Patellofemoral pain is one of the most common musculoskeletal disorders and affects one in four youth and adults.1 The prognosis is not always favourable with over 50 % of people reporting pain up to 8 years after diagnosis.2 Patellofemoral pain negatively impacts daily activities, physical activities and causes withdrawal from sports participation and overall quality of life.3

Patellofemoral pain is a clinical diagnosis, characterized by non-traumatic onset and diffuse anterior knee pain that is aggravated by knee-loading activities.4 The most reported painful activities are squatting, stepping-up and down, and prolonged sitting and/or kneeling.5 Clinicians often use these pain-provoking activities during the examination and track them over a course of treatment.6,7,8,9 In clinical practice, functional status can be assessed using both self-reported outcome measures and performance-based evaluations. While self-reported measures capture an individual’s perceived limitations in the context of their condition, performance-based tests quantify these limitations through standardized tasks such as squatting, stair ascent/descent, and single-leg hops.10,11

Despite the common use of these clinical functional tests, the reporting guide for patellofemoral pain (REPORT-PFP)12 does not include these condition defining activities. The REPORT-PFP recommends reporting a measure of pain severity and a condition-specific patient reported outcome–suggesting the Anterior Knee Pain Scale (Kujala Patellofemoral Scale)13 and the Knee Injury and Osteoarthritis Score Patellofemoral subscale (KOOS-PF)14 for the latter. This discrepancy highlights an important gap: although tests such as squatting and step-up/down are widely employed to assess functional performance in individuals with PFP, their relationship with validated PROMs remains poorly understood. Only a few studies have explored the association between clinical functional tests and PROMs in this population. For example, Nunes et al.15 investigated the relationship between stair climbing, chair stand, step-down, forward and side hop tests with Kujala and KOOS-PF scores, and found that only stair climbing time was moderately correlated with lower Kujala scores. Similarly, Zamboti et al.9 found significant association only between the 30-seconds sit-to-stand test and AKPS, despite evaluating other tests, including sitting-rising, stair-climb, stair-descent, and the six-minute step tests.

Function-based assessments offer valuable, low-cost, and time-efficient means to complement patient-reported data, providing a more comprehensive understanding of functional impairment.6,10 Understanding whether these commonly used clinical functional tests relate to patient-reported improvements would help clinicians monitor progress and adapt interventions more effectively throughout the treatment. Therefore, the aim of this study was to investigate how changes in clinical functional tests of step-up, step-down and squat are associated with changes in patient-reported outcomes among individuals with PFP.

Of the 218 participants with data available, 61 were excluded because participants achieved a maximum of 25 repetitions on step-up, step-down or squat tests at both the baseline and 12-week follow-up (Fig. 1), without onset or increase in symptoms, representing a ceiling effect. The number of participants who reached 25 repetitions for each test was as follows: only step-up: 24; only step-down: 3; only squat: 3; both step-up and down: 18; both step-up and squat: 6; and all three tests: 7 participants.

Fig. 1.

Participant flow chart showing exclusions for this secondary analysis (adapted from the original RCT16).

The remaining 157 participants were included in the data analysis. Descriptive baseline data characteristics for the entire sample are summarized in Table 1. In addition, Table 2 summarizes data from the follow-up measurement, mean difference and the standardized response mean.

Tables 3 and 4 show the results for the overall model significance and tests of between-subjects effects from the multivariate regression analysis, respectively. The multivariate tests showed that among the independent variables, step-up presented a significant effect (F = 2.515, p = 0.027), with a contribution of 13.8 % (Partial Eta Squared = 0.138) on the dependent variables. On the other hand, step-down (F = 1.566, p = 0.166) and squat (F = 1.147, p = 0.342) did not have significant effects on the dependent variables, with Partial Eta Squared values of 0.091 and 0.068, respectively. This indicates that these variables had a marginal impact on the dependent variables, explaining a smaller proportion of the variation.

The between-subjects effects analysis examined how each independent variable contributed individually to the variation in the dependent variables. The proportion of variance explained by the multivariate regression models ranged from 11.2 % to 29.6 %, with the KOOS Sport and Recreation model showing the highest explanatory power (29.6 %), followed by Kujala (21.6 %), KOOS ADL (18.1 %), KOOS Pain (15.9 %), KOOS QoL (15.3 %), and KOOS Symptoms (11.2 %). Among the predictor variables, performance in the step-up test was the most consistent and significant contributor across models. It was significantly associated with KOOS subscales of Symptom (p = .021; 5.3 % of variance explained), Pain (p = .018; 5.5 %), Sport & Recreation (p = .001; 11.1 %) and Quality of Life (p = .046; 4.0 %). The step-down test showed significant effects on Kujala (p = .044; 4 %) and KOOS Sports & Recreation (p = .030; 4.7 %). The squat test did not show significant effects on any of the dependent variables (all p > 0.20). Both the multivariate analysis and the between-subjects effects analysis agree in identifying step-up as the most significant factor, with a stronger impact on the set of dependent variables.

For the binary logistic regression, only the step-up test showed a significant association (p = .001) between its change scores and the dichotomized GROC, explaining 16.0 % of the variance.

Out of 218 participants from a previous RCT comparing foot orthoses and hip exercises, we had complete data from 157 to include in this secondary analysis. We found that step-up test was associated with all PROMs, except KOOS ADL subscale and Kujala. The relationship was positive, meaning that an increase in step-up was associated with improvements in these PROMs. The squat test appears to be less informative and can’t be recommended based on the current data. Our findings suggest that the step-up test may be a useful in-clinic measure for tracking improvements in function.

Treatment with either hip exercises or a combination of foot orthoses and foot exercises improved all the outcomes reported herein. The association between the improvements in step tests and patient-reported outcomes could be explained in several ways. One possibility is that the treatment improved the participant’s neuromuscular system which may account for the observed improvements in clinical functional tests and patient-reported outcomes. A previous study also reported similar improvements in functional tests following a physical therapy intervention; which consisted of mainly neuromuscular exercises targeting the gluteal and quadricep muscles.19 The effect of the treatment on muscle capacity might account for these findings, as another study reported that reduced hip muscle strength is associated with lower functional performance (fewer step-down repetitions and shorter forward hop for distance).15 The pain reduction and improvements in physical function observed in our study may also be explained by neurophysiological mechanisms, such as exercise-induced hypoalgesia and the release of endogenous opioids.30

The finding that step-up and down tests similarly improved over the 12-weeks alongside the finding that only step up was associated with PROMs requires exploration. This stronger association with the step-up test may be because it places lower patellofemoral joint reaction force and stress compared to other tasks like the forward step down.31 Additionally, the clinical functional tests used are pain threshold tests, and participants were able to perform nearly twice the number of repetitions on step-up at baseline compared to the step-down and squat tests. This suggests that improvement in their capacity to perform the test was more substantial, given their already higher baseline performance. It is also possible that, due to increased load during step-down, participants felt higher pain intensities during this test, meaning that even with similar change scores, the improvement had a smaller effect on the PROMs. It is also important to note that most of the 61 participants excluded for reaching the maximum of 25 repetitions did so on the step-up test. Therefore, it is possible that their improvements in change scores may have been greater but were not captured in the analysis.

Interestingly, the change score for the squat test did not contribute to any of the models. One possible explanation could be linked to the nature of the movements. While the squat is an important functional movement, in questionnaires like the KOOS and Kujala, many questions are associated with activities more directly related to locomotion and daily functionality (e.g., walking, climbing stairs, carrying weight, or changing/moving positions), which may align more closely with the step-up and step-down movement than with the squat.

The models with clinical functional tests included in the analysis explained from 11 to 30 % of the variance, with generally weak to moderate association, consistent with the multifactorial nature of PFP and highlighting that changes in functional tests explain only part of the variance in patient-reported improvements. There are other variables not included in the analysis that play a role in explaining changes in the PROMs. For example, psychological factors, which we did not assess in this study, might influence changes in PROMs. As an example, kinesiophobia is associated with self-reported pain and disability in women with PFP, while patellofemoral joint loading variables are not.32 Other studies have shown a positive association between kinesiophobia, pain catastrophizing, anxiety, depression and fear of movement with pain, and a negative association with function, meaning that higher levels of these psychological traits are associated with increased pain and reduced function.33–35 Self-efficacy, in particular, has been identified as a strong predictor of successful outcomes for patients with persistent pain, irrespective of the intervention provided.36 Furthermore, higher self-efficacy is linked to better function and health-related quality of life.37 Thus, it is possible that in addition to improving the number of repetitions in the clinical functional tests, the treatment provided by the RCT also positively influenced the patient’s self-efficacy (their belief in their ability to perform an activity or overcome a challenge)36 and contributed to improved function in clinical tests and better self-reported outcomes. This is all purely speculative, as this is the first study testing the association between change scores in the pain threshold tests of squatting and stepping-up and down with PROMs in individuals with PFP.

Since functional test performance may be dependent on factors such as neuromuscular and biomechanical factors,15,38 future studies should explore whether changes in these domains are associated with improvements in PROMs. For example, King et al.,39 investigated lower-limb joint moment impulses during walking and their association with KOOS subscales in people with patellofemoral osteoarthritis. They reported that these associations varied depending on the anatomical plane, joint compartment involved, and disease severity, highlighting the complex relationship between joint mechanics and symptoms, that may inform treatment strategies. Furthermore, approximately 30 % of the data were excluded due to ceiling effects, suggesting that the functional tests employed — particularly the step-up— may not have been sufficiently challenging or demanding38 for individuals with higher baseline function. This limited sensitivity likely contributed to the weak to moderate associations observed, indicating that while the step-up may have limited discriminatory capacity at the higher end of functional performance, it retains clinical utility for detecting meaningful differences in patients with lower to moderate functional levels. In clinical practice, functional tests should avoid substantial ceiling effects (i.e., many participants achieving the highest possible scores) or floor effects (many scoring the lowest), and should demonstrate sensitivity to change and responsiveness.40 Future studies should consider more demanding and provocative functional assessments (i.e., single leg vertical and lateral hopping tests), and investigate what constitutes a meaningful change for commonly used tests.40 Additionally, test selection should be informed by the physical activity level of the population being assessed,38 to better reflect the heterogeneity in functional capacity among individuals with PFP and improve the detection of clinically relevant associations.

There are limitations to this study that should be considered. We stopped the tests after a maximum of 25 repetitions was reached without the onset of pain. This decision aimed to standardize data collection by limiting the duration of the testing and avoiding exacerbating the symptoms. To mitigate this ceiling effect, we excluded participants who achieved 25 repetitions at both the baseline and 12-week follow-up. This exclusion was necessary because when a significant proportion of participants reaches the maximum possible score (ceiling), the clinical test of function may not effectively measure change in performance.41 We did not measure knee flexion angle and there is a possibility that the amount of knee flexion achieved by each individual participant, and thus the joint load and possibly pain onset, would be different between participants. We suggest that this does not impact on our findings because we analysed the change in numbers of repetitions of steps/squats, which would mitigate inter-subject variability in this regard. Whilst the potential impact of test order should be acknowledged, as pain and fatigue could affect performance, given the step-up test was first performed, we proposed test-order influence to be minimal. This study involved an exploratory secondary analysis of data from a RCT. While the sample size was determined based on the primary RCT objectives, we propose that the number of participants was adequate for the regression models, which included only three independent variables per model.

Improvements in the clinical functional tests of step-up and step-down, but not in the squat test, were associated with improvements in PROMs. The step-up test showed the strongest association. Clinicians using these tests as in-clinic are likely to expect changes reflected in PROMs. This reinforces the clinical value of combining in-clinic functional tests with PROMs to monitor and guide patient’s progress throughout rehabilitation.

The authors declare no competing interest.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

We confirm that ethical approval was granted by the University of Queensland Medical Research Ethics Committee (2013000981) and by the Ethics Comittee in the North Denmark Region (N-20140022), and the participants gave informed consent to the work.