Quantifying CrossFit®: Potential solutions for monitoring multimodal workloads and identifying training targets

Abstract

The design of high-intensity functional training (HIFT; e. g., CrossFit®) workouts and targeted physiological trait(s) vary on any given training day, week, or cycle. Daily workouts are typically comprised of different modality and exercise combinations that are prescribed across a wide range of intensities and durations. The only consistent aspect appears to be the common instruction to maximize effort and workout density by either completing “as many repetitions as possible” within a time limit (e.g., AMRAP, Tabata) or a list of exercises as quickly as possible. However, because effort can vary within and across workouts, the impact on an athlete's physiology may also vary daily. Programming that fails to account for this variation or consider how targeted physiological systems interrelate may lead to overuse, maladaptation, or injury. Athletes may proactively monitor for negative training responses, but any observed response must be tied to a quantifiable workload before meaningful changes (to programming) are possible. Though traditional methods exist for quantifying the resistance training loads, gymnastic movements, and cardiorespiratory modalities (e.g., cycling running) that might appear in a typical HIFT workout, those methods are not uniform, and their meaning will vary based on a specific exercise's placement within a HIFT workout. To objectively quantify HIFT workloads, the calculation must overcome differences in measurement standards used for each modality, be able to account for a component's placement within the workout and be useful regardless of how a workout is commonly scored (e.g., repetitions completed vs. time-to-completion) so that comparisons between workouts are possible. This review paper discusses necessary considerations for quantifying various HIFT workout components and structures, and then details the advantages and shortcomings of different methods used in practice and the scientific literature. Methods typically used in practice range from being excessively tedious and not conducive for making comparisons within or across workouts, to being overly simplistic, based on faulty assumptions, and inaccurate. Meanwhile, only a few HIFT-related studies have attempted to report relevant workloads and have predominantly relied on converting component and workout performance into a rate (i.e., repetitions per minute or second). Repetition completion rate may be easily and accurately tracked and allows for intra- and inter-workout comparisons. Athletes, coaches, and sports scientists are encouraged to adopt this method and potentially pair it with technology (e.g., linear position transducers) to quantify HIFT workloads. Consistent adoption of such methods would enable more precise programming alterations, and it would allow fair comparisons to be made between existing and future research.

Normative Scores for CrossFit® Open Workouts: 2011–2022

“…Such efforts may ultimately prove to be redundant and unnecessary. Recently, it was suggested that relationships might exist amongst workouts and/or workout components (e.g., the pull-up component of “Fran”) [ 21 ]. If true, workout components or entire workouts might be classified, and normative scores may only be necessary for symbolic representations of workout types or classifications.…”

“…In each instance, the modified workout typically programmed variants of Rx exercises, prescribed different repetition counts (per exercise), and/or different intensity loads when applicable [ 3 ]. Because these differences alter the assigned workload, equating different CrossFit ® -style workouts is inherently difficult [ 21 ], and verifying modified workloads may not be possible, only Rx performances were considered for this study. Cases were also excluded if the reported age was not between 18 and 54 years due to the lack of clarity on the leaderboard about which workout version these athletes completed.…”

“…(kg)). Whenever the score could be officially quantified in multiple units (e.g., CFO 17.1 could be quantified as TTC or repetitions if the workout was not completed within the time limit), all scores were converted into a repetition completion rate (i.e., repetitions completed divided by TTC or workout duration; repetitions · minute −1 ) as previously described [ 21 , 22 ]. In these instances, the calculated repetition completion rate was used for all statistical analyses and to present sex differences, whereas the original scoring format was used to present normative scores.…”

Normative scores for CrossFit® Open Workouts: 2011 - 2022

“…Such efforts may ultimately prove to be redundant and unnecessary. Recently, it was suggested that relationships might exist amongst workout and/or workout components (e.g., the pull-up component of "Fran") [21]. If true, workout components or entire workouts might be classified, and normative scores may only be necessary for symbolic representations of workout types or classifications.…”

Muscular performance analysis in “cross” modalities: comparison between “AMRAP,” “EMOM” and “RFT” configurations

“…Within the realm of research, there exist inquiries aimed at scrutinising the immediate impacts of these training configurations ( Gómez-Landero and Frías-Menacho, 2020 ), with a particular emphasis on factors such as fatigue levels associated with these methodologies ( De-Oliveira et al, 2022 ). Despite the difficulty in defining and controlling the training stimulus, given the variable characteristics of the programs ( Mangine and Seay, 2022 ), we will focus on the strength exercises that are usually part of these programs. In this regard, the vast majority of the studies analyse the effects on similar absolute loads or, failing that, relative loads established with respect to the value of 1RM ( Maté-Muñoz et al, 2018 ; Silva-Grigoletto et al, 2020 ; De-Oliveira et al, 2021 ; De-Oliveira et al, 2022 ), which can present serious limitations ( Sanchez-Medina and González-Badillo, 2011 ; Elvar-Heredia et al, 2021 ; Maté-Muñoz et al, 2022 ).…”