Biomechanics of the Foot in Diabetes Mellitus。 Some Theoretical Considerations

Abstract

Although diabetes mellitus is a biochemical disease, it has biomechanical consequences for the lower extremity. Numerous alterations occur in the function of the foot and lower extremity in people with diabetes. This article evaluates biomechanical alterations of the foot in the presence of neuropathy in patients with diabetes in the context of several theoretical concepts. Further study of these hypotheses will result in a better understanding of how diabetes causes elevated plantar pressures and the potential of strategies to prevent these changes so that the burden of diabetic foot disease can be reduced.

Diabetes mellitus imposes a large economic burden on society and the individual. Complications of the foot are a significant contributor to this burden. [1] Diabetes mellitus is a biochemical disease, but a large number of lower-extremity complications of the disorder are due to biomechanical dysfunction. Diabetes not only alters the biomechanics of the lower extremity, it also complicates any preexisting biomechanical dysfunction. Reviews of the biomechanics of the foot in patients with diabetes are available elsewhere, [2,3,4,5,6,7] so the discussion here will focus on reviewing recent data and thought on foot biomechanics in patients with diabetes in the context of a number of theoretical considerations.

A key feature of the foot in people with diabetes is that dynamic plantar pressures are higher than in those without diabetes. This feature is independent of body weight. [8] Issues concerning measurement of plantar pressures have been adequately dealt with elsewhere [9,10] The causes of higher plantar pressure [11] are generally assumed to be bony deformity, clawing of the toes, pes cavus, lack of soft-tissue cushioning, callus formation, and limited joint mobility. A loss of protective sensation in the foot due to peripheral sensory neuropathy permits undetected injury from mechanical insults to occur to the foot when higher plantar pressures are present. Brand [12] first elucidated the mechanical causes of plantar ulcers in neuropathic feet.

Limited Joint Mobility

Limited joint mobility has been widely documented in patients with diabetes. [13,14,15] It is probably due to a nonenzymatic glycosylation of collagen from the chronic hyperglycemia, resulting in a stiffening of the joint ligaments and other structures around joints. Delbridge et al [16] found a significant decrease in the range of subtalar joint motion in diabetic patients with a history of foot ulceration compared with control groups. Fernando et al [17] showed that diabetic patients with limited joint mobility had higher plantar pressures.

In 1949, Hiss [18] suggested that a limitation in joint mobility results in an alteration of the progression of forces through the foot, which alters weightbearing. It is generally assumed that limited joint mobility increases plantar pressures owing to restriction of pronation at the subtalar joint [16] based on the traditional understanding of the foot [19] as a “mobile adaptor.” Pronation at the subtalar joint when the foot contacts the ground is assumed to allow the foot to absorb shock. However, recently several of the concepts of Root et al [19] have been questioned, [20,21] and one possible alternative, the sagittal-plane facilitation of motion model, has been proposed. [22] It is also difficult to accept that mobility of the subtalar joint allowing the foot to absorb shock during the contact period of the gait cycle could be related to increased pressures under the forefoot during propulsion. Recently Yingling et al [23] demonstrated that there was no increase in the impulse wave at the level of the tibia when subtalar joint pronation was restricted, which raises a question about subtalar joint pronation absorbing impact shock.

The sagittal-plane facilitation of motion model [22] suggests that normal foot function is dependent on an adequate range of motion at the first metatarsophalangeal joint during dynamic function (which is independent of the range of motion of the joint during clinical examination) so that the windlass mechanism [24] can be established. This allows the foot to adequately resist the stress applied to the foot during the propulsive phase of gait. If this mechanism and other autosupportive mechanisms [22] are not established owing to a functional hallux limitus or an inappropriate direction of “weight flow” through the foot or to limited joint mobility affecting dorsiflexion of the first metatarsophalangeal joint and plantarflexion of the first ray, a number of compensatory mechanisms are predicted by the model to occur. [22] One of these is an “off-loading” of the medial side of the forefoot and an increase in weightbearing under the lateral side of the forefoot. While other mechanisms or structural changes may increase plantar pressures under the first metatarsophalangeal joint, such as a limited range of motion of the first ray, [25] the model offers another alternative theoretical explanation for the correlation between limited joint mobility and increased plantar pressures.

A restricted range of motion at the first metatarsophalangeal joint has been shown to be more common in patients with diabetes compared with controls. [26] One of the findings of Stokes et al [27] was a more lateral distribution of plantar pressures in the diabetic group compared with controls. Stess et al [28] showed that plantar pressures increased more in those with a history of diabetic ulcers compared with the diabetic control group, with the highest pressure noted on the lateral side of the plantar forefoot. St Zimney et al [29] also reported a greater increase in plantar pressures under the fourth and fifth metatarsal heads. These reported findings are consistent with what would be predicted by the sagittal-plane model; however, Ctercteko et al [30] showed a medial shift in the increase of plantar pressures. Schaff and Cavanagh [31] reported that a rocker-bottom shoe reduced pressures by over 30% under the medial forefoot, but increased pressures on the lateral side of the forefoot. The sagittal-plane facilitation of motion model would predict that the rocker sole would prevent the autosupportive mechanisms of the foot from being established owing to a lack of dorsiflexion of the first metatarsophalangeal joint, resulting in an increase in lateral plantar pressures.

The presence of limited joint mobility in patients with diabetes raises the possibility of therapeutic interventions to increase the mobility of the joints of the foot, with the aim of reducing plantar pressures. In a small uncontrolled trial, Curran et al [32] used physical therapy on a group of diabetic patients with limited joint mobility for 6 weeks; the result was an increase in the range of motion of the subtalar joint from 14.5° to 24.5° and an increase in that of the midtarsal joint from 17.6° to 31.4°. The peak plantar pressures decreased from 10.1 to 7.2 kg/cm². It is unclear what exercises were used and which other joints were mobilized, but this hypothesis is promising and needs further investigation by means of prospective clinical trials.

Sensory-Attenuation Hypothesis

The sensory-attenuation hypothesis, originally proposed by Robbins et al [33,34,35,36] in relationship to athletic footwear, states:

In humans, avoidance of uncomfortable or painful but locally innocuous plantar cutaneous tactile stimuli moderates shock on subsequent impacts when humans walk, run, or jump repetitively. This feedback control circuit is optimized in terms of protection for mechanical interaction of the bare foot and natural surfaces. Eventually learning allows anticipatory avoidance. Modern athletic footwear is unsafe because it attenuates plantar sensations that induce the behavior required to prevent injury. [33] (p218)

This hypothesis is not without controversy, [37,38] but it suggests that some plantar discomfort on impact is required for optimal shock absorption, which is attenuated by soft footwear. When impact is sensed, some impact-moderating behavior takes place. If impact is attenuated by soft footwear or if sensory neuropathy is present, there is a perceptual illusion of less plantar impact than actually exists, which reduces impact-moderating behavior.

As a consequence, the model predicts that more weight is borne by the metatarsal heads and less by the toes, which was demonstrated by Robbins and Hanna [39] in those who habitually wear soft running shoes. Ctercteko et al [30] showed a similar effect in patients with sensory neuropathy with a reduced toe loading and increased metatarsal head pressures. Similar findings have been reported by Stokes et al [27] and Boulton et al. [35] Frykberg [5] attributed these reported changes to intrinsic muscle atrophy and claw toe development in patients with diabetic neuropathy; however, it is proposed here that the sensory-attenuation hypothesis offers an alternative explanation.

If this hypothesis can be demonstrated by prospective studies to apply to the lower extremities of those with sensory neuropathy, it raises a question about the use of soft cushioning on orthoses or in footwear in those who have not yet developed a sensory neuropathy or are in the early stages of one. If plantar sensations are further attenuated, this may alter foot morphology and function in a detrimental way. Further prospective investigations are needed to test this hypothesis.

Structural Changes

The putative effects of motor neuropathy and amputations on foot function are adequately dealt with elsewhere [6] and are not considered here. The significance of structural changes was demonstrated by Cavanagh et al, [40] who demonstrated that static structural variables are significant predictors of dynamic foot function in walking. Measurements from standardized radiographs explained approximately 35% of the variance of dynamic plantar pressures under the first metatarsal and the heel. The two most dominant factors in the predictions of pressures for both regions were soft-tissue thickness and inclination of the first metatarsal. For each 1° increase in first metatarsal inclination, heel pressures increased by 17 kPa and first metatarsal head pressure increased by 47 kPa.

In the hand, limited joint mobility causes a contracture of the palmar aponeurosis that results in the characteristic “prayer sign,” in which the patient cannot oppose the palmar surfaces. It has been generally assumed that this contracture does not occur in the foot because of the weightbearing forces on the plantar aponeurosis. However, in a dynamic situation, Kidd and Kidd [41] hypothesized that glycosylation of the collagen-rich plantar aponeurosis would result in a contracture of the plantar truss structures, leading to an accentuated arch height via first ray plantarflexion and the windlass mechanism. The result would be a more rigid, high-arched foot with prominent metatarsal heads. In support of this theory, pes cavus has been shown to be more common in those with diabetes than in nondiabetic controls, [26] whereas the prevalence of pes planus was the same in the two groups.

The characteristic “intrinsic minus foot,” [7] with claw toes, anterior advancement of the plantar foot pads, and depressed metatarsal heads, has generally been considered to result from intrinsic muscle atrophy and dominance of the long flexor muscles. However, the hypothesis proposed by Kidd and Kidd and the sensory-attenuation hypothesis discussed above could also account for these structural changes.

Mechanism of Tissue Damage

It has been well documented that elevated pressures are associated with tissue damage, [42] but the mechanism of the tissue damage has not been clearly elucidated. Landsman et al [43] proposed a cellular model that shows that high rates of tissue deformation may cause elevated intracellular calcium concentrations, resulting in cellular death, while comparable loads gradually applied do not. They hypothesized that ulcer formation is a result of high-strain-rate deformation rather than the actual peak threshold. They further hypothesized that a reduction in the rate of loading, rather than the reduction of peak loads, will lead to healing and minimize the recurrence of foot ulceration. Strength in the anterior compartment of the leg will be reduced because of the motor neuropathy, resulting in a “foot slap” and a higher rate of forefoot loading. They proposed that an ankle-foot orthosis will reduce foot slap. No data can be found linking the velocity of foot strike and anterior-compartment weakness in those with plantar ulcers. Anderson et al [44] found a 21% reduction in strength of the ankle dorsiflexion muscle group in insulin-dependent diabetic patients compared with controls, which would support the hypothesis presented above. However, Anderson and Mogensen [45] showed that peak velocity at the ankle was lower in patients who had been insulin-dependent for a long time, presumably owing to a slower cadence.

Most of the literature has focused on elevated plantar pressure, probably because of the limitations inherent in the equipment used to measure shear stress under the foot. The consequences of combined shear and vertical pressures are likely to be much greater than those of vertical pressures alone. Ulbrecht et al [9] discussed the issues surrounding the variability in the reported threshold of vertical pressures at which tissue damage occurs. Some of the variability may be due to shear stresses. Dinsdale [46] was able to induce an ulcer in the skin of swine with a verticle pressure of 290 mm Hg, but a pressure as low as 45 mm Hg was sufficient when friction was present. Davis [47] proposed the existence of a “wrinkled carpet” effect, with the tissues being either “bunched up,” stretched, or sheared, depending on the direction of slippage. Brand [48] showed in a study using pressure pads on pigs that when the pressure was exerted over 5 to 7 hours, there was some tissue destruction. The pressure was greatest at the center of the pressure pads, but the greatest tissue damage was at the periphery of the pads, where there is increased shear stress between the area under pressure and the area not under pressure. Brash et al [49] demonstrated a signal void (“dropout”) on magnetic resonance imaging (MRI) evaluation indicating a microscopic hemorrhage in the subcutaneous fat in the region of prior ulceration in diabetic subjects. Shear stress may be responsible for the type of tissue damage demonstrated by the MRI. Bojsen-Moller and Lamoreux [50] showed that dorsiflexion of the toes tightens the connective-tissue framework of the ball of the foot, which enables shear forces to be transferred to the skeleton. Limited joint mobility will restrict this process, and the glycosylation of tissues will also interfere with this mechanism, further increasing the potential for tissue damage.

Conclusion

This article has proposed several new theoretical concepts regarding biomechanical alteration of the foot in the presence of neuropathy in patients with diabetes mellitus. As yet, these concepts are unsupported. Further prospective study of these hypotheses will result in a better understanding of how diabetes causes elevated plantar pressures and the potential of strategies to prevent these changes and the subsequent tissue damage. The proposed hypotheses are not necessarily inconsistent with one another, but each has different consequences for therapeutic interventions. Appropriate prospective clinical investigations are needed to test the interventions. Strategies that need to be implemented should be done so in the wider context of appropriate intervention strategies for the diabetic foot. [51]