Participants aged between 18 and 70 years were recruited from the community between July 2017 and March 2019 in two groups: those who presented with clinical signs of TPT and asymptomatic controls. Participants were screened online and via phone prior to attending one physical screening (completed by a physical therapist; MHR) and subsequent testing session at The University of Queensland. Participants were eligible for the TPT group if they presented with medial foot/ankle pain on most days for at least three months, with an average pain in the previous week greater than 2/10 on a Numeric Rating Scale (NRS) anchored with “no pain” at 0 and “worst pain imaginable” at 10, and pain or inability to perform a single leg heel raise (SLHR) on physical screening. Selection criteria were based on a clinical definition of tendinopathy from the International Scientific Tendinopathy Symposium Consensus (ICON) statement.10 Participants in the TPT group could have pain in other locations provided that their medial foot/ankle pain was their predominant pain.

Asymptomatic control participants who were of same sex and similar age (i.e., within 5 years) were eligible providing they had no lower extremity pain in the previous three months. Participants with a history of lower limb surgery in the previous year and systemic, neurological, or arthritic diseases were excluded from both groups. The study was part of a larger body of work and was approved by The University of Queensland Human Research Ethics Committee. Each participant provided written informed consent prior to participation.

One researcher (MHR) screened participants and those who were eligible completed the physical data collection session and completed questionnaires online. The online questionnaires included the Foot Function Index-Revised (FFI-R), Active Australia Survey (AAS), Pain Catastrophization Scale (PCS), Tampa Scale of Kinesiophobia (TSK), and Assessment of Quality of Life (AQoL). These outcome measures were selected based on previous TPT literature,5 and where measures were lacking, selected commonly used measures in musculoskeletal health.

Foot-related structure and function were measured using the pain and stiffness subscales of the FFI-R (e.g., how severe was your foot pain/stiffness before you get up in the morning?). Scores are derived by summing responses on a 5-point scale for 7 items in each subscale and expressed as a percentage of the maximum possible score, with higher scores representing a worse state. The FFI-R (and its subscales) has been shown to have good clinimetric properties in assessment of patients with foot problems.11,12

Activity limitations due to foot problems were evaluated using the difficulties subscale of the FFI-R (e.g., how much difficulty did your foot problems cause you climbing stairs?) (11 items). Scores were calculated as above. To further characterise activity limitations, during the physical data collection session, participants were asked to descend and then ascend a flight of 20 stairs (not assisted by placing hands on the railing), and time taken to complete the test was recorded. The intensity of pain before, during, and after the test (using the NRS) was recorded. Timed stair-climb tests have been shown to adequately assess physical function with excellent reliability.13

The AAS was used to measure total number of minutes of physical activity performed in the preceding week, and to classify participants as inactive (0 minutes) or participating in insufficient (<150 minutes) or sufficient (>150 minutes) amounts of moderate to vigorous physical activity.14 The self-administered AAS demonstrates acceptable levels of test-retest reliability14 and validity.15

Finally, the independent living domain (4 items) of the AQoL was used to obtain a measure of participants’ quality of life relating to living independently. The AQoL has six domains: independent living, relationships, mental health, coping, pain, and senses (the latter 2 were not used in this study due to using the FFI-R pain and stiffness subscales) and demonstrates appropriate levels of construct validity16 and reliability.17 Domains are scored separately and combined for an overall score with higher scores representing greater impairment in quality of life.16,18

The social restrictions (e.g., how much of the time did you limit social activities due to foot problems?) (6 items) and activity limitations (e.g., how much of the time did you limit your outdoor activities because of foot problems?) (3 items) subscales of the FFI-R were used to measure foot-related participation restrictions. Participation in personal relationships was evaluated with the relationships subscale (3 items) and the AQoL (total) was used to obtain an overall indication of general health-related quality of life. In addition, we included the total FFI-R score as a measure of foot-health related quality of life.

Personal factors, including the mental health (4 items) and coping (3 items) domains of the AQoL, were used to obtain an indication of personal contextual factors that influence how individuals experience a condition. To further contextualise the individual pain experience, we asked participants to complete the PCS and the TSK. The PCS is a valid and reliable measure,19 comprised of 13 items pertaining to three domains; rumination, magnification, and helplessness.20 Total scores range from 0 to 52 where 52 represents a high level of catastrophizing.20 The TSK is a valid and reliable measure of pain related fear of movement or (re)-injury – it applies only to those in pain (i.e., the TPT group).21 Total scores range between 17 and 68. Higher scores indicate greater levels of fear of movement.22 A total score >37 indicates significant fear of movement.22

Individual participant data were plotted in box and whisker plots for visual inspection – given the small sample size. All statistical analyses were undertaken in R version 4.0.3 (R Core Team, Vienna, Austria).23 Independent t-tests were used to determine differences in demographic characteristics between groups. The Hodges-Lehmann method was used to estimate median differences between groups with 90% confidence intervals (CIs)24 and Cliff's delta (d) was used to determine the standardised difference between groups for each outcome measure. Cliff's d provides a robust non-parametric effect size (ES) measure, which is an estimate of the standardised difference between groups,25 and allows for comparison across measures with different scales and distributions. Values of Cliff's d range from 1 to -1, with values of 1 indicating all values from k > l, and values of -1 indicating all values from k < l. Cliff's d (and 95% confidence interval, CI) was calculated using the effsize package,26 with values interpreted as small 0.147, moderate 0.33, and large 0.474.27

Twenty-two participants (19 [86%] females) met the selection criteria for TPT and were compared to 27 (25 [93%] females) asymptomatic controls (Fig. 1).

Fig. 1.

Participant flow through the study.

(0.47MB).

The two groups were similar in age, sex composition, and height, but body mass and body mass index (BMI) were higher in those with TPT (mean difference (95% CI) in body mass ∼17 (7, 27) kg) (Table 1). Number of SLHR repetitions was less in those with TPT compared to controls (mean difference (95% CI) ∼11 (-14, -8 repetitions)). There were no differences between groups with regards to pain medication use in the previous 48 hours, regular medication use, diabetes, hormonal status, or arch height index.

Participants in the TPT group had a median (IQR) symptom duration of 13 (6 to 51) months. Five participants (23%) with TPT had pain in another location (which was not their predominant pain) and 3 (13%) had bilateral medial ankle pain (Table 1).

Individual participant data plots showed substantially higher levels of pain and stiffness in individuals with TPT (Fig. 2A-B) (large ES > -0.47; Table 2).

Fig. 2.

Box and whisker plots for body structure and function outcomes (A to B) and personal contextual factors (C to E). Circles indicate individual participants (control; blue, TPT; red). Abbreviations: FFI-R, foot function index – revised (scale 0-100); TPT, tibialis posterior tendinopathy; pain catastrophising (scale: 0-52).

(0.27MB).

Individual participant data plots showed that compared to controls, individuals with TPT had substantially higher levels of foot-related functional difficulties and difficulties with independent living (Fig. 3A, B) (large ES > -0.47; Table 2). Individuals with TPT took approximately 4.5 seconds longer than controls to complete the stair descent/ascent task (Fig. 3C) (large ES >-0.47; Table 2). During the stairs task, 21 (95%) of participants with TPT reported pain, with a median (IQR) pain severity of 3 (1.9 to 4). There were no observable differences between groups in physical activity over the previous week (Fig. 3D) or the proportion of participants who were sufficiently active (p=0.43) (Table 2).

Fig. 3.

Box and whisker plots for activity outcomes (A to D). Circles indicate individual participants (control; blue, TPT; red). Abbreviations: AQoL, Assessment of Quality of Life (scale: independent living (0-22); FFI-R, Foot function index – revised (scale: 0-100); TPT, tibialis posterior tendinopathy.

(0.33MB).

Individual participant data plots show that social restrictions, activity limitations, overall quality of life, and self-reported foot function (overall) were substantially poorer (represented by higher scores) in individuals with TPT compared to controls (Fig. 4A-D) (large ES >-0.47; Table 2). Participation in personal relationships did not appear to be different between groups (Fig. 4E).

Fig. 4.

Box and whisker plots for participation outcomes (A to E). Circles indicate individual participants (control; blue, TPT; red). Abbreviations: AQoL, Assessment of Quality of Life (scale: mental health (0-20), relationships (0-13), coping (0-15), total (0-99)); FFI-R, Foot function index – revised (scale: 0-100); TPT, tibialis posterior tendinopathy.

(0.3MB).

Plots show that individuals with TPT had higher levels of mental health concerns (Fig. 2C) (moderate ES >0.33; Table 2), but there were no demonstrable differences between groups on the coping domain of the AQoL (Fig. 2D). There were no differences between groups for pain catastrophising (Fig. 2E), but eight participants (36.4%) in the TPT group exhibited a high degree of fear of movement or (re)-injury (>37 on the TSK).