Study design

This study used a repeated measures mixed study design consisting of a combination of between and within subject factors [21]. The between-subjects independent variable was group allocation (i.e., control group and PFP group), where group assignment was non- randomised (due to presence/absence of PFP). Within-subjects independent variables included time (repeated measures) and extremity (injured [or non-dominant] and uninjured [or dominant] limb).

Participants

A total of 35 participants volunteered for the study, of which 17 were part of the control group (female [n = 10] and male [n = 7]) and 18 were part of the PFP group (female [n = 15] and male [n = 3]. Given the requirements of the study design, a minimum total sample size of 24 was calculated based on a repeated measures analysis of variance (ANOVA) design that incorporated (i) a within-between interaction, (ii) a moderate effect size (f = 0.25), (iii) a type-I error rate of 5% (α = 0.05), (iv) a type-II error rate of 20% (β = 0.20), (v) 2 groups (control [n = 17] and PFP [n = 17]), (vi) 2 repeated measurements, (vii) an anticipated dropout of 10%, and (viii) a minimum expected correlation of 0.50 among repeated measurements [22]. Accounting for a potential drop-out of 20%, the minimum sample size for adequate statistical power was 29 participants. Two participants were lost to follow-up, both in the PFP group (female [n = 1] and male [n = 1]), resulting in a final sample size of 33 participants.

Participants were recruited by dispersing electronic flyers via social media platforms (e.g., Facebook, WhatsApp), inviting prospective participants to contact the researcher for an information letter which outlined the details of the study. Participants who were willing to take part in the study were screened for eligibility (see inclusion and exclusion criteria below) and received an informed consent form to read through and were given 72 hours to sign and submit the form to an independent person affiliated with the study.

The inclusion criteria for the PFP group of this consisted of the following: (i) aged between 18 and 35 years, (ii) could be male or female, (iii) had to have retropatellar and/or peripatellar pain aggravated by at least one activity that loads the patellofemoral joint during weight bearing on a flexed knee, such as squatting, stair climbing, jogging/running, and hopping/jumping, and (iv) had to participate in some form of rehabilitation program. The control group had similar inclusion criteria, with the exception for points (iii) and (iv).

The specific exclusion criteria for both the PFP and control group consisted of the following: participants should not (i) have had previous patellar dislocation or subluxation, (ii) have had previous injury or surgery to the knee, and (iii) have had recent (within the last 6 months) injury to the lower limbs (ex. Achilles tendinopathies, ankle sprain, etc.). All data collection occurred between 9^th January 2023 to 22^nd September 2023.

Ethics approval was granted by the Human Research Ethics Committee of the university (NWU-00163-22-A1), and all participants completed the informed consent forms before participation. An independent researcher not affiliated with the study served as an independent witness and collected the signed consent forms. All ethical procedures conformed to the requirements of ethical conduct set forth in the Declaration of Helsinki.

Instruments

A general demographic questionnaire was used to obtain the participant’s contact details, e-mail address, age, sex, involved limb, dominant limb, previous injuries and PFP symptoms. The anterior knee pain scale (AKPS), also known as the Kujula patellofemoral scoring system, was completed electronically to capture knee-related pathologies where the total score out of 100 was captured. The AKPS was the chosen questionnaire, as it was developed specifically for evaluation of pain and disability in individuals with PFP [23], showed high test-retest reliability (intraclass correlation coefficient [ICC] = 0.95), and exhibited moderate responsiveness to clinical change, which implies that the score will likely reflect meaningful changes in a patient’s condition over time [24]. Body mass and stature were measured to calculate body mass index (BMI) [25]. Body mass was measured with an electronic scale (Seca 874, Seca, Germany) to the nearest 0.1 kg and stature with a portable stadiometer (Holtain Ltd., U.K.) to the nearest 0.01 m. A cycle ergometer (Wattbike Pro, Wattbike Ltd, Nottingham, UK) was used for a ten-minute warm-up before the testing on the ID for optimal muscle performance and reduced risk of injury. The ID (Cybex II, CSMi, Stoughton, MA, USA) was used as the gold standard for evaluating isometric muscle strength and a SG (Crane & Hanging Scale, Micro Mini CS300, Border Scales & Labels) served as the low-cost alternative to measure isometric quadriceps strength. Strength was evaluated with the knee joint at 60° of knee flexion, where peak force of the quadriceps is usually generated [26]. A digital goniometer (EasyAngle, Meloq AB, Stockholm, Sweden) was used to measure the knee angle in all instances to ensure true validity and replicability.

Procedures

The demographic questionnaire was completed first, and only at the baseline testing. Thereafter the participants completed the AKPS questionnaire verbally, where the score was calculated out of 100 and recorded electronically. Stature and body mass were measured during barefoot standing and with minimal clothing. Lower leg length was measured from the lateral condyle of the femur to the lateral malleolus of the tibia to use as the lever length in the formula to convert force to torque for the SC measurements. Prior to testing participants completed a ten-minute warm-up on a cycle ergometer at lowest resistance and a comfortable speed (rating of perceived exertion [RPE] < 2 on the modified Borg scale). Participants completed baseline testing in two sessions separated by 24 hours, and then repeated the same testing eight weeks later. During the first session, isometric strength of the quadriceps muscles was measured first with the ID, and thereafter with the SG. During the second session, only the testing on the SG was completed for the reliability analysis. Reliability analysis was done both at baseline testing and at follow-up testing conducted after 8 weeks to ensure a stronger reliability score.

For the evaluation of isometric quadriceps strength on the ID, all variables related to the set-up were recorded to replicate the exact position for follow-up testing (Fig 1A). The dominant limb (defined as the preferred kicking limb) in the control group and the uninjured limb in the PFP group were measured first. The participants were placed in a seated position, stabilized with upper body straps and an upper leg strap just above the knee joint. The knee was positioned at a 60° angle, using the machine angle provided by the dynamometer. The lateral femoral epicondyle was aligned with the axis of rotation of the dynamometer, and the resistance pad was positioned anterior to the distal tibia just superior to the lateral and medial malleoli. A gravity correction was performed to account for any potential additional torque induced on the attachment. The participant completed a warm-up round consisting of three repetitions, with the instruction to do one repetition at approximately 25%, 50% and 75% of perceived maximum effort before commencing of the test to assist with familiarisation of the test. A 10 second rest period was allotted between trial repetitions. A one-minute rest period was granted after the warm-up, after which the participant completed repetitions of five second maximal extension contraction with ten seconds of rest between repetitions. Verbal encouragement was given throughout the whole procedure. Pain levels were monitored by asking the participant to rate pain on a scale of 1–10 throughout the procedure (where 1 = pain free, 2 = very mild, 3 = discomforting, 4 = tolerable, 5 = very distressing, 6 = intense, 7 = very intense, 8 = utterly horrible, 9 = excruciating unbearable, 10 = unimaginable unspeakable) [27], and participants were permitted to stop the test if pain levels were above bearable levels (≥4). For the evaluation of isometric quadriceps torque using the SG, all positional measurements were recorded for replicability. Participants were seated on a standard chair, and the SG was attached to a strap that was attached to a fixed surface behind the chair (Fig 1B). The strap was fastened to the participant’s ankle just above the lateral and medial malleoli where the distance between the lateral malleolus and the start of the ankle strap was measured to more accurately calculate the lever arm length in the evaluation of torque from the SG. The straps were adjusted until the knee was at 60° of flexion, measured with a digital goniometer. The participant completed the same warm-up and testing protocol as with the ID. The tests were repeated in the second session for the reliability analysis, and all procedures were repeated after 8-weeks. The conversion of raw kilogram values from the SG were converted to torque values using the following equation: [Eq 1]

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1.

Experimental setup for the isokinetic dynamometer (panel A) and the strain gauge (panel B) for the right leg of a representative participant. The inset of panel C is a close-up of the commercially available strain gauge used in the study.

https://doi.org/10.1371/journal.pone.0298570.g001

Statistical analyses

All data were evaluated for normality using the Shapiro-Wilk test, with deviations from normality being accepted at p < 0.05. All data are presented as mean ± standard deviation (SD) unless otherwise stated. To determine the reliability of the SG, the intraclass correlation coefficient (ICC, two-way mixed effects, absolute agreement) between two measurements on separate occasions was used. The ICC values were interpreted as follows: poor: < 0.50; moderate: 0.50–0.75; good: 0.75–0.90; excellent: >0.90 [28]. Additional measures of reliability included: (i) typical error of measurement (TEM) (Eq 1), (ii) TEM% (Eq 3), and (iii) clinically meaningful difference (CMD) (Eq 4) [29]: [Eq 2] [Eq 3] [Eq 4]

The TEM% scores were qualitatively interpreted as: excellent:<5%; good: 5–10%; poor: >10%. The CMD was used to evaluate the smallest detectable difference that would need to be overcome to conclude that a true change has occurred. A generalised linear model was used to evaluate mean differences between groups (2 levels: Con vs. PFP), and devices (2 levels: ID vs. SG) where each participant was treated as a random effect in the model. Post-hoc analyses entailed the use of paired sample t-tests with a Holm correction to adjust for multiple comparisons. Standardised mean differences were calculated as Hedges’ g, the magnitude of which was qualitatively interpreted as: trivial: <0.2; small: 0.2–0.6; moderate: 0.6–1.2; and large: > 1.2 [30].

The concurrent validity was determined by using linear regression where the Pearson correlation coefficient (r), coefficient of determination (r²), standard error of the estimate (SEE), and the slope of the regression lines were evaluated. The magnitude of the correlation coefficients were qualitatively interpreted as follows: negligible: 0.00–0.10; weak: 0.10–0.39; moderate: 0.40–0.69; strong: 0.70–0.89; and very strong: 0.90–1.00 [31]. Bland-Altman analyses were used to determine the systematic bias between the SG (reference measure) and the ID (criterion measure) [32]. For both the regression and Bland-Altman analyses, the point estimates were evaluated for potential outliers using Cook’s distance where potential outliers (PO) were flagged when the Cook’s distance exceeded a given threshold calculated as: 4/n (where n is the number of observations) [33]. All statistical analyses were completed using R [34].

Reliability

The SG showed excellent relative and absolute reliability in both the control and PFP groups (ICC>0.90 and TEM <10%) in all instances apart from the uninjured limb of the PFP group. The large ICC values obtained in the present study compared favourably with those from a separate study where belt-stabilisation was used with an HHD (ICC = 0.62 to 0.96) [35] as well as study focusing on the development of a novel affordable strain gauge (ICC = 0.99) [20]. Similarly, the TEM% scores of the present study were substantially better those reported by Martins et al. [35] who recorded TEM% of 12% for a belt-stabilised HHD and Chamorro et al. [18] who yielded values ranging between 4–15% for measurements without stabilisation, both for knee extension strength. The ICC point estimates showed narrow confidence intervals in which the upper and lower limits fell within the margin defined as excellent in most instances. However, longer confidence intervals were observed in the PFP group during the initial set of testing indicated that the PFP group had greater variability in torque measures in the injured as well as the uninjured limbs compared to the control group. Torque variability for knee extensors is present in other knee injuries as well [36], and in this case the variability may be accounted for in the PFP group on the basis that PFP patients tend to have impaired quadriceps function often ascribed to impaired vastus medialis oblique firing [37, 38] and reduced eccentric control [39]. However, the variability improved notably in the uninjured limb following the intervention period, possibly due to enhanced motor learning which tends to occur in injured populations [40].

It should be noted that the reliability of the SG, as with any device, is dependent on the set-up and therefore clinicians should be vigilant in following the set-up instructions set out in the procedures section and keep the set-up consistent when doing testing and re-testing in clinical settings, to ensure reliable results. Belt-stabilisation seems to improve reliability in SG-based devices and is recommended for use by practitioners [26, 35].

Clinically meaningful difference

The CMD within the present study ranged between 12–45 Nm, implying that the differences in torque production by the quadriceps musculature between limbs or over the course of rehabilitation can be detected with the SG. Previous research has shown that an HHD exhibited comparable, although marginally lower CMD values of 17–27 Nm [41], with the ID yielding the lowest and most consistent CMD of approximately 27 Nm [42].

Group differences in torque production

An intriguing finding of the present study relates to the simple interaction effects regarding within-group (e.g., ID vs. SG) and between-group (e.g., control vs. PFP) differences for quadriceps torque (see Fig 3A and 3B). The mean differences in torque production between devices and groups are classified as trivial-to-moderate, and typically range -2.28 Nm to 27.83 Nm (p = 0.113–0.999). In most instances the point estimates of the mean difference were not significantly different from zero largely due to the longer confidence intervals. Such a result generally implies, at least in principle, similarities in torque production between groups and between devices. The standardised effect sizes (together with the 95% CI) show that, despite a lack of statistical significance, the magnitude of the mean differences are likely to be meaningful, especially for the PFP group which consistently yielded higher torque values on the ID compared to the SG (M_diff = 43.53 Nm, p < 0.001). The interaction effects should be of interest to practitioners on the basis that the interchangeability of device measurements must be considered when interpreting results and making clinical inferences [42]. Given that, at least in some instances, practitioners might receive isokinetic results for a specific PFP patient or a referring clinician, these results should be verified when using the SG in the clinical setting so that more precise interpretations can be made when evaluating temporal changes in quadriceps strength. Moreover, the same device should be used by the same clinician to ensure adequate consistency in readings and insights of whether true changes have accrued.

Validity

The results of the preset study suggest that the concurrent validity of the SG, when compared to the ID, appears to be strong to very strong, although a few discrepancies are noteworthy. The torque values from the SG are typically lower than those from the ID, with the exception of torque values exceeding ~300 Nm (see Figs 4 and 5). It is therefore reasonable to state that the SG might be of better use in injured and non-athletic populations that will most likely produce lower torque values compared to healthy, uninjured, or very athletic participants that might produce substantially higher torque values. The validity of the SG was greater than that of an HHD which yielded poorer scores even when stabilised with a belt, with correlations ranging between r = 0.3–0.8 [18, 43]. The SG is not a perfectly valid tool given the bias in torque readings and moderate-to-large limits of agreement (LoA: 72–86 Nm) and is therefore unlikely to replace the ID for absolute values. More specifically, there is evidence of a systematic bias in favour of the ID (see Fig 5), confirming that the SG typically measures lower torques than the ID on average. These results correspond with Martins et al. [35], who showed that knee extension tested with a belt-stabilised HHD, exhibited similar mean difference in torque production, with bias towards the ID. The HHD also shows, on average, wide LoA values (33.59%, CI_95% [23.91%, 43.26%]) for knee extension in other literature [18], which was echoed by Oranchuk et al. [44] who also showed fairly wide LoA for their experimental strain gauge (LoA: 150–200 Nm) during isometric quadriceps testing in a group of healthy participants. The wide LoA therefore support the notion that the SG is not valid enough to replace values obtained from the ID but given the excellent reliability, coupled with affordability, would still be an exceptional tool for the evaluation of isometric quadriceps strength. It is also important to highlight that isokinetic norms should not be used to make clinical interpretations when comparing these to values derived from the SG. It would be important to develop independently generated normative data for SG-derived torque values to facilitate decision-making across different joints and population groups.

Temporal changes in torque

Finally, although the mean peak torque values did not change significantly from the baseline to follow-up testing (M_diff = 0.77–14.63 Nm, p = 0.531–0.969), it should be noted that there were substantial individual improvements both within- and between groups (see Fig 6). The variability in individual responses underscores the potential inadequacy of the intervention programmes followed by the groups which were beyond the control of the current study. Whether more focused and targeted interventions or prolonged rehabilitation timelines (> 8 weeks) would elicit more favourable outcomes would require further research, especially in those with PFP. It should however be noted that there were consistencies in the measured values between devices across time, implying that the SG indeed has exceptional utility as a measurement tool for evaluating isometric quadriceps strength.

Limitations

Given the strengths of the present study, it is also important to underscore some limitations. Firstly, participants could follow any rehabilitation program at any practitioner of their choice, and therefore could not control the details associated with targeted quadriceps strengthening as this was beyond the scope of the present study. Secondly, the ratios of males versus females in the groups differed, although consistency within groups was more important for this study than consistency between groups. Thirdly, participants were not divided into different groups according to the magnitude of quadriceps strength deficits. This latter point could provide more nuanced information on strength improvements over time in future studies.