Pes cavus

Principal authors: Louise Crawford, Jim Barrie

Latest evidence check March 2010

In most people with pes cavus, the deformity appears to develop as a result of muscle imbalance, although the imbalance and its cause may be difficult or impossible to detect objectively. The exceptions are those who develop a cavus-type deformity as a result of trauma or previous surgery (and even then imbalance may play a part).

There is relatively little objective data on muscle strength and balance in neurological foot deformities, particularly following changes in strength over a period in a population. Therefore the pathogenetic mechanisms must be viewed, even more than in many foot conditions, as hypotheses that may need to be revised as better data becomes available.


This is probably the simplest imbalance to understand. It is particularly associated with polio, in which the long plantarflexor muscles, particularly gastrocnemius/soleus, are paralysed while the dorsiflexors are at least relatively normal. The heel drops down and the midfoot is pulled up by tibialis anterior prodicing a dorsiflexed hindfoot and reciprocally plantarflexed forefoot.

Cavus and cavovarus

Evidence for imbalance

A number of imbalances probably work together to produce these deformities. Duchenne emphasised intrinsic muscle imbalance causing the elevated arch, as did Sabir and Lyttle (1984). Dwyer (1975), on the other hand, felt that there was not enough evidence for a significant role of the intrinsic muscles. Most likely a combination of the intrinsic and extrinsic muscles is the cause of the imbalance.

Mann (1988) described the pathogenesis of pes cavus in patients with CMT disease using the muscle agonist/antagonist theory – most muscles in the foot have two antagonists. Tibialis anterior is antagonised by tibialis posterior & peroneal muscles. Peroneus longus pulls harder than the weak anterior tibials, causing plantarflexion of the first ray and forefoot valgus. The posterior tibialis pulls harder than the weak peroneus brevis, causing hindfoot adduction.  Muscle power loss is not predictable – they don’t deteriorate at the same rate and so each deformity must be individually assessed.

Specific imbalances

Intrinsics vs long flexors/extensors

Intrinsic weakness leads to loss of balance between long extensor and flexor muscles. In the intrinsic-minus foot the force of the long extensors is exerted mainly on the MTP joints, leaving th long flexors to flex the IP joints. the result is a claw deformity with flexion of the IP joints and hyperextension at the MTP joint (Sarrafian). Price et al (1993) used CT to look at patterns of muscle wasting. They performed CT on 26 patients with HMSN and found earlier and more severe involvement of the intrinsic muscles compared with the extrinsics. The most consistent early degeneration occurred in the pedal lumbricals  and interossei (most distally innervated). Gallardo (2006) repeated this using MRI and found the same pattern of abnormality. The intrinsic weakness leads to an ‘intrinsic minus’ foot and clawing of the toes and a combination of the windlass effect and shortening of the intrinsic may then lead to shortening of the longitudinal arch. However, Dwyer (1975) considered that cavus deformity can progress without clawing of the toes and therefore felt that intrinsic weakness itself is not a major factor in pes cavus formation.

Long extensors vs long flexors with weak tibialis anterior

Anterior tibial weakness leads to loss of dorsiflexion power during swing phase so that EHL and EDL work harder to provide enough dorsiflexion to clear the floor. Mann and Missarian found that tibialis anterior was, on average, weaker than EHL. The force the long extensors exert to dorsiflex the ankle also tends to dorsiflex the MTP joints. Olson (2003), in a cadaver study, found that relative EHL overpull produced hyperextension at the hallux MTP joint.

Peroneus longus vs tibialis anterior

working against a weak tibialis anterior allows plantarflexion of the first ray. Olson's (2003) cadaver study supported relative peroneus longus overpull as the most important force in producing first ray plantarflexion and increased pressure under the first MT head. Peroneus longus is relatively well preserved in HMSN. Helliwell (1995) studied muscle biopsies and MR imaging and found tibialis anterior atrophy and peroneus longus hypertrophy in all cases.

Tibialis posterior vs peroneus brevis

Peroneus brevis muscle weakness allows relative overpull of tibialis posterior resulting in hindfoot varus (Mann and Missarian 1988). The deep posterior compartment is relatively spared in the degenerative process (Price 1993, Gallardo 2006), and the tibialis posterior is relative much stronger than peroneus brevis so can lose more of its working fibres before weakness becomes apparent (Mann and Missarian).

Structural effects

As well as the muscle imbalances, a number of factors related to static foot shape and structure contribute to the deformity. Initially contracture of the plantar fascia was considered the most important. However, the work of Coleman highlighted the interaction between forefoot pronation and hindfoot varus, and this is the most important factor in adult patients.

Plunger effect

As the toes adopt an intrinsic-minus clawed position, extension of the MP joints forces the metatarsal heads plantarward and the metatarsals into plantarflexion. When the proximal phalanx comes to lie on the metatarsal head, the deep transverse metatarsal ligament and plantar plates become locked in the dorsal position, trapping the metatarsal heads. Stainsby (1997) called this the "plunger effect".

Windlass effect

The plantar plates of all the MTP joints are attached to the longitudinal bands of the plantar fascia. MTP joint extension tightens the plantar fascia. As the anterior and posterior ends of the arch are drawn together, the apex of the arch rises. Hicks described this as "the windlass effect". As the medial band of the plantar fascia is the largest and inserts relatively medially on the calcaneum, the windlass effect produces both cavus and hindfoot varus.

Tripod effect

Normally the heel, first metatarsal head and fifth metatarsal head form a balanced "tripod" - with all three points on the ground the foot is plantigrade and neither valgus nor varus. If the first ray is plantarflexed, the three points can only rest on the ground if the hindfoot tips into varus. This phenomenon was well-described by Coleman (1977) who described an elegant clinical test (the block test) to differentiate between hindfoot varus which was flexible and created mainly by the tripod effect, and varus which had become fixed.

Bony deformity

As with any deformity which develops gradually through childhood, bone growth will be affected by the deforming forces and altered bone shape will form a fixed part of the deformity. This was studied by Aktas and Sussman (2000) who looked at lateral weight bearing xrays of 26 cavus feet (all with CMT disease). They found a mean lateral talo-first metatarsal angle of 17.8deg plantarflexion (normal is 0deg). Interestingly, they also found that both the talus and calcaneus were in significant dorsiflexion.