., duration, speed) as the physical stimulus, which, in turn, leads to internal responses (e.g., heart rate derived indices) becoming the most quantified outcome measures ^[2].

Internal responses: Heart rate derived indices

Heart rate derived indices are the most common outcome measures to assess athletes’ internal responses, and can be collected either during or following the activity (Figure 2)^[4,5].

Exercise HR (HRex) stands out as the most research backed and widely used outcome measure, with approximately 80% of practitioners reporting it as their preferred measure ^[6]. HR recovery (HRR) is also commonly employed, often together with HRex ^[2].

I have never used HR variability (HRV) indices in the context of SMFT in team sports. HRV indices have remarkably high measurement error ^[4], require longer recording time (3–5 minutes post exercise), and need multiple measurement points across the week to ensure valid and usable data ^[4]. Plus, too many factors on the field can influence them.

Exercise heart rate: First among outcome measures

Deciding on your primary outcome measure addresses most of your protocol related questions. HRex is a trustworthy proxy of athletes’ fitness levels and adaptive responses, where a chronically lower HRex suggests improved fitness, while an increase may indicate a negative response, such as detraining ^[7]. It provides a good marker of exercise intensity when measured relative to athletes’ maximum heart rate (HRmax) ^[2,4]. The underlying mechanisms are mainly due to the linear relationship between HR response and oxygen demands during a wide range of steady state activities.

Therefore, when selecting your protocol, ensure it induces a stable HR response, preferably by administering continuous-fixed protocols (Video 1). Athletes tend to reach a stable HR response in about 2–2.5 minutes. A 3-4 minute protocol where HRex is the mean HR across the last 30–60 seconds is the optimal method (Figure 3).

Does the SMFT intensity matter?

A common question among researchers and practitioners relates to the role of exercise intensity. While there is a prevailing belief that higher submaximal intensities may offer better insights into athletes’ fitness levels, the current evidence (together with unpublished observations) shows that the validity and reliability of HRex in surfacing changes in aerobic fitness remain consistent, irrespective of how hard you push your athletes ^[7].

That means you can basically choose your desired exercise intensity, knowing that it is unlikely to meaningfully compromise the accuracy of tracking fitness changes.

From a real world viewpoint, considering that continuous-fixed protocols are primarily used at the beginning of the session and often serve as the general warm up phase, I usually recommend targeting intensities that elicit between 70–85% of HRmax at preseason baseline. This facilitates the warm up, ensures the athletes will neither perceive the exercise as too easy nor too hard, and lets them pace themselves comfortably. Depending on the sex and level of your athletes, this typically comes with running speeds between 10-14 km·h⁻¹.

A range of 70–85% of HRmaxmight still seem quite broad, but that’s what makes SMFT so attractive. They offer the flexibility and adaptability to suit your or your athletes’ needs. Think about scenarios where lower standardised intensities could be beneficial, such as return to performance protocols after injury, when athletes need to build a gradual recovery process.

HR recovery: Time for a rethink?

The rate at which HR declines to a standardised external stimulus can serve as a marker of athletes’ ability to recover following the activity, with a faster HR drop typically indicating increased fitness levels ^[5].

Although some practitioners use longer collection timeframes (60 vs. 180 s) and different body positions (e.g., athletes lying supine) to improve HRR accuracy ^[8,9], the most realistic, long term approach is to collect HRR for a maximum of 60 seconds while athletes are standing upright.

Generally, the research supports the use of HRR ^[2], and I almost always collect it. But be aware of certain factors that may undermine your data.

First, the intensity of the preceding exercise influences HRR response. Any changes in HRex may impact HRR interpretation. Second, HRR tends to display a larger error of measurement compared with HRex ^[2,7], and the error tends to get larger in lower versus higher intensities (~ 80% vs. 90% HR_max) ^[10]. Third, after collecting HRR numerous times, I still find it challenging to maintain a highly standardised environment, particularly ensuring that my athletes refrain from changing their body positions or interacting with each other.

Despite these challenges, I typically advocate for collecting HRR alongside HRex, as they may offer insights into different aspects of cardiovascular function ^[4]. Among many analysis options available, team sport research tends to support analysing HRR independently of HRex, that is, the mean or minimum HR over the last 5–10 s of the recovery (mean of the 50^th to 60^th second) ^[10].

What’s the place for other protocols?

Shorter versions of intermittent-incremental protocols

Practitioners also commonly use shorter versions of established intermittent incremental field tests, like the 30-15 intermittent fitness test (30–15IFT, see Video 2) ^[11]. This could stem from the absence of clear guidelines in the past, or follow from using protocols that are more specific to the intermittent nature of team sports.

However, these tests involve frequent recovery periods, resulting in a more variable HR response, which may subsequently impact the accuracy of HRex (Figure 4).

If you decide to stick with this choice, it might be crucial to adapt your analysis accordingly. For instance, in the 30–15IFT, rather than analysing the mean HR over the last 30 seconds of the fourth minute, which includes 15 s of recovery and 15 s of activity, you could calculate HRex from a selected period that does not involve recovery. One approach could be the mean HR between 3:45 – 4:15 minutes (stage at 12.5 km·h⁻¹).

Intermittent-variable protocols: Should we just play?

Besides addressing pragmatic challenges such as time constraints and human resources, a benefit of SMFT is their ability to alleviate both the physical and mental burdens associated with isolated physical “testing.” They can be seamlessly integrated into the training session without adding extra strains on our athletes.

That’s why the use of intermittent-variable protocols such as “ball involved drills” have also received some attention from practitioners ^[12,13].

While such drills may involve maximal activities (e.g., repeated accelerations, changes of direction), it is important to remember that they are an integral part of the regular training plan, and would be performed anyway.

Unlike more generic SMFT, like the ones introduced before, the decision on the protocol lies mainly on which drills you can expect the athletes to perform frequently enough (at least weekly), and how open you are to embracing variability in your athletes’ locomotor outputs. Consider the parallel to Matt Taberner’s excellent work on the “control–chaos continuum” in rehabilitation phases.

In this continuum, you may adopt drills that create a relatively controlled environment, such as passing drills. Alternatively, you may incorporate drills that are on the extreme end, prompting a more “chaotic” environment such as small sided games (SSGs). You also want to consider other contextual factors, like the number of players, pitch dimension and constraints ^[14].

Matthew Lacome recently embraced a more flexible approach by including a wide range of team sizes (5 – 10 players a side), pitch dimensions, and game objectives (score vs. possession). Using a regression analysis technique for controlling the external load parameters during SSGs, they were able to predict HRex response during these drills (90% Pearson’s correlation range between 0.74 – 0.84). This predictive ability extended to changes in HRex measured during a continuous-fixed protocol involving running for four minutes at 12 km·h⁻¹ (correlation range: 0.50 – 0.82).

This illustrates how you can ensure daily monitoring throughout regular practice, eliminating the necessity for formal testing. On the other hand, this approach puts more variability in external loads into the mix, which may, in turn, compromise the sensitivity and increase uncertainty.

My go-to choice is more conservative, at least in most situations. I usually try to pick 1-2 passing training drills that the coaches frequently use and that we know yield more consistent external load outcomes. These often come early in the training session, minimising other factors such as cardiac drift, and potentially erroneous HR data due to contacts and shocks that may happen in game situations.

Figure 5. demonstrates HRex responses across the preseason, gathered from both continuous-fixed protocol and a passing drill (Y shape, four minutes) administered weekly on MD+3.

Monitoring while intervening: Example from an intermittent-fixed protocol

Programming “off ball running” exercises into the training routine (e.g., tempo runs, high intensity interval training), whether it’s scheduled during the preseason, rehabilitation return to performance phases, or “top ups” during the season stands out among the best interventions to enhance athletes’ performance ^[15].

Such interventions involve exercising at an intensity that can be fixed (i.e., absolute) for all athletes, or anchored to each individual athlete’s capacity (e.g., a fraction of maximal aerobic speed) ^[15].

Monitoring changes in HR indices across the intervention may also provide practitioners with valuable insights into their athletes’ changes in fitness without formally testing them ^[16].

The example below (Figure 6) shows a football (soccer) player’s response to an intermittent-fixed protocol in weeks 1, 3 and 5 of preseason, involving the first four bouts at 18 km·h⁻¹ across 50 m linear runs (10 seconds “in”, 20 seconds “out”; Video 3). I primarily use these runs to gain information on mechanical variables via inertial measurement units (Table 1), but that’s for a different time.

The graph clearly shows that the player exhibits lower internal response (HR) to the same standardised stimulus across the activity. Here, I also calculated HRex as the mean HR across 4 × 10 second activity bouts, excluding recovery periods. This indicates that the player is becoming more efficient in responding to the high intensity intervals, reflecting improved aerobic fitness.

What’s the role of HRex in monitoring fatigue?

The utility of HR indices to gauge transient fatigue is uncertain and not straightforward, warranting the need for further research in this area. The physiological mechanisms supporting this rationale involve temporary changes in the autonomic nervous system (the balance between cardiac sympathetic and parasympathetic systems), and fluctuations in plasma volume ^[2,5].

For example, in ball sports, consecutive days of accumulated training loads (3-4 days) have shown varied effects on HRex. Moreover, endurance athletes experiencing training induced fatigue and functional overreaching (over longer periods, e.g., 10-20 days), sometimes respond in the opposite direction than expected, with lower SMFT HRex and higher SMFT HRR ^[2].

With this in mind, consider interpreting the trends in your results in relation to your training aims and specific context. Likewise, due to the potential sensitivity of HR indices to training loads, when the goal is monitoring changes in fitness, rather than fatigue, aim to maintain training load exposure heading into the assessment. Try scheduling it after the day off, either following matches or after a mid-microcycle break.

How to maximize submaximal fitness tests

Master your programming variables

If it wasn’t evident by this point, let me emphasise that running protocols are my top recommendation for maintaining specificity and feasibility. Just like when monitoring maximal aerobic fitness, you’d probably opt to do it through running based protocols.

Shuttle distances or track courses (e.g., oval, rectangle) are among the many course protocols that work for continuous-fixed SMFT. If you opt for shuttle courses, keep in mind that the multiple changes of direction can increase neuromuscular loading, potentially influencing the cardiovascular response. To minimize the impact on the neuromuscular system, program longer distances as applicable. In sports with oval fields, like Australian Rules Football, implementing track course protocols may be easier.

Instead of fussing with different equipment on the pitch like poles or cones (as I still do), simplify the process by using existing pitch markings to denote distances for the tests, in conjunction with the poles and cones. Slightly adjust running speed to ensure the athletes complete the shuttle in the desired time to maintain a constant speed.

Table 2 has a range of examples demonstrating continuous-fixed protocols customised for various sports, all using existing pitch markings. Remember, you have the flexibility to adapt the distances and the time for completing the shuttle, thus adjusting the speed as needed.

Sport	Marking	Distance	Shuttle time	Speed	Actual speed*	Number of shuttles
Football (soccer)	Box to box	72m	22s	11.8 km·h−1	12.7 km·h−1	~8/11
Rugby league	Try line to halfway	50m	15s	12.0 km·h−1	12.6 km·h−1	12/16
Rugby union	Try line to halfway	50m	15s	12.0 km·h−1	12.6 km·h−1	12/16
Australian football	Track (oval)	50m	15s	12.0 km·h−1	12.0 km·h−1	–
Netball	Touchlines (length)	30.5m	9s	12.2 km·h−1	13.2 km·h−1	18/24
Basketball	Touchlines (length)	28m	8.5s	11.9 km·h−1	12.9 km·h−1	~21/28
*Actual speed is based on the assumption that completing a 180° change of direction after each shuttle takes approximately 0.7 s. Number of shuttles are derived from a duration of 3-4 minutes, separated by a slash.

Don’t forget the influencing factors

Like with any test, many contextual and individual factors can influence data quality and accuracy, particularly given the integration of SMFT into training sessions. Considering that team sports often take place outdoors, environmental conditions like ambient temperature and humidity can impact cardiovascular response. Additionally, individual variables such as athletes’ diet pre-assessment (e.g., caffeine) should also be taken into account and, ideally, standardised ^[7].

Real changes: What are they?

The measurement error of HRex is ~1.5% ^[7]. These values refer to the raw values (i.e., percentage point of HRmax) rather than actual percentages. For instance, if an athlete’s baseline HRex is 85% of HRmax, a 1.5 percentage point margin of error indicates an expected range of variability between 83.5 – 86.5% of HRmax. The minimum detectable meaningful change in HRex is about three percentage points.

Having spent considerable time in research and on the field, my current approach considers a margin of 4-5 percentage points.

We have found that a five percentage point reduction in HRex is associated with improvements in maximal aerobic speed by 0.55 – 0.65 km·h⁻¹ (approximately an additional 200 m between tests). Future research should strive to determine the real smallest worthwhile change in HRex.

SMFT offer a practical tool to assess an athletes’ physiological state in a short duration and non-exhaustive manner. Herat rate derived indices, notably HRex and HRR, are prevalent and effective for tracking aerobic fitness changes.

To attain a stable HR response, start with continuous-fixed protocols lasting 3-4 minutes. You have the flexibility to administer other intermittent protocols (e.g., fixed, incremental). However, because of the alternating periods of activity and recovery, this might lead to a more variable HR response, requiring potential adjustments to your HR analysis timeframe.

It is also essential that you consider factors linked to protocols’ programming variables, including the running course, cues for maintaining running speed, and potential influencing factors like environmental conditions. This approach ensures accurate measurements and meaningful monitoring outcomes, reinforcing the adaptability of SMFT.