Time to return to sports in football players recovering from lower limb soft tissue injuries: a systematic review with meta-analysis Authors: Jente Wagemans, Elaheh Ahmadi, Jan Taeymans, Jeroen Van Eetvelt, Geert Van Neylen, Peter Catteeuw, Heiner Baur, Dirk Vissers Published: 2026 ABSTRACT AbstractObjectivesDepict football players’ average times to return to sports (RTS) after non-contact lower limb soft-tissue injuries.DesignSystematic review with meta-analysis.Data sourcesPubMed, WoS, Scopus, PEDro, Google Scholar and Cochrane Central Register of Controlled.Eligibility criteria for selecting studiesWe included records investigating time to return to sports in football players suffering from non-contact lower limb soft-tissue injuries.Data synthesisTime to return to sports was the main outcome. We appraised the quality of the included studies with the QualSyst tool for quantitative research. A random-effects model was wielded for the meta-analyses.Results39 studies were included. Time to return to match (RTM) for ligament injuries ranged from 43 days (lateral collateral ligament) to 249.6 days (ACL rupture and repair (ACLR); RTM for muscle injuries ranged from 21 days (hamstring muscle injury) to 69.1 days (soleus); RTM for tendon injuries ranged from 29.4 days (patella tendinopathy) to 220.4 days (Achilles tendon rupture). Time to return to training (RTT) for ligament injuries is also longest for ACLR (212.9 days) and lowest for lateral ankle sprain (19.9 days); RTT for muscle injuries ranged from 8 days (hamstring muscle injury grade 0 to 65.3 days (HMI grade 3)); popliteus tendinopathy (11 days) had lowest RTT among tendon injuries, groin pain had the longest (45 days).ConclusionsThis review provides an overview of RTS times—expressed as RTT and RTM—for lower limb injuries in football players. Our results also underscore the existing variability in wielded RTS definitions and injury classifications.Keywords: Injury, Rehabilitation, Football, Sports medicine WHAT IS ALREADY KNOWN ON THIS TOPIC Many studies report return to sports (RTS) times per injury. Many studies report return to sports (RTS) times per injury. WHAT ARE THE NEW FINDINGS An overview of pooled RTS times for lower limb injuries in football players. An overview of pooled RTS times for lower limb injuries in football players. This study discusses nuances around RTS times concerning healing processes, rehabilitation phases, uniformity regarding RTS definitions, grading or categorisation. This study discusses nuances around RTS times concerning healing processes, rehabilitation phases, uniformity regarding RTS definitions, grading or categorisation. Introduction Football is the most popular sport in the world with up to 450 million people playing football worldwide1, with still rising popularity.2 The lifetime injury rate in football players is 1000 times higher than in non-athletes.3 A professional football player sustains on average two to four time-loss injuries per season.4 5 Lower extremity injuries reportedly accounted for 86% of all injuries during the UEFA EURO 2024 and CONMEBOL Copa America football tournaments6, of which injuries to the thigh region have the highest overall incidence rate (1.8 injuries/1000 hours of exposure) among lower limb injuries, with the hamstring muscle being the most commonly involved.5 About 37% of all muscle injuries in football are hamstring muscle injuries (HMI), followed by adductor muscles (23%) and quadriceps muscle (19%)5 7 8 injuries—most frequently muscle strains. The second injury location in terms of incidence is the knee (1.2 injuries/1000 hours exposure).5 Among those, reported incidence rates are 0.51 per 1000 hours of exposure (men) and 0.87 per 1000 hours of exposure (women) for ACL injuries9, 0.33 per 1000 hours of exposure for medial collateral ligament (MCL) injuries. Ankle injuries in football comprise 1.1 injuries per 1000 hours of exposure5, with the ankle sprain as the most common ankle injury (0.9 per 1000 hours of exposure).10 These numbers highlight the clinical burden of lower limb injuries in football. These lower-limb injuries reduce the amount of time football players are available for training or participation in matches, which is reported as time to return to sports (RTS). The absence of usable RTS timelines limits the ability for substantiated decision-making after an injury based on tissue healing, physical and mental capacity and contextual factors (eg, timing, pressure, financial burden due to direct and indirect costs).11 12 Understanding the average time to RTS following a lower limb injury in football players can help manage expectations for athletes, medical professionals and coaches within the club involved in rehabilitation.13 Realistic goals and timelines can reduce anxiety and uncertainty in athletes.14 Such uncertainty may lead to premature RTS—increasing the risk of re-injury—or overly conservative rehabilitation—increasing the burden on the team and the player.13 Knowing RTS timelines can be used to guide individual treatment plans for football players.13 The current literature body on the topic of time to RTS is large, however diverse and scattered. This reduces quality and comparability in research where studies may use heterogeneous and poorly defined data, making it difficult to compare outcomes across cohorts. This may yield reduced external validity and reproducibility of scientific output.15 Therefore, the aim of this present study is to provide an overview of the existing literature regarding RTS definitions and time, and to meta-analyse time to RTS in football players after soft-tissue injuries of the lower extremities. Methods This systematic review with descriptive meta-analysis was carried out according to the updated Preferred Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines and PRISMA in Exercise, Rehabilitation, Sport medicine and SporTs science guidelines (PERSiST).16 17 We registered the protocol of the systematic review a priori with Open Science Framework (https://doi.org/10.17605/OSF.IO/24F59). We systematically searched PubMed, Web of Science, Scopus, Physiotherapy Evidence Database (PEDro), Google Scholar and Cochrane Central Register of Controlled Trials to obtain eligible articles from inception throughout March 2026. By using the PICOS (Population, Intervention, Comparison, Outcome, Study design) model, the search strategy was established for a MEDLINE search using Boolean operators (‘and’ and ‘or’), free search terms, truncation and Medical Subject Headings. Online supplemental file 1 provides the integral search strategy. Subsequently, the search strategy was adapted accordingly to the specific requirements of each database. Additionally, reference lists of articles that were identified during the searches and grey literature were reviewed to find further eligible articles. Studies that included non-contact lower limb injuries in football players were eligible for inclusion. table 1 depicts the detailed eligibility criteria. Table 1Eligibility criteriaInclusion criteriaExclusion criteriaStudy designRandomised controlled trialsCase–controlled studiesProspective cohort studiesRetrospective studiesCross-sectional studiesParticipantsFootball players with soft-tissue injuries to the lower extremities:Ligament, muscle, tendonOther sports than football, other injuries than lower limb soft-tissue injuries, contact injuriesIntervention//Outcome measuresTime to return to sportsAll other outcomes than time to return to sportsComparison// Eligibility criteria All identified articles were transferred to Rayyan screening tool for removal of duplicates and both screening phases.18 Screening for eligibility was performed in two phases: title and abstract screening, and full text screening; two independent reviewers (GVN and JVE) performed the eligibility screening double-blinded. We consulted a third reviewer (EA) in case of disagreement to reach consensus. online supplemental file 2 lists the citations excluded during full-text screening. Quality of the included articles was assessed double-blinded by two independent reviewers (GVN and JVE) using the ‘Qualsyst tool for quantitative research’.19 This tool was chosen because of its suitability for evaluating case–control and cohort studies as well as randomised controlled trials (RCTs).19 The QualSyst tool includes 14 items (online supplemental file 3), each of which must be answered with either ‘yes’ or ‘partial’ or ‘no’ or ‘N/A’. The following data were extracted independently by two reviewers (GVN and JVE): author, publication year, study design, language, participants demographics, inclusion/exclusion criteria of the study, sample size, intervention (if applicable), duration of the study, duration of follow-up, time to RTS, definition of RTS, level of evidence. Time to RTS was categorised in time to return to training (RTT) or time to return to match (RTM) while injuries of the lower limb were categorised based on ligament, muscle or tendon involvement. Results from studies investigating the same measures of the RTS construct (ie, RTT or RTM) among football players with lower limb injuries are synthesised and presented in different forest plots per type of lower limb injury (ie, ligament, muscle and tendon injuries). If, for example, in the subgroup RTM of the studies on tendons, five studies investigated RTM among athletes after Achilles tendon rupture and repair, these five results were first meta-analysed. The calculated weighted mean RTM value together with its corresponding 95% CI was then included as a single study result in the forest plot together with RTM results from other studies investigating RTM among football players suffering from other types of lower limb tendon lesions. For reasons of lacking plausibility, authors decided to omit further RTT or RTM subgroup meta-analyses. Hence, in the forest plots depicted in the results section of this present work, diamonds representing the overall weighted mean subgroup RTT or RTM values are not shown. If studies reported central tendencies and variation in terms of medians and IQR instead of means and SD, these median values were used as means while SD was calculated as: SD=IQR/1.35 (https://training.cochrane.org/handbook). If studies reported variation in terms of range with highest and lowest observed values, SD was calculated following the guidelines of Wan et al using the online application (math.hkbu.edu.hk/~tongt/papers/median2mean.html).20 In the case of multiple study arms in RCTs, time to RTS values from both arms were included separately. To avoid over-representation of the number of participants, we corrected for the exact sample size per group.21 For this systematic review and descriptive meta-analyses, we allowed a priori a broad variety of study designs. Hence, a high amount of heterogeneity could be expected. Pooling of data were conducted using a random-effects model with inverse variance method of DerSimonian and Laird.22 If a meta-analysis was conducted, heterogeneity across time to RTS values included in the meta-analysis was assessed. First, a χ2 test was conducted to test the assumption of homogeneity across study results. χ2 values with their corresponding df and p values were calculated and presented. Second, Higgins’ I2 statistic was calculated to evaluate the portion of total observed variation that can be explained by the true between study variation and was expressed in per cent. For the interpretation of the I2 statistic, benchmarking values as proposed by Higgins were used: I2 values around 25%, 50% and 75% or more represent low, moderate and high heterogeneity, respectively.21 In case observed heterogeneity was statistically significant and moderate or high, one or more subgroup analyses were conducted in an attempt to explain at least a part of this heterogeneity. All meta-analyses calculations and development of the forest plots were conducted using the Comprehensive Meta-Analysis V.2.0 software (Biostat, USA). Results Study selection We screened a total of 2677 records, of which we retained 154 for full-text screening, rendering 44 articles that were eligible for inclusion in our qualitative analysis (figure 1). Preferred Items for Systematic Reviews and Meta-analyses (PRISMA) flow diagram. Study characteristics Online supplemental file 4 summarises characteristics of the 44 included studies. 18 studies310 23,38 were prospective cohort studies, 1339,51 were retrospective studies, 852,59 epidemiological studies, 2 RCTs60 61, 1 survey study62 and 2 case-controlled studies63 64. 9 studies3 10 23 25 26 35 38 50 51 only reported the number of included injuries instead of participants (range: 207–5603 injuries). Sample sizes of the included studies varied from 12 to 671 participants. The majority of studies (38/44) included only elite football players; two studies45 61 a combination of elite, semi-professional and amateur football players; two studies32 37 reported on youth players; one60 on both elite and sub-elite football players; and another study (62) on amateur players. Only five studies included female football players.(31, 34, 38, 59, 61) 17 studies reported return to sports times after muscle injuries(23-25, 28-30, 33, 35, 36, 38, 41, 42, 46, 50-52, 60, 61), 20 studies after ligament injuries(3, 10, 26, 31, 34, 35, 38-40, 44, 45, 48, 49, 53, 56, 58, 59, 62-64) and 11 studies did so after tendon injuries(27, 32, 35, 37, 38, 43, 47, 54, 55, 57). Several studies overlapped in terms of the included injured structures. The rate of players returning to sports ranged from 63% to 100%; this was not specified by 23 of the 44 records. 12 studies emphasised return to training in terms of return to sports (3, 10, 23-25, 27, 32, 38, 39, 42, 46, 52, 61); 17 studies focused on return to match/ competition (34-36, 45, 48-51, 53, 54, 57-59, 63); 5 studies (26, 30, 55, 62, 64) included both return to training and match separately. Quality appraisal of the included studies Table 2 illustrates the scores assigned to each individual study by the QualSyst tool for quantitative research. Three studies scored maximally across all criteria.41 45 63 All but one study47 defined outcome and exposure measures and reported means of assessment. Also, all but one record34 reported sufficient detail about the results, while all but three studies26 37 53 described the research question/objective sufficiently. All but three studies23 34 59 reported some estimate of variance for the main result. About 85% (37/44) of the included studies had evident and appropriate study designs and had appropriate and well-described analytical methods. Also, in 85% (38/44) of the included records, the conclusion was supported by the results. Selection criteria were properly described in 64% (28/44) of the studies. 30 out of the 44 included studies (73%) described subject characteristics sufficiently. We included four studies with an intervention, of which two studies37 60 described or reported the items regarding interventions properly. Sample size was appropriate in almost 64% of the studies (28/44). Almost 40% (17/44) of the included records controlled adequately for potential confounding factors according to the QualSyst tool. 4 out of the 39 included studies (10%) were RCTs, and hence were scored for items 5, 6 and 7. These items were not applicable for the other studies. Two37 60 out of the four RCTs (50%) were scored as ‘appropriate’ for these three items. Table 2Quality appraisal with the ‘Qualsyst tool for quantitative research’19AuthorsCriteria as described in online supplemental file 31234567891011121314Ekstrand et al23Waldén et al10Erikson et al63Lundblad et al3Hallen et al25Waldén et al26Ekstrand et al24Mendiguchia et al60Crema et al52Gouttebarge et al27Pezzotta et al28Luiso et al62Krutsch et al53Sochacki et al54Grassi et al55Jiménez-Rubio et al29D’Hooghe et al48Valera-Garrido et al30Leiderer et al39Brophy et al31Forsythe et al40Lavoie-Gagne et al56Lavoie-Gagne et al41Shamji et al42Eberbach et al32Forlenza et al43Jiménez-Rubio et al33Hong et al 202234Bonanzinga et al44Mazza et al58Mazza et al57Szymski et al45Krutsch et al35Valle et al46Shimozaki et al36Gonzalez-Iglesias et al61Guarneri et al47Niering et al37Abed et al59Battaglia et al49Chamari et al 2025Ekstrand et al38Forsythe et al50Forsythe et al50Grassi et al64Percentage per item93%84%64%73%50%50%50%98%64%91%93%39%97%86%Colour backgrounds represent the following: Green:Yes,Yellow: partially,Red:No. Quality appraisal with the ‘Qualsyst tool for quantitative research’19 Colour backgrounds represent the following: Green:Yes,Yellow: partially,Red:No. Return to sport In football players recovering from different types of lower limb injuries (ie, ligament injuries, muscle injuries and tendon injuries), different proxies (ie, RTT, RTM) for the time to RTS construct were analysed. online supplemental file 5 provides a detail of RTM and RTT times for all injuries, from all separate studies. Ligament injuries Figure 2 RTM in football players after ACL rupture and repair (ACLR) was reported in 13 studies,2631 35 40 44 45 49 53 58 59 62,64 including 1499 athletes. Meta-analysis of these 13 time to RTM values found a weighted mean TRM of 249 (95% CI 219 to 280) days. Heterogeneity was statistically significant (I2=796, df=12, p<0.001) and high (I2=98.5%). A subgroup analysis for different study design (prospective vs retrospective) was conducted. Retrospective studies26 31 35 49 53 58 59 62 63 (k=9; n=1199) showed a weighted mean time to RTM value of 258 days while prospective studies40 44 45 64 (k=4; n=248) found a weighted mean time to RTM value of 230 days. The between-subgroup difference was statistically not significant (p=0.477). I2 decreased to 86% in the group of prospective studies but remained similar in the retrospective studies group. Time to RTM among football players after 14 other types of lower limb ligament injuries was investigated always in one single study. Hence, no meta-analysis was conducted for these studies, but extracted time to RTM values of the 11 individual studies are depicted in the forest plot. Time to RTM after a partial ACL rupture takes 99 (95% CI 0 to 211) days on average. Time to RTM after lateral collateral ligament (LCL) injury takes 43 (95% CI 19 to 67) days—shortest among ligament injuries in the knee, and overall ligament injuries. In terms of ankle ligament injuries, mean time to RTM is longest for a deltoid ligament rupture (125.5 days; 95% CI 65.7 to 185.3), followed by a syndesmosis rupture with surgical fixation (103 days; 95% CI 48.1 to 157.9). Forest plots of ligament injuries. ACL(R), ACL (reconstruction); k, number of studies; LCL, lateral collateral ligament, MCL, medial collateral ligament; n, sample size; PCL, posterior cruciate ligament; RTM, return to match; RTT, return to training. Time to RTT of football players after lateral ankle sprain, MCL injury or ACLR was evaluated by always two reports (with n ranging from 215 to 604 players) and hence these RTT values were meta-analysed per injury type. Meta-analysis of the two studies investigating RTT of football players after ACLR showed a mean Time to RTT value of 212.9 (95% CI 113.6 to 228.9) days. Heterogeneity was statistically significant (I2=425, df=2, p<0.001) and high (I2=99.5%). This pooled time to RTT value for lateral ankle sprain and MCL injury was 19 (95% CI 14 to 26) days and 27 (95% CI 20 to 34) days, respectively. Heterogeneity across the lateral ankle sprain studies was statistically significant (χ2=28.9, df=2, p<0.001) and high (I2=93.1%). Heterogeneity across the MCL injury studies was statistically not significant (χ2=4.9, df=2, p=0.088) and moderate (I2=58.9%). Heterogeneity in these three meta-analyses was not further analysed because of the small number of studies included (all k=2). Time to RTT among football players after seven other types of lower limb ligament injuries was investigated always in one single study. Hence, no meta-analysis was conducted, but extracted time to RTT values from the seven individual studies are depicted in the forest plot. Mean time to RTT after a syndesmosis injury with surgical fixation was 72 (95% CI 19.1 to 124.9) days—longest among ankle ligament injuries. Shortest time to RTT among knee ligament injuries is after an LCL injury (weighted mean 23 days; 95% CI 0 to 74.0); shortest time to RTT among ankle ligament injuries is 20 (95% CI 4 to 26) days, after a lateral ankle sprain. Muscle injuries figure 3 depicts (weighted) mean TRS values and corresponding 95% CIs among football players after different lower limb muscle injuries. Forest plots of muscle injuries. ACL(R), ACL (reconstruction); BAMIC, British Athletic Muscle Injury Classification; k, number of studies; MMIC, Munich Muscle Injury Classification; n, sample size; RTM, return to match; RTT, return to training. Time to RTM among football players after rectus femoris muscle injury was analysed in three studies, totalling 52 injured players. Meta-analysis of these three studies yielded an overall weighted mean time to RTM of 27.4 (95% CI 14.7 to 40.2) days. Heterogeneity was statistically significant (χ2=50, df=2, p<0.001) and high (I2=95.9%). Also, three studies reported time to RTM after adductor muscle injuries, with meta-analysis showing weighted mean time to RTM of 32 days (95% CI 13 to 51 days; heterogeneity: χ2 114, df=2, p=<0.001; I2=98.2%). Two studies (n=36) reported on RTM times among football players after soleus muscle injury. Mean weighted time to RTM value was 69.1 (95% CI 12.3 to 125.9) days. Heterogeneity was statistically significant (χ2=28, df=1, p<0.001) and high (I2=96.4%). Three reports of a further two studies showed RTM time data among football players after grade 1 HMI with a mean weighted time to RTM of 31.0 (95% CI 20.5 to 41.4) days. Heterogeneity was statistically significant (χ2=14, df=2, p=0.001) and high (I2=85.3%). Because of the low number of studies included in these meta-analyses, observed heterogeneity was not further assessed. One study reported on time to RTM in football players after quadriceps muscle tear injury and HMI grade 1–3 with the results also depicted in the forest plot. Time to RTT among football players after grade 1 HMI was assessed in four studies, totalling 477 injured athletes. Meta-analysis of these four studies yielded an overall weighted mean time to RTT value of 17.1 (95% CI 15.6 to 18.6) days. Heterogeneity was statistically not significant (χ2=4.6, df=3, p=0.203) and low (I2=34.8%). Another four studies reported on RTT among football players after grade 2 hamstring muscle injury (n=234). Meta-analysis of these four studies yielded an overall weighted mean time to RTT value of 23 (95% CI 22 to 24) days. Heterogeneity was statistically not significant (χ2=2.1, df=3, p=0.554) and low (I2=0%). A further two studies were found, analysing RTT among football players after grade 3 HMI (n=14). Meta-analysis of these two studies found an overall weighted mean time to RTT value of 65.3 (95% CI 0 to 95.8) days. Heterogeneity was statistically not significant (χ2=0.17, df=1, p=0.682) and low (I2=0%). Also, time to RTT after calf muscle and quadriceps muscle injuries were reported in each of the two studies. Meta-analyses yielded weighted mean time to RTT times of 17 days (95% CI 2 to 31; heterogeneity: χ2=29, df=1, p=<0.001; I2=99.5%) and 17 days (95% CI 0 to 39 days; heterogeneity: χ2=61, df=1, p=<0.001; I2=98.4%) time to RTT among football players after 13 other types of lower limb muscle injuries was investigated in a further 13 individual studies. Hence, no meta-analysis was conducted, but the extracted RTT values and corresponding 95% CIs from these individual studies are depicted in the forest plot. The weighted mean time to RTT in football athletes recovering from HMI ranges from 8 (95% CI 2 to 13.9) days after grade 0 injury to 65.3 (95% CI 34.8 to 95.8) days after a grade 3 hamstring muscle injury. When looking at the British Athletic Muscle Injury Classification (BAMIC), time to RTT ranges from 8.0 days (BAMIC 1a—95% CI 0 to 30.4) to 49.3 (BAMIC 2c—95% CI 0 to 208.3) days. Average terms of return to training (TRT) after hamstring injury based on the Munich Muscle Injury Classification (MMIC—3b) take 29.4 (95% CI 24.2 to 34.6) days. Average time to RTT after a rectus femoris injury is 15.6 days (95% CI 12.1 to 19.2), while it takes football players, on average, 13.1 days (95% CI 10.3 to 15.9) to RTT after an adductor muscle injury (figure 3). Tendon injuries Figure 4 shows (weighted) mean Time to RTS values and corresponding 95% CIs among football players after different lower limb tendon injuries. Forest plots of tendon injuries. k, number of studies; n, sample size; RTM, return to match; RTT, return to training. Time to RTM among football players after Achilles tendon rupture and repair was analysed in five studies, totalling 313 injured players. Meta-analysis of these five studies yielded an overall weighted mean TRM value of 220.4 (95% CI 167.4 to 273.4) days. Heterogeneity was statistically significant (χ2=101, df=4, p<0.001) and high (I2=96.0%). Meta-analysis of three reports from two studies evaluating time to RTM among football players after patella tendinopathy (n=43) found a weighted mean RTM value of 29.4 (95% CI 0 to 59.2) days. Heterogeneity was statistically significant (χ2=50, df=2, p<0.001) and high (I2=96.0%). Because of the low number of studies included in these meta-analyses, observed heterogeneity was not further assessed. RTM among football players after three other types of lower limb tendon injuries was investigated in a further three individual studies. Hence, no meta-analysis was conducted, but the extracted RTM values and corresponding 95% CIs from these individual studies are depicted in the forest plot. Time to RTT among football players suffering from groin pain was assessed in three studies, totalling 64 athletes. Meta-analysis of these two studies yielded an overall weighted mean time to RTT value of 45 (95% CI 18 to 72) days. Heterogeneity was statistically significant (χ2=67.1, df=2, p<0.001) and high (I2=97%). Because of the low number of studies included in this meta-analysis, observed heterogeneity was not further assessed. Time to RTT among football players after six other types of lower limb tendon injuries was investigated in a further three individual studies. Hence, no meta-analysis was conducted, but the extracted time to RTT values and corresponding 95% CIs from these individual studies are depicted in the forest plot. Discussion The aim of this systematic review with descriptive meta-analysis was to provide clinicians working with athletes an overview of the different times to RTS among football players recovering from lower limb musculoskeletal injuries. Within lower limb soft-tissue injuries, we distinguished ligament injuries, muscle injuries and tendon injuries. Similarly, we considered different perspectives on RTS time—RTM or RTT—within each of these structural categories. In this limited set of studies out of the universe of studies describing time to RTS among football players after lower limb soft tissue injuries, evidence suggests that time to RTS (both RTM and RTT) after ACL rupture and repair is longest among all lower limb musculoskeletal injuries (RTM: mean 249.6 days, 95% CI 218.9 to 280.2; RTT: mean 212.9 days, 95% CI 253.7 to 272.1). Within the included studies, a grade 0 hamstring injury takes fewest number of days (mean 8 days; 95% CI 2.1 to 13.9) time to RTT, while a rectus femoris muscle strain reportedly takes fewest number of days (mean 27.4; 95% CI 14.7 to 40.2) to RTM. RTS is subject to a broad spectrum of variables: tissue healing, tissue stress and load capacity and tolerance modifiers.11 Within this decision-making continuum evaluating tissue health is the first step. In the included studies, RTS times (ie, TRT and TRM) are highest in football athletes recovering from ACL ruptures and reconstruction. To the best of our knowledge, data concerning RTS times after ACL rupture without reconstruction in football players is lacking. The time for the ACL graft to fully heal can take several years.65,68 MRI studies found hyper-intense signalling 2 years post-surgery, indicating the increase in type III collagen.69 This long tissue healing time might explain why athletes returning to sports before 9 months post-surgery have a sevenfold increased risk for a re-rupture.70 Although time to RTS after other ligament injuries (without graft procedure) is considerably shorter, remodelling can also take up to 1 year after the inciting event.71 Average time to RTT after a lateral ankle sprain is 19.88 (95% CI 13.51 to 26.26) days, while many athletes still have residual impairments and limitations in terms of activity and functioning 2 weeks after the injury event.72 Moreover, some patients with a lateral ankle sprain have persistent mechanical laxity and perceived instability up until 1-year follow-up, highlighting the active remodelling of (connective) tissue in the ligament.73 These short RTS-time windows might explain the high recurrence rates ranging from 19% to 46% in patients recovering from ankle sprain.74 When looking at contractile soft tissue, it appears that tendon injuries heal more slowly than muscle injuries. Understanding the anatomy and physiological healing processes of tendon tissue, as opposed to muscle tissue, is critical to determining the appropriate rehabilitation modalities. Tendons serve as a vital link for the transmission of muscle contraction force to the skeletal system, enabling movement.58 Tendons have a higher collagen content (which accounts for 70% of the dry weight of the tendon) than muscles (which typically account for between 1% and 10% of the dry weight of muscle mass) and have a parallel fibre orientation to provide the tensile strength, essential for their role in the transmission of force between muscle and tendon.58 Studies have shown that tendon tissue heals differently than muscle tissue due to its lower vascularisation, which necessitates collagen synthesis and remodelling of the tissue within an extracellular matrix scaffold.75,79 Müller et al75 observed that collagen content steadily increases during tendon healing, with the repair callus reaching its maximum size at about the fifth week. Furthermore, collagen turnover in tendinopathy can take up to 3 months.80 Conversely, muscle tissue proliferation typically peaks in the second to third week of the healing process. These distinct pathoanatomical differences between muscle and tendon emphasise the importance of considering them in the rehabilitation process, as well as prognostic factors that could influence return to play.75 79 81 Second in the RTS decision-making process comes evaluating tissue stress and load capacity. Rehabilitation affects RTS of athletes recovering from sports injuries. Understanding the average time to RTS following a lower limb injury in football players can mitigate rehabilitation in terms of establishing goals and criteria.13 14 It can be argued that knowing the time to RTS might underpin a time-contingent rehabilitation approach instead of a criteria-based approach, tailored to phases of tissue healing and the injury, while a criteria-based approach aims to tailor rehabilitation to the athlete’s individual needs and the sport’s demands.82 Rehabilitation of athletes after ACLR is well developed towards a criteria-contingent approach that guides the athlete through key rehabilitation milestones with consideration of the specific demands of the athlete towards return to sports.83,85 Also, rehabilitation of athletes after a hamstring muscle injury, adductor muscle injury and Achilles tendinopathy finds the way to a criteria-based approach.13 Mendiguchia et al found that using an individualised, criteria-based approach reduced the risk of a recurrent hamstring muscle injury in athletes.60 Serner et al also found fewer re-injuries by applying a criteria-based approach when treating athletes after adductor muscle injuries.86 Griffin et al published their protocol of a clinical trial comprising an individualised, criteria-based rehabilitation programme for patients with mid-portion Achilles tendinopathy.87 Conversely to these common injuries among football players, literature regarding ankle ligament injuries lacks empirically derived criteria to inform rehabilitation82, rendering clinicians to rely on guidelines and frameworks substantiated by expert consensus.88,90 Changes are afoot in the ankle region: prospective research is ongoing to determine key clinical rehabilitation criteria for the treatment of athletes after acute ankle sprains91, while Picot et al recently published their test-battery to determine RTS in patients suffering from ankle sprain.92 93 Rehabilitation modalities can vary among practitioners and depending on the type of injuries; other types of interventions can be wielded as part of the entire spectrum of the recovery process. The rehabilitation process—a possible confounding factor—is not investigated or described in the majority of the papers included in this study; the objective of these studies is to provide information to manage expectations that can drive realistic goals and guide individual strategies for the specific injuries investigated.13 14 We included four RCTs where different types of interventions were described, and the effect of these interventions on RTT or RTM times.37 54 56 60 We included all arms of these RCTs to align with the other studies, with the purpose of illustrating RTT or RTM times in real-world situations. We excluded the effect (sizes) of interventions as the purpose of this study was to provide an overview of RTS times across all lower limb soft tissue injuries in football players, which aligns with the objectives of the individual studies included in this study. Nonetheless, the absence of clarity about content, intensity and progression could undermine interpretability of RTS times due to its effect on athletes’ recovery.13 Moreover, lack of transparency increases heterogeneity across studies, again hampering interpretability. Particularly with the shift towards criteria-based rehabilitation, where physical, functional, neurocognitive and mental readiness are in the epicentre, both clinicians and researchers require transparency to avoid misinterpretation and, hence, wrong expectations. The scarcity of transparency undermines external validity and clinical applicability; clinicians remain in the abyss regarding which rehabilitation intervention is effective, and how content ought to be structured and progressed to facilitate optimal recovery and RTS. This emphasises the necessity for standardised and detailed reporting of rehabilitation—and by extend all potential confounders. The recently developed Ankle-Go test battery and corresponding score to determine RTS in athletes after ankle sprain includes sensorimotor outcomes and two specific questionnaires to assess psychological readiness and perceived ankle function in such patients. 93 Ankle sprains occur in athletes during high-velocity, reactive phases of play with neurocognitive challenges, while performing changes of direction, cutting and landings. 94 Although the Ankle-Go test battery is a step in the right direction, it fails to expose the injured athletes to their specific tasks and demands. 95 This phenomenon is also present in RTS testing of athletes after ACLR, Achilles tendon rupture and repair and hamstring muscle injury. RTS test batteries for the assessment of athletes after ACLR comprise of muscle strength testing and single-leg hop tests where the decision is made based on limb symmetry indices of >90%. 83 84 96 97 RTS decision-making for athletes after hamstring muscle injury mostly relies on pain and subjective clinical and performance assessments,98 while RTS in athletes suffering from Achilles tendon rupture is mainly determined by testing the heel-raise function. 99 The mechanisms of the respective injuries as well as the psychological and neurocognitive demands that athletes need to be able to bear during sports performance are not replicated in these assessments. Although such test batteries, and the passing thereof, indeed show significant risk reductions of a subsequent injury,92 100 their validity raises questions.101 Re-injury rates of most common musculoskeletal injuries are high in football players (lateral ankle sprains=19%74, ACLR=17.8%102, Achilles tendon rupture=10%103, hamstring muscle injuries=22–34%104). Preparing athletes to sport-specific demands—physical, neurocognitive and mental—is paramount to prevent recurrent injuries105, as well as implementing these demands in the RTS decision-making process.11 13 RTS-criteria for many injuries included in this review are—to the best of our knowledge—non-existent: we failed to find any empirically-derived RTS-criteria for MCL and LCL injuries of the knee, rectus femoris injuries and deltoid ligament, midtarsal sprains and syndesmosis injuries of the ankle; this emphasises that RTS is still mainly time-contingent instead of using average timings as a guide to establish rehabilitation progression, goals and criteria for advancement. Therefore, we suggest that more prospective research is warranted to establish empirically driven criteria for RTS among football players recovering from lower limb tissue injuries. None of the included studies considered confounding factors regarding RTS decision modifiers. These refer to contextual factors that influence the conclusive decision-making process.11 Neglecting these factors in research might render a reductionist approach where RTS is solely underpinned by biological healing, stress and load tolerance and functional healing. Such modifiers can both prolong or shorten time to RTS, or impact re-injury risk.106 Absent transparency can—like with rehabilitation—hamper external validity, interpretability and clinical applicability of research findings. Without such contextual information driving RTS decision-making, clinicians could experience difficulties to translate research findings as a tool to mitigate recovery and rehabilitation in terms of establishing goals and criteria.107 We elaborated on time to RTS among football players after lower limb tissue injuries in TRT or RTM play, based on clinical applicability for clinicians; to render uniformity when considering RTS decision-making in football. Yet, the given definitions of RTS lack unequivocally across the included studies, with various definitions wielded: from readiness to start training to training participation, from match appearance and full match performance. The 2016 consensus statement of Ardern et al13 provides an RTS continuum, highlighting three distinct phases; the definitions of these phases were subsequently updated by Tassignon et al: return to participation, return to sports and return to performance.82 Return to sports decision-making for injured athletes comprises various factors interacting with each other.11 82 106 The complexity of this decision, and hence, the RTS-continuum—where the stages flow into each other instead of having clearly distinguished phases—might yield equivocality in terms of definitions used in clinical practice and research. Uniformity is also the common denominator concerning reported injury severity grading. The included studies reported three different types of severity grading for hamstring muscle injuries (MRI grading, BAMIC score and MMIC score). Furthermore, where hamstring muscle injuries are comprehensively—yet inconsistently—categorised, other injuries excel in the lack thereof; whether ligament injuries, muscle injuries or tendon injuries, these are mostly distinguished as ruptured or not in the included studies. Lateral ankle sprains, among ligament injuries, are commonly categorised in three grades of severity: ligament elongation (grade 1), partial ligament tear (grade 2), full ligament rupture (grade 3).71 Also, tendinopathy can be delineated into grades of severity within a continuum: reactive tendinopathy, tendon disrepair and degenerative tendinopathy.108 Reporting injury severity—and consistency thereof—is warranted to obtain clarity and unequivocality regarding rehabilitation and subsequent RTS decision-making. The 2015 Doha Agreement presents a classification of possible involved structures or causes of groin pain.109 Yet, the included studies mainly used the umbrella term of groin pain. Clinical implications The foregoing paragraphs delineated the nuances around RTS times in terms of tissue healing, tissue stress and load tolerance, RTS decision-making, contextual decision modifiers and uniformity regarding RTS-definitions, grading or categorisation. RTS, as well as injuries themselves, is a comprehensive process where various factors influence this process, each contributing with their own weight. How various confounding factors affect RTS, how the complex-systems approach plays a role in recovery, rehabilitation and the RTS decision-making process and time is still unexplored territory in contemporary research.106 The stack of new emerging research regarding data-driven approaches and algorithms to predict and prevent injuries is increasing by the minute, although still in development.110 111 Future research should also shift its attention to data-driven models, including complex-systems analysis for RTS. Another field of unexplored territory is RTS times in female football players; only 5 of the 44 studies evaluated in this systematic review included the female population. Both FIFA and UEFA are already increasingly investigating injuries in women’s football.112 With increasing injury rates, more research regarding women’s injuries and RTS is warranted. Our findings highlight substantial variability in RTS timelines across lower limb soft-tissue injuries. This underscores the importance of personalised, injury-specific and criteria-based rehabilitation and decision-making that considers all confounders affecting RTS; where RTS timelines are a tool to substantiate this approach and where the distinction between RTT and RTM further supports a phased RTS process. Clinicians should therefore prioritise individualised, criteria-based progression over fixed timelines, using objective measures aligned with the specific injury and athlete context to guide return-to-sport decisions. Additionally, clearly distinguishing between RTT and RTM enables a structured, phased approach with defined checkpoints, reducing the risk of premature return and reinjury. Limitations This systematic review with descriptive meta-analysis has its limitations. First, given the broad variation in study designs, populations, competitive levels, operationalisation of time to RTS and unclear (rehabilitation) interventions, substantial heterogeneity is likely. We were able to evaluate heterogeneity for multiple subgroups. However, explaining heterogeneity might be tricky in subgroups with low number of studies (≥10). Such likely heterogeneity complicates interpretation when describing effect sizes. This descriptive study, however, aims to provide an overview of RTT and RTM times among football players after the respective injuries without addressing true effect sizes. More research is warranted to tackle this issue for future reference. Especially studies describing such RTS times after injury among amateur and semi-professional as well as female football players are underrepresented. Although the injury rate in professional football players is 1000-fold the injury rate in non-professionals, only a select group of footballers can turn professional. More prospective research in a non-elite football population is advocated to guide strategies, goals and criteria when they need rehabilitation. Second, RTS remains an ambiguous term. It is a continuum that lacks specific thresholds. Hence, the definitions provided between the included studies varied considerably. We were able to derive which reports focused mainly on training and match play, but some remained unclear. We were also unable to derive RTS timings for both training and match play for each reported injury. Additionally, pooling RTS data under RTM or RTS groups—also without knowing all RTS rates—may have introduced misclassifications and contributed to bias in the pooled estimates, which may limit interpretability and generalisability of the findings. Third, most studies reported on hamstring muscle injuries and ACLR. Even though we found RTS timings in the other soft-tissue injuries as well, some were uniquely reported and might need verification by further research. Lastly, we aimed to review non-contact lower limb injuries in football. However, we were unable to verify whether all included studies in this review solely described non-contact injuries. Conclusions This systematic review provides an overview of weighted mean RTS times for all lower limb injuries in football, distinguished by ligament, muscle or tendon and by RTM or training. Time to RTS (both RTM and RTT) after ACL rupture and repair is longest among all lower limb musculoskeletal injuries, while a grade 0 hamstring injury takes fewest number of days to RTT and a rectus femoris muscle strain reportedly takes fewest number of days to RTM. Current RTS definitions and injury severity classifications remain inconsistent across studies, underscoring the necessity for greater uniformity to improve clinical decision-making and criteria-based rehabilitation protocols in football. Key points Return to match and training takes football players longest after ACL injury and reconstruction and Achilles tendon rupture. Return to match and training takes football players longest after ACL injury and reconstruction and Achilles tendon rupture. Football players with muscle injuries, such as hamstring or adductor strains, return to sports quickest, with recovery time increasing with injury severity. Football players with muscle injuries, such as hamstring or adductor strains, return to sports quickest, with recovery time increasing with injury severity. Other ligament injuries, like ankle or collateral knee sprains, show shorter but highly variable times to return to match and training. Other ligament injuries, like ankle or collateral knee sprains, show shorter but highly variable times to return to match and training. Supplementary material Acknowledgements We gratefully acknowledge the contributions of doctoral researcher Tom Van Deuren and Professor Dr Tim Verdonck in helping steer the direction of this paper. Furthermore, we want to acknowledge the contributions of Thibaut Smeyers, Jonas Busschots, Jens Smet and Noah Maes, students of Rehabilitation Sciences and Physiotherapy at the University of Antwerp, for their assistance in screening and extracting data. Footnotes Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors. Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Ethics approval: Not applicable. Data availability statement All data relevant to the study are included in the article or uploaded as supplementary information. References Associated Data Supplementary Materials Data Availability Statement All data relevant to the study are included in the article or uploaded as supplementary information.