Study design
A within-subject, randomized, crossover design in a biomechanical laboratory was used in this study. Three conditions included the shod condition, customized insole without medial wedges condition, and customized insole with medial wedges condition. Participants were additionally asked to use the customized insole with medial wedges and were followed up for 3 months. The research protocols were approved by the center of Ethical Reinforcement for Human Research of Mahidol University (COA No. MU-CIRB 2020/285.2109). All experiments were performed in accordance with relevant guidelines and regulations.
Study Participants
At least 27 participants had a power of 80% and a significance level of 5% to detect biomechanical differences among the three conditions with a medium effect (Cohen’s d = 0.5). Inclusion criteria were: (i) age between 18 and 60 years old; (ii) specific criteria of PF [1]; (iii) symptom of heel pain at least 6 weeks, indicating the chronic condition [17]; (iv) average pain intensity during last week at least 30 mm on a 100-mm visual analog scale (VAS). The exclusion criteria were as follows: (i) body mass index (BMI) greater than 30 kg/m2; (ii) leg length difference greater than 1 cm; (iii) positive sciatica test, indicating L5-S1 nerve root irritation; (iv) history of lower-extremity fracture; and (v) diagnosis of at least one disease as follows: gout, diabetic neuropathy, rheumatoid arthritis, systemic lupus erythematosus (SLE), cancer, infectious disease and tumor. Informed consent was obtained from all participants prior to data collection.
Customized Insole
Participants were given one pair of customized insoles (Fig. 1A) from the 1st physical therapist who had 7 years of experience in the use of insoles for musculoskeletal problem management. It was a 3-quarter-length insole with three layers, which included one layer of 1.2-mm genuine leather to increase comfort and two layers of 1-mm polyvinyl chloride (PVC) to increase strength. It incorporated a heat-molding process, which could be set within approximately three minutes, to adjust the individual foot shape in the sitting position. Medial wedges with full length of a 3-mm soft foam layer (Fig. 1B) were added under the customized insole (Fig. 1C). The present study developed 3-degree, 6-degree, and 8-degree wedges for the rearfoot and forefoot. There were three sizes for these products, which included small, medium and large sizes, following the foot length of participants (Fig. 1D). The forefoot angle was used to determine the degrees of the rearfoot and forefoot varus wedges. Previous studies suggested 50% correction of the forefoot angle, up to a maximum of 6 degrees for the use of rearfoot varus wedge, and 60% correction of the forefoot angle, up to a maximum of 8 degrees for the use of forefoot varus wedge [18, 19].
Gait Assessment
A three-dimensional motion analysis system (10 cameras, Vicon, Vantage V5 series, Oxford, UK) was used to track the lower extremities and multi-segment foot motions during gait at a sampling rate of 100 Hz. The cameras were synchronized with two force plates (AMTI, model OR6-7, USA), which were set to have a sampling rate of 1000 Hz on an 8-m walkway. All participants were instrumented with 42 retroreflective markers following the Plug-In-Gait (PIG) model and the Oxford Foot Model (OFM) by the 2nd physical therapist. Then, they were asked to walk with a pair of commercially athletic shoes (Adidas, Model: Duramo SL) and a shoe with two types of customized insoles, which included an insole with medial wedges and without medial wedges, in randomized orders. Before data collection in each condition, participants were asked to walk for approximately one minute to familiarize with each condition. Data were collected for 3–5 successful gait trials per condition with a self-selected speed. The comfort level was assessed after walking in each condition ranging from 0 to 10 points, and a higher score represented more comfort. Data from the motion analysis system were collected once at the beginning of the study.
Clinical And Us Assessment
The visual analog scale (VAS), foot function index (FFI), and ultrasonographic assessment were collected at baseline and at the 3-month follow-up. A previous study reported minimally clinically important differences (MCIDs) of each standard questionnaire among patients with foot pain, with an MCID of 9 mm for VAS and an MCID of 6.5 points for FFI [20].
An ultrasound machine (RUSI, model Affiniti 50, Philips, NV, USA) with a broadband linear array (Model L12-3) was additionally used by the 2nd physical therapist to assess edema of the plantar fascia on the symptomatic side. Participants were evaluated in the prone lying with a neutral position of the ankle joint. A probe that was adjusted at a depth of 3 cm was placed at the plantar side over the medial tubercle of the calcaneus [21, 22]. Then, the examiner captured the plantar fascia at 5 millimeters from the anteroinferior aspect of the calcaneus and measured its thickness and echogenicity [17]. The apparent US and uncertain US groups were additionally classified in this study. The former group included persons with plantar fascia thickness greater than 3.8 mm and hypo-echogenicity in the plantar fascia that represented those with plantar fascia edema, and the latter included persons without these criteria [23].
Data Processing
Joint kinematics were tracked using Nexus software (version 2.8.1) to determine the peak angles and motions of the pelvis, hip, knee, rearfoot (hindfoot relative to tibia), forefoot (forefoot relative to hindfoot), and hallux. The kinematic and kinetic data were filtered by the 4th -order zero-lag, low-pass Butterworth technique at cutoff frequencies of 6 Hz and 30 Hz, respectively. Initial contact and toe-off events for each foot were identified using vertical ground reaction force (GRF) data with a 10 N threshold. The stance phase of each foot was then normalized over a gait cycle by using custom MATLAB software (R2017a). Joint motion was defined as the difference between the maximum and minimum joint angles within each subphase of the stance, including the contact phase, midstance phase, and propulsive phase [24]. Two peaks of normalized GRF were determined in three directions, including anteroposterior, mediolateral, and vertical directions.
Statistical analysis
Characteristic data from the symptomatic limb are shown as the mean ± standard deviation (SD). Three inferential statistics were analyzed in the present study. First, repeated-measures ANOVA was used to compare the peak angles and the joint motions of the lower extremity and multi-segment foot as well as two peaks of ground reaction force during gait among three conditions, including the shod condition, customized insole with medial wedges condition, and customized insole without medial wedges condition. Bonferroni post hoc test was used for pairwise comparison. Regarding the nonparametric data, the Friedman test and Wilcoxon signed-rank test were used to compare such outcomes. Second, either the paired sample t test or the Wilcoxon signed-rank test was used to determine pre-post changes in pain intensity, foot function, and plantar fascia thickness after using the insole for 3 months. Finally, the chi-square test was used to compare the number of persons with apparent US and hypo-echogenicity pre- and posttreatment. All statistical analyses were performed using SPSS software version 22.0 (IBM Statistics, USA), with a statistical significance level of P < 0.05. Effect sizes using Cohen’s d were calculated for all variables [25].