A PRISMA-guided systematic review of musculoskeletal modelling approaches in lower-limb cycling biomechanics

Abstract

Cycling is commonly employed in sports performance, rehabilitation, and clinical contexts, while musculoskeletal (MSK) simulations enable the investigation of internal biomechanics that cannot be measured experimentally. Despite growing use, the application, validation, and standardisation of MSK simulations in cycling remain unclear. This review aimed to systematically characterise the application, validation strategies, modelling assumptions, and reporting practices of musculoskeletal simulations in lower-limb cycling biomechanics. Searches were performed in Scopus, PubMed, IEEE Xplore, and Web of Science on 1 August 2024, covering studies from January 2010 to July 2024. Peer-reviewed English-language journal articles applying MSK simulations to lower-limb cycling were included; inverse kinematics-only was excluded. No protocol was registered, and no formal risk-of-bias assessment was conducted, as there were no intervention effects and no quantitative synthesis. Twenty-eight studies met the inclusion criteria. Most of them investigated bicycle-rider configuration, neuromuscular coordination, or electrical stimulation control, with participant cohorts overwhelmingly composed of young men and minimal female representation (272 total). Model reporting was often incomplete, with wide variation in anatomical scope, inconsistent descriptions of degrees of freedom, and limited sharing of models or code. Use of experimental data was uneven across studies: while all incorporated kinematic measurements, only two-thirds included kinetic data, and only one study reported physiological measures. Model validation was generally based on literature values. Seventy-eight per cent of studies used optimisation, mainly with effort-based cost functions, and parameter variations were exploratory rather than systematic. The evidence base is limited by small, predominantly male cohorts, inconsistent reporting standards, and limited physiological validation. These results consolidate current practices and highlight the need for more transparent and open reporting, sex-balanced and clinically diverse participant representation, stronger validation, and more rigorous sensitivity analysis to enhance reproducibility and practical relevance. This review was funded by AGAUR (Spain), CAPES (Brazil) and FAP-DF (Brazil). Author summaryCycling is widely used in sports training, rehabilitation, and clinical practice, and musculoskeletal simulations are increasingly used to study how muscles and joints work during cycling. These simulations allow us to estimate internal biomechanical variables that cannot be directly measured in experiments, such as muscle forces and joint loading. However, it is currently unclear how consistently these simulations are applied, validated, and reported across the literature. In this study, we systematically reviewed research published over the past 15 years that used musculoskeletal simulations to analyse lower-limb cycling. We identified 28 relevant studies and examined their modelling choices, experimental inputs, optimisation strategies, and validation approaches. We found substantial variability in model complexity, limited transparency in reporting, and a strong reliance on simplified literature-based validation methods. Most studies focused on narrow participant groups and explored modelling parameters in an ad hoc rather than a systematic way. Our findings highlight important gaps in current practice and point to clear opportunities for improvement. We provide an overview of common approaches and their limitations, and outline key recommendations to enhance the transparency, reproducibility, and practical relevance of musculoskeletal simulations in cycling research.