Impaired hemodynamics of the patella in patients with patellofemoral pain: A case–control study

Abstract Purpose According to the homeostasis model, patellofemoral pain (PFP) arises as a consequence of disturbed homeostasis of anterior structures of the knee due to vascular insufficiency. Near‐infrared spectroscopy (NIRS) allows to measure changes of concentrations (µmol/cm2) of (de)‐oxygenated hemoglobine (HHb and O2Hb). The aim was to study differences in patellar hemodynamics between patients and healthy controls. Methods Hemodynamics of patients (n = 30 [female = 20, age = 21.5, BMI = 22.9]) and controls (n = 30 (female = 18, age = 21.4, BMI = 22.4]) were evaluated for two activities (‘Prolonged Sitting’ and ‘Stair Descent’). Blinding for health status was implemented. Results During ‘Prolonged Sitting’, PFP patients exhibited smaller decreases in mean changes for HHb (PFP [M = −1.5 to −1.9], healthy controls [M = −2.0 to −2.3]) and O2Hb (PFP [M = −2.0 to −3.2], healthy controls [M = −3.4 to −4.1]). However, these differences were statistically non‐significant (p = 0.14–0.82 and p = 0.056–0.18, respectively). Conversely, for ‘Stair Descent’, PFP patients showed statistically significant smaller decreases in mean changes for HHb (PFP [M = −1.9, SD = 1.8], healthy controls [M = −2.5, SD = 1.7], p = 0.043) and O2Hb (PFP [M = −3.2, SD = 3.2], healthy controls [M = −4.9, SD = 2.7], p = 0.004). Conclusions The differences suggest potential impairment in patellar hemodynamics in PFP patients, providing support for the homeostasis model. Evidence‐based treatment strategies targeting patellar hemodynamics should be further refined and subjected to evaluation in clinical trials. Level of Evidence Level III.


BACKGROUND
The patellofemoral maltracking (PFM) model is a simultaneously acknowledged and debated model explaining the onset of patellofemoral pain (PFP) [36].Patellar maltracking refers to altered kinematics of the patellofemoral joint, arising from compromised muscle function and impaired soft tissue flexibility [26,41].Methodological concerns regarding the validity of this pathomechanical model, originating in the 1980s, have been expressed recently [17,25].
The homeostasis model represents an underexplored alternative model compared to the PFM model.According to this model, PFP arises as a consequence of the disturbance of homeostasis in both osseous and soft tissue within the peripatellar region due to supraphysiologic loading [12,36].Supporting evidence from observational studies comprises morphologic changes of the retinacula [24,41], including neovascularisation and hyperinnervation [37,38], and increased intraosseous water content and pressure of the patellar bone [20,28].Some authors propose that these changes are hypoxia-induced due to peripatellar anastomotic ring vascular insufficiency [15,37,39,42].However, only a few clinical studies evaluated blood flow of the patella in PFP patients, suggesting reduced drainage time of the venous system, reduced pulsatile blood flow and statistically nonsignificant differences in blood perfusion [2,19,29].
Near-infrared spectroscopy (NIRS), an optical, noninvasive method using a light-source anddetector allows a continuous bone hemodynamics assessment [5,27].The patellar bone has never been studied with NIRS, but the results of a recent study indicate that NIRS measurements are sufficiently reliable to compare real-time bone hemodynamics in PFP patients and healthy controls [34].This opens a window of opportunity to evaluate if vascular insufficiency of the peripatellar anastomotic ring plays a role in PFP patients.
The objective of the current study was therefore to study potential differences in patellar hemodynamics between PFP patients and healthy controls in clinically relevant positions.

Participants
This observational study was conducted according to the Declaration of Helsinki [46].The study protocol was approved by the Medical Research Ethics Committee of Amsterdam UMC location University of Amsterdam (NL77408.018.21) and has been registered (ISRCTN 90377123) prior to the start of the data collection.
A convenient sample was recruited from (1) a private physical therapy clinic, and (2) the HAN University of Applied Sciences and (3) the Radboud University in Nijmegen, the Netherlands between February and May 2022.A senior physical therapist (MO) screened subjects based on inclusion and exclusion criteria (Table 1) through history taking and standardised physical examination.
After inclusion, participants' demographic data including gender, age, body mass index (BMI), current smoking status and hours of sports per week were collected.Blood pressure was measured (Omron M6; TA B L E 1 Criteria for inclusion and exclusion.
• Previous or current clinical diagnosis of serious pathology (such as malignancy).• Previous or current other clinical diagnosis of specific knee conditions (such as patellar instability or dislocation, jumpers knee, meniscus tears, or other ligament injury).• Previous surgery (ankle, knee, hip, or lower back).

PFP patients
• Pain: ∘ experienced around and/or behind the patella.∘ aggravated by one or more of the following activities: Squatting, stair ambulation, jogging/running, hopping/ jumping.∘ lasting for three months or longer.∘ not as a result of trauma.
• Experience worst pain levels of at least 3/10 on a Visual Analogue Scale (VAS-W) during previous week.

Healthy controls
• Previous diagnosis of PFP.
• Complaints of ankle, knee, hip or lower back over the past 6 months requiring attention from a health care professional (physician, physical therapist), or resulting in missing more than one game, competition or training.
Omron Healthcare) [1].Furthermore, all participants completed the Tegner score and the Anterior Knee Pain Scale (AKPS) [23,44].The Tegner score evaluates the current physical activity level (0-10) [9,21], with higher scores representing higher activity levels.The Dutch Tegner score is reliable (ICC = 0.93-0.97)and valid with an internal consistency of r = 0.73-0.83[13].The AKPS is a 13-item questionnaire to quantify subjective symptoms and functional disabilities, with higher scores corresponding to fewer symptoms and disabilities [23].The Dutch version of the AKPS is reliable (ICC = 0.98) and valid with an internal consistency of r = 0.78-0.80[45].Participants were instructed to cease any sports activity 12 h before the NIRS measurement.

NIRS
Near-infrared light penetrates human tissue superficially up to a depth of 4 cm [4].The NIRS device used was the PortaLite (Artinis Medical Systems).The PortaLite is a continuous-wave device and the sensor (13.4 cm 2 ) consists of three light emitting diodes arranged at 30, 35 and 40 mm distance from the detector.The diodes transmit near-infrared light at 760 and 820 nm.These wavelengths have specific absorption characteristics for deoxygenated (HHb) and oxygenated hemoglobine (O 2 Hb).Considering the NIRS measurement depth (roughly half the sensor-detector distance [3]), the device assesses concentrations of HHb and O 2 Hb at depths ranging from 15 to 20 mm within the tissue.The assessment of HHb and O 2 Hb concentrations through NIRS relies on the adaptation of the Beer-Lambert law, incorporating the 'differential pathlength factor' (DPF) [3,11].This adjustment is specifically designed to accommodate measurements conducted within biological tissue.Since no research with NIRS is done to examine patellar hemodynamics, the DPF was estimated with the following equation: ( ) which takes the absorption (μ a ) and reduced scattering (μ s ) coefficients of the human skull into account [40], resulting in a DPF of 8.66.
Data were sampled at a frequency of 10 Hz.Moreover, NIRS measures no absolute concentrations of HHb and O 2 Hb, but only relative changes of concentrations compared to a baseline.Therefore, all measurements were preceded by a 3-min baseline and relative changes in HHb and O 2 Hb between baseline and experimental measurements (Δ) were calculated in μmol/cm 2 [3].The NIRS sensor was affixed to the skin with medical transparent double-sided adhesive tape (2181 disks; 3M) and placed according to the standardised procedure as described previously [34].The NIRS sensor was covered with an opaque cloth to reduce ambient light influence.Furthermore, the laboratory room had constant light and was maintained at a room temperature between 21 and 23°C.
'Prolonged Sitting' and 'Stair Descent' Patellar bone hemodynamics were evaluated during two clinically relevant activities for PFP patients: Prolonged sitting with the knees flexed and stair descent.
Activity 1 'Prolonged Sitting' After a 15-min rest period where the participant's knees were flexed in 20 degrees, a 3-min baseline was established in the same position, once stability of the NIRS signal was visually confirmed by examiner AW.Subsequently, the knees were flexed to 90 degrees for a period of 30 min (Figure 1a), and extended again to 20 degrees of flexion for 5 min.Accuracy of knee angles were evaluated with an extendable goniometer.

Activity 2 'Stair Descent'
Stair descent was replicated through the utilization of the previously developed Decline Step-Down Test (DSDT) [32].Following a 15-min rest period in standing position and stable NIRS signal (visual inspection), a 3min baseline was established, while the participant was standing on the decline step-down set-up with extended knees.Then, the participant performed a motion simulating stair descent to 45 degrees of knee flexion and maintained this for 1 min (Figure 1b).Accuracy of this knee angle was evaluated with an extendable goniometer.After a 3-min rest, the procedure was repeated for the other leg.The order of the starting leg was randomised, resulting in half of the participants started with the left leg and the others with the right leg.
Ophey et al. [34] reported moderate to almost perfect agreement of patellar NIRS measurements during Activity 1 'Prolonged Sitting' [34].For HHb concentrations, ICCs ranged from 0.51 to 0.75, with the smallest detectable change (SDC) ranging from 1.8 to 3.3.Similarly, for O 2 Hb concentrations, reported ICCs ranged from 0.56 to 0.95, with SDCs ranging from 0.6 to 1.5.These values are applicable to the 30 mm optode.Furthermore, Ophey et al. [34] noted moderate to substantial agreement during Activity 2 'Stair Descent'.For HHb concentrations, ICCs ranged from 0.50 to 0.68, with SDCs ranging from 0.5 to 1.0.For O 2 Hb concentrations, the reported ICCs ranged from 0.51 to 0.62, with SDCs ranging from 0.8 to 1.8.These values are pertinent to the 35 mm optode.It is important to highlight that the reliability of other optodes | 3 of 12 was found to be more variable, potentially attributed to the smaller sample size in the study by Ophey et al. [34].Consequently, our decision was to exclusively analyse hemodynamics of the patellar bone using the aforementioned optodes.

Visual Analogue Scale (VAS)
The VAS offers a continuous scale (0-10) from 'no pain' to 'extreme pain', thereby providing a subjective pain assessment [22].For Activity 1 'Prolonged Sitting', pain levels were assessed at baseline (after 1.5 min), every two minutes while maintaining seating position with the knees flexed, and during recovery period (after two and four minutes).During Activity 2 'Stair Descent', participants were queried about their pain level at baseline (after 1.5 min) and during the experimental measurement (after 0.5 min).

Patella skinfold and width
For a more comprehensive assessment of anthropometric characteristics of the anterior knee, prepatellar skinfold thickness and patella width were measured before placement of the NIRS sensor.Adherence to the guidelines of the International Society for the Advancement of Kinanthropometry (ISAK) was maintained for these anthropometric measurements [43].
Following training administered by a certified and ISAK (level 1) registered dietitian, skinfold thickness was measured at the centre of the patella using a skinfold caliper (Harpenden; Baty International), while patella width was measured at the broadest section of the patella using a slide caliper (Innovare; Cescorf).The execution of these measurements followed established protocols as previously detailed [34].

Lower limb range of motion (LLROM)
The LLROM test assesses the flexibility of soft tissues over multiple joints within the kinetic chain [31].Soft tissue flexibility of the anterior and lateral parts of the kinetic chain is evaluated by performing two passive movements [33], (1) maximal knee flexion with combined hip extension and upper body extension/rotation, lengthening the quadriceps and iliopsoas muscles (Figure 2a) and (2) maximal hip adduction with the same upper body position, lengthening the iliotibial tract, gluteal, abdominals and quadratus muscles in a combined fashion (Figure 2b).The cumulative score for total lower limb range of motion (total ROM) is derived by adding the scores for knee flexion and hip adduction.The LLROM test, previously described and established as reliable [31], was performed subsequent to the completion of NIRS measurements.

Blinding
Blinding with respect for health status (PFP patient or healthy control) was implemented.Examiner MO was tasked with participant inclusion and exclusion.Examiner AW conducted the measurements without knowledge of the participant's health status.

Sample size
Due to the absence of available literature on anticipated effect sizes in NIRS studies focusing on the hemodynamics of the patellar bone, our study adopted a pragmatic approach.The sample size was determined to be as large as possible within the constraints of our research group's available resources.We aimed to recruit a convenient sample comprising 30 individuals with PFP and 30 healthy controls for the study.

Statistical analysis
The normality of data distribution was assessed through visual inspection and the Shapiro-Wilk test.Descriptive statistics, including means (M), standard deviations (SD) for continuous variables and percentages (%) for dichotomous variables, were employed to participants' baseline characteristics.PFP patients and healthy controls were not matched for any baseline characteristic.Continuous baseline characteristics were compared between PFP patients and healthy controls using student's t test (normally distributed data) or Mann-Whitney U test (not normally distributed data).Fisher's exact test was utilized for analysing differences in categorical baseline characteristics.
The NIRS sensor was placed on both knees of each participant, and knees were included separately for the statistical analysis.For PFP patients, only the symptomatic knees were included, while the nonsymptomatic knees were excluded from the statistical analysis.In contrast, for healthy controls, data from both knees were included in the analysis.
For  using SPSS (version 28.0) with a significance level set at p < 0.05.

RESULTS
A total of 61 subjects underwent screening for participation, with sixty ultimately meeting the inclusion criteria.Unfortunately, one subject had to be excluded due to personal time constraints (Figure 3

Activity 1 'Prolonged Sitting'
While PFP patients exhibited smaller decreases in mean changes of HHb and O 2 Hb in PFP patients compared to healthy controls during Activity 1 'Prolonged Sitting', these differences between groups were not statistically significant for all parameters, with p values ranging from 0.14 to 0.82 and from 0.06 to 0.18, respectively (Table 3).Plotted means of HHb and O 2 Hb during Activity 1 'Prolonged Sitting' are presented in Figure 4.

DISCUSSION
This was the first study to evaluate hemodynamics of the patella in positions clinically relevant for PFP patients.The main finding was smaller decreases in concentrations of (de-)oxygenated hemoglobine (HHb and O 2 Hb) of the patellar bone during Activity 2 'Stair Descent' among PFP patients when compared to healthy controls.This suggests a potential impairment in hemodynamics of the patella in PFP patients, and opens up a new window of opportunities to develop more effective evidence-based treatment strategies for PFP patients.
In the current study, nearly half (43%) of PFP patients reported experiencing some knee pain during Activity 1 'Prolonged Sitting'.This aligns with previous reports indicating that 55% of PFP patients encountered 'problems with prolonged sitting', as assessed by item 8 of the AKPS [10].The current study is the first to evaluate patellar bone hemodynamics during an extended sitting period.Although decreases in concentrations of HHb and O 2 Hb were observed to be smaller in PFP patients during Activity 1 'Prolonged Sitting', the between-group comparisons did not reach statistical significance, possibly due to insufficient power.Näslund et al. [29] found a reduction in patellar bone blood flow in PFP patients measured by photoplethysmography (PPG) during 5 min of knee flexion in a supine position.Since PPG measures the amplitude of pulsatile blood flow in bone tissue based on attenuation changes in the reflected signal [27], it is biologically plausible that a reduced amplitude in the PPG signal corresponds to smaller decreases in concentrations of HHb and O 2 Hb when evaluated with NIRS.
This study marked the first real-time monitoring of pain during prolonged sitting in PFP patients using a VAS.As the sitting period extended, an increasing number of patients began reporting knee pain, with pain intensity gradually escalating during sitting (Figure 5).Interestingly, following the initial decrease in the first 10 min of sitting with the knees flexed, concentrations of HHb and O 2 Hb also showed a gradual increase during the sitting period (Figure 4).This lends support to the vascular insufficiency theory and suggests the presence of a homeostatic pain mechanism as an explanation for the occurrence of sitting pain in PFP patients [36].
In the current study, a significant majority (83%) of PFP patients reported pain during Activity 2 'Stair Descent', aligning with the previously reported 88% of  PFP patients experiencing pain during stair ambulation [35].The simulated stair descent in this activity induces an eccentric action of the quadriceps muscles, involving simultaneous contraction and stretch, leading to increased patellofemoral contact forces [6].Notably, the current study is the first to investigate patellar bone hemodynamics during active loading of the patellofemoral joint.Arnoldi [2] hypothesised that quadriceps muscle contraction amplifies compression at the base of the patella, impacting venous outflow of the patella's anastomotic ring.The absence of quadriceps muscle contraction during Activity 1 'Prolonged Sitting' may result in less vascular insufficiency, potentially explaining fewer differences in patellar hemodynamics between PFP patients and healthy controls during this activity.Given the cross-sectional nature of the current study, it is not possible to draw conclusions about impaired patellar hemodynamics as a cause for the onset of PFP.However, the association between impaired hemodynamics and ongoing anterior knee pain in PFP patients appears plausible.A previous study observed an increase in HHb and O 2 Hb concentrations of the patella during venous occlusion with a thigh cuff in healthy controls [34].Consequently, smaller decreases in HHb and O 2 Hb concentrations in symptomatic knees of PFP patients (as identified during activity 2 'Stair Descent') might be linked to such venous occlusion.In essence, a reduced decrease in concentrations of (de-)oxygenated hemoglobine may signify venous occlusion of the peripatellar anastomotic ring in PFP patients.
The posterior surface of the apex patellae serves as the most important exit point of the venous system in distal posterior direction, draining into the saphenous and popliteal vein [2].Additionally, clusters of veins exit the basis patellae in proximal direction, with veins leaving the anterior, medial and lateral aspect of the patella also documented [2].We hypothesise that venous outflow is compromised not only by quadriceps muscle contraction but also by lower soft tissue flexibility.This aligns with Arnoldi's early 1990s proposition [2], where he discussed reduced venous outflow as a potential pain mechanism of PFP, involving 'extraosseous compression by the skin and subcutaneous tissue during stretching of these structures'.In the current study, we observed lower soft tissue flexibility (LLROM) in PFP patients, and interventions targeting LLROM in PFP patients led to an immediate reduction of pain and disability in the short-term [33].
Van der Heijden et al. [19] investigated patellar bone perfusion using dynamic contrast-enhanced magnetic resonance imaging (MRI) in PFP patients and healthy controls.This MRI investigation was conducted with the participant's knees in extension and unloaded in a supine position.This specific positioning could account for their observation of no differences in patellar bone perfusion between PFP patients and healthy controls.
The observed prepatellar skinfold thickness in PFP patients and healthy controls was 9.1 and 7.5 mm, respectively, resulting in skin thickness (skinfold divided by two) of 4.6 and 3.8 mm, respectively.When the superficial layer (e.g., skin) is less than 4 mm, it has a nonsignificant confounding contribution on NIRS measurements [14].Only when the superficial layer exceeds 13 mm does it predominantly influence the measurements [14].Since the reported thickness of the skin of both PFP patients and healthy controls was close to or below the recommended 4 mm, we conclude that skin thickness did not significantly affect our measurements.Additionally, the difference in prepatellar skinfold thickness between PFP patients and healthy controls was statistically significant.PFP patients had a lower Tegner score, participated in sports fewer hours per week, and had a higher BMI, although the differences of the latter two were not statistically significant.Thus, one explanation of the observed difference in skinfold thickness may be a reduced activity pattern in PFP patients with associated changes in body composition.Although lower activity levels and increased BMI of PFP patients have been reported in the literature [16,18], the design of the current study does not allow for causal inferences.Additionally, the skin thickness difference between the two groups was <1 mm, which may (at least partially) have been the result of measurement error.Before the experiments, we were trained according to the ISAK criteria, and for measurements, we used the highquality Harpenden caliper.Thus, we aimed to minimise measurement variability.Given that the observed difference between both groups was <1 mm, and skin compression does not affect NIRS measurements [34], we assumed that this difference did not affect the NIRS measurement in the current study.
Näslund et al. [30] demonstrated an increase in patellar bone blood flow during contraction of the quadriceps muscle in healthy participants.Given that contraction of the quadriceps muscle increases patellar bone blood flow in healthy individuals and venous outflow is decreased in PFP patients, the commonly recommended quadriceps-oriented exercise therapy [8] to reduce PFM may potentially exacerbate symptoms of PFP, as reported by many clinicians.Näslund et al. [30] also showed that intermittent contractions (with 2 s of rest between repetitions) reduce patellar bone blood flow in healthy participants.Therefore, we propose that a multimodal treatment strategy, considering impaired hemodynamics of the patella, should include interventions to improve LLROM involving soft tissue, hip-oriented exercises, intermittent quadricepsoriented exercise therapy and education on homeostatic pain mechanisms.Assessing the impact of such a multimodal treatment strategy on the improvement of patellar hemodynamics and its effectiveness in reducing pain and disability should be explored in future research.
We acknowledge certain limitations in the current study.The pragmatic enrolment of participants, recruiting 30 PFP patients and 30 healthy controls, may have led to an inappropriate sample size.Specifically, for Activity 1 'Prolonged Sitting', the failure to detect statistically significant differences may be attributed to only 43% of PFP patients experiencing pain during prolonged sitting.A post hoc power analysis indicated that a sample size of n = 110 for each group would be needed to reach a power of 0.80.Therefore, future investigations on patellar hemodynamics, especially those assessing pain during prolonged sitting, should consider defining pain during this activity as an inclusion criterion and aim to increase the number of participants accordingly.The findings from the current study can serve as a basis for calculating sample sizes in future research.
We acknowledge another limitation related to the design of Activity 2 'Stair Descent'.While this activity was intended to simulate descending stairs, the necessity of a 1-min quadriceps muscle contraction in 45 degrees of knee flexion emerged as a requirement for generating reproducible NIRS measurements.
Unfortunately, this aspect diminishes the validity of the set-up in mimicking a real-world situation of descending stairs.Given the sensitivity of the NIRS device employed in the current study, it is improbable that a more dynamic simulation of stair descent would yield NIRS data of satisfactory quality.

CONCLUSIONS
The current study indicates potential impairment in the hemodynamics of the patella in PFP patients when compared to healthy controls.The findings provide support for the homeostasis model as a contributing factor to the existence of PFP, possibly attributed to vascular insufficiency in the peripatellar anastomotic ring.Recommendations for evidence-based treatment strategies targeting the hemodynamics of the patella should be further refined and subjected to evaluation in clinical trials.

F
I G U R E 1 (a) Activity 1 'Prolonged Sitting' and (b) activity 2 'Stair Descent' (absence of the opaque cloth to visualise the set-up, during measurements the sensor was covered).
Activity 1 'Prolonged Sitting' four changes in HHb and O 2 Hb were assessed: (1) change between baseline and first 10 min of sitting with the knees flexed (ΔBas_Sit0-10), (2) change between baseline and second 10 min of sitting with the knees flexed (ΔBas_Sit10-20), (3) change between baseline and third 10 min of sitting with the knees flexed (ΔBas_Sit20-30), (4) change between third 10 min of sitting with the knees flexed and five minutes of recovery time (ΔSit20-30_Rec).For Activity 2 'Stair Descent', the change in HHb and O 2 Hb between baseline and 45 degrees of knee flexion (ΔBas_Step) was assessed.Mean and standard deviation in HHb and O 2 Hb changes between baseline and experimental measurement were calculated for the 30 mm optode during Activity 1 'Prolonged Sitting' and the 35 mm optode during Activity 2 'Stair Descent'.The first and last three seconds were removed to mitigate the influence of movement artifacts in the data.Differences between PFP patients and healthy controls were analysed using student's t test (normally distributed data) or Mann-Whitney U test (not normally distributed data).Additionally, the changes of HHb and O 2 Hb were visualised with plots.Effect size (ES) was calculated using Cohen's d to present the magnitude of the change between baseline and experimental measurement, with an ES of 0.2 considered small, 0.5 medium and ≥0.8 large[7].The NIRS measurements were conducted using Oxysoft, version 3.2.72 (Artinis).Subsequently, the Oxysoft files underwent conversion into MATLAB files (R2020a version 9.8.0.1323502;Mathworks).In MA-TLAB, a moving average filter was applied to the raw values, and changes between baseline and experimental measurements were calculated.The resulting outcomes were then entered into Excel (Microsoft Office version 16).Statistical analyses were carried out F I G U R E 2 (a) Assessment of lower limb range of motion of knee flexion and (b) hip adduction (Ophey and colleagues).

Activity 2 '
Stair Descent' The decreases in mean changes of HHb and O 2 Hb in PFP patients were statistically significantly smaller than those in healthy controls, with (t[106] = −2.05;p = 0.043, d = 0.39) and (t[106] = −2.95;p = 0.004, d = 0.57), respectively (Table 3).Plotted means of HHb F I G U R E 3 Flowchart of the inclusion process.and O 2 Hb during Activity 2 'Stair Descent' are illustrated in Figure 4.

7 ) − 2 .
Abbreviations: df, degrees of freedom; ES, effect size (Cohens 'd'); HHb, deoxygenated hemoglobine; O 2 Hb, oxygenated hemoglobine; SD, standard deviation.a Change between baseline and first 10 min of sitting with the knees flexed.b Change between baseline and second 10 min of sitting with the knees flexed.c Change between baseline and third 10 min of sitting with the knees flexed.d Change between third 10 min of sitting with the knees flexed and 5 min of recovery time.e Change between baseline and 45 degrees of knee flexion.*p < 0.05; **p < 0.01.

F
I G U R E 5 VAS during activity 1 'Prolonged Sitting'.
TA B L E 2 Baseline characteristics.
TA B L E 3 Changes in HHb and O 2 Hb (in µmol/cm 2 ) during activity 1 'Prolonged Sitting' and activity 2 'Stair Descent'.