Disclosure: This project was supported by FAPESP Grant 2004/09585-2

The loss of protective sensation due to the consequences of the diabetic peripheral neuropathy (DPN) is one major risk factor for ulceration in diabetic feet.1–5 The prevalence of DPN is highly correlated to the duration of the diabetes, aging, and metabolic control.6 Approximately 20% of patients with diabetes will develop clinically significant DPN within 10 years of onset,7 and this frequency can increase to 50% after 10 or 15 years.8 Ulceration and amputation are the most common complications of DPN and are responsible for the majority of hospitalizations among patients with diabetes.9 Furthermore, these amputations cost R$18.2 million (Brazilian Reais) to the Brazilian health system annually, which is equivalent to $8.4 million (US).10
Ulcerated feet have decreased plantar sensitivity.11 Research shows that patients with an ulcer due to diabetes have had a longer duration of diabetes and are older than their nonulcerated counterparts with diabetes.12 In the same study,12 ulcerated feet had higher vibration perception threshold values and poor sensitivity to the monofilament test. The 10-g monofilament is considered a good instrument to assess the loss of protective sensation in all patients with diabetes. The International Consensus on the Diabetic Foot recommends its use.13 It has high reproducibility and specificity,3 and it can predict risk for ulceration and amputation.5,14,15 The motor nerve conduction velocity test can also predict ulceration and mortality in people with diabetes.15 Other authors demonstrated that the induced plantar insensitivity can lead to higher mechanical loads under the foot during gait.16,17 Therefore, the more insensitive a foot becomes, the more predisposed it is to ulceration. Such patients may also present other symptoms such as, numbness, prickling, muscle cramps, and burning sensation, indicating loss of protective sensation.
The new approach of this study is to determine if patients with a history of plantar ulcers due to the DPN have more symptoms, a longer duration of disease, and poorer foot sensitivity. These 3 parameters have not been previously studied together in ulcerated patients with diabetes.
Many health professionals who care for patients with diabetes use tools such as, the Michigan Neuropathy Screening Instrument questionnaire [MNSI-q]),18 to evaluate symptoms and signs of neuropathy. Ordinarily, these forms are effective and easy to apply and can serve as the primary screening indicator for assessment of neuropathy. It can, therefore, be confirmed that symptoms are related to the loss of sensitivity in ulcerated patients— this would be an important indication in the process of diagnosing patients at risk of foot ulceration; otherwise professionals should carry on with electrophysiological tests to confirm this risk.
The purpose of this study was to compare the duration of disease and the prevalence of diabetic neuropathic symptoms. This includes investigating plantar insensitivity and its prevalence by assessing sensitive chronaxie, tactile and thermal sensitivities, among diabetic neuropathic subjects with and without previous history of plantar ulcers.This will determine if ulcerated subjects with diabetic neuropathy have an increased sensory deficit in their feet, which may lead to ulcer formation. In addition, the authors correlated the symptoms of neuropathy with the sensory tests results to determine their relationship. It is hypothesized that the ulcerated group will have decreased plantar sensitivity, longer duration of diabetes, and an increase in neuropathic symptoms.

Subjects. Sixty-six subjects volunteered to participate in this prospective study.Twenty-five were healthy subjects and 41 subjects had diabetes. Fourteen subjects with diabetes reported previous history of foot ulceration. Eleven subjects with diabetes with no history of ulceration were excluded from the study because they did not present with all inclusion criteria as discussed below. Moreover, in an attempt to match the groups in anthropometric and demographic characteristics such as age, mass, and gender, the authors also excluded 6 subjects from the control group and 5 from the ulcerated diabetic group.
This prospective study involved 44 voluntary adult men and women divided into 3 groups: control group (CG) (n = 19), diabetic neuropathic group (DG) (n = 16), and diabetic neuropathic group with previous history of plantar ulceration (UDG) (n = 9). Ethics approval was obtained from the Education and Research Committee, School of Medicine, University of São Paulo.The subjects included in the CG were employees of the School of Medicine, University of São Paulo, Brazil where the study was conducted; some were accompanying the patients.
Physicians diagnosed all neuropathic subjects (DG and UDG).The following inclusion criteria was adopted: the patients must have type 1 or type 2 diabetes and be 5 years post onset with at least 2 plantar areas with tactile insensitivity to the 10-g monofilament,5,12,19 and a score minimum of 6 on the Michigan Neuropathy Screening Instrument questionnaire (MNSI-q).18,20–22 The MNSI-q is a validated instrument for screening the symptoms related to the diabetic neuropathy. In addition, the UDG should have had at least 1 case of plantar ulceration in the previous 2 years. The exclusion criteria adopted for all studied groups included: subjects older than 65 years; subjects with partial or total amputation of the foot; the presence of peripheral or central neurological disease out of the diabetes etiology; presence of peripheral vascular disease prior evaluated by UniversityHospital team; history of alcohol abuse; traumatic ulceration that could be immediately realized by the subject; and plantar ulcers present at the initial evaluation.
Procedures. This research was part of a project granted by a state financial Agency (FAPESP n. 2004/09585-2) and it was developed at the Laboratory of Movement and Human Posture of the Physical Therapy, Speech and Occupational Therapy Department, School of Medicine, University of São Paulo, Brazil. The procedures were explained to the subjects and they signed an informed consent form before participating in the study. A preliminary investigation was completed to verify all inclusion criteria. Next, the subjects were distributed into the 3 study groups.
The MNSI questionnaire was used to assess the characteristics of neuropathy and contained 15 “Yes” or “No” questions concerning symptomatology. This questionnaire has been used previously in patients with diabetic neuropathy.20–24 Examples of questions were: “Are your legs and/or feet numb?” and “Do you ever have any burning pain in your legs and/or feet?”The maximum score was 13 points following the MNSI instructions.18 The 2 questions that were not computed in the total score concern a measure of impaired circulation and a measure of general asthenia—“Do you get muscle cramps in your legs and/or feet?” and “Do you feel weak all over most of the time?”The other 13 questions investigate only typical sensory symptoms of diabetic neuropathy such as numbness, paresthesia, pain, burning,and some questions involving history of the presence of open wounds in feet, ulceration, amputation, and dry skin.
Additionally, a modified MNSI form was used to investigate the foot’s dermatologic and anthropometric characteristics, looking for evidence of excessively dry skin, callus formation, fissures, frank ulceration, or deformities such as flat feet, hammer toes, overlapping toes, hallux valgus, joint subluxation, and prominent metatarsal heads. The ankle reflexes (Achilles tendon reflex) were also examined by the MNSI form using an appropriate reflex hammer. The original MNSI form includes vibration and tactile perception, which were excluded from the authors’ screening because the vibration tests could not be performed, and the tactile perception with 10-g monofilament was conducted in a second part of the study not reported here. Accordingly, the score on the MNSI form was based on a total of 6 points.
The evaluation of plantar sensitivity was composed of 3 sensory tests: sensitive chronaxie, and thermal and tactile sensitivities.The plantar sensitivity was measured in both feet in 5 areas of the plantar surface: medial forefoot, concerning first and second metatarsal heads (medial plantar nerve); lateral forefoot, concerning third, fourth, and fifth metatarsal heads (lateral plantar nerve); midfoot (principal ramification of lateral plantar nerve); heel (calcaneus nerve).Ten total areas were measured.

Sensitive chronaxie was determined in the plantar surface. Subjects were asked to lie on a clinical table in a prone position while their feet rested on a cushion. Using a universal pulse generator, NeMESys (Quark, Piracicaba, Brazil; Figure 1), electrical pulses were applied with twice the rheobase intensity in order to find the minimum pulse duration that the subjects couldperceive in the selected areas.25 The electric current applied to determine the rheobase was monopolar with electric pulses of 1000 milliseconds (ms) by 2000 ms during the pulse intervals (Figure 2).
The patient was informed of the possible sensations that he or she might feel. Subjects were instructed to inform the examiner as soon as the stimulus was felt in the area surrounded by the pulse generator pen. In the first step, the generator provided results related to the intensity of stimulus (the rheobase). Secondly, with the pulse generator, the examiner will measure chronaxie using the rheobase (the first measure given) as a reference. All the procedures were conducted keeping the pen in contact with the foot on the same area applying the same pressure in order to avoid disturbance in foot sensation.
The mean values of sensitive chronaxie were compared among groups, and for each subject in each group, the number of areas that achieved a sensitive chronaxie value higher than 0.30 ms was recorded.The number of subjects per group who achieved an abnormal value in each plantar area was also measured. The value of 0.30 ms was chosen based on previous studies26,27 and represents the normal value for asymptomatic subjects.

The tactile sensitivity based on cutaneous perception threshold was determined using a set of 3 Semmes- Weinstein monofilaments (Sorri Bauru, Bauru, Brazil) (4.17/5-g; 5.07/10-g; 6.10/300-g; Figure 3B). In each area, the frequency of subjects who could not perceive the pressure stimulus from the 10-g monofilament was recorded.The right foot was analyzed separately from the left.The number of areas with deficient perception to the 10-g monofilament was also included in the 10 total areas (both feet).This number was expressed using the median for further comparison among groups.
The thermal sensitivity was measured to verify if the subjects could differentiate cold and hot temperatures. Two metallic, 0.5-cm diameter, pen-shaped instruments were applied to each of the 10 plantar areas (Figure 3A). The object was warmed on a candle until a safe temperature was achieved to avoid damaging the skin.The other object was cooled with iced water. The same object at ambient temperature was applied over the plantar surface to demonstrate a neutral reference to the subjectbefore tests. For each area of the foot, the number of subjects unable to discriminate temperature differences was counted. As in the tactile sensitivity test, the right foot was analyzed separately from the left. The number of areas with deficit sensitivity to both temperatures was also observed, and the median was calculated for each group.
Statistical analysis. All the statistic analyses were performed using the Statistica software package. After checking data normal distributions by the Shapiro Wilks test and homoscedasticity by Levine’s test, the inferential analysis of the chronaxie among groups and among areas was made using an ANOVA 2-way test matching the 3 groups and the 5 areas of each foot analyzed. Age and body mass index (BMI) were also analyzed using an ANOVA 1-way test. The Scheffé test was used as post-hoc comparisons when significant differences were observed in the ANOVA tests. When only the 2 diabetic groups were compared, such as in the comparison of the MNSI score, the glycemic level, and duration of diabetes, a t-test was used. Categorical variables (gender, type of diabetes) were compared using Chisquared test for multiple responses. The median number of areas with loss of sensitivity per group was compared using the Kruskal Wallis test because of its ordinal nature. The Spearman correlation test was used to correlate MNSI score between sensitive chronaxie, and tactile and thermal sensitivities. An alpha of 5% was adopted.


General characterization of sample and plantar sensitivity. Table 1 shows anthropometric and demographic data descriptions for the 3 analyzed groups. The neuropathic groups (DG, UDG) were matched in age, gender distribution, and BMI with the control group.
The neuropathic groups were not different in relation to diabetic characteristics and presented a similar number of symptoms according to the score obtained on the MNSI questionnaire described in Table 2. Only 1 subject from DG had type 1 diabetes.
The results regarding the sensitive chronaxie for all groups are shown in Table 3. The subjects with diabetic neuropathy and previous ulcers (UDG) achieved higher values for this sensory parameter and had significantly different results from the other groups. This delay in sensing the electrical stimulus is higher in the UDG. However, the UDG presented significantly higher mean values on the midfoot (P = 0.0001) and medial forefoot (P = 0.0001) in comparison to the CG. The number of areas with sensitive chronaxie values higher than 0.30 ms did not show any difference among groups (P = 0.593). However, the UDG had more areaswith values higher than the threshold. The results were: 2 areas for CG; 3 areas for DG; 5 areas for UDG.

A significant difference (P = 0.00) was found among groups regarding the number of areas (10) where the 10- g of pressure was not felt. The neuropathic groups had more areas than the CG (P = 0.000), and the UDG had more areas compared with GC (P = 0.000) and with DG (P = 0.047). The medians of the number of areas with deficit of tactile sensitivity were: no areas for CG; 2 areas for DG, and 6 areas for UDG. Five subjects from UDG (55.5%) were unable to feel the 10-g monofilament pressure in more than 5 areas of the feet. Three subjects from the DG (18.7%) had the same deficit in more than 5 plantar areas. A significant difference was found among the 3 groups (P = 0.000) regarding the 10 areas where subjects could not distinguish hot and cold temperatures. Both neuropathic groups presented significant differences in comparison to CG (P = 0.000).There was no difference for the number of insensitive areas between neuropathic groups (P = 0.249).The medians for the number of areas with inability to discriminate hot and cold temperatures were: zero areas for CG, 2 areas for DG, and 3 areas for UDG.
The MNSI total score achieved a weak and insignificant correlation with sensitive chronaxie (r = -0.17; P > 0.05) and with thermal sensitivity (r = -0.11;P > 0.05) for all areas. Only in areas of the midfoot and heel were moderate and inverse proportional correlations found to the tactile sensitivity (midfoot: r = -0.62, P = 0.001;heel: r = -0.42, P = 0.038).Thus, the higher the MNSI score is and the more neuropathic symptoms that patients report, the worse the tactile sensitivity.

This study investigated the presence of symptoms related to DPN, the disease duration in 16 subjects with diabetic neuropathy, and 9 subjects’ diabetic neuropathy and history of plantar ulcers. Plantar sensitivity was measured to determine if ulcerated subjects would exhibit a decreased sensitivity during sensory tests.
Generally it was found that the ulcerated group did not present longer disease duration or had more symptoms than subjects without a history of plantar ulcers. It is known that DPN is a consequence of diabetes mellitus and its progression and the risk of ulceration increases with time, and therefore, with increasing age.8,22 This was not found to be a contributing factor in this study, as the duration of disease did not have a relationship to ulcer appearance. Results similar to a previous study12 were expected. The study found that ulcerated subjects with diabetes presented a longer duration of diabetes in comparison with those without ulcers: 17 ± 9.5 versus 12 ± 10.8 years.
Surprisingly, there was no difference regarding the MNSI score for symptoms and signs of diabetic neuropathy between neuropathic groups. As disease duration increased, an increase in neuropathy symptoms was expected. The presence of ulcers may be indicative of disease progression, but in the current study, the nonulcerated and ulcerated subjects had similar MNSI scores and disease duration. These variables did not indicate severity and could not differentiate the analyzed groups. It was already found a positive correlation between the MNSI score and duration of diabetes (Pearson’s correlation coefficient, r = 0.563;P = 0.005).23 Conversely,another study evaluated diabetic neuropathy symptoms of patients with type 1 and type 2 diabetes using the MNSI instrument.22 Their results showed that type 1 subjects with diabetes with longer disease duration had fewer symptoms compared to subjects with type 2 diabetes.22 These findings demonstrate that disease duration and the number of symptoms cannot always be considered as differential factors for chronic complications, such as the formation of plantar ulcers.The authors believe that the type of diabetes, glycemic control, and the different treatments, such as insulin administration and dietetic control, etc., are relevant to the diabetic neuropathy etiology.
The pathogenesis of ulceration is complex, and involves peripheral vascular disease and DPN with complications such as sensory, autonomic, and motor dysfunctions. This may lead to overloading the feet and may contribute to sores.28,29 In addition, metabolic control is also an important factor in predicting complications in people with diabetes. Poor glycemic control is one contributing factor to the development of DPN.30 According to these results, the ulcerated group showed a trend toward higher glycemic levels. This hyperglycemia may increase sorbitol concentrations31 that may cause alterations on the nerve conduction velocity.32
The sensory tests performed could differentiate ulcerated from nonulcerated subjects.The sensitive chronaxie could assess the loss of somatosensorial sensitivity evaluating indirectly the peripheral nerve conduction velocity. An alteration in this parameter can be caused by diffuse axonal loss or demyelinization.This will lead to higher values of sensitive chronaxie. Higher values were observed in the neuropathic groups in the present study.
The sensitive chronaxie increased in the ulcerated group in comparison to the nonulcerated group in all areas of the foot,and was significantly higher in the medial forefoot and hallux.The loss of sensation on the forefoot and the hallux can interfere with the propulsion phase of the gait cycle, leading to a decrease in normal function. Some authors discussed an alteration in the roll over process of the foot due to insensitivity, which the hallux presents at low loads.17 The authors also discuss the possibility of redistributing loads from regions with sensory deficit to other areas that are more sensitive. They also suggest that the mechanoreceptors of the plantar surface play an important role in gait.17 These data suggest that the ulcerated group had a lower nerve conduction velocity.A decrease in nerve conduction velocity is indicative of risk of plantar ulceration and increases the mortality in people with diabetes.15 Tactile insensitivity measured by the ability to feel the pressure of 10-g monofilament was also significantly higher in the ulcerated group, and agrees with other studies.11,12,33 It was evident from all modalities tested that the ulcerated subjects had a greater deficit in tactile sensation, sensitive chronaxie, and thermal sensation.This result concerning thermal insensitivity was expected due to abnormalities in hot and cold threshold detection is typically the first evidence of peripheral neurological deficits.34 Additionally, the ulcerated group had more plantar areas presenting deficits in sensitivity for the 3 sensory tests.
Moderate incidence of callus was found during initial inspection in the plantar areas of the subjects’ feet. In the ulcerated group 55% of subjects had callus under the hallux, and 33% had callus under the first and second metatarsal heads.The neuropathic subjects without history of ulcers had an incidence of 34% and 31% for the same areas, respectively, whereas the control group for the same areas had an incidence of 13% and 16%, respectively. Based on this information, the authors agree it is feasible that plantar hyperkeratosis has an influence on tactile sensitivity. However, the authors are aware that such alteration is not exclusive to subjects with neuropathy. Even healthy feet, as demonstrated by the CG, can present callus.
The DPN can be present under the subclinical stage on the first phase during the development of the diabetic neuropathy.35 Presently, only electrophysiological measures are able to detect abnormal neurological signs.36,37 Thus, the real deficit of sensitivity cannot be based only on the reported symptoms. As observed, the UDG presented the same level of symptoms in comparison to DG according to the MNSI score. Although the first group also had striking differences regarding insensitivity of their feet,they also had a high sensitivity deficit according to their quantitative sensory tests.This may be because subjects with ulcers have vascular and autonomic alterations, which added to the motor and sensory dysfunctions that lead to the ulceration. Such autonomic alterations were not assessed in the present study. In general, the MNSI may be used as a tool to predict patients with DPN.20 A moderate correlation was shown between MNSI score and tactile sensitivity.However, the MNSI could not discriminate the severity of DPN, since its score did not differentiate neuropathic groups with and without history of plantar ulcers. In the literature it has been assumed that the presence of ulceration in the clinical history of a patient may be an indicator of increased progression of DPN.
This work had a few limitations. The difficulties in recruiting patients with neuropathy and who had a healed ulceration could be due to the high incidence of amputation in patients with previous ulceration, or due to the extensive amount of time plantar ulcers take to heal. In addition, the patients were recruited from hospitals and health centers that did not have a specialized podiatric center for diabetic foot care.Another hypothesis is that some patients that could be included in the present investigation were undergoing an ulcer cicatrization process or were in another outpatient unit,such as endocrinology and metabolism, or plastic surgery, to which the authors did no have access during the study.
A type II error, due to the small sample size of the ulcerated group, could have influenced the study results. This error could partially explain the reason for the lack of difference in duration of diabetes and in MNSI scores between neuropathic groups.A possible beta error could also be hiding the likelihood of a difference in these parameters. However, this limitation did not interfere with the capacity to differentiate both neuropathic groups concerning sensitivity. Other researchers support the idea that a single neuropathic symptom does not form the basis for diagnosis of DPN,but that MNSI should be implemented as a short screening program during the patient’s daily routine.22
The results found in the present study support the use of the MNSI instrument as a first tool to screen the patient with diabetic neuropathy. The MNSI results can help inform the decision as to whether electrophysiological tests should be performed.38 The MNSI form is an easy, fast, and reproducible in 88.8% betweenobservers,39 and shows relative specificity in the range of 75%39 to 83%.38 The MNSI score may indicate the presence of diabetic neuropathy, mainly when a high score is obtained. It has been shown that the MNSI score decreases in the presence of subclinical DPN, which was confirmed after electrophysiological studies.38 Nevertheless, when the scores are low, the clinician should reevaluate the patient to check for possible early stadiation of the diabetic neuropathy or to perform quantitative electrophysiological testing as a follow up strategy to investigate subclinical forms of the disease.
The scores obtained from the MNSI questionnaire and form suggested the presence of diabetic neuropathy as was expected, according to the inclusion criteria. Only the sensitive chronaxie, as an electrophysiological parameter, and the others sensory tests performed could differentiate neuropathic groups and showed that the subjects with ulcers have worse progression of DPN.

The data indicate that patients with diabetic neuropathic ulceration had decreased sensitivity in their feet, which was confirmed by the high number of insensitive plantar areas, and high values of sensitive chronaxie. These findings were not related to the duration of diabetes onset, nor to the symptoms reported by the patients. However, the symptoms may be an indicator of decreased foot sensitivity, but the symptoms alone are not able to differentiate between groups with variation in progression of diabetic neuropathy.

The authors thank FAPESP for the Grant 2004/09585- 2 and Joshua Allen Michael MacNeil for his editorial assistance in preparing this manuscript.