The basic biomechanical abnormality for transfer metatarsalgia is manifested as an alteration of forefoot loading pattern. A previous study by Geng et al.  on post-osteotomy patients showed that the risk of metatarsalgia increased substantially when the plantar loading ratio of the central rays reached 55% in the push-off phase during gait. Therefore, this study focuses on the effect of magnitudes of the first metatarsal shortening on the plantar pressure distribution during push-off. Moreover, by using the forefoot loading ratio of 55% as a threshold value, we might determine the optimal value for the first metatarsal shortening to reduce risks.
Standardized FE simulations for metatarsal shortening procedure were performed in a healthy foot rather than a diseased one. In this situation, we assume all the alignment and other structural deformities have been corrected, so the confounding factors due to different deformities that may potentially influence model results are excluded. Furthermore, our study is not intended to evaluate the efficacies of certain osteotomy procedures for hallux valgus correction. Rather, we carefully designed a method that is not applied during clinical routine, but can purely simulate the effects of the first metatarsal shortening. All of these ensure this study to focus on potential relationships between the level of metatarsal shortening and alterations in plantar pressure patterns.
The three-dimensional finite element model established here contains relatively complete tissue structures. The soft tissue is set as a hyperelastic material ensuring the simulation and structural completeness of the model. In addition, the model reflects an actual forefoot push-off phase, and both the dorsiflexion angles of the ankle and the first metatarsophalangeal joint during push-off are reproduced. Validation with plantar pressure results is promising.
Our results showed that the weight-bearing ratio of the first ray was gradually decreased with the increased shortening length when the first metatarsal is shortened; the central rays and the lateral rays’ plantar pressure increased. We speculate that the contact area of the first ray during push-off decreases when the first metatarsal is shortened. This leads to more weight transfer to the central rays and lateral rays. These results meet our expectations.
The loading ratio of the central rays increased up to 54.8% when the first metatarsal was shortened by 6 mm. This is close to the 55% that is thought to be a risky threshold based on previous studies . In other words, when the first metatarsal shortening is less than 6 mm, the increased loading ratio of the central rays does not reach the critical value for transfer metatarsalgia. To avoid simultaneous elevation of the first metatarsal head, the shortening direction is parallel to the sole of the foot. In addition, since the first metatarsal of the model volunteer is at relatively the same length as the second, we can conclude that the first metatarsal should be controlled to be 6 mm shorter than the second metatarsal.
In this study, the central rays loading ratio reached 59.5% when the shortening reached 8 mm. This resulted in a high risky foot with postoperative transfer metatarsalgia. However, we noticed that if the distal end of the first metatarsal is depressed further (“push down motion”) by 3 mm, then its plantar pressure could be restored considerably, i.e., the central rays’ loading ratio again dropping to 47.6%. This suggests that if a larger magnitude of shortening of the first metatarsal is necessary during surgery, then we shall compensate for its loss of weight-bearing function by appropriately depressing its distal bone segment.
Carr and Boyd  argued that the degree of the first metatarsal shortening in the treatment of hallux valgus should not exceed 4 mm. In this study however, the weight-bearing ratio of the central rays only increased to 51.6% when the first metatarsal was shortened by 4 mm. This gap may be associated with the shortening method in their study—it simultaneously elevates the distal end while also leading to shortening—both of these impair the weight-bearing function of the first ray during the push-off phase. However, the shortening done here is parallel to the sole of the foot, and it does not involve any elevation in the sagittal plane. Thus, the safe range of shortening is relatively larger.
Zhang et al.  reported that the first metatarsal was shortened on average by 1.8 mm with the maximum shortening no more than 6 mm. They found that the incidence of postoperative transfer metatarsalgia increased with the degree of shortening, but they failed to describe sagittal elevation or depression of the distal end. Their measurement method connected the central points of the proximal to the distal articular surfaces of the first metatarsal; thus, they focused on absolute rather than relative length, which we believe has less clinical relevance.
Toth et al.  showed an average shortening of 3.8 mm of the first metatarsal in their operation, while the average depression of the distal end was 2.7 mm. They concluded that the shortening of the first metatarsal was significantly associated with postoperative subjective pain scores. However, the authors failed to record whether the patients with postoperative pain were accompanied by recurrence of hallux valgus, malunion, dislocation of metatarsophalangeal joint, or hammer toe deformity—this could also cause postoperative pain despite less shortening.
There have been previous reports of a relatively large degree of first metatarsal shortening. Klosok et al.  reported an average shortening of the first metatarsal by 10 mm after surgery for hallux valgus. Keogh et al.  followed up hallux valgus cases and found an average shortening of the first metatarsal by 5 mm. Pouliart et al.  reported an average shortening of 8.5 mm after the operation. However, none of these studies found any definite correlation between the degree of shortening and the incidence of transfer metatarsalgia. From a biomechanical viewpoint, our simulation provides a theoretical basis for these clinical reports because when the first metatarsal distal end height stays unchanged, the central rays’ loading ratio will not increase to a relatively risky range until the shortening exceeds 6 mm. Furthermore, even larger length of shortening might be allowed if its distal end is pushed down.
The factors affecting the distribution of forefoot weight-bearing are certainly not due only to the length of the first metatarsal. In this study, however, we try to rule out the poor alignment of first ray by modeling a normal foot including the first tarsometatarsal joint relaxation, sesamoid dislocation, and other related factors. We focused solely on the length change of the first metatarsal to provide guidelines for osteotomy surgical strategy. With more research, patient-specific models may emerge to improve this complicated procedure.
There are some limitations to this study. The simulation of osteotomy is rare in the literature. It is designed, however, according to the direction of the longitudinal axis of the second metatarsal and a reference plane that assures the anatomical axis of the first metatarsal remained unchanged. Also, this biomechanical study is still based on a quasi-static model that merely focuses on a certain moment during push-off. Finally, while our FE model was validated by plantar pressure data, it cannot replace in vivo studies. The inter-individual variability and other in vivo compensatory mechanisms are not considered. Therefore, quantitative results from this study only provide general guidelines for surgeons to perform first metatarsal shortening procedures during the hallux valgus reconstruction surgery. Further clinical studies on patient populations are needed to validate these conclusions, as well as to determine whether and how they can be precisely applied to different populations.