The effects of insole configurations on forefoot plantar pressure and walking convenience in diabetic patients with neuropathic feet

1. Introduction 

Foot orthosis therapy is a common treatment used to relieve plantar pressure, in order to prevent (re)ulceration and alleviate pain in rheumatic and neuropathic feet (Cavanagh, 2004, Kavlak et al., 2003). Decreased sensation, in combination with high plantar pressures, has been identified as a prime etiological factor in the cause of plantar neuropathic ulceration (Veves et al., 1992, Stess et al., 1997). Foot orthoses are widely prescribed in an attempt to decrease elevated plantar pressures in areas of actual or potential ulceration at the shoe–foot interface. Only a few studies that evaluated the effectiveness of custom foot orthoses on plantar pressure have been performed. There is very little scientific evidence to support the efficacy of their use (Landorf and Keenan, 2000, Rendall and Batty, 1998, Ball and Afheldt, 2002, Lee, 2001). It is thought that the effect of foot orthoses is primarily determined by their shape and the characteristics of the materials from which the orthoses are made (Ball and Afheldt, 2002). We found no studies that have examined the relationship between plantar pressure and various orthosis components, such as experimental shape variations, for ‘off-the-shelf’ shoes in low risk diabetic patients. The objective of this study was to evaluate the effects of a metatarsal dome, a varus and a valgus wedge and two arch supports on plantar pressures and walking convenience in patients with diabetic neuropathy.

2. Methods 

2.1. Patients 

Twenty male diabetic patients with flexible, non-deformed neuropathic feet and elevated bare foot plantar pressure were selected from an outpatient clinic. The Research Ethical Committee of the University Hospital Maastricht approved the study. Before the start of the study, patients were informed about all study procedures and their possible risks. Vibration perception threshold was determined with a biothesiometer at the apex of the hallux (Bloom et al., 1984) (Biomedical, Newbury, OH, USA). A vibration perception threshold of 25V was used as diagnostic criterion for peripheral neuropathy. Range of motion of the foot and ankle joints were evaluated according to the guidelines of the American Academy of Orthopaedic Surgeons (Engelberg, 1988). An EMED SF-4® pressure platform (Novel, Munich, Germany) was used to screen for elevated bare foot plantar pressures and measurements were performed according to a two-step protocol (Bus and Lange, 2005). Bare foot peak pressure was estimated per foot, by calculating the mean over the readings of five steps. A bare foot peak pressure higher than 700kPa was considered to be elevated(Armstrong et al., 1998b) and was used as an inclusion criterion. Three patients were excluded because of corrupted plantar pressure data due to technical failure. Eventually, data of 17 patients were used for analysis (Table 1).

Table 1. Patient characteristics (n=17)

		Median	Minimum	Maximum
	Age (years)	64.0	44.0	78.0
	Body mass index (kg/m2)	25.6	19.3	46.0
	HBa1c (%)	8.0	6.4	9.8
	VPT (V)	50.0	25.3	50.0
	Bare foot peak pressurea (kPa)	840	703	1253
	Walking speed (km/h)	2.5	2.1	5.0

HbA1c, hemoglobin A1c; VPT, vibration perception threshold; kPa, kiloPascal.

a Bare foot plantar pressure (⩾700kPa) was only used for patient selection. Median values of VPT and peak pressure are calculated over all feet.

2.2. Construction of basic insole 

For this study, one qualified and experienced orthopedic shoemaker made all the casts and insole constructions. An insole was obtained through a commonly used method for semi-weight bearing foam box casting (Guldemond et al., 2006). The insole was made following standardized construction procedures, using identical materials. A three layer construction was made of: 5-mm Lunalastik and 8-mm Lunasoft SL, top and bottom, respectively (NORA® Freudenberg GmbH, Weinheim, Germany), and 1.1-mm Rhenoflex 3208® (Rhenoflex GmbH, Ludwigshafen, Germany) as an internal reinforcement layer. Materials with a relatively high stiffness [Shore A higher than 60 (Charanya et al., 2004)] were used in order to minimize the influence of ‘cushioning’ on plantar pressure loading. After construction, 5mm of the ‘original arch’ produced through the casting method was removed resulting in a lower arch support. This construction was defined as the basic insole. Two 5-mm Lunalastik arch ‘inserts’ could be placed on the basic insole, resulting in three support heights: no support (basic insole), standard support and extra support, where the standard and the extra arch support were higher than the basic insole by 5 and 10mm, respectively. The ‘standard’ and ‘extra’ arch supports were proportionally tailored according to the original curves of the plaster mould. For the dome conditions, a standard manufactured metatarsal foam rubber (Shore A 28) dome was positioned 5mm behind the 2nd to 4th metatarsal heads on the basic insole: 11mm height, 15 shore type A. The locations of the metatarsal heads were determined through a dynamic pressure sheet footprint (Silvino et al., 1980). Custom-made full-length 5 degrees varus and valgus ‘posts’ or ‘wedges’ made of cork were placed underneath the basic insole to facilitate supination and pronation, respectively. All insole components (Fig. 1) were detachable because of the use of a weak adhesive (Talens® Rubber Cement, Ref. 95306500. Apeldoorn, Netherlands).

View Large Image Download to PowerPoint

Fig. 1. Right angular front view of basic insole with components. (1) Basic insole, (2) metatarsal dome, (3) ‘normal’ arch support, (4) ‘extra’ arch support, and (5) wedge (medial).

2.3. Measurement of plantar pressure and walking convenience 

Eleven insole configurations were compared with the basic insole (Table 2). The sequence of evaluation was determined by using two 12×12 ‘latin squares’. The corresponding test sequences were assigned randomly. All insole configurations were evaluated while patients walked on a treadmill in standard socks and Oxford type shoes (Van Lier®: outer sole shore type A: 86, N°814 Loon op Zand, Netherlands). Patients had the opportunity to accommodate to treadmill walking and individually chose their preferred walking speed (Table 1), which was kept the same for all subsequent measurements. The pressure data were sampled with a frequency of 50Hz per sensor with the Pedar Insole-system® (Novel Munich, Germany). A single trial took ≈50s and the produced ASCII-files were processed through Matlab routines. Twenty-eight sensors located in the forefoot area of the Pedar-insole were assigned to four anatomical forefoot regions, which are the locations with the highest plantar ulcer incidence (Reiber et al., 2001, Armstrong et al., 1998a): the lateral, central and medial forefoot and the big toe. The relation of the assigned regions and the individual foot skeletons were verified through anterior–posterior radiographs (Weijers et al., 2003). For each region, the sensor with the highest values for peak pressure and pressure time integral measured during the baseline condition (basic insole) was used to calculate the difference between the experimental conditions. To evaluate the subjective appreciation of each insole configuration, patients scored the walking convenience separately for the left and right insole on a 10-point rating scale (0–10: i.e. 0=‘bad’ to 10=‘excellent’).

Table 2. Means and standard deviations of plantar pressure data

		Peak pressure (kPa)						Pressure time integral (kPa)s
		Basic		Standard arch support		Extra arch support		Basic		Standard arch support		Extra arch support
		Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD
	Lateral
	Basic	135	43.2	138	43.5	136	39.8	63	27.8	64	27.7	63	26.2
	Varus	135	37.1	141	40.2	133	35.8	62	22.9	67	24.4	63	25.5
	Valgus	132	35.6	134	39.9	130	37.0	63	22.5	65	26.2	64	25.6
	Dome	121	40.0	124	39.8	123	38.6	55	27.3	58	27.5	59	30.7
	Central
	Basic	210	58.4	190	61.6	164	63.5	88	34.7	77	32.8	60	29.2
	Varus	206	43.3	188	50.8	154	70.0	89	35.6	82	34.7	60	32.4
	Valgus	206	66.4	186	57.5	162	65.2	90	36.7	77	32.4	63	29.4
	Dome	172	67.1	163	60.4	128	66.8	68	30.2	62	29.1	48	31.9
	Medial
	Basic	231	58.9	216	50.8	192	53.0	93	32.5	85	28.1	70	23.7
	Varus	225	57.2	211	46.5	178	52.6	90	30.0	85	26.2	68	25.1
	Valgus	226	68.7	217	54.5	191	52.8	91	32.0	85	28.4	73	26.8
	Dome	187	58.2	177	53.2	150	65.0	71	29.0	67	26.9	56	28.3
	Big toe
	Basic	185	83.8	181	96.5	170	92.7	68	33.3	64	35.3	59	36.2
	Varus	182	8.4	170	95.3	146	98.1	63	35.8	61	39.7	51	39.2
	Valgus	170	77.5	169	75.1	171	97.7	58	27.4	57	30.1	58	35.2
	Dome	170	101.6	180	110.9	165	102.2	57	32.2	64	38.3	65	40.0

Means and standard deviations (SD) of peak pressure and pressure time integrals for the basic insole and the 11 configurations in two 4×3 matrices per region (data averaged over left and right fore feet: means). Basic conditions are in printed bold.

3. Results 

The demographic characteristics of the patients are presented in Table 1. In the overall F-ratio test the repeated measures ANOVA showed no statistically significant difference between left and right feet for the effect on peak pressures and pressure time integrals (P⩾.194). Therefore, we averaged values for each region over both left and right feet for reporting of the results, although both feet were analyzed separately. Table 2 show means and standard deviations of peak pressure and pressure time integrals for the basic insole and the 11 configurations in two 4×3 matrices per region. The values of the basic condition compared to the experimental conditions (individual cell means, Table 2, Table 3) can be calculated from the values in the tables, while the main or overall effects (based on marginal means) are cited in this section.

Table 3. Scores of walking convenience

		Basic			Arch support			Extra arch support
		Mean	SD	min–max	Mean	SD	min–max	Mean	SD	min–max
	Basic	7.5	1.00	5.5–9.5	7.5	1.10	5.5–9.0	6.7	1.43	3.0–9.0
	Varus	7.6	1.21	5.0–9.0	7.2	1.53	5.0–9.5	6.5	1.55	4.5–9.0
	Valgus	7.7	1.46	5.0–10.0	7.4	1.38	4.8–9.5	7.0	1.43	5.0–9.0
	Dome	7.1	1.03	5.0–9.0	7.3	1.33	5.0–10.0	5.9	1.57	3.0–8.5

Means and standard deviations (SD) and range (min–max) of walking convenience. Basic condition is in printed bold.

There were no statistically significant effects of the insole configurations on peak pressure in the lateral forefoot region, except for the main or overall effect of a metatarsal dome: −14kPa, (P⩽.001). For the central forefoot region, the main effects of a metatarsal dome: −32kPa, standard arch support −17kPa, and extra arch support −45kPa, were statistically significant (P⩽.001). The effect of extra arch support was also statistically significant when compared with standard arch support: −28kPa (P⩽.001). For the peak pressure in the medial forefoot region, the effects of a varus wedge −9kPa, metatarsal dome −42kPa, standard arch support −12kPa, extra arch support −38 kPa, and the difference between both types of arch support −26kPa, were statistically significant (P⩽.036).

There were no statistically significant effects of the insole configurations on peak pressure in the big toe, except for the effect of the combination of an extra arch support and a varus wedge: −52kPa (P=.017).

There were no statistically significant effects of the insole configurations on pressure time integrals in the lateral forefoot region, except for the overall effect of a metatarsal dome: −6(kPa)s (P=.016). For the central forefoot region, the effects of a metatarsal dome: −17(kPa)s, standard arch support −9(kPa)s, and extra arch support −27(kPa)s, were statistically significant (P⩽.001). Also the difference between both types of arch support 18(kPa)s, was statistically significant (P⩽.001). For the pressure time integrals in the medial forefoot region, the effects of a metatarsal dome −19(kPa)s, standard arch support −6(kPa)s, extra arch support −20(kPa)s, and the difference between both types of arch support 14(kPa)s, were statistically significant (P⩽.001). There were no statistically significant effects of the insole configurations on pressure time integrals in the big toe region.

The mean walking convenience scores of a basic insole with a varus and a valgus wedge received the highest ratings: 7.6 and 7.7, respectively, (Table 3). A discrete metatarsal dome was, on average, less appreciated than a basic insole: 7.1 versus 7.5 (P=.037). The basic insole and a standard arch support received equal ratings, whereas the ratings of the combination of a standard arch support with other components lowered to 7.2. The self-reported walking convenience appeared to deteriorate with application of an extra arch support and gradually worsened by adding subsequently a varus wedge and a metatarsal dome: 6.7, 6.5 and 5.9, respectively, whereas the application of a valgus wedge resulted in a slight improvement: 7.0. The effects of the extra arch support and the metatarsal dome on walking convenience were statistically significant (P⩽.037). There was no interaction from both components and there were no statistically significant effects from both wedges.

4. Discussion 

We studied the effects of insole configurations on plantar peak pressure, pressure time integral and walking convenience in diabetic patients with neuropathic feet. A repeated measures ANOVA approach was used to resolve the problem of left and the right feet in one patient. First-order interactions between each experimental factor of interest and the factor ‘side’ (left and right feet) turned out to be statistically not significant. Therefore, we consider that the experimental effect in one foot is not different from that in the other.

A metatarsal dome and an extra arch support significantly reduced peak pressures and pressure time integrals in the central and medial forefoot regions compared to the basic insole. The largest reductions were accomplished with a metatarsal dome in combination with extra arch support. The reductions achieved with a combination of components seem to result from an additive effect of the independent components. Only for the big toe region an ‘extra effect’ superimposed on the additive effect was achieved. The effects of the insole components in the lateral and big toe regions were smaller and, except for the effect of dome on the lateral region, not statistically significant. For the big toe region, this may be due to the larger variation of the plantar pressure data. Both varus and valgus wedges resulted in minor, statistically non-significant effects, apart from the effect of the varus wedge in the medial region for peak pressures. The effects of the components on pressure time integral showed a tendency similar to that of the effects on peak pressures.

Walking convenience is important for the patient’s acceptance of foot orthoses for daily use. The results of our study indicate that walking convenience was best with basic insoles. An increase of the arch height and the application of a metatarsal dome lowered walking convenience. The varus and valgus wedges had no statistically significant influence on walking convenience, although some patients reported that the application of a wedge improved gait steadiness. Walking convenience was judged immediately after the measurement session. Therefore, it remains unclear whether the patient’s appraisal will be the same after getting used to the orthosis.

Most previous studies on the effectiveness of foot orthoses therapy on plantar pressure were focused on high-risk patients and evaluated the effectiveness of total contact casts, orthopedic or therapeutic shoe wear and special devices like cast walkers and half-shoes (Chantelau, 2001, Chantelau et al., 1990, Lobmann et al., 2001, Mueller et al., 1989, Praet and Louwerens, 2003, Birke et al., 2004, Bus et al., 2004, Armstrong et al., 2001, Busch and Chantelau, 2003, van Schie et al., 2000). Consequently, most information about effectiveness of off-loading therapy is related to the healing of ulcers and prevention of re-ulceration in high-risk diabetic patients. Contrary to treatment and secondary ulcer prevention, primary ulcer prevention has received little attention. This is remarkable, because prescription of foot orthoses to relief local plantar pressure could be a practical, inexpensive and acceptable preventive treatment for a large population of neuropathic patients who wear comfort or ‘off the shelf’ shoes. The few studies that evaluated orthoses for comfort shoes, could not provide clear guidelines due to the heterogeneity in study populations, outcome measures and orthosis design (Cavanagh, 2004, Spencer, 2001, Wooldridge et al., 1996, Reiber et al., 2002).

There are six studies that evaluated insole configurations for the reduction of plantar pressure. In subjects with normal feet, Abu-Faraj found an effect of an arch support on peak pressures in the big toe and the heel regions, while Hayda and Chang showed an effect of a dome on forefoot pressure (Chang et al., 1994, Hayda et al., 1994, Abu-Faraj et al., 1996). Also, in studies on rheumatic patients and patients with metatarsalgia it was found that a metatarsal dome or pad reduced forefoot peak pressures (Holmes and Timmerman, 1990, Hodge et al., 1999). However, in both feet of diabetic patients with unilateral big toe amputation, no statistically significant effects of arch supports or metatarsal domes on plantar pressure were found (Ashry et al., 1997). Furthermore, studies on the effects of wedges on plantar pressure reported statistically significant reductions for discrete forefoot areas (Milani et al., 1995, Rose et al., 1992, Van Gheluwe and Dananberg, 2004). These studies were performed with either heel and forefoot wedges while we used full-length wedges. During the rollover process, the torsion between the rear- and forefoot could be less with full-length wedges, leading to different loading patterns than ‘partial’ wedges. This may explain why we did not find an effect of wedges.

We did find a pressure reducing effect of both arch support and a metatarsal dome. Unfortunately, there is no absolute or relative standard for valuing the reduction of inshoe plantar pressure. As a ‘rule of thumb’, 20% pressure reduction in relation to the base line condition is considered as relevant (Jackson et al., 2004, Lobmann et al., 2001, Albert and Rinoie, 1994, Postema et al., 1998). In this study, a dome in combination with an extra arch support accomplished about 39% peak pressure reduction compared to de basic insole condition. Optimization of shapes, height, location and material characteristics may lead to better results. For example, a wider dome, similar to a metatarsal bar, could be expected to achieve larger reduction in the lateral and medial regions. Application of an insole material under the metatarsal heads with different mechanical properties could also enhance metatarsal pressure relief (Cavanagh et al., 1996).

The results of our study are a step towards developing an evidence-based algorithm for the construction of optimal orthoses in therapeutic shoe design, as advocated by Boulton and Dahmen (Dahmen et al., 2001, Boulton and Jude, 2001). Further research is required to evaluate the effects of insole materials and combinations of insole components for pressure reduction and walking convenience in different types of shoes.

5. Conclusions 

For non-deformed flexible neuropathic feet, both varus and valgus wedges resulted in minor, statistically non-significant, effects. Larger reductions of peak pressures and pressure time integrals were achieved in the central and medial forefoot regions through application of a metatarsal dome and/or an arch support or an extra arch support. The dome and the arch supports have a considerable plantar pressure reducing effect, but the combination of a dome and extra arch support resulted in the best plantar pressure reducing effect. Unfortunately, this combination of insole components was not very well appreciated by the patients. Therefore, walking convenience must be taken into account when designing insoles with a metatarsal dome and/or arch support.