When starting their training journey, people begin to learn many of the fundamental principles of programming and progression. Terms like “adaptation”, “recovery”, and “specificity” come into focus in the context of the person’s individual goals, whether they are focused on strength, power, cardiorespiratory endurance, or a specific skill.
The commonly cited principle of “Progressive Overload” holds that the body must be challenged by progressively greater training stimuli in order to generate fitness adaptations over time. This concept can be interpreted in two different ways, leading to different conclusions and practical strategies in training.
In one view, increasing the training stimulus itself stimulates adaptation and performance improvements. For example, lifting a heavier load directly drives a subsequent increase in strength.
In the other view, as strength increases from a given stimulus, subsequent load increases are necessary in order to “keep up” with this improvement. In other words, as we adapt and get stronger, we then become able to increase the load.
In this article, we’ll try to provide clarity on the sequence and practical implications of this principle in training.
We begin by introducing and defining the key terms for discussing exercise programming.
The training stimulus is the sum of all exercise activity performed, where exercise activity is planned, repetitive, and structured to improve or maintain health or fitness. The training stimulus is therefore defined by the program, including things like exercise selection, repetition and loading schemes, proximity to failure, rest periods, and other variables that can either be explicitly manipulated, or may be implicit to the programming plan.
The training stimulus produces a training stress that is unique to the individual, based on the program’s variables, the person’s current fitness and performance, the environment, and many other factors. This unique training stress is also known as the “internal load”, which can be measured by Rating of Perceived Exertion (RPE), Repetitions in Reserve (RIR), Ratings of Fatigue (ROF), or other metrics such as barbell velocity or heart rate, when applicable. The important takeaway is that the same general training program will generate different training stresses for different individuals, based on a variety of factors.
The training stress resulting from a given session or series of sessions produces both fitness adaptations and fatigue. “Fitness adaptations” collectively describe the positive changes induced by training such as increased muscular strength and size, improved cardiorespiratory endurance, reduced resting blood pressure, psychological changes, and others. “Fatigue” describes the subjective experience of negative exercise-induced changes such as muscular soreness, reduced force production, tiredness, and others.
A person’s performance potential at a given time is primarily determined by the balance of relevant fitness adaptations and fatigue, although other factors certainly play a role too. These include things like the environment, an individual’s mood, motivation, psychological arousal, and many others. If all other variables are constant, having more fitness adaptations relative to fatigue should allow a person to demonstrate improved performance.
Putting it all together, we’d like a training program to produce the desired fitness adaptations with a manageable amount of fatigue. If done correctly, many of these fitness adaptations should be demonstrable via an improvement in performance. If done incorrectly, for example if the training program provides an inappropriate type or amount of training stress, then fitness adaptations and performance potential are less likely to improve.
Fitness Adaptations and Performance
Many fitness adaptations can be measured through physical performance or physiological assessment. Adaptations measurable through performance include things like strength and cardiorespiratory endurance. Physiological assessments can be useful for metrics like resting blood pressure and bone mineral density.
Fatigue can play a major role in the measurement of adaptations that rely on physical performance, such as a 1-repetition maximum deadlift or a 100-meter sprint. In contrast, many physiological assessments like muscle mass, bone mineral density, or resting blood pressure are not as affected by training-related fatigue.
Physical performances have much more day-to-day variability than physiological assessments due to the numerous factors that influence performance, such as the environmental context (where you are, competition, temperature, time of day, etc.), motivation, levels of psychological arousal, mood state, and many others.
This variability makes predicting physical performance on any given day very difficult – if not impossible. This has important implications for generating the appropriate stress through load selection in training, and how we evaluate the effectiveness of a program that aims to improve physical performance.
Putting Adaptations on the Clock
Regardless of which interpretation of the progressive overload principle someone holds, the relationship between the training stress generated by the training stimulus and how quickly it generates fitness adaptations is important to understand.
Fitness adaptations accrue gradually over time. The time required for these adaptations varies between specific adaptations (strength, hypertrophy, endurance, skill, etc.), between different training programs, between individuals, and across people’s training careers.
Many progression models assume a predictable rate of adaptation and a constant performance potential — or at least minimal performance variability from session to session. For example, a program might prescribe a given training stimulus on day 1 with a predetermined increase in loading for day 2. This assumes 1) sufficient strength adaptations will have taken place between these sessions, and 2) the individual’s relative performance potential will be the same between both sessions, allowing them to express this adaptation.
Unfortunately, neither of these are safe assumptions. Humans are not robots, where defined inputs reliably lead to predictable outputs. Expectations that adaptations will necessarily occur, and reliably manifest in performance improvements on a particular predetermined timeline, are unrealistic. This is due to differences in adaptive rates between people, differences in individuals’ responsiveness to a given training program, and fluctuations in all the variables that can impact day-to-day performance.
The Optimization Problem
The contrasting views presented at the beginning of this article boil down to a question of sequencing. Should one increase the training stimulus and stress prior to any indication that fitness adaptations have actually accrued, i.e. before an increase in performance, in order to drive this improvement? Or, should one wait to increase the training stimulus, holding training stress constant, until an increase in fitness is evident, i.e., with a demonstrable improvement in performance?
Put more simply: does lifting more weight make someone stronger, or does someone get stronger and then lift more weight? The answer to this has significant implications for how we should go about our training and progression.
In exercise programming, the evidence suggests that there is a minimum threshold for weight (e.g. intensity) that is needed to improve strength performance. From existing evidence, we can ballpark a minimum threshold for strength training of approximately 65-75% 1RM for multiple-repetition work and greater than approximately 85-90% 1RM for single-repetition efforts. Schoenfeld 2017 Mangine 2015 Androulakis-Korakakis 2018 Pairing these intensity minimums with a rep scheme that is similar to the desired outcome, e.g. maximal strength/1RM and not strength endurance/20RM, lays the groundwork for an exercise program that is specifically targeting maximal strength performance.
Once the appropriate training intensity has been selected, there seems to be a relationship between training volume and strength improvement. Radaelli 2015 Rhea 2002 Ralston 2017 As training volume increases, so do improvements in strength, provided the fatigue is manageable. So, at a given intensity, how do we manage the fatigue being generated? This brings us to the concept of proximity to failure.
Training to muscular failure or near-failure (compared with staying further from failure) tends to generate more fatigue. Moran-Navarro 2017 Dos Santos 2021 In practice, a decrease in barbell velocity over the course of a set reflects decreasing force production due to the accumulation of fatigue within the set. Fuglevand 1999 Nocella 2011 Training closer to failure produces more stress than staying further away from failure.
Recent research shows that training closer to failure, as evidenced by greater decreases in barbell velocity within each set, does not produce greater strength gains than a smaller (10-20%) loss in velocity. Pareja-Blanco 2016 Pareja-Blanco 2016 Similarly, outside of untrained individuals doing isolation exercises (e.g., biceps curls), training to failure does not produce more muscle growth than training in the range of 2-4 repetitions in reserve. Carroll 2019 Helms 2018 These illustrate how “higher stress” workouts do not always produce greater fitness adaptations than “lower stress” workouts.
Matching Stimulus & Performance
Collectively, these findings suggest that the second view of progressive overload is likely the more correct interpretation: an increase in performance potential should be met with an increase in training stimulus to maintain the desired training stress, rather than increasing training stress to directly drive improvements in performance. We get “bigger, stronger, or faster,” and are then able to lift more weight, handle more volume, or complete conditioning tasks more quickly.
Many people believe things occur in the reverse sequence due to a mis-interpretation and, ultimately, a mis-application of the “overload” principle. In the strength training world, this is the idea that increasing weight itself drives the adaptations in strength. For example, an individual already struggling to lift a given load may be reluctant to decrease or repeat this load during a subsequent session, due to a belief that the only way to continue driving strength adaptation is to keep increasing the load – and that failing to do so represents a waste of time. This may be an unfortunate consequence of the term “overload”, rather than simply describing a process of “progressive loading”.
This interpretation of the overload principle leads people to feel like they have far more control over the process of adaptation than they really do, and to attempt to force adaptation by increasing the stimulus more quickly than their rate of adaptation allows. This thought process seems to be less prevalent in the conditioning world; it is analogous to telling a 100-meter sprinter to simply run their 100 meters in less time in order to drive the necessary adaptations to get faster — an assertion that is silly on its face.
Increasing the training stimulus too quickly out-paces the individual’s level of adaptation and generates excessive fatigue. This lowers performance potential, increases injury risk, and — based on the data we’ve described above — may not actually produce greater fitness adaptations in the end. Conversely, if the increase in stimulus occurs too slowly, fitness adaptations will not develop at their maximal rates, if at all. Fortunately, this is a fairly wide-range; over a sufficiently long training career it is not critically important to “optimize” rates of adaptation, nor is there any way to ensure such “optimization” given the number of variables involved.
Ultimately, if the training stimulus is not matched to the person’s current fitness and performance potential, the individual won’t get the best result. The person’s adaptation rate, as reflected by their performance potential, should determine the training stimulus on a given day. Providing the right dosage of stimulus during a particular training session is easier said than done, but the rate at which this stimulus increases over time should approximate how quickly the adaptations are occurring – not by how quickly we want them to occur. In this way, the weight loaded on the bar should be reactive, not proactive.
Practicing Progressive Loading
Putting things together, we can propose basic criteria for a general strength program, from which further individualization can occur depending on the person’s demonstrated response. These criteria include:
- Exercises specific to the desired adaptations, e.g., specific movements, range of motion, movement velocity, contraction type, individual preferences, etc.
- A minimum threshold for strength training of ~65-75% 1RM for multiple-repetition work and greater than ~85-90% 1RM for single-repetition efforts.
- Rep schemes that are similar to the desired outcome(s) being tested.
- Total training volume that is sufficient to drive adaptations at an acceptable rate, yet is also well-tolerated.
- Proximity to failure that generates relatively small decreases in velocity and does not routinely expose the individual to muscular failure.
- Autoregulation tools to guide load selection to match the current fitness and performance level, while maintaining an appropriate proximity to failure.
- A progression model that supports the criteria above.
As the individualization process proceeds, people may deviate from these criteria, but these should be decisions informed by their unique response to the program. Several of the criteria above pertain to weight selection, which highlights the importance of this factor. So, how do we pick the right weight to get the best results, and what happens if we go wrong?
We need within-workout tools that can reliably tell us whether the training is appropriate or not. We think the best tools currently available involve measures of internal load; depending on the sport / task, these include Ratings of Perceived Exertion (RPE), Repetitions in Reserve (RIR), barbell velocity (for barbell training), heart rate (for conditioning training), and/or combinations of these.
An individual training for aerobic endurance, such as running or swimming, may be programmed to complete a 5 km run or swim at a given heart rate target. Monitoring this heart rate in real time allows them to determine whether their performance potential on that day allows for a given pace. If their heart rate is too high during training, the pace needs to slow down in order to provide the appropriate training stress without excessive fatigue. If the heart rate is too low during training, their performance potential is higher that day and they can push a faster pace to hit the intended training stress. This illustrates the value of using metrics of internal load to regulate training intensity.
In the gym, consider the example of an individual whose workout calls for squatting 5 reps @ RPE 8 with two subsequent lighter sets of 5 reps at 75% of their estimated 1-Repetition Maximum (1RM).
On this particular day, they work up a set of 5 repetitions at 315 pounds. They rate the set RPE 8, estimating that they had 2 reps left in reserve (2 RIR). From there, they estimate their 1RM at around 389 pounds using an 1RM estimation calculator. Then they perform two sets of 5 reps at 292 pounds (75% of their estimated 1RM for the day). By using RPE during the workout, the lifter got real-time feedback on their current fitness and performance level, which allowed them to adjust loading and generate the desired training stress from the session.
The next week the lifter has the same workout programmed again. The lifter feels fine and their warm-up is roughly the same as the previous week’s with respect to bar speed, exertion, and technical proficiency. What should they do today?
Option 1: The Full Send
In this scenario, the individual believes that they must force the adaptation to occur by adding more weight to the bar. They think, “If it’s not heavier, it’s not going to work, right?” This lifter increases to 320 pounds for their set of 5 reps and rates it RPE 9 (1 RIR).
Option 2: The Repeater
In this scenario, the individual believes they are performing about the same as the previous week. If they are trying to match the training to performance for this session, and last week’s performance was the same, it seems appropriate to repeat the weight. This lifter does 315 pounds for their top set of 5 reps and rates it RPE 8 (2 RIR).
Option 3: The Undershoot
In this scenario, the individual also believes they are performing roughly the same as they were the previous week. However, they aren’t as sure that 315 pounds for 5 reps would be RPE 8 this time around. Cautiously, this lifter does 305 pounds for their top set of 5 reps and rates it RPE 7 (3 RIR).
The Repeater most likely got it right, by matching their training load and performance potential. While they weren’t able to showcase an improvement in performance in a week’s time, they continued to stack useful training stress (since they are still training within a useful intensity range and proximity to failure), that will cumulatively drive fitness adaptations over a slightly longer period of time, without generating too much fatigue.
The Full Send was not actually stronger, as evidenced by their inability to complete the same task at a heavier weight. While they lifted 5 pounds more, it was harder and slower. In this way, using a qualitative metric (RPE, RIR, or bar speed) can be helpful in measuring actual progress. Folks who are actually getting stronger should be able to increase weight on the bar without significant increases in relative effort (RPE) or changes in other variables. A load increase with a commensurate increase in RPE (or a decrease in bar velocity, if more objective measures are used) shows that adaptation had not yet taken place, AND that the lifter generated more fatigue. As discussed above, training closer to failure does not necessarily produce greater adaptations and may confer a higher risk of developing aches, pains, and injury if done consistently.
The Undershoot also didn’t accurately match their training load and performance potential for the day. However, importantly, they did not generate excess fatigue. Since they are still within the useful range of intensity and proximity to failure, their training still provided productive training stress, while mitigating the risks described above. Over the course of an entire training career it is preferable to “undershoot” more often rather than to “overshoot”. The perception that training must be maximally time-efficient, increasing loads as quickly as possible in order to avoid “wasting time”, often ends with frustration, burnout, or pain and injury.
The converse situation — chronically “undershooting” so severely that the individual never increases load at all — can be an issue for some trainees as well. This may benefit from some analysis and encouragement from a coach if the individual’s goals are to get the most performance improvement possible out of their training. On the other hand, if an individual is simply training for general health, then absolute loading is never of primary concern, but rather we are satisfied with simply meeting general physical activity guidelines alongside any other individual goals.
This “matching” of loading to performance potential is not an exact science, and there is no way to definitively prove that we have perfectly matched these variables in training. However, using metrics of internal load (such as RPE/RIR, bar velocity, or — in conditioning tasks — heart rate), and being honest with ourselves can best help us adjust the intensity of our training in order to deliver a “good enough” stimulus that isn’t “too much”.