Advertisement

Cartilage recovery in runners with and without knee osteoarthritis: A pilot study

Published:August 18, 2019DOI:https://doi.org/10.1016/j.knee.2019.07.011

      Abstract

      Objective

      Running is an easy way of meeting physical activity recommendations for individuals with knee osteoarthritis (KOA); however, it remains unknown how their cartilage reacts to running. The objective of this pilot study was to compare the effects of 30 min of running on T2 and T1ρ relaxation times of tibiofemoral cartilage in female runners with and without KOA.

      Methods

      Ten female runners with symptomatic KOA (mean age 52.6 ± 7.6 years) and 10 without KOA (mean age 52.5 ± 7.8 years) ran for 30 min on a treadmill. Tibiofemoral cartilage T2 and T1ρ relaxation times were measured using magnetic resonance imaging prior to and immediately after the bout of running. Repeated-measures analyses of covariance (ANCOVA) were conducted to examine between-group differences across scanning times.

      Results

      No Group × Time interactions were found for T2 (P ≥ 0.076) or T1ρ (P ≥ 0.288) relaxation times. However, runners with KOA showed increased T2 values compared with pre-running in the medial and lateral femur 55 min post-running (5.4 to 5.5%, P < 0.022) and in all four tibiofemoral compartments 90 min post-running (6.9 to 11.1%, P < 0.01). A significant group effect was found for T1ρ in the medial femur, with greater values in those with KOA compared with controls.

      Conclusion

      While Group × Time interactions in T2 and T1ρ relaxation times remained statistically insignificant, the observed significant increases in T2 in runners with tibiofemoral osteoarthritis TFOA may suggest slower and continuing changes in the cartilage and thus a need for longer recovery after running. Future research should investigate the effects of repeated exposure to running.

      Keywords

      1. Introduction

      Knee osteoarthritis (KOA) is globally considered to be one of the leading causes of disability, and its prevalence is projected to increase considerably over the next few decades [
      • Cross M.
      • Smith E.
      • Hoy D.
      • Nolte S.
      • Ackerman I.
      • Fransen M.
      • Bridgett L.
      • Williams S.
      • Guillemin F.
      • Hill C.L.
      • Laslett L.L.
      • Jones G.
      • Cicuttini F.
      • Osborne R.
      • Vos T.
      • Buchbinder R.
      • Woolf A.
      • March L.
      The global burden of hip and knee osteoarthritis: estimates from the Global Burden of Disease 2010 study.
      ]. Maintaining an active lifestyle that includes regular physical activity helps to relieve symptoms and optimize function of individuals with KOA [
      • Nelson A.E.
      • Allen K.D.
      • Golightly Y.M.
      • Goode A.P.
      • Jordan J.M.
      A systematic review of recommendations and guidelines for the management of osteoarthritis: the chronic osteoarthritis management initiative of the U.S. bone and joint initiative.
      ]. Running represents an easy and accessible way of meeting physical activity recommendations; however, it remains unknown how cartilage of those with pre-existing KOA reacts to high-impact activities such as running.
      Quantitative magnetic resonance imaging (MRI) techniques such as T2 and T1ρ relaxation mapping can be used to study cartilage properties and its immediate response to loading. T2 relaxation represents the time constant of the molecular motion of water in cartilage, which is influenced by the composition of collagen and specifically reflects changes to the extracellular matrix [
      • Mosher T.J.
      • Dardzinski B.J.
      Cartilage MRI T2 relaxation time mapping: overview and applications.
      ]. In contrast, T1ρ provides an indication of glycosaminoglycan concentration in cartilage [
      • Gold G.E.
      • Chen C.A.
      • Koo S.
      • Hargreaves B.A.
      • Bangerter N.K.
      Recent advances in MRI of articular cartilage.
      ]. Previous research has suggested that a single bout of running could measurably affect cartilage properties, as assessed using these MRI techniques [
      • Hoessly M.L.
      • Wildi L.M.
      Magnetic resonance imaging findings in the knee before and after long-distance running — documentation of irreversible structural damage? A systematic review.
      ]. For example, tibiofemoral T2 relaxation time decreased immediately following 15 to 45 min of running [
      • Subburaj K.
      • Kumar D.
      • Souza R.B.
      • Alizai H.
      • Li X.
      • Link T.M.
      • Majumdar S.
      The acute effect of running on knee articular cartilage and meniscus magnetic resonance relaxation times in young healthy adults.
      ,
      • Behzadi C.
      • Welsch G.H.
      • Laqmani A.
      • Henes F.O.
      • Kaul M.G.
      • Schoen G.
      • Adam G.
      • Regier M.
      The immediate effect of long-distance running on T2 and T2* relaxation times of articular cartilage of the knee in young healthy adults at 3T.
      ,
      • Mosher T.J.
      • Liu Y.
      • Torok C.M.
      Functional cartilage MRI T2 mapping: evaluating the effect of age and training on knee cartilage response to running.
      ,
      • Gatti A.A.
      • Noseworthy M.D.
      • Stratford P.W.
      • Brenneman E.C.
      • Totterman S.
      • Tamez-Pena J.
      • Maly M.R.
      Acute changes in knee cartilage transverse relaxation time after running and bicycling.
      ], while T1ρ relaxation time decreased after running for 30 min in young, healthy individuals. Together, these results suggest that water is expelled from cartilage secondary to high compression forces sustained during the stance phase of running, which can reach as much as 10.4 bodyweights [
      • Messier S.P.
      • Legault C.
      • Schoenlank C.R.
      • Newman J.J.
      • Martin D.F.
      • DeVita P.
      Risk factors and mechanisms of knee injury in runners.
      ].
      Nevertheless, healthy cartilage is known to fully recover even after long periods of running. For instance, despite significant decreases in tibiofemoral cartilage volume observed immediately after 20 km of running, Kessler et al. reported a full return to pre-run values one-hour post-run. Similarly, talocrural cartilage has been shown to remain intact – and even partially regenerate – under extreme ultra-endurance loading conditions [
      • Schütz U.H.W.
      • Ellerman J.
      • Schoss D.
      • Wiedelbach H.
      • Beer M.
      • Billich C.
      Biochemical cartilage alteration and unexpected signal recovery in T2* mapping observed in ankle joints with mobile MRI during a transcontinental multistage footrace over 4486 km.
      ]. However, it must be noted that all of the aforementioned studies were conducted in individuals without radiological or clinical signs of osteoarthritis. Tibiofemoral cartilage in those with pre-existing KOA may not tolerate forces applied during running as well as healthy cartilage. Despite greater baseline T2 and T1ρ relaxation times in individuals with KOA compared with healthy controls [
      • Souza R.B.
      • Kumar D.
      • Calixto N.
      • Singh J.
      • Schooler J.
      • Subburaj K.
      • Li X.
      • Link T.M.
      • Majumdar S.
      Response of knee cartilage T1rho and T2 relaxation times to in vivo mechanical loading in individuals with and without knee osteoarthritis.
      ], greater reductions in T1ρ in those with KOA under sustained simulated weightbearing conditions suggest a reduced capacity to retain water and dissipate loads. Thus, such deficits would hypothetically translate to decreased ability of osteoarthritic cartilage to normalize water content from loads applied during running. This could have important implications for sports medicine clinicians providing recommendations to runners with KOA.
      The objective of this pilot study was to compare the effects of 30 min of running on T2 and T1ρ relaxation times of tibiofemoral cartilage in female runners with and without KOA. It was hypothesized that both T2 and T1ρ would significantly decrease in both groups immediately following running, but that a greater decrease would be observed in those with KOA. Additionally, it was hypothesized that values would return to baseline values in controls but remain depressed in runners with KOA up to 90 min after running.

      2. Materials and methods

      2.1 Participants

      Ten female runners with symptomatic tibiofemoral osteoarthritis (TFOA group) and 10 healthy age-matched runners (controls) were recruited through advertisements in local sports medicine clinics, running stores, and online through social media (Table 1). To be included, all participants had to: [
      • Cross M.
      • Smith E.
      • Hoy D.
      • Nolte S.
      • Ackerman I.
      • Fransen M.
      • Bridgett L.
      • Williams S.
      • Guillemin F.
      • Hill C.L.
      • Laslett L.L.
      • Jones G.
      • Cicuttini F.
      • Osborne R.
      • Vos T.
      • Buchbinder R.
      • Woolf A.
      • March L.
      The global burden of hip and knee osteoarthritis: estimates from the Global Burden of Disease 2010 study.
      ] be aged ≥40 years; [
      • Nelson A.E.
      • Allen K.D.
      • Golightly Y.M.
      • Goode A.P.
      • Jordan J.M.
      A systematic review of recommendations and guidelines for the management of osteoarthritis: the chronic osteoarthritis management initiative of the U.S. bone and joint initiative.
      ] have run at least 10 km/week during the previous two years; and [
      • Mosher T.J.
      • Dardzinski B.J.
      Cartilage MRI T2 relaxation time mapping: overview and applications.
      ] be comfortable running on a treadmill for 30 min. Runners were included in the TFOA group if they showed radiographic signs of tibiofemoral osteoarthritis characterized by Grade ≥ 2 on the Kellgren-Lawrence scale as shown by weightbearing X-rays [
      • Kellgren J.H.
      • Lawrence J.S.
      Radiological assessment of osteo-arthrosis.
      ]. Additionally, they had to report regular knee pain of ≥3/10 on a numerical pain rating scale (0 = no pain; 10 = worst pain imaginable) during running, as well as during other activities of daily living such as going up or down stairs, kneeling or squatting. Runners were included in the control group if they showed no radiographic signs of tibiofemoral osteoarthritis (Grade 0 on the Kellgren–Lawrence scale) and were free of knee pain for the previous six months [
      • Kellgren J.H.
      • Lawrence J.S.
      Radiological assessment of osteo-arthrosis.
      ].
      Table 1Study participants.
      TFOA (n = 10)Controls (n = 10)P
      Demographics
      Age (years)52.6 ± 7.652.5 ± 7.81.00
      Body mass index (kg/m2)23.0 ± 3.423.5 ± 2.50.684
      Kellgren-Lawrence gradeGrade 2 (n = 8)

      Grade 3 (n = 2)
      Grade 0 (n = 10)
      Symptom duration (months)78.1 ± 118.2N/A
      KOOS score (0–100)69.6 ± 14.495.4 ± 4.5<0.001
      Running experience (years)11.1 ± 9.013.1 ± 8.70.481
      Average running distance (km/week)20.2 ± 9.925.3 ± 13.60.481
      30-minute run
      Pain level during the run (0–10)2.0 ± 2.30.2 ± 0.40.030
      Treadmill speed (km/h)7.5 ± 0.97.8 ± 1.20.353
      Step rate (steps/min)166.9 ± 6.7158.6 ± 8.80.015
      Running shoes Minimalist index (%)21.2 ± 9.319.8 ± 5.80.631
      Foot strike pattern (%)0.587
       Rearfoot8070
       Midfoot2020
       Forefoot010
      Data presented as Mean ± SD
      TFOA, tibiofemoral osteoarthritis group; KOOS, Knee Osteoarthritis Outcome Score
      Potential participants were excluded from both groups if they reported: any history of traumatic knee injury or knee surgery, presence of a neurological or inflammatory arthritic condition, current lower limb pain during running (other than knee pain in the TFOA group), use of any oral or injected corticosteroids or viscosupplementation during the preceding six months, and presence of any contraindication for MRI (e.g. metal implants). This study was approved by the Institutional Clinical Research Ethics Board (H16-02059) and all participants signed a detailed informed written consent form.

      2.2 Study design

      The study consisted of a single session of data collection, during which quantitative knee MRI measurements were taken before and after 30 min of treadmill running (Figure 1).
      Figure 1
      Figure 1Timeline for the data collection session. Post-run times represent the time elapsed between the end of the run and the different sequences.
      Participants were instructed to avoid any running during the 24 h prior to data collection. Upon arrival, they were asked to rest in a seated position for 15 min while they completed questionnaires about demographics, running habits (experience, weekly running distance) and knee symptoms (Knee Osteoarthritis Outcome Score; KOOS) [
      • Roos E.M.
      • Lohmander L.S.
      • Roos E.M.
      • Lohmander L.S.
      The Knee injury and Osteoarthritis Outcome Score (KOOS): from joint injury to osteoarthritis.
      ]. In total, participants unloaded their knee cartilage for approximately 30 min (filling-in questionnaires, preparation for MRI, survey scans) before collection of MRI data started, as per previous studies [
      • Subburaj K.
      • Kumar D.
      • Souza R.B.
      • Alizai H.
      • Li X.
      • Link T.M.
      • Majumdar S.
      The acute effect of running on knee articular cartilage and meniscus magnetic resonance relaxation times in young healthy adults.
      ,
      • Gatti A.A.
      • Noseworthy M.D.
      • Stratford P.W.
      • Brenneman E.C.
      • Totterman S.
      • Tamez-Pena J.
      • Maly M.R.
      Acute changes in knee cartilage transverse relaxation time after running and bicycling.
      ].
      Thereafter, participants were scanned in a supine position with knees in 20° flexion. In the TFOA group, the most symptomatic knee was considered as the study knee if both knees were suitable for inclusion. A randomly chosen knee was selected for the control group.
      Following baseline imaging, participants were transported to the laboratory by wheelchair so that knee cartilage loading was isolated to the bout of running. They were provided with one minute of walking to get accustomed to the treadmill before starting the 30-min run. In order to replicate habitual individual training conditions, participants self-selected their running speed and wore their usual running shoes. Step rate [
      • Willy R.W.
      • Meardon S.A.
      • Schmidt A.
      • Blaylock N.R.
      • Hadding S.A.
      • Willson J.D.
      Changes in tibiofemoral contact forces during running in response to in-field gait retraining.
      ], foot strike pattern [
      • Esculier J.F.
      • Silvini T.
      • Bouyer L.J.
      • Roy J.S.
      Video-based assessment of foot strike pattern and step rate is valid and reliable in runners with patellofemoral pain.
      ], and the level of minimalism of running shoes (rated using the Minimalist Index) [
      • Esculier J.F.
      • Dubois B.
      • Dionne C.E.
      • Leblond J.
      • Roy J.S.
      A consensus definition and rating scale for minimalist shoes.
      ] were documented to control for potential confounders between groups. Participants also rated their average level of knee pain during the 30-min run on a numerical pain rating scale (0 = no pain; 10 = worst pain imaginable).
      Immediately after completing 30 min of running, participants sat on the wheelchair and were transported back to the MRI department and prepared for the post-run assessment, which consisted of the same scanning sequences as the pre-run MRI. Three additional T2 scanning sequences were conducted (T2postA, T2postB, T2postC). Two additional T1ρ scanning sequences were run (T1ρpostA, T1ρpostB) in between T2 sequences (Figure 1). Participants remained supine in the MRI apparatus during the entire post-run imaging acquisition period.

      2.3 Magnetic resonance data acquisition

      All data were collected on a 3T MRI scanner (Philips Achieva) equipped with an eight-channel SENSE knee coil. T2 relaxation time was measured using a multi-slice, multi-echo, spin echo scan with seven echoes (TE1 = 13 ms, echo spacing = 13 ms, TR = 4477 ms, acquired voxel size = 0.4 × 0.4 × 3 mm3, reconstructed voxel size = 0.31 × 0.31 × 3 mm3). Sagittal images were acquired using SENSE in two directions (anterior–posterior factor 2, foot-head factor 1.3). T1ρ mapping was performed using a three-dimensional (3D) spoiled gradient echo sequence with sagittal readout and five spin lock durations (TSL = 1/10/20/30/40 ms). The spin lock frequency was 500 Hz, as determined by the duration and angle of the spin lock pulses (TE 1.37 ms, TR 4.3 ms, acquired voxel size 0.6 × 0.8 × 3.0 mm3, reconstructed voxel size 0.48 × 0.48 × 1.5 mm3). A fat suppression (SPIR) and inversion pre-pulse (delay 1700 ms) were applied to avoid signal shifts due to the presence of fat and free fluid. A total of 26 slices was acquired for T2 and 66 slices for T1ρ. A 3D sagittal water selective fluid scan (WATSf) was acquired for segmentation purposes. This scan employed a fat suppression pulse (ProSet 1331) and was acquired at 0.6 × 0.6 × 1.5 mm3, reconstructed to 0.29 × 0.29 × 1.5 mm3 with TE = 4.6 ms, TR = 20 ms. Parallel imaging was used in the anterior-posterior direction, with a SENSE factor of 2.

      2.4 Data processing

      After data collection, image files were anonymized and assigned a random number. Data analysis for T2 and T1ρ relaxation times was performed by a member of the research team (MJ, physicist experienced in MRI analyses) who was blinded to group and scanning time. Four regions of interest (ROI) were considered in the analyses: lateral femur, lateral tibia, medial femur and medial tibia. The segmentation of these ROIs was manually performed slice-by-slice on the WATSf scans in the sagittal plane. On each slice, the weightbearing cartilage was bounded in the posterior–anterior plane by the margins of the meniscus. In the medial-lateral plane, ROIs were bounded by the point of separation of the tibial and femoral cartilage, and the intercondylar fossa. A threshold value of 200 ms was used to exclude voxels containing tissues other than cartilage.
      T2 relaxation maps were obtained directly from the scanner and based on voxel-wise, mono-exponential fitting of the decay curves. Similarly, T1ρ maps that were computed offline employed a mono-exponential, voxel-wise fitting procedure. Image processing was performed in Python using the Nipy suite of analysis tools along with custom code [
      • Gorgolewski K.
      • Burns C.D.
      • Madison C.
      • Clark D.
      • Halchenko Y.O.
      • Waskom M.L.
      • SS G.
      Nipype: a flexible, lightweight and extensible neuroimaging data processing framework in Python.
      ].
      T2 maps were aligned to the WATSf image space using SPM 12's coreg function with default parameters in Matlab (The MathWorks Inc., Natick, MA) [
      • Penny W.
      • Friston K.
      • Ashburner J.
      • Kiebel S.
      • Nichols T.
      Statistical Parametric Mapping: The Analysis of Functional Brain Images.
      ]. T1ρ maps were aligned using ANTs, a non-linear transformation tool [
      • Avants B.B.
      • Tustison N.J.
      • Song G.
      • Cook P.A.
      • Klein A.
      • Gee J.C.
      A reproducible evaluation of ANTs similarity metric performance in brain image registration.
      ], also with default parameters. Registrations were visually checked for accuracy and manual corrections to the cartilage segmentations were made as necessary.

      2.5 Outcome measures

      The primary outcomes of this study were tibiofemoral cartilage T2 and T1ρ relaxation times, which were averaged over the full thickness. T2 relaxation is a recommended imaging outcome for clinical trials in individuals with KOA, as determined by the Osteoarthritis Research Society International (OARSI) [
      • Hunter D.J.
      • Altman R.D.
      • Cicuttini F.
      • et al.
      OARSI clinical trials recommendations: knee imaging in clinical trials in osteoarthritis.
      ]. Such measurements have been shown to be valid and reliable assessments of tibiofemoral cartilage integrity [
      • Li X.
      • Wyatt C.
      • Rivoire J.
      • Han E.
      • Chen W.
      • Schooler J.
      • Liang F.
      • Shet K.
      • Souza R.
      • Majumdar S.
      Simultaneous acquisition of T1rho and T2 quantification in knee cartilage: repeatability and diurnal variation.
      ,
      • Li X.
      • Wyatt C.
      • Rivoire J.
      • et al.
      Simultaneous acquisition of T1rho and T2 quantification in knee cartilage: repeatability and diurnal variation.
      ]. Shorter T2 relaxation times indicate less water in cartilage [
      • Souza R.B.
      • Kumar D.
      • Calixto N.
      • Singh J.
      • Schooler J.
      • Subburaj K.
      • Li X.
      • Link T.M.
      • Majumdar S.
      Response of knee cartilage T1rho and T2 relaxation times to in vivo mechanical loading in individuals with and without knee osteoarthritis.
      ]. Though less established than T2, T1ρ relaxation time is also advocated by OARSI guidelines [
      • Hunter D.J.
      • Altman R.D.
      • Cicuttini F.
      • et al.
      OARSI clinical trials recommendations: knee imaging in clinical trials in osteoarthritis.
      ], and has shown low inter-session coefficients of variation (5.3%) with no diurnal variation between morning and afternoon scans [
      • Li X.
      • Wyatt C.
      • Rivoire J.
      • et al.
      Simultaneous acquisition of T1rho and T2 quantification in knee cartilage: repeatability and diurnal variation.
      ]. Shorter T1ρ relaxation times after activity-related compression correspond to areas of increased glycosaminoglycan and collagen concentration [
      • Gold G.E.
      • Chen C.A.
      • Koo S.
      • Hargreaves B.A.
      • Bangerter N.K.
      Recent advances in MRI of articular cartilage.
      ,
      • Menezes N.M.
      • Gray M.L.
      • Hartke J.R.
      • Burstein D.
      T2 and T1ρ MRI in articular cartilage systems.
      ].

      2.6 Statistical analyses

      Statistical analyses were conducted by a member of the research team (JFE) who was not involved in data processing. Mann–Whitney U (continuous variables) and Chi-square (foot strike pattern) tests were used to compare both groups' demographics and running habits. Time separating the end of the run and T2 and T1ρ sequences were compared between groups using independent sample t-tests. The effects of running on tibiofemoral cartilage were assessed using repeated-measures ANCOVA (General Linear Model). Given the known influence of speed on knee joint forces during running [
      • Orendurff M.S.
      • Kobayashi T.
      • Tulchin-Francis K.
      • Herring Tullock A.M.
      • Villarosa C.
      • Chan C.
      • Strike S.
      A little bit faster: Lower extremity joint kinematics and kinetics as recreational runners achieve faster speeds.
      ], treadmill speed was included as a covariate in T2 and T1ρ relaxation time analyses. Step rate was also added as a covariate, since it was significantly different between TFOA and controls, and because it has been shown to influence tibiofemoral contact forces [
      • Willy R.W.
      • Meardon S.A.
      • Schmidt A.
      • Blaylock N.R.
      • Hadding S.A.
      • Willson J.D.
      Changes in tibiofemoral contact forces during running in response to in-field gait retraining.
      ]. Alpha level was set at 0.05, and Bonferroni corrections were applied to adjust for multiple comparisons. All statistical analyses were conducted using the Statistical Package for Social Sciences (SPSS) version 22 (IBM Corporation, Armonk, NY, USA).

      3. Results

      All participants completed the study without adverse effects. The TFOA group reported a significantly greater level of knee pain than the control group during the bout of running (P = 0.03; Table 1). Time elapsed between the end of the run and the post-run MRI data acquisition was not significantly different between groups (P ≥ 0.08; Table 2).
      Table 2Time (in minutes) elapsed between the end of the run and MRI sequences.
      TFOA

      (n = 10)
      Controls

      (n = 10)
      T2postA20.9 ± 6.316.9 ± 3.6
      T1ρpostA33.6 ± 6.329.3 ± 3.7
      T2postB56.4 ± 6.652.8 ± 3.7
      T1ρpostB68.9 ± 6.564.7 ± 3.6
      T2postC92.6 ± 6.988.6 ± 3.7
      Data presented as Mean ± SD.
      MRI, magnetic resonance imaging; TFOA, tibiofemoral osteoarthritis group
      A visual representation of pre-run and post-run T2 relaxation time in a runner from the TFOA group is presented in Figure 2. No main effects for group or time were found for T2 relaxation time in any ROI (Table 3 and Figure 3). Non-statistically significant trends were found for group × time interaction effects (P = 0.076–0.089; Table 3). However, TFOA showed significantly higher T2 relaxation times at T2postB, than at T2pre for lateral (+5.4%, P = 0.003) and medial femur (+5.5%, P = 0.022). At T2postC, TFOA values were higher than at baseline for all four ROIs (lateral femur, +7.0%, P = 0.003; lateral tibia, +9.1%, P = 0.01; medial femur, +6.9%, P = 0.005; medial tibia, +11.1%, P = 0.01; Table 3). In the control group, none of the post-run T2 values were significantly different from pre-run (P ≥ 0.06).
      Figure 2
      Figure 2Example of T2 relaxation time (A) T2pre; (B) T2postA; (C) T2postB; (D) T2postC
      Table 3T2 relaxation values at the different time points, and mean difference with pre-running values.
      T2 relaxation time values (ms)Mean difference with T2pre (ms)
      T2preT2postAT2postBT2postCT2postAT2postBT2postC
      Lateral femur
      TFOA50.4

      [47.6, 53.2]
      50.3

      [46.6, 53.9]
      53.1

      [49.7, 56.6]
      53.9

      [50.5, 57.4]
      0.2

      [−2.5, 2.1]
      2.8

      [0.6, 4.8]
      3.5

      [1.1, 6.0]
      Group: 0.402

      Time: 0.120

      Interaction: 0.076
      Controls49.6

      [46.8, 52.4]
      48.7

      [45.0, 52.3]
      49.6

      [46.1, 53.0]
      52.0

      [48.5, 55.4]
      0.9

      [−3.2, 1.4]
      0.0

      [−2.1, 2.1]
      2.4

      [−0.1, 4.8]
      Lateral tibia
      TFOA31.6

      [28.5, 34.7]
      31.3

      [28.6, 33.9]
      33.6

      [30.6, 36.6]
      34.5

      [31.7, 37.3]
      0.4

      [−2.4, 1.7]
      2.0

      [−0.4, 4.4]
      2.9

      [0.6, 5.2]
      Group: 0.515

      Time: 0.190

      Interaction: 0.088
      Controls33.8

      [30.7, 36.9]
      33.6

      [30.9, 36.3]
      33.7

      [30.7, 36.7]
      35.2

      [32.4, 38.0]
      0.2

      [−2.2, 1.9]
      0.1

      [−2.5, 2.3]
      1.4

      [−0.8, 3.7]
      Medial femur
      TFOA51.1

      [48.1, 54.1]
      51.3

      [48.2, 54.3]
      53.8

      [51.2, 56.6]
      54.6

      [51.7, 57.5]
      0.2

      [−2.4, 2.7]
      2.8

      [0.3, 5.3]
      3.5

      [0.9, 6.1]
      Group: 0.051

      Time: 0.345

      Interaction: 0.078
      Controls48.5

      [45.5, 51.5]
      47.1

      [44.0, 50.1]
      48.6

      [45.9, 51.2]
      50.0

      [47.2, 52.9]
      1.5

      [−4.0, 1.1]
      0.0

      [−2.4, 2.5]
      1.5

      [−1.1, 4.1]
      Medial tibia
      TFOA34.8

      [31.5, 38.1]
      34.6

      [30.9, 38.3]
      38.0

      [34.3, 41.7]
      38.6

      [34.9, 42.3]
      0.2

      [−3.6, 3.2]
      3.2

      [−1.2, 7.5]
      3.8

      [0.8, 6.9]
      Group: 0.226

      Time: 0.638

      Interaction: 0.089
      Controls34.1

      [30.8, 37.4]
      31.5

      [27.8, 35.2]
      32.6

      [28.9, 36.3]
      35.8

      [32.1, 39.5]
      2.5

      [−5.9, 0.8]
      1.4

      [−5.8, 2.9]
      1.7

      [−1.3, 4.8]
      Results presented as mean [95% CI] after controlling for treadmill speed and step rate.
      TFOA, tibiofemoral osteoarthritis group; T2pre, T2 relaxation time before the run; T2postA, T2 relaxation time at first post-run scan; T2postB, T2 relaxation time at second post-run scan; T2postC, T2 relaxation time at third post-run scan.
      Bold values indicate significant within-group changes based on 95% CI limits.
      Figure 3
      Figure 3Mean T2 relaxation time (ms) before and after running for (A) lateral femur; (B) medial femur; (C) lateral tibia; (D) medial tibia. Error bars represent SD. TFOA, tibiofemoral osteoarthritis group; T2pre, T2 relaxation time before the run; T2postA, T2 relaxation time at first post-run scan; T2postB, T2 relaxation time at second post-run scan; T2postC, T2 relaxation time at third post-run scan.
      As for T1ρ, a significant group effect was found in the medial femur (P = 0.016), indicating greater values in TFOA than in controls (Table 4 and Figure 4). A trend for a group effect was also found for the lateral femur (P = 0.099). T1ρ after running was significantly different from baseline in controls: T1ρpostA was 4.7% lower in the lateral femur (P = 0.015) and 4.5% lower in the medial femur (P = 0.029). No main effect of time or group × time interaction effects were found for T1ρ in any ROI (Table 4).
      Table 4T1ρ relaxation values at the different time points, and mean difference with pre-running values.
      T1ρ relaxation time values (ms)Mean difference with T1ρpre (ms)
      T1ρpreT1ρpostAT1ρpostBT1ρpostAT1ρpostB
      Lateral femur
      TFOA52.5 [49.7, 55.4]51.4 [49.1, 53.7]52.5 [49.9, 55.1]1.1 [−3.1, 0.8]0.0 [−2.1, 2.1]Group: 0.099

      Time: 0.192

      Interaction: 0.288
      Controls50.4 [47.6, 53.3]48.0 [45.7, 50.3]48.8 [46.2, 51.4]2.4 [−4.3, −0.4]1.6 [−3.7, 0.5]
      Lateral tibia
      TFOA33.8 [31.0, 36.6]32.8 [30.7, 34.9]33.3 [31.1, 35.6]1.0 [−2.9, 0.9]0.5 [−2.8, 1.9]Group: 0.326

      Time: 0.234

      Interaction: 0.693
      Controls35.8 [33.0, 38.7]34.3 [32.3, 36.4]34.6 [32.4, 36.9]1.5 [−3.4, 0.4]1.2 [−3.6, 1.1]
      Medial femur
      TFOA53.4 [50.6, 56.2]51.7 [49.1, 54.3]53.6 [51.4, 55.8]1.7 [−3.7, 0.3]0.2 [−1.9, 2.2]Group: 0.016

      Time: 0.675

      Interaction: 0.882
      Controls49.1 [46.4, 51.9]46.9 [44.4, 49.5]49.0 [46.8, 51.2]2.2 [−4.2, −0.2]0.1 [−2.2, 1.9]
      Medial tibia
      TFOA30.9 [28.0, 33.9]31.0 [28.4, 33.6]30.7 [28.2, 33.3]0.0 [−3.5, 3.5]0.2 [−4.1, 3.7]Group: 0.681

      Time: 0.901

      Interaction: 0.863
      Controls32.1 [29.1, 35.1]31.7 [29.1, 34.3]30.9 [28.3, 33.4]0.4 [−3.9, 3.1]1.2 [−5.1, 2.7]
      Results presented as mean [95% CI] after controlling for treadmill speed and step rate.
      TFOA, tibiofemoral osteoarthritis group; T1ρpre, T1ρ relaxation time before the run; T1ρpostA, T1ρ relaxation time at first post-run scan; T1ρpostB, T1ρ relaxation time at second post-run scan.
      Bold values indicate significant within-group changes based on 95% CI limits.
      Figure 4
      Figure 4Mean T1ρ relaxation time (ms) before and after running for (A) lateral femur; (B) medial femur; (C) lateral tibia; (D) medial tibia. Error bars represent SD. TFOA, tibiofemoral osteoarthritis group; T1ρpre, T1ρ relaxation time before the run; T1ρpostA, T1ρ relaxation time at first post-run scan; T1ρpostB, T1ρ relaxation time at second post-run scan.

      4. Discussion

      This is the first study to compare the effects of running on T2 and T1ρ relaxation times in the tibiofemoral cartilage of runners with and without KOA. The relatively long post-run scanning time used in this pilot study is novel and allowed detection of potential differences in cartilage recovery in runners with KOA, which will help inform future research. Based on previous studies in healthy runners, it was hypothesized that the TFOA group would experience a greater immediate decrease along with delayed normalization of relaxation times after running.
      Unlike previous studies, no statistically significant changes were observed in T2 relaxation time in either group immediately following 30 min of running (i.e. T2postA). Behzadi et al. reported reductions of 12.3–13.2% immediately after a 45-minute run [
      • Behzadi C.
      • Welsch G.H.
      • Laqmani A.
      • Henes F.O.
      • Kaul M.G.
      • Schoen G.
      • Adam G.
      • Regier M.
      The immediate effect of long-distance running on T2 and T2* relaxation times of articular cartilage of the knee in young healthy adults at 3T.
      ], while Subburaj et al. noted decreased T2 relaxation times after 30 min of running, albeit only in the medial tibiofemoral compartment (5.4%) [
      • Subburaj K.
      • Kumar D.
      • Souza R.B.
      • Alizai H.
      • Li X.
      • Link T.M.
      • Majumdar S.
      The acute effect of running on knee articular cartilage and meniscus magnetic resonance relaxation times in young healthy adults.
      ]. Gatti et al. also found that 15 min of running was sufficient to reduce tibial T2 values by 6.1% [
      • Gatti A.A.
      • Noseworthy M.D.
      • Stratford P.W.
      • Brenneman E.C.
      • Totterman S.
      • Tamez-Pena J.
      • Maly M.R.
      Acute changes in knee cartilage transverse relaxation time after running and bicycling.
      ]. Interestingly, these previous studies were all conducted in healthy young individuals (mean age < 30 years). While age has not been found to significantly influence T2 response to running in previous studies [
      • Mosher T.J.
      • Liu Y.
      • Torok C.M.
      Functional cartilage MRI T2 mapping: evaluating the effect of age and training on knee cartilage response to running.
      ,
      • Cha J.G.
      • Lee J.C.
      • Kim H.J.
      • Han J.K.
      • Lee E.H.
      • Kim Y.D.
      • Jeon C.H.
      Comparison of MRI T2 relaxation changes of knee articular cartilage before and after running between young and old amateur runners.
      ], perhaps a combination of age and running experience could have influenced the current findings. Indeed, the cohort comprised experienced recreational runners (average 12.1 years of experience, running 22.7 km/week). It has been previously suggested that greater baseline levels of physical activity could lead to smaller changes in tibiofemoral T2 values after loading [
      • Subburaj K.
      • Kumar D.
      • Souza R.B.
      • Alizai H.
      • Li X.
      • Link T.M.
      • Majumdar S.
      The acute effect of running on knee articular cartilage and meniscus magnetic resonance relaxation times in young healthy adults.
      ,
      • Gatti A.A.
      • Noseworthy M.D.
      • Stratford P.W.
      • Brenneman E.C.
      • Totterman S.
      • Tamez-Pena J.
      • Maly M.R.
      Acute changes in knee cartilage transverse relaxation time after running and bicycling.
      ]. Cartilage adaptation to loading could potentially lead to less T2 variations after a bout of running, and perhaps a longer run would have been necessary to observe reductions in T2 values in this cohort of experienced runners.
      Another potential explanation for the absence of immediate effects of running on T2 is the time elapsed between the end of the run and the post-run scans. While every effort was made to commence MRI as soon as possible, T2postA started nearly 20 min after the run. In comparison, studies that reported significant decreases in T2 commenced data collection <4 min after running [
      • Behzadi C.
      • Welsch G.H.
      • Laqmani A.
      • Henes F.O.
      • Kaul M.G.
      • Schoen G.
      • Adam G.
      • Regier M.
      The immediate effect of long-distance running on T2 and T2* relaxation times of articular cartilage of the knee in young healthy adults at 3T.
      ,
      • Gatti A.A.
      • Noseworthy M.D.
      • Stratford P.W.
      • Brenneman E.C.
      • Totterman S.
      • Tamez-Pena J.
      • Maly M.R.
      Acute changes in knee cartilage transverse relaxation time after running and bicycling.
      ]. The longer period in the current study could have allowed water to migrate back into the cartilage before measurements took place. Although decreased T2 values were still observed in young active males 30 min after running for 30 min [
      • Karanfil Y.
      • Babayeva N.
      • Dönmez G.
      • et al.
      Thirty minutes of running exercise decreases T2 signal intensity but not thickness of the knee joint cartilage: A 3.0-T magnetic resonance imaging study.
      ], cartilage of young healthy runners recovered in as little as 60 min even after 20 km of running [
      • Kessler M.A.
      • Glaser C.
      • Tittel S.
      • Reiser M.
      • Imhoff A.B.
      Recovery of the menisci and articular cartilage of runners after cessation of exercise: additional aspects of in vivo investigation based on 3-dimensional MRI.
      ]. Therefore, it cannot be excluded that this prevented detection of true immediate changes immediately after running.
      In the current study, decreased T1ρ relaxation time was observed solely in the controls' femoral cartilage immediately following the run, thereby only partially confirming the first hypothesis. Using similar 30-minute treadmill running protocols, previous studies have reported significant T1ρ reductions in both the medial and lateral tibiofemoral compartments [
      • Subburaj K.
      • Kumar D.
      • Souza R.B.
      • Alizai H.
      • Li X.
      • Link T.M.
      • Majumdar S.
      The acute effect of running on knee articular cartilage and meniscus magnetic resonance relaxation times in young healthy adults.
      ,
      • Chen M.
      • Qiu L.
      • Shen S.
      • Wang F.
      • Zhang J.
      • Zhang C.
      • Liu S.
      The influences of walking, running and atair activity on knee articular cartilage: Quantitative MRI using T1 rho and T2 mapping.
      ]. The absence of changes in TFOA and in the controls' tibias could be due to the lower running speed in the current study (average 7.7 km/h) compared to previous studies (average 10.4–10.7 km/h). However, altered cartilage behaviour as a consequence of KOA may provide a better explanation for the current findings. Between-group differences in femoral T1ρ values (significant group effect for medial femur, trend for lateral femur) could be indicative of alterations in proteoglycan structure or content in TFOA [
      • Gold G.E.
      • Chen C.A.
      • Koo S.
      • Hargreaves B.A.
      • Bangerter N.K.
      Recent advances in MRI of articular cartilage.
      ]. Since T1ρ is also thought to be influenced by both collagen composition and cartilage hydration [
      • Li X.
      • Cheng J.
      • Lin K.
      • Saadat E.
      • Bolbos R.I.
      • Ries M.D.
      • Horvai A.
      • Link T.M.
      • Majumdar S.
      Quantitative MRI using T1ρ and T2 in human osteoarthritic cartilage specimens: Correlation with biochemical measurements and histology.
      ], the absence of changes in TFOA after loading may reflect reduced capacity for fluid movements, which are necessary for maintaining homeostasis [
      • Houard X.
      • Goldring M.B.
      • Berenbaum F.
      Homeostatic mechanisms in articular cartilage and role of inflammation in osteoarthritis.
      ].
      The current findings also did not confirm the second hypothesis, which stated that values in the TFOA group would remain lower after running in comparison with controls. Instead, statistical analyses indicated trending group × time interaction effects for increased post-run T2 values in all four ROIs only in TFOA. Interestingly, increased T2postB (femur, tibia) and T2postC values compared with baseline values are indicative of greater water content in TFOA [
      • Gold G.E.
      • Chen C.A.
      • Koo S.
      • Hargreaves B.A.
      • Bangerter N.K.
      Recent advances in MRI of articular cartilage.
      ]. Given that participants from the TFOA group were all symptomatic during habitual running, it is possible that the treadmill run triggered an acute overload response in the knee joint, thereby increasing cartilage fluid. However, the amount of time needed for T2 values to recover to baseline is unknown. Impaired ability of the knee joint to recover from running has previously been reported in individuals with post-anterior cruciate ligament repair. In addition to greater pre-run T2 values in the medial femur compared with controls, which may be indicative of increased cartilage fluid, Van Ginckel et al. showed that cartilage volume had still not recovered fully 45 min after running [
      • Van Ginckel A.
      • Verdonk P.
      • Victor J.
      • Witvrouw E.
      Cartilage status in relation to return to sports after anterior cruciate ligament reconstruction.
      ]. A trend towards increased bone marrow edema as much as 48 h following a marathon has also been reported in eight runners with a previous anterior cruciate ligament reconstruction, in comparison with their healthy knee [
      • Leiter J.R.
      • MacDonald L.
      • McRae S.
      • Davidson M.
      • MacDonald P.B.
      Intrinsic stresses on bone and cartilage in the normal and anterior cruciate ligament-reconstructed knee before and after a half marathon: a magnetic resonance imaging analysis.
      ]. Therefore, future studies in runners with KOA should consider a sequential series of scans to better understand the time-course of cartilage recovery. This could provide more information on the readiness of cartilage to sustain subsequent impact, which would help provide specific training frequency and volume recommendations to those with symptomatic KOA. Even if impact exercise or maintaining a running program has not been found to be deleterious for KOA progression [
      • Bricca A.
      • Juhl C.B.
      • Steultjens M.
      • Wirth W.
      • Roos E.M.
      Impact of exercise on articular cartilage in people at risk of, or with established, knee osteoarthritis: a systematic review of randomised controlled trials.
      ,
      • Lo G.H.
      • Musa S.M.
      • Driban J.B.
      • Kriska A.M.
      • McAlindon T.E.
      • Souza R.B.
      • Petersen N.J.
      • Storti K.L.
      • Eaton C.B.
      • Hochberg M.C.
      • Jackson R.D.
      • Kwoh C.K.
      • Nevitt M.C.
      • Suarez-Almazor M.E.
      Running does not increase symptoms or structural progression in people with knee osteoarthritis: Data from the Osteoarthritis Initiative.
      ,
      • Chakravarty E.F.
      • Hubert H.B.
      • Lingala V.B.
      • Zatarain E.
      • Fries J.F.
      Long distance running and knee osteoarthritis. A prospective study.
      ], little is known about optimal running parameters in that population.
      This study has limitations. First, the relatively small sample size of this pilot study may not have been sufficient to detect between-group differences in cartilage recovery. Second, only female runners were included. Thus, results may not be generalizable to male runners with similar characteristics. Recruiting only one sex allowed the influence of potential systemic and hormonal confounders to be minimized, and is an approach common in previous studies investigating the effects of running on knee cartilage [
      • Behzadi C.
      • Welsch G.H.
      • Laqmani A.
      • Henes F.O.
      • Kaul M.G.
      • Schoen G.
      • Adam G.
      • Regier M.
      The immediate effect of long-distance running on T2 and T2* relaxation times of articular cartilage of the knee in young healthy adults at 3T.
      ,
      • Gatti A.A.
      • Noseworthy M.D.
      • Stratford P.W.
      • Brenneman E.C.
      • Totterman S.
      • Tamez-Pena J.
      • Maly M.R.
      Acute changes in knee cartilage transverse relaxation time after running and bicycling.
      ,
      • Cha J.G.
      • Lee J.C.
      • Kim H.J.
      • Han J.K.
      • Lee E.H.
      • Kim Y.D.
      • Jeon C.H.
      Comparison of MRI T2 relaxation changes of knee articular cartilage before and after running between young and old amateur runners.
      ,
      • Karanfil Y.
      • Babayeva N.
      • Dönmez G.
      • et al.
      Thirty minutes of running exercise decreases T2 signal intensity but not thickness of the knee joint cartilage: A 3.0-T magnetic resonance imaging study.
      ,
      • Kessler M.A.
      • Glaser C.
      • Tittel S.
      • Reiser M.
      • Imhoff A.B.
      Recovery of the menisci and articular cartilage of runners after cessation of exercise: additional aspects of in vivo investigation based on 3-dimensional MRI.
      ,
      • Van Ginckel A.
      • Baelde N.
      • Almqvist K.F.
      • Roosen P.
      • McNair P.
      • Witvrouw E.
      Functional adaptation of knee cartilage in asymptomatic female novice runners.
      ]. Third, analyzing mean values for the different ROIs irrespective of cartilage depth may have limited interpretation of results. It is possible that superficial and deep layers recovered differently after running [
      • Mosher T.J.
      • Liu Y.
      • Torok C.M.
      Functional cartilage MRI T2 mapping: evaluating the effect of age and training on knee cartilage response to running.
      ]; however, the resolution of the current images did not allow depth-specific analyses. Fourth, the interpretation of results was limited by the absence of detailed data on running biomechanics. It is believed that only Kersting et al. paired biomechanical data with cartilage analysis in healthy runners [
      • Kersting U.G.
      • Stubendorff J.J.
      • Schmidt M.C.
      • Bruggemann G.P.
      Changes in knee cartilage volume and serum COMP concentration after running exercise.
      ]. Given that they reported significant associations between joint contact force and reductions in cartilage volume following a bout of running, future studies should consider including a detailed biomechanical analysis to supplement cartilage behaviour data in those with KOA. In addition to clarifying how running mechanics affect cartilage, this would provide a mechanistic basis for potential biomechanical interventions in this population.

      5. Conclusion

      Despite the absence of statistically significant differences in cartilage recovery following 30 min of running between female runners with and without symptomatic KOA, a delayed increase in T2 relaxation times in those with KOA suggests that cartilage may need more time to recover in that population. Further research should investigate the effects of repeated exposure and running mechanics to aid in formulating evidence-based clinical recommendations to older runners with KOA.

      Acknowledgements

      Funding for this study was obtained from the Physiotherapy Foundation of Canada, the Canadian Academy of Sports and Exercise Medicine, and the Natural Sciences and Engineering Research Council of Canada. JFE is supported by a Research Fellowship from the Canadian Institutes of Health Research. MAH is supported by a Michael Smith Foundation for Health Research Scholar Award and a New Investigator Award from the Canadian Institutes of Health Research. The authors would like to acknowledge all runners who took part in this study, the UBC MRI Research Centre and its technologists, Guillaume Gilbert from Philips and Jane Desrochers, PhD for providing feedback on the manuscript. AR is supported by Canada Research Chairs Program.

      Declaration of Competing Interest

      The authors report no conflict of interest relevant to this work.
      Anthony Gatti is the founder of NeuralSeg Ltd., a provider of medical image analyses that support research conducted within academic and industry settings.

      References

        • Cross M.
        • Smith E.
        • Hoy D.
        • Nolte S.
        • Ackerman I.
        • Fransen M.
        • Bridgett L.
        • Williams S.
        • Guillemin F.
        • Hill C.L.
        • Laslett L.L.
        • Jones G.
        • Cicuttini F.
        • Osborne R.
        • Vos T.
        • Buchbinder R.
        • Woolf A.
        • March L.
        The global burden of hip and knee osteoarthritis: estimates from the Global Burden of Disease 2010 study.
        Ann Rheum Dis. 2014; 73: 1323-1330
        • Nelson A.E.
        • Allen K.D.
        • Golightly Y.M.
        • Goode A.P.
        • Jordan J.M.
        A systematic review of recommendations and guidelines for the management of osteoarthritis: the chronic osteoarthritis management initiative of the U.S. bone and joint initiative.
        Semin Arthritis Rheum. 2014; 43: 701-712
        • Mosher T.J.
        • Dardzinski B.J.
        Cartilage MRI T2 relaxation time mapping: overview and applications.
        Semin Musculoskeletal Radiol. 2004; 8: 355-368
        • Gold G.E.
        • Chen C.A.
        • Koo S.
        • Hargreaves B.A.
        • Bangerter N.K.
        Recent advances in MRI of articular cartilage.
        Am J Roentgenol. 2009; 193: 628-638
        • Hoessly M.L.
        • Wildi L.M.
        Magnetic resonance imaging findings in the knee before and after long-distance running — documentation of irreversible structural damage? A systematic review.
        Am J Sports Med. 2017; 45: 1206-1217
        • Subburaj K.
        • Kumar D.
        • Souza R.B.
        • Alizai H.
        • Li X.
        • Link T.M.
        • Majumdar S.
        The acute effect of running on knee articular cartilage and meniscus magnetic resonance relaxation times in young healthy adults.
        Am J Sports Med. 2012; 40: 2134-2141
        • Behzadi C.
        • Welsch G.H.
        • Laqmani A.
        • Henes F.O.
        • Kaul M.G.
        • Schoen G.
        • Adam G.
        • Regier M.
        The immediate effect of long-distance running on T2 and T2* relaxation times of articular cartilage of the knee in young healthy adults at 3T.
        Br J Radiol. 2016; 89: 20151075
        • Mosher T.J.
        • Liu Y.
        • Torok C.M.
        Functional cartilage MRI T2 mapping: evaluating the effect of age and training on knee cartilage response to running.
        Osteoarthr Cartil. 2010; 18: 358-364
        • Gatti A.A.
        • Noseworthy M.D.
        • Stratford P.W.
        • Brenneman E.C.
        • Totterman S.
        • Tamez-Pena J.
        • Maly M.R.
        Acute changes in knee cartilage transverse relaxation time after running and bicycling.
        J Biomech. 2017; 53: 171-177
        • Messier S.P.
        • Legault C.
        • Schoenlank C.R.
        • Newman J.J.
        • Martin D.F.
        • DeVita P.
        Risk factors and mechanisms of knee injury in runners.
        Med Sci Sports Exerc. 2008; 40: 1873-1879
        • Schütz U.H.W.
        • Ellerman J.
        • Schoss D.
        • Wiedelbach H.
        • Beer M.
        • Billich C.
        Biochemical cartilage alteration and unexpected signal recovery in T2* mapping observed in ankle joints with mobile MRI during a transcontinental multistage footrace over 4486 km.
        Osteoarthr Cartil. 2014; 22: 1840-1850
        • Souza R.B.
        • Kumar D.
        • Calixto N.
        • Singh J.
        • Schooler J.
        • Subburaj K.
        • Li X.
        • Link T.M.
        • Majumdar S.
        Response of knee cartilage T1rho and T2 relaxation times to in vivo mechanical loading in individuals with and without knee osteoarthritis.
        Osteoarthritis Cartilage. 2014; 22: 1367-1376
        • Kellgren J.H.
        • Lawrence J.S.
        Radiological assessment of osteo-arthrosis.
        Ann Rheum Dis. 1957; 16: 494-502
        • Roos E.M.
        • Lohmander L.S.
        • Roos E.M.
        • Lohmander L.S.
        The Knee injury and Osteoarthritis Outcome Score (KOOS): from joint injury to osteoarthritis.
        Health Qual Life Outcomes. 2003; 1: 64
        • Willy R.W.
        • Meardon S.A.
        • Schmidt A.
        • Blaylock N.R.
        • Hadding S.A.
        • Willson J.D.
        Changes in tibiofemoral contact forces during running in response to in-field gait retraining.
        J Sport Sci. 2016; 34: 1602-1611
        • Esculier J.F.
        • Silvini T.
        • Bouyer L.J.
        • Roy J.S.
        Video-based assessment of foot strike pattern and step rate is valid and reliable in runners with patellofemoral pain.
        Phys Ther Sport. 2018; 29: 108-112
        • Esculier J.F.
        • Dubois B.
        • Dionne C.E.
        • Leblond J.
        • Roy J.S.
        A consensus definition and rating scale for minimalist shoes.
        J Foot Ankle Res. 2015; 8: 42
        • Gorgolewski K.
        • Burns C.D.
        • Madison C.
        • Clark D.
        • Halchenko Y.O.
        • Waskom M.L.
        • SS G.
        Nipype: a flexible, lightweight and extensible neuroimaging data processing framework in Python.
        Front Neuroinform. 2011; 5: 13
        • Penny W.
        • Friston K.
        • Ashburner J.
        • Kiebel S.
        • Nichols T.
        Statistical Parametric Mapping: The Analysis of Functional Brain Images.
        1st ed. Academic Press, 2007: 656pp
        • Avants B.B.
        • Tustison N.J.
        • Song G.
        • Cook P.A.
        • Klein A.
        • Gee J.C.
        A reproducible evaluation of ANTs similarity metric performance in brain image registration.
        Neuroimage. 2011; 54: 2033-2044
        • Hunter D.J.
        • Altman R.D.
        • Cicuttini F.
        • et al.
        OARSI clinical trials recommendations: knee imaging in clinical trials in osteoarthritis.
        Osteoarthr Cartil. 2015; 23: 698-715
        • Li X.
        • Wyatt C.
        • Rivoire J.
        • Han E.
        • Chen W.
        • Schooler J.
        • Liang F.
        • Shet K.
        • Souza R.
        • Majumdar S.
        Simultaneous acquisition of T1rho and T2 quantification in knee cartilage: repeatability and diurnal variation.
        J Magn Reson Imaging. 2014; 39: 1287-1293
        • Li X.
        • Wyatt C.
        • Rivoire J.
        • et al.
        Simultaneous acquisition of T1rho and T2 quantification in knee cartilage: repeatability and diurnal variation.
        J Magn Reson Imaging. 2014; 39: 1287-1293
        • Menezes N.M.
        • Gray M.L.
        • Hartke J.R.
        • Burstein D.
        T2 and T1ρ MRI in articular cartilage systems.
        Magn Reson Med. 2004; 51: 503-509
        • Orendurff M.S.
        • Kobayashi T.
        • Tulchin-Francis K.
        • Herring Tullock A.M.
        • Villarosa C.
        • Chan C.
        • Strike S.
        A little bit faster: Lower extremity joint kinematics and kinetics as recreational runners achieve faster speeds.
        J Biomech. 2018; 71: 167-175
        • Cha J.G.
        • Lee J.C.
        • Kim H.J.
        • Han J.K.
        • Lee E.H.
        • Kim Y.D.
        • Jeon C.H.
        Comparison of MRI T2 relaxation changes of knee articular cartilage before and after running between young and old amateur runners.
        Korean J Radiol. 2012; 13: 594-601
        • Karanfil Y.
        • Babayeva N.
        • Dönmez G.
        • et al.
        Thirty minutes of running exercise decreases T2 signal intensity but not thickness of the knee joint cartilage: A 3.0-T magnetic resonance imaging study.
        Cartilage. 2018; (Epub ahead of print)
        • Kessler M.A.
        • Glaser C.
        • Tittel S.
        • Reiser M.
        • Imhoff A.B.
        Recovery of the menisci and articular cartilage of runners after cessation of exercise: additional aspects of in vivo investigation based on 3-dimensional MRI.
        Am J Sports Med. 2008; 36: 966-970
        • Chen M.
        • Qiu L.
        • Shen S.
        • Wang F.
        • Zhang J.
        • Zhang C.
        • Liu S.
        The influences of walking, running and atair activity on knee articular cartilage: Quantitative MRI using T1 rho and T2 mapping.
        PLoS ONE. 2017; 12e0187008
        • Li X.
        • Cheng J.
        • Lin K.
        • Saadat E.
        • Bolbos R.I.
        • Ries M.D.
        • Horvai A.
        • Link T.M.
        • Majumdar S.
        Quantitative MRI using T1ρ and T2 in human osteoarthritic cartilage specimens: Correlation with biochemical measurements and histology.
        Magn Reson Imaging. 2011; 29: 324-334
        • Houard X.
        • Goldring M.B.
        • Berenbaum F.
        Homeostatic mechanisms in articular cartilage and role of inflammation in osteoarthritis.
        Curr Rheumatol Rep. 2013; 15: 375
        • Van Ginckel A.
        • Verdonk P.
        • Victor J.
        • Witvrouw E.
        Cartilage status in relation to return to sports after anterior cruciate ligament reconstruction.
        Am J Sports Med. 2013; 41: 550-559
        • Leiter J.R.
        • MacDonald L.
        • McRae S.
        • Davidson M.
        • MacDonald P.B.
        Intrinsic stresses on bone and cartilage in the normal and anterior cruciate ligament-reconstructed knee before and after a half marathon: a magnetic resonance imaging analysis.
        Clin J Sport Med. 2012; 22: 439-442
        • Bricca A.
        • Juhl C.B.
        • Steultjens M.
        • Wirth W.
        • Roos E.M.
        Impact of exercise on articular cartilage in people at risk of, or with established, knee osteoarthritis: a systematic review of randomised controlled trials.
        Br J Sports Med. 2019; 53: 940-947
        • Lo G.H.
        • Musa S.M.
        • Driban J.B.
        • Kriska A.M.
        • McAlindon T.E.
        • Souza R.B.
        • Petersen N.J.
        • Storti K.L.
        • Eaton C.B.
        • Hochberg M.C.
        • Jackson R.D.
        • Kwoh C.K.
        • Nevitt M.C.
        • Suarez-Almazor M.E.
        Running does not increase symptoms or structural progression in people with knee osteoarthritis: Data from the Osteoarthritis Initiative.
        Clin Rheumatol. 2018; 37: 2497-2504
        • Chakravarty E.F.
        • Hubert H.B.
        • Lingala V.B.
        • Zatarain E.
        • Fries J.F.
        Long distance running and knee osteoarthritis. A prospective study.
        Am J Prev Med. 2008; 35: 133-138
        • Van Ginckel A.
        • Baelde N.
        • Almqvist K.F.
        • Roosen P.
        • McNair P.
        • Witvrouw E.
        Functional adaptation of knee cartilage in asymptomatic female novice runners.
        Osteoarthr Cartil. 2010; 18: 1564-1569
        • Kersting U.G.
        • Stubendorff J.J.
        • Schmidt M.C.
        • Bruggemann G.P.
        Changes in knee cartilage volume and serum COMP concentration after running exercise.
        Osteoarthr Cartil. 2005; 13: 925-934