Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Nov 1.
Published in final edited form as: J Orthop Res. 2021 Feb 3;39(11):2340–2352. doi: 10.1002/jor.24994

Characterizing the transient response of knee cartilage to running: Decreases in cartilage T2 of female recreational runners

Hollis A Crowder 1,2, Valentina Mazzoli 2, Marianne S Black 2, Lauren E Watkins 2,3, Feliks Kogan 2, Brian A Hargreaves 2,3,4, Marc E Levenston 1,2, Garry E Gold 2
PMCID: PMC8295402  NIHMSID: NIHMS1718949  PMID: 33483997

Abstract

Cartilage transmits and redistributes biomechanical loads in the knee joint during exercise. Exercise-induced loading alters cartilage hydration and is detectable using quantitative magnetic resonance imaging (MRI), where T2 relaxation time (T2) is influenced by cartilage collagen composition, fiber orientation, and changes in the extracellular matrix. This study characterized short-term transient responses of healthy knee cartilage to running-induced loading using bilateral scans and image registration. Eleven healthy female recreational runners (33.73 ± 4.22 years) and four healthy female controls (27.25 ± 1.38 years) were scanned on a 3T GE MRI scanner with quantitative 3D double-echo in steady-state before running overground (runner group) or resting (control group) for 40 min. Subjects were scanned immediately post-activity at 5-min intervals for 60 min. T2 times were calculated for femoral, tibial, and patellar cartilage at each time point and analyzed using a mixed-effects model and Bonferroni post hoc. There were immediate decreases in T2 (mean ± SEM) post-run in superficial femoral cartilage of at least 3.3% ± 0.3% (p = .002) between baseline and Time 0 that remained for 25 min, a decrease in superficial tibial cartilage T2 of 2.9% ± 0.4% (p = .041) between baseline and Time 0, and a decrease in superficial patellar cartilage T2 of 3.6% ± 0.3% (p = .020) 15 min post-run. There were decreases in the medial posterior region of superficial femoral cartilage T2 of at least 5.3 ± 0.2% (p = .022) within 5 min post-run that remained at 60 min post-run. These results increase understanding of transient responses of healthy cartilage to repetitive, exercise-induced loading and establish preliminary recommendations for future definitive studies of cartilage response to running.

Keywords: cartilage, running, T2, transient

1 |. INTRODUCTION

Knee cartilage transmits and redistributes a variety of biomechanical loads during normal activities like exercise.1 Cartilage stiffness supports loading and is influenced by cartilage extracellular matrix (ECM).2,3 Differences in cartilage ECM due to aging, injury, or disease, like reduced proteoglycan and increased water contents, affect cartilage loading response.4 Therefore, characterizing cartilage loading response may be a way to monitor cartilage health. Additionally, cartilage loading response bears consideration when designing longitudinal studies on joint cartilage. If there are detectable, sustainable differences in cartilage properties postexercise, efforts should be made to mitigate loading history effects during subject screening.

Quantitative magnetic resonance imaging (MRI) provides a noninvasive tool for studying cartilage properties. T2 relaxation time (T2) correlates with changes in cartilage collagen composition, orientation, and water content.1,5 Running and walking involve repetitive biomechanical loads in the knee, and T2 may detect immediate effects of exercise-induced loading on tibial and femoral articular cartilage.3,6 Reduced T2 relaxation times (T2) have been observed in healthy knee cartilage post-run and post-walk (reductions are larger in magnitude post-run7) with limited temporal resolution (one or two time points).7–9 This suggests compression forces on the knee while running (up to nine times bodyweight during knee extension/stance phase10) temporarily force water out of cartilage and alter collagen fiber orientation, affecting cartilage T2.

Cartilage T2 response to running varies between femoral, tibial, and patellar cartilage and within cartilage tissue. The responses are also depth-dependent, indicating tissue-, region-, and depth-dependent analysis is important.8,11,12 Prior work is limited to one knee, disregarding effects of gait asymmetries between knees.13 Sex-based differences may also affect cartilage loading response. Cartilage in women has decreased volume and thickness and sees less medial loading compared to men, even controlling for joint size and body mass index (BMI).14 Therefore, it is important to investigate possible differences in healthy cartilage T2 loading response within cartilage tissue, between different cartilage tissues, between right/left knees, and to consider sex-based differences.

Previous work shows healthy cartilage T2 response to exercise-induced loading is time-dependent. Healthy cartilage T2 can recover within minutes of exercise even after long-distance running (>20 km) indicating changes in collagen fiber orientation and water content may be temporary.15,16 However, the time evolution of these changes is poorly understood. Previous studies examined either immediate or delayed effects of running on cartilage T2 (within 4 min, 20 min, or 48 h of running, respectively)9,16,17 but not time points in between. Cartilage T2 measurements limited to one snapshot may miss subtle changes that transient analysis could uncover.

Historically, longer scan times characteristic of quantitative MRI (8–20 min, depending on the sequence) have restricted transient analysis of running-induced cartilage T2 changes.16 Transient analysis has become possible due to recent developments in 3D scan sequences, which allow for observation of short-term transient cartilage behaviors.18,19 Novel, time-efficient (5 min) sequences like bilateral quantitative 3D double-echo in steady-state (qDESS) make transient analysis possible. These scans allow for cartilage T2 measurement closer to exercise completion, revealing immediate cartilage response to loading and enabling more frequent cartilage T2 measurements to track transient cartilage behavior. Bilateral qDESS scanning also enables tracking left/right knee differences.18 Post-scan, image registration techniques increase image processing efficiency because registration allows the use of one manual binary segmentation mask (segmentation) on multiple images and time points. These advancements allow the characterization of short-term transient cartilage T2 response to exercise-induced loading.

This study is a first proof-of-concept characterizing short-term (within 60 min) transient responses of healthy knee cartilage in female recreational runners to running-induced loading compared to healthy female controls to inform future designs of larger definitive studies on the short-term transient response of healthy cartilage to running. This study observed short-term transient responses of femoral, tibial, and patellar cartilage T2 to running-induced loading, and to layer- and regional-specific recover patterns following exercise.

2 |. METHODS

This prospective cohort (Level II) study was performed according to Institutional Review Board approved procedures with signed informed consent obtained from all study participants.

2.1 |. Subjects

Eleven healthy female recreational runners (age, 33.73 ± 4.22 years; height, 1.72 ± 0.02m, weight, 63.37 ± 2.04 kg, and BMI, 21.35 ± 1.38 kg/m2) running –925 miles per week (mean mileage 16.19 ± 1.45 miles/week; median mileage 15 miles/week) were included in the runner group. Four healthy female controls (age, 27.25 ± 1.37 years; height, 1.64 ± 0.01m, weight, 59.52 ± 2.33 kg; BMI, 21.85 ± 0.89 kg/m2) running 0 miles per week were included in the control group. No participants had knee pain or prior knee trauma, known joint disorders, or history of orthopedic surgery.

2.2 |. Participants and study design

Participants abstained from running long distances (>13 miles) or engaging in strenuous activity outside their typical activity level the week before participation, and activity was minimized on scanning day to reduce effects of preceding physical activities on imaging results.20 Demographic data (age, weight, height, and BMI) and average miles run per week for the last month were collected before scanning.

All participants received a pre-activity bilateral scan (baseline) laying in supine position with padding around each knee coil and lower leg to reduce knee rotation and translation. Subsequently, each runner participant ran for 40 min on flat outdoor terrain at a comfortable, sustained effort. Non-runner controls rested for 40 min in a chair after the baseline scan. Immediately (2 min) after completing the activity each participant was scanned again (Time 0 min), with scans repeated every 5 min for 60 min (Time 5 min, Time 10 min, …, Time 60 min) post-activity (Figure 1A) for a total of 14 bilateral scans per participant.

FIGURE 1.

FIGURE 1

Open in a new tab

Timeline of (A) experimental procedure and magnetic resonance (MR) image acquisitions; and (B) image analysis procedure with scan quality verification, manual segmentation, and T2 relaxation data projections

2.3 |. MR image acquisition

All imaging experiments were performed on a 3.0T GE Premier scanner (GE Healthcare). Knees were scanned simultaneously using two 16-channel flexible phased-array, receive-only, medium-sized extremity coils (NeoCoil).18 During scanning participants lay supine with knees in an unlocked, fully extended, and neutrally rotated position. The MR protocol consisted of a repeated bilateral qDESS sequence with parameters outlined in Table 1 acquired at every time point.18,21

TABLE 1.

Summary of MR image acquisition parameters

MR Parameter Specification
Repetition time (TR) 15.5 ms
Echo time 1 5.2 ms
Echo time 2 25.8 ms
Slice thickness 1.4 mm
Matrix size 256 × 256
Number of slices 500
Field of view (FOV) 160 × 160 mm
Flip angle 20
Total scan time 65 min
Open in a new tab

2.4 |. MR image analysis

Every scan was visually inspected, and T2 relaxation values were estimated pixel-by-pixel from qDESS echo images.21–23 T2 values over 100 ms were assumed to be associated fluid and were discarded. Femoral, tibial, and patellar cartilage for each knee were manually segmented using the first echo of the Time 0 min (Time 0) scan using Amira 6.7.0 (Thermo Fisher Scientific)21 (Figure 1B).

Scans for each subject were registered to their respective Time 0 scans. Differences in knee-flexion angle between baseline and post-activity scans were accounted for by registering baseline scans to Time 0 scans using piece-wise registration, with femoral, tibial, and patellar segmentations as regions of interest using rigid registration (Figure 2A–D).19

FIGURE 2.

FIGURE 2

Open in a new tab

Representative registration process demonstrating the registration of a Time 60 min post-activity scan to a Time 0 min post-activity scan using the femoral Time 0 segmentation as the region of interest

Registration transformations were transferred to corresponding T2 maps at each time point. Padding was assumed to minimize differences in knee-flexion angle between sequential post-activity scans. Post-activity scans were rigidly registered to Time 0 scans using femoral cartilage as the region of interest. All registration steps were implemented using Elastix.19 Manual cartilage segmentations of baseline and Time 60 min (Time 60) images validated registration algorithms. Image registration error was quantified by comparing baseline cartilage T2 maps created using unregistered baseline scans and baseline manual segmentations to baseline cartilage T2 maps created using registered baseline scans and Time 0 segmentations, and similarly for Time 60. Cartilage T2 maps created using manual segmentation baseline and Time 60 scans were designated as ground truth.

2.5 |. Cartilage T2 maps

Femoral, tibial, and patellar cartilage T2 values were extracted using registered images and corresponding Time 0 segmentations. Femoral cartilage T2 maps were automatically split into superficial/deep layers at 50% cartilage thickness relative to bone-cartilage interfaces and cartilage surfaces.24 The voxel size for this study was 0.625 × 0.625 × 1.4mm3 and cartilage in the knee has been reported to have an average thickness of approximately 2mm.25 Dividing at 50% thickness ensured a thickness of at least one voxel would be included in both the superficial and deep layers. For visualization purposes, 3D segmentation volumes of femoral cartilage T2 maps were projected onto a sagittal plane, fit cylindrically using a least squares method, and unrolled.26,27 Tibial and patellar cartilage T2 projections were created by projecting 3D segmentation volumes of tibial and patellar cartilage T2 maps onto an axial or coronal plane, respectively. Tibial and patellar cartilage were divided into superficial/deep layers at 50% of cartilage thickness similar to femoral cartilage.24

Femoral cartilage regions were designated as follows26: Posterior regions were designated as 44 degrees from vertical to the posterior edge of the segmentation and anterior regions as 40 degrees from vertical to the anterior edge of the segmentation to ensure the central region comprised femoral cartilage in contact with the meniscus. Medial/lateral regions were designated by dividing the maximal width of femoral cartilage in two at each anterior, central, and posterior division, for six femoral anatomical regions, with two depths (superficial/deep) totaling twelve femoral regions. Regions of tibial cartilage were designated by dividing the maximal width of tibial cartilage by two and dividing maximal length (anterior to posterior) of each side by 3 to designate anterior/central/posterior regions, for six anatomical regions, with two depths (superficial/deep) totaling twelve tibial regions. Patellar cartilage regions were designated by dividing the maximal width of patellar cartilage in two, creating medial/lateral anatomical regions, with two depths (superficial/deep) totaling four regions. Outcomes were reported for each cartilage structure as both global (entire cartilage area) and regional T2 relaxation times.

2.6 |. Statistical analysis

Registration error was quantified using Lin's Concordance Correlation Coefficient (LCCC) and evaluated using Bland-Altman analysis.28,29 A two-sided mixed-effects model with a restricted maximum likelihood estimation method and Bonferroni post-hoc correction was used to compare mean T2 times between control and runner groups at each time point and to compare mean T2 times of post-activity to baseline scans within runner and control groups. Subjects were designated as random effects and nested within condition (control or runner). Knee (left or right) was nested within subject, and depth, region, and time were fixed. Statistical significance was set at α = .05. Statistical analysis was performed with Minitab (version 19.2.0.0, Minitab, LLC).

3 |. RESULTS

3.1 |. Patient characteristics

Visual inspection revealed insufficient image quality of one knee for three participants in the runner group (two right knees, one left knee) leaving 19 knees across 11 participants and eight knees across four participants in the runner and control groups, respectively.

3.2 |. Registration validation

Visual inspection revealed subjects shifted position between scans, and rigid registration was used to account for this. To evaluate piece-wise rigid registration of baseline scans to Time 0 scans, baseline T2 values extracted using Time 0 manual segmentations and registered baseline T2 data (registered T2 data) were compared to baseline cartilage T2 values extracted using baseline manual segmentations (manual T2 data) using Lin's Concordance Correlation Coefficient for femoral, tibial, and patellar cartilage.28 Similar analysis was performed at Time 60 to evaluate rigid registration of post-activity scans to Time 0 scans. The highest correlation between baseline registered T2 data and baseline manual T2 data was observed served in femoral (0.97) and tibial (0.97) cartilage, the lowest correlation in patellar cartilage (0.94). Mean biases (difference in mean T2) between Time 60 registered and manual T2 data were −0.68, −0.66, and −0.21 ms for femoral, tibial, and patellar cartilage, respectively. Higher correlation between Time 60 registered T2 data and Time 60 manual T2 data was observed in femoral cartilage (0.98) and patellar cartilage (0.96) and lower correlation in tibial cartilage (0.95). Mean biases between Time 60 registered and manual T2 data were −0.72, 0.13, and −0.55 ms for femoral, tibial, and patellar cartilage, respectively. These results are summarized in Table 2.

TABLE 2.

Summary of comparison between T2 values generated by segmentation (seg) and registration (reg) of femoral, tibial, and patellar cartilage, for baseline and Time 60 time points

Cartilage Time point Mean T2 seg (ms) Mean T2 rega (ms) Bias (95% CIb) SD c LOAd LCCCe
Femoral Baseline 31.32 30.64 −0.68 (−1.17, −0.19) 1.26 (−3.16, 1.79) 0.97
Tibial Baseline 26.75 26.08 −0.66 (−1.08, −0.24) 0.91 (−2.45, 1.12) 0.97
Patellar Baseline 29.41 29.20 −0.21 (−0.88, 0.47) 1.76 (−3.66, 3.24) 0.94
Femoral Time 60 31.73 33.76 −0.72 (−1.04, −0.4) 3.80 (−5.42, 9.48) 0.98
Tibial Time 60 26.29 26.43 0.13 (−0.52, 0.79) 1.70 (−3.20, 3.47) 0.95
Patellar Time 60 29.91 29.36 −0.55 (−1.04, −0.07) 1.26 (−3.02, 1.91) 0.96
Open in a new tab
a

Registration.

b

Confidence interval.

c

Standard deviation.

d

Limits of agreement.

e

Lin’s concordance correlation coefficient.

Differences between the two methods are summarized with Bland-Altman plots29 in Figure 3A–F. More variation was observed in mean T2 differences in tibial and patellar cartilage than femoral cartilage and in the runner group compared to the control group in tibial cartilage at both time points (Figure 3B,E).

FIGURE 3.

FIGURE 3

Open in a new tab

Bland-Altman plots of mean T2 values at baseline in (A) femoral, (B) tibial, and (C) patellar cartilage and at Time 60 in (D) femoral, (E) tibial, and (F) patellar cartilage for both runners and controls

3.3 |. T2 analysis

Figure 4A–C shows representative projections of femoral, tibial, and patellar cartilage T2 maps for a runner at each time point. Superficial cartilage T2 maps show elevated T2 values compared to deep cartilage T2 maps, and clusters of elevated cartilage T2 at baseline in femoral cartilage (Figure 4A), tibial (Figure 4B), and patellar cartilage (Figure 4C) diminish immediately post-run before recovering by Time 60.

FIGURE 4.

FIGURE 4

Open in a new tab

T2 projection maps for (A) femoral, (B) tibial, and (C) patellar cartilage at each time point

Generally, there were no significant differences in mean cartilage T2 between runner and control groups at any time point. No changes in T2 in the superficial or deep mean global cartilage T2 in femoral, tibial, and patellar cartilage were observed in the control group over time. In all cases, mean superficial cartilage T2 remained consistently higher than mean deep cartilage T2. Numerical results for mean T2 are presented as mean ± SEM with units of ms. Percent differences in mean T2 are presented as (mean percent difference – SEM) – (mean percent difference + SEM) with baseline mean T2 for reference.

There were significant decreases within the runner group in global mean cartilage T2 post-run compared to baseline, with a decrease in mean superficial femoral cartilage T2 of 3.2%–4.6% between baseline (37.32 ± 0.70 ms) and Time 0 (35.86 ± 0.43 ms) (p = .002). Significant decreases in T2 between baseline and post-run scans continued from Time 0 to Time 25, with a maximum percent T2 decrease of 4.8%–5.4% between baseline (37.32 ± 0.70 ms) and Time 10 (35.41 ± 0.56 ms) (p = .002) (Figure 5A). There was a decreasing trend in mean superficial femoral cartilage T2 of 3.0%–3.8% between baseline (37.32 ± 0.70 ms) and Time 30 (36.04 ± 0.52 ms) that was not significant (p = .072). There was a significant decrease in mean superficial tibial cartilage T2 within the runner group of 2.6%–3.3% between baseline (33.00 ± 0.89 ms) and Time 0 (32.03 ± 0.99 ms) (p = .041) (Figure 5B). There was a significant decrease in mean superficial patellar cartilage T2 within the runner group of 3.3%–3.8% between baseline (34.85 ± 0.85 ms) and Time 15 (33.62 ± 0.90 ms) (p = .020) (Figure 5C).

FIGURE 5.

FIGURE 5

Open in a new tab

Mean global T2 of control and runner groups at each time point in (A) femoral, (B) tibial, and (C) patellar cartilage. Error bars designate 95% confidence intervals

There were also significant differences in mean regional cartilage T2 between baseline and post-run scans within the runner group, with significant decreases in mean T2 of superficial medial posterior femoral cartilage between baseline and every time point post-run except Time 0, with a maximum decrease of 5.8%–6.1% between baseline (39.18 ± 0.76 ms) and Time 60 (36.90 ± 0.62 ms) (p = .001) (Figure 6C). There were trends of decreased mean T2 post-run in superficial femoral cartilage in the lateral anterior (p = .181) and lateral posterior (p = .133) regions of the runner group (Figure 6D,F) but neither were significant.

FIGURE 6.

FIGURE 6

Open in a new tab

Mean regional T2 relaxation times of (A) medial anterior, (B) medial central, (C) medial posterior, (D) lateral anterior, (E) lateral central, and lateral posterior (F) femoral cartilage at the superficial and deep layer for each time point of the control and runner groups. Error bars designate 95% confidence intervals

There were no significant differences in regional mean T2 between baseline and post-run scans in superficial or deep tibial cartilage within the runner group (Figure 7A–F).

FIGURE 7.

FIGURE 7

Open in a new tab

Mean regional T2 relaxation times of (A) medial anterior, (B) medial central, (C) medial posterior, (D) lateral anterior, (E) lateral central, and (F) lateral posterior tibial cartilage at the superficial and deep layer for each time point of the control and runner groups. Error bars designate 95% confidence intervals

There were trends of decreased mean regional T2 in the medial superficial patellar cartilage of 7.6%–9.7% between baseline (34.89 ± 0.93ms) and Time 15 (31.88 ± 1.22ms) (p =.052) (Figure 8A) and in the lateral superficial patellar cartilage of 5.1%–6.6% between baseline (34.76 ± 0.85ms) and Time 15 (32.73 ± 1.06ms) (p =.068) within the runner group that were not significant (Figure 8B).

FIGURE 8.

FIGURE 8

Open in a new tab

Mean regional T2 relaxation times of (A) medial and (B) lateral patellar cartilage at the superficial and deep layers for each time point of the control and runner groups. Error bars designate 95% confidence intervals

Finally, there were significant differences in mean cartilage T2 of global femoral cartilage between right (30.39 ± 0.32 ms) and left knees (30.75 ± 0.32 ms) (p = .022), in global tibial cartilage between right (25.63 ± 0.37 ms) and left knees (26.30 ± 0.35 ms) (p = .029), and in global patellar cartilage between right (28.82 ± 0.32 ms) and left knees (28.85 ± 0.27 ms) (p = .013).

4 |. DISCUSSION

This study is a first proof-of-concept in characterizing short-term (within 60 min) transient responses of healthy knee cartilage in female recreational runners to running-induced loading compared to healthy female controls. No change in knee cartilage T2 of female male controls was observed and running-induced differences in transient responses of knee cartilage T2 of female recreational runners were detected. Running-induced decreases in mean cartilage T2 of femoral, tibial, and patellar cartilage were observed immediately post-run until 25 min post-run, while running-induced decreases in the medial posterior region of superficial femoral cartilage were observed immediately post-run and continued for 60 min post-run.

LCCC values of femoral cartilage (0.97 for baseline and 0.98 for Time 60) calculated during registration validation were lower than prior results of 0.996 in a study using 3D rigid registration on femoral cartilage segmentations.30 Differences between LCCC values of femoral/tibial cartilage and patellar cartilage at baseline and between femoral and tibial/patellar cartilage at Time 60 may be due to decreased cross-sectional area, as decreases in agreement between rigidly registered scans have been found as tissue cross-sectional area decreases.31 Tibial and patellar cartilage are smaller than femoral cartilage, which could explain decreased agreement in tibial cartilage at Time 60, and in patellar cartilage at both time points. Differences between LCCC values of tibial/patellar cartilage and femoral cartilage at Time 60 may be explained by using femoral cartilage segmentations as the target registration ROI before copying those femoral-based transforms for tibial and patellar cartilage analysis. Additionally, compressive, running-induced loads decrease cartilage thickness, which rigid registration may not account for.32

In Bland-Altman analysis, knee femoral cartilage had a maximum mean bias magnitude (±95% CI) less than 0.68 ± 0.49 and 0.72 ± 0.32 ms at baseline and Time 60, respectively. Therefore the smallest detectable difference accounting for error introduced during registration rather than true differences in cartilage T2 in femoral cartilage at any time point was 0.72 ± 0.32 ms. Similarly, the smallest detectable difference in cartilage T2 was 0.66 ± 0.42 and 0.55 ± 0.49 ms for tibial and patellar cartilage, respectively.

Cartilage T2 is influenced by multiple factors, including collagen fiber organization and water content, which are layer-dependent.1 Temporary changes in superficial cartilage T2 indicate differences in collagen fiber orientation and water content of that layer. Repetitive, running-induced loading of knee cartilage may change collagen fiber alignment and push water out of superficial cartilage (which is more compressible than deep cartilage) affecting T2.1 Significant decreases in global superficial femoral cartilage T2 were not observed after 25 min, indicating short-term impacts on collagen fiber orientation and water content across the cartilage area post-run. Conversely, significant initial decreases in mean T2 of the medial posterior region of superficial femoral cartilage were sustained for 60 min, indicating region-specific and sustained changes in cartilage collagen fiber orientation and water content post-run in that region.

Additionally, superficial cartilage T2 was consistently higher than deep cartilage T2 in both the runner and control groups, and changes in mean cartilage T2 between baseline and post-run scans in the runner group were limited to superficial cartilage. Collagen fibers in superficial cartilage (10%–20% of cartilage nearest to the surface) are oriented differently (parallel to the articular surface) than in deep cartilage (30%–40% of cartilage furthest from the surface) in which collagen fibers are oriented perpendicular to the articular surface, producing different responses to biomechanical loading.1 Specifically, superficial cartilage compresses more than deep cartilage, which may partially account for differences in superficial cartilage T2 post-run that were not present in deep cartilage T2.1 For practical reasons, this study divided cartilage into superficial and deep layers at 50% of cartilage thickness, meaning part of the 40%–60% of cartilage in the middle zone where there are collagen fibers aligned parallel and perpendicular to the cartilage surface was included in both the superficial and deep-layer designations, limiting the conclusions that can be drawn about the effect of collagen fiber orientation.1 Even so, differences were observed between the response of the superficial and deep layers of cartilage in this study, indicating that averaging cartilage T2 measurements across full cartilage thickness may obscure layer-specific cartilage responses and that future running studies would benefit from a layer-specific analysis.

Decreases in superficial femoral cartilage T2 (3.3%–4.9%) between baseline and Time 0 of young female recreational runners in this study were consistent with immediate decreases in T2 (5.4%–5.7%) in superficial femoral cartilage of young healthy male volunteers who ran for 30 min and were scanned within 4 min of running, indicating little difference between the immediate running-induced cartilage T2 response of female and male runners.9 Additionally, decreases in femoral cartilage were smaller than other studies that found decreases in T2 relaxation times of 5.4%–6.0% and 7.1% in superficial femoral cartilage after runs as short as 15 min.6,20 However, runners in those studies ran on a treadmill, whereas runners in this investigation ran over-ground. Running on a treadmill influences gait characteristics compared to over-ground running, changing the biomechanics and force distribution normally experienced by cartilage while running, affecting cartilage T2.33,34 These studies also used different sequences to acquire T2, making direct T2 comparisons problematic due to the sequence-dependent nature of T2 measurements.6,20,33,34

Other studies observed no significant changes in femoral or tibial cartilage T2 in female runners following 30 min of running, though scanning commenced approximately 20 min post-run and analysis was limited to full-thickness cartilage.16 Based on our results, layer-specific cartilage transient responses and decreases in femoral and tibial cartilage T2 may occur within 5 min post-run, indicating 20 min between the end of running and start of scanning may be too long to capture T2 changes. To observe immediate exercise-induced changes in cartilage T2, our results suggest scanning within 5 min of the end of exercise. Our results also demonstrate femoral, tibial, and patellar cartilage may have different transient responses to exercise-induced loading, and regional cartilage T2 responses may behave differently over time compared to each other and global cartilage T2, indicating future large-scale running studies may benefit from transient, regional, and layer-specific analyses.

Measuring postexercise cartilage strain using repeated DESS scans and 3D joint modeling has also been investigated as a noninvasive method to evaluate cartilage health.35 Full surface tibial cartilage strain was measured using repeated 10-min DESS scans after 30 min of walking on a treadmill in young, healthy male and female participants with an average recovery time to baseline strain levels of 25.2 min.35 Though there were differences between this study design and our study design (subject sex, mode and duration of exercise, tissue of interest, and scan duration) it is interesting to note that initial decreases in cartilage T2 superficial femoral cartilage in our study were not observed after 25 min. Further investigation is necessary to determine whether a relationship exists between transient strain response and transient T2 response of healthy cartilage to exercise-induced loading.

Additionally, there were small but statistically significant differences between right and left knees. Registration validation determined a smallest detectable difference of 0.72 ± 0.32, 0.66 ± 0.42, and 0.55 ± 0.49 ms for femoral, tibial, and patellar cartilage T2, respectively. Therefore, statistically significant differences between right and left knees of 0.34 ± 0.32, 0.67 ± 0.36, and 0.03 ± 0.30 ms in global femoral, tibial, and patellar cartilage T2, respectively, are too small to be practically relevant. Further investigation is required to determine whether there are both statistically and practically significant differences in cartilage T2 loading response between right and left knees of young healthy female runners.

Ultimately, this study demonstrates the importance of including layer- and region-specific analyses when characterizing cartilage changes due to loading during exercise. These results also have practical implications for future definitive running studies: Such studies may benefit from starting postexercise scans as soon as possible after exercise (within 5 min, ideally) and included repeated scans for at least 60 min postexercise to investigate the transient nature of cartilage response to running.

4.1 |. Limitations

While layer- and regional-specific transient analyses increase information about cartilage T2 response to running, there are several limitations. This was a pilot study of young healthy female participants, which could explain the lack of between-group differences in cartilage T2. Additionally, the mechanism to explain the delayed significant decrease in superficial patellar cartilage T2 at Time 15 min is unclear, which may indicate that result may not be practically significant. Furthermore, cartilage layer-specific analysis requires high-resolution imaging, requires longer scan times, and compromises temporal resolution, possibly contributing to a lack of T2 differences. T1ρ has a larger dynamic range than T2 to detect changes postexercise, however T1ρ takes longer to acquire and does not extend easily to bilateral scanning, so cannot be used to evaluate asymmetry between knees nor subtle regional cartilage differences.16,20

Data collection in a larger female population with similar characteristics is necessary to expand these results. Additional investigation is also necessary to determine the transient cartilage T2 response in male and/or older populations. Neither biomechanical joint loading nor gait analysis data were collected, so the relationship between those factors and cartilage T2 was not assessed. Importantly, these results apply to healthy female recreational runners but do not extend to cartilage with different compositional and structural properties like reduced proteoglycan and increased water contents due to aging, injury, or disease. Running was the only exercise evaluated here and involves high strain magnitudes and strain rates, making it unsuitable for patient groups with compromised health or low fitness levels.36 Deep knee bends are an alternative, lower-impact exercise that could be used to observe transient cartilage T2 in patient groups for which running is inappropriate.

5 |. CONCLUSION

Overall, this study provided a proof-of-concept for characterizing short-term transient response of cartilage T2 to running-induced loading in young healthy female runners. Decreases in cartilage T2 in femoral, tibial, and patellar cartilage post-run varied over time, and each tissue had a unique layer- and region-specific transient response. These results provide new information about the timing of post-run scanning, highlight the role of transient analysis in the investigation of cartilage response to biomechanical loading, and provide preliminary data for consideration in the design of larger, definitive running studies.

ACKNOWLEDGMENTS

The study was supported by the NIH (Grant nos. NIH R01EB002524, NIH K24AR062068, NIH R01AR07449201, R01AR077604) and GE Healthcare.

Funding information

National Institutes of Health, Grant/Award Numbers: K24AR062068, R01AR07449201, R01AR077604, R01EB002524

Footnotes

CONFLICT OF INTERESTS

The authors decalre that there are no conflict of interests.

REFERENCES

  • 1.Sophia Fox AJ, Bedi A, Rodeo SA. The basic science of articular cartilage: structure, composition, and function. Sports Health. 2009;1(6):461–468. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Wilkins RJ, Hopewell B, Urban JPG. Fixed charge density and cartilage biomechanics. Many Faces Osteoarthr. 2002:387–395. [Google Scholar]
  • 3.Heijink A, Gomoll AH, Madry H, et al. Biomechanical considerations in the pathogenesis of osteoarthritis of the knee. Knee Surg Sports Traumatol Arthrosc. 2012;20(3):423–435. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Kleemann RU, Krocker D, Cedraro A, Tuischer J, Duda GN. Altered cartilage mechanics and histology in knee osteoarthritis: relation to clinical assessment (ICRS Grade). Osteoarthr Cartil. 2005;13(11): 958–963. [DOI] [PubMed] [Google Scholar]
  • 5.Dunn TC, Lu Y, Jin H, Ries MD, Majumdar S. T2 relaxation time of cartilage at MR imaging: comparison with severity of knee osteoarthritis. Radiology. 2004;232(2):592–598. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Gatti AA, Noseworthy MD, Stratford PW, et al. Acute changes in knee cartilage transverse relaxation time after running and bicycling. J Biomech. 2017;53:171–177. 10.1016/j.jbiomech.2017.01.017 [DOI] [PubMed] [Google Scholar]
  • 7.Chen M, Qiu L, Shen S, et al. The influences of walking, running and stair activity on knee articular cartilage: quantitative MRI using T1 rho and T2 mapping. PLOS One. 2017;12(11):e0187008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Hoessly ML, Wildi LM. 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(5):1206–1217. [DOI] [PubMed] [Google Scholar]
  • 9.Behzadi C, Welsch GH, Laqmani A, et al. The immediate effect of long-distance running on T2 and T2 * relaxation times of articular cartilage of the knee in young healthy adults at 3.0 T MR imaging 1. Br J Radiol. 2016;89:20151075. 10.1259/bjr.20151075. Accessed December 11, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.DĽima DD, Fregly BJ, Patil S, Steklov N, Colwell CW. Knee joint forces: prediction, measurement, and significance. Proc Inst Mech Eng Part H J Eng Med. 2012;226(2):95–102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Mosher TJ, Liu Y, Torok CM. Functional cartilage MRI T2 mapping: evaluating the effect of age and training on knee cartilage response to running. Osteoarthr Cartil. 2010;18(3):358–364. 10.1016/j.joca.2009.11.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Stehling C, Liebl H, Krug R, et al. Patellar cartilage: T2 values and morphologic abnormalities at 3.0-T MR imaging in relation to physical activity in asymptomatic subjects from the osteoarthritis initiative. Radiology. 2010;254(2):509–520. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Forczek W, Staszkiewicz R. An evaluation of symmetry in the lower limb joints during the able-bodied gait of women and men. J Hum Kinet. 2012;35(1):47–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Kumar D, Souza RB, Subburaj K, et al. Are there sex differences in knee cartilage composition and walking mechanics in healthy and osteoarthritis populations? Clin Orthop Relat Res. 2015;473(8): 2548–2558. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Kessler MA, Glaser C, Tittel S, Reiser M, Imhoff AB. Recovery of the menisci and articular cartilage of runners after cessation of exercise: additional aspects of in vivo investigation based on 3-dimensional magnetic resonance imaging. Am J Sports Med. 2008;36(5):966–970. [DOI] [PubMed] [Google Scholar]
  • 16.Esculier JF, Jarrett M, Krowchuk NM, et al. Cartilage recovery in runners with and without knee osteoarthritis: a pilot study. Knee. 2019;26(5): 1049–1057. 10.1016/j.knee.2019.07.011 [DOI] [PubMed] [Google Scholar]
  • 17.Luke AC, Stehling C, Stahl R, et al. High-field magnetic resonance imaging assessment of articular cartilage before and after marathon running: does long-distance running lead to cartilage damage? Am J Sports Med. 2010;38(11):2273–2280. [DOI] [PubMed] [Google Scholar]
  • 18.Kogan F, Levine E, Chaudhari AS, et al. Simultaneous bilateral-knee MR imaging. Magn Reson Med. 2018;80(2):529–537. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Klein S, Staring M, Murphy K, Viergever MA, Pluim J. Elastix: a toolbox for intensity-based medical image registration. IEEE Trans Med Imaging. 2010;29(1):196–205. [DOI] [PubMed] [Google Scholar]
  • 20.Subburaj K, Kumar D, Souza RB, et al. 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(9): 2134–2141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Sveinsson B, Chaudhari AS, Gold GE, Hargreaves BA. A simple analytic method for estimating T2 in the knee from DESS. Magn Reson Imaging. 2017;38:63–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Jordan CD, McWalter EJ, Monu UD, et al. Variability of CubeQuant T1ρ, quantitative DESS T2, and cones sodium MRI in knee cartilage. Osteoarthr Cartil. 2014;22:1559–1567. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Matzat SJ, McWalter EJ, Kogan F, Chen W, Gold GE. T2 Relaxation time quantitation differs between pulse sequences in articular cartilage. J Magn Reson Imaging. 2015;42(1):105–113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Wirth W, Maschek S, Roemer FW, Eckstein F. Layer-specific femorotibial cartilage T2 relaxation time in knees with and without early knee osteoarthritis: data from the osteoarthritis initiative (OAI). Sci Rep. 2016;6(September):1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Shepherd DET, Seedhom BB. Thickness of human articular cartilage in joints of the lower limb. Ann Rheum Dis. 1999;58(1):27–34, [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Black M, Yoon D, Young K, et al. Detecting early changes in ACL-reconstructed knee cartilage: cluster analysis of T2 relaxation times in superficial and deep cartilage and ADC analysis. ISMRM Annu Meet. 2020;2020:2738. [Google Scholar]
  • 27.Monu UD, Jordan CD, Samuelson BL, Hargreaves BA, Gold GE, McWalter EJ. Cluster analysis of quantitative MRI T2 and T1ρ relaxation times of cartilage identifies differences between healthy and ACL-injured individuals at 3T. Osteoarthr Cartil. 2017;25: 513–520. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Lin LI. A concordance correlation coefficient to evaluate reproducibility. Biomatrics. 1989;45(1):255–268. [PubMed] [Google Scholar]
  • 29.Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1: 307–310. [PubMed] [Google Scholar]
  • 30.Duarte A, Ruiz A, Ferizi U, et al. Diffusion tensor imaging of articular cartilage using a navigated radial imaging spin-echo diffusion (RAISED) sequence. Eur J Radiol. 2019;29(5):2598–2607. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Newbould RD, Miller SR, Toms LD, et al. T2* measurement of the knee articular cartilage in osteoarthritis at 3T. J Magn Reson Imaging 2015;35(6):1422–1429. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Desrochers J, Yung A, Stockton D, Wilson D. Depth-dependent changes in cartilage T2 under compressive strain: a 7T MRI study on human knee cartilage. Osteoarthr Cartil. 2020;(June):1–10. 10.1016/j.joca.2020.05.012 [DOI] [PubMed] [Google Scholar]
  • 33.Hatchett A, Armstrong K, Parr B, Crews M, Tant C. The effect of a curved non-motorized treadmill on running gait length, imbalance and stride angle. Sports. 2018;6(3):58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Van Hooren B, Fuller JT, Buckley JD, et al. Is Motorized treadmill running biomechanically comparable to overground running? A systematic review and meta-analysis of cross-over studies. Sport Med. 2020;50(4):785–813. 10.1007/s40279-019-01237-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Cutcliffe HC, Davis KM, Spritzer CE. The characteristic recovery time as a novel, noninvasive metric for assessing in vivo cartilage mechanical function. Ann Biomed Eng. 2020;48:2901–2910. 10.1007/s10439-020-02558-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Kaab MJ, Ito K, Clark JM, Notzli H. Deformation of articular cartilage collagen structure under static and cyclic loading. J Orthop Res. 1998;16:743–751. [DOI] [PubMed] [Google Scholar]

RESOURCES