2.1

The inclusion criteria were women above 18 years old with hallux valgus with clinical instability for the first group and without clinical instability for the second group. Since the disease is most prevalent in the women gender, with a ratio of up to 9:1, we opted to exclude males. ^,,,,,

For the non-inclusion criteria we opted to use the fact if a patient had other foot and ankle diseases, and diabetes, rheumatoid arthritis, gout, neuromuscular diseases, to minimize the confusion factor of the associated disease. Also, we did not accept juvenile hallux valgus and traumatic hallux valgus, since these are different diseases. Finaly, patients with a disability/contraindication to perform a CT scan in the weight-bearing CT scanner (the main contraindication was pregnancy) were not included.

2.2

The participant underwent a clinical examination using the Morton test, conducted by two physicians. (Appendix 1) In the event of disagreement, the test was jointly evaluated by both physicians until a consensus was reached. This determined whether the participant was hypermobile or not. The physicians conducting this examination were orthopedic specialists with qualifications in foot and ankle. One had two years of specialization, while the other three had over five years of training in the subspecialty. As shown in the appendix, the physicians performed a McNemar’s test, which yielded a Chi-squared Statistic of 0.0 between them for the division of cases in both groups. The group with patients with instability was denominated Group 1, while the group with patient without instability was Group 2.

2.3

We conducted the first exam with the participant seated in a chair, with their feet resting on the WBCT ( Fig. 2 A and B). The positioning of the ankle was neutral and without applying load, except for simple contact. Additionally, during the positioning of the foot, we rotated the lower limb along the long axis of the foot (from the second metatarsal to the calcaneus), following the linear markings on the WBCT support platform. We referred to this position as Position-1.

TCCC device with the positioning of the volunteer (A, B and C). Volunteer seated without load, with the limb rotated to follow the alignment of the device markings (D, E and F). Volunteer in the orthostatic position with full load. Figures from the author’s collection with the limb rotated to follow the alignment of the device markings. TCCC: weight-bearing computed tomography.

For the second scan, we instructed the participant to maintain an orthostatic position, evenly distributing weight on both feet. A new acquisition of images was then performed, this time with load ( Fig. 2 C and D). In this second exam, the ankle joint remained in a neutral position, and we rotated the limb to follow the markings of the CT scanner. And we referred to this position as Position-2.

2.4

We analyzed the WBCT using the Disior® program from Paragon 28 (Helsinki, Finland). The measurements that were evaluated were directly correlated with the hallux: hallux valgus angle (HVA); 1st-2nd intermetatarsal angle (IMA), and first tarsometatarsal joint angle on the anteroposterior radiograph (TMTA) ( Fig. 3 ). These angles were measured only in the weight-bearing exam, as described in the literature.

Radiographic measurements of foot angulation. A) Hallux valgus angle on the anteroposterior radiograph (HVA). B) 1st-2nd intermetatarsal angle on the anteroposterior radiograph (IMA). C) First tarsometatarsal joint angle (TMTA).

To analyze the mobility of the first ray, a second analysis process of the WBCT data was conducted, in which we sent the DICOM files to the company Paragon 28, where the calculations were continued. For this part, the company re-segmented the bones in positions 1 and 2, and three planes were described with a base axis to characterize the mobility of the first metatarsal and the medial cuneiform, following the protocol described initially by Geng et al. and reproduced by Kimura et al. ^,,, The bones were described by the angle formed by their long axis ( Fig. 4 ). The angular movement of these bones was evaluated comparatively by subtracting the seated position from the orthostatic position (position 2– position 1) to determine the mobility of each bone ( Figs. 5–7 ).

Image demonstrating the axes for bone analysis. A) Axial tomographic image with the axial axis of the bones (in blue), acquired by artificial intelligence through the midpoint of the proximal and distal joints, compared with the Y axis (in green). B) Sagittal tomographic image with the axial axis of the bones (in blue), acquired by artificial intelligence through the midpoint of the proximal and distal joints, compared with the Y axis (in green). C) Coronal tomographic image with the axial axis of the first metatarsal, the middle cuneiform, the navicular, and the talus in the determined order (in purple), acquired by artificial intelligence through the midpoint of the proximal and distal joints, compared with the X axis (in red).

Illustrative images of how the subtraction of angles works to define the bone movement in the axial plane. A) Angle of the first metatarsal in position 1 in the axial plane, acquired through the protocol described previously; B) Angle of the first metatarsal in position 2 in the axial plane, acquired through the protocol described previously; C) Subtraction of angle 2 from angle 1, showing the movement of the first metatarsal in the axial plane.

Illustrative images of how the subtraction of angles works to define the bone movement in the sagittal plane. A) Angle of the first metatarsal in position 1 in the sagittal plane, acquired through the protocol described previously; B) Angle of the first metatarsal in position 2 in the sagittal plane, acquired through the protocol described previously; C) Subtraction of angle 2 from angle 1, showing the movement of the first metatarsal in the sagittal plane.

Illustrative images of how the subtraction of angles works to define the bone movement in the coronal plane. A) Angle of the talus in position 1 in the coronal plane, acquired through the protocol described previously; B) Angle of the talus in position 2 in the coronal plane, acquired through the protocol described previously; C) Subtraction of angle 2 from angle 1, showing the movement of the talus in the coronal plane.

2.5

We used the article by Kimura et al. to calculate the sample size. They analyzed the hypermobility of the first metatarsal in the sagittal plane using CT, similar to our main hypothesis. The average mobility in this plane was 3.6° ± 2.3° of dorsiflexion in the hallux valgus group and 2.0° ± 1.3° of dorsiflexion in the control group (p = 0.037). The mobility of the first tarsometatarsal (TMT) joint in the sagittal plane was used with the null hypothesis that there was no difference in mobility between the two groups. We used the G*Power software to analyze the data, and it demonstrated that we needed 18 participants in each group for comparison.

Overall, the study analyzed 56 feet; however, only 36 of them were allowed to continue in the study, with 18 in the instability group and 18 in the normal group. Ten of the excluded feet were excluded because, after the WBCT, the patient was diagnosed with arthritis. Six had Juvenile hallux valgus and only informed us after the exam, and four feet opted to leave the research.

2.6

Numerical variables were described using the mean and standard deviation. Categorical variables were described by absolute numbers and percentages within the group. Numerical comparisons were conducted between the group of participants with hallux valgus with hypermobility and the group of hallux valgus without clinical hypermobility. For this purpose, the Shapiro-Wilk test was used to assess the homogeneity of the sample, and the Student’s t -test was applied for homogeneous samples, while Welch’s test was used for non-homogeneous samples. The JASP 0.18.3 (Jeffreys’s Amazing Statistics Program) was used to perform the calculations.