Introduction

In adults, symptoms of chronic neurological disorders are one of the most common reasons for physician visits [1]. Evaluation of sensory symptoms is one of the five most common reasons to visit to a neurologist. The prevalence of neuropathy in general population is 2.4% and increases to 8% in those older than 55 years [2, 3].

Diabetes mellitus is a global health issue affecting people of all age groups with prevalence gradually increasing to epidemic frequency [4]. This increase marks a parallel increment in the incidence of its micro- and macroangiopathic complications and the economic burden posed by these. Of these complications, diabetic neuropathy tends to be the most disabling and is an important therapeutic challenge for diabetologists and primary care physicians [5]. Diabetes is one of the most common causes of peripheral neuropathy. Neuropathic prevalence in diabetes is 8% at the time of diagnosis [6, 7] and increases up to 66% with increased disease duration.

Chronic hyperglycemia-induced vascular and metabolic derangements are postulated to be the cornerstone for the pathogenesis of typical distal sensorimotor diabetic polyneuropathy, while intercurrent, painful neuropathy involving small fibers constitutes the chronic idiopathic axonal polyneuropathy (CIAP). Approximately one-third of the total patients with diabetes mellitus are thought to develop diabetic sensorimotor polyneuropathy and its subsequent complications [8]. Previously published literature reports chronic hyperglycemia to account for only a third of the total variance in severity of microangiopathic diabetic complications [9]. Impairment of lower-extremity sensations in addition to pain results in significant balance and movement disorders in a sizeable diabetic population. Neuropathy can predate the development of overt hyperglycemia or can occur at any point over the dysglycemia continuum.

Hyperglycemia, age, duration of diabetes, tobacco use, dyslipidemia, hypertension (especially diastolic), increased height, presence of cardiovascular disease (CVD), presence of severe ketoacidosis and presence of microalbuminuria (indicating early renal dysfunction) [10–12] are thought to be independent predictors for development of diabetic neuropathy, but there appears to be scarcity of literature studying the correlation of these factors to severity of diabetic peripheral neuropathy and the strength of such association.

Physicians have to deal with the specific challenge of effectively screening for diabetic neuropathy at an early stage, generally with 2–3 min tests, to stratify patients with symptoms, decide about specialty referral and treat the distressing symptoms accordingly. Several clinical and diagnostic techniques are used to confirm the presence of diabetic neuropathy and gauge its severity. Of these, a notable technique is the Michigan Neuropathy Screening Instrument (MNSI) which is a simple, cost-effective and validated bedside tool for detection of diabetic neuropathy. One study found it to have a sensitivity, specificity, accuracy and kappa value (over vibration perception threshold (VPT) and 10 g monofilament examination) of 78.15, 88.43, 83.33 and 0.67%, respectively [13]. The study employs this tool to gauge the presence and severity of diabetic polyneuropathy.

Aims and objectives

  1. 1.

    To study the prevalence of diabetic neuropathy in type 2 diabetics and describe its severity and frequency of different manifestations.

  2. 2.

    To study the correlation between severity of diabetic neuropathy and patient-related (age, gender, history of smoking, co-morbid illnesses, BMI, waist-to-hip ratio, lifestyle habits, family history of diabetes) and disease-related variables (duration of diabetes, control of diabetes, random and fasting blood sugar readings) and determine whether one or more of these can be used as predictors of severity of this disabling diabetic complication.

Materials and methods

Over a study period of March–June, 2015, all diabetic patients presenting to the outpatient department of KRL General Hospital and Centre for diabetes and liver diseases, Islamabad, were chosen to enter this study using convenient sampling. A total of 800 such patients were then screened for the presence or absence of diabetic neuropathy. All patients were required to provide informed consent. The hospital ethical committee approved the study. All patients were asked simple affirmative/negative questions relating to the presence of diabetic neuropathy using standardized tools (MNSI Performa provided in Annexure 1), and an MNSI score of >2 was used to diagnose diabetic neuropathy. Patient-related variables were noted. A thorough physical examination was performed on the index visit according to MNSI protocol.

It was a cross-sectional study.

Data were analyzed using SPSS v16.0. Qualitative variables were reported using percentages and quantitative variables by using mean ± standard deviation. Pearson and Spearman correlation coefficients were used to determine the presence and strength of correlation between outcome measure and parametric and nonparametric variables, respectively, and multivariate linear regression applied to determine predictors of diabetic neuropathy and reported using ANOVA (Analysis of variance). A p value of <0.05 was considered to be statistically significant.

Results

Demographic data

Out of the 800 diabetic patients screened for neuropathy, 90 were found to have diabetic neuropathy (11.25%). The mean age of the patients with diabetic neuropathy was 54.47 ± 10.87 years (range 27–86 years). Forty-one (45.5%) patients were males and 49 (54.4%) females. 71% of the population did not smoke. Coexistent morbidities included hypertension in 47.8% of the patients, ischemic heart disease in 7.8%, history of prior cerebrovascular accidents in 2.3% and chronic kidney disease (diabetic or otherwise) in 2.3%. 78.6% of patients with diabetic neuropathy had a family history of diabetes mellitus. The mean duration of diabetes was 11.12 ± 9.8 years in the population with diabetic peripheral neuropathy (minimum—less than a year, maximum—62 years). All of the included patients were type 2 diabetics. 42.5% had a sedentary lifestyle, and 57.5% led a relatively active life. 76.5% did a weekly self-monitoring of blood glucose (SMBG) and 23.5% kept no such records. 34% were on insulin therapy for blood sugar controls. Average height was noted to be 161.67 ± 17.42 cm (range 106–249 cm), and the average weight was 85 ± 8.9 kg (range 51–98 kg). The average BMI was 29.59 ± 4.26 (range 22–41). Average waist/hip ratio was 0.98. Among the group, the average fasting blood sugar was 172 ± 71 mg/dL (minimum 72, maximum 415) and random measurements ranged between 95 and 537 mg/dL (average 256 ± 106). The average HbA1c value was 9.129 ± 2.4% (range 5.4–14%).

Symptomatology

63.6% patients had numbness in their legs and/or feet and 52.3% patients reported a pricking sensation in legs and feet. 23.3% patients had an enhanced sensitivity to touch in the affected body parts. 51.2% patients reported as having burning or electric current-like sensation in the peripheries. Muscle cramps were reported by 53.8% patients. 54.4% patients had legs that hurt on walking. 84.8% could perceive the difference between hot and cold using affected parts of the body.

61.5% patients said that they had generalized body weakness for most part of the day, and a total of 50% patients had worsening of the above-noted symptoms at night. 32.5% patients had dry, cracked feet. 14.3% reported a prior history of open wound/ulcer over their feet. 4.7% had had an amputation. Only 13.3% of the patients had previously been informed by their physicians about them having diabetic neuropathy as a complication of their disease (Table 1). The average MNSI score was 4.72 (range 2–10). 21.4% had been educated about foot care by their physicians. 11.8% had had prior access to health care. 7.1% reported social isolation due to a variety of medical and non-medical conditions. 28.6% reported bare foot walking; 37.2% had unsteady gait.

Table 1 Neuropathy symptomatology reported by the patients
Full size table

Physical examination

Of the sample population, 66.7% had had a single previous complete physical assessment and 33.3% had had two. 84% patients had normal feet on physical examination, 14.7% had deformities on either of the two feet and 38.7% had dry skin/callus. 10.7% had an infected wound on either of the two feet and 6.8% had ulcerations.

Ankle jerks were positive in 78.7% of the patients without reinforcement, present with reinforcement in 16% of the patients and absent in 5.3%. 6.8% of patients had an absent sense of vibration, 20.3% had reduced and 72% had a normal sense of vibration. Fine touch (monofilament) was intact in 63.2% of the patients reduced in 23.7% and absent in 13.2%. Average MNSI score on physical examination was 4.5 (range 0.5–9). 8.8% of these patients had asymmetrical findings. 5.1% had loss of mobility in either of the two feet.

7.2% of the patients had a severely reduced quadriceps femoris strength (MRC 0–2), 66.7% had moderately reduced (MRC 3) and 26.1% had a mildly reduced strength. 1.4% had severely reduced tibialis anterior strength, 68.6% had moderately reduced strength and 30% had mildly reduced strength.

Peripheral vascular disease

Right-sided dorsalis pedis pulse was impalpable in 1.2% of the patients, reduced in intensity in 13.6% and normal in 85.2%. Left dorsalis pedis was impalpable in 1.2%, reduced in 15.9% and normal in 82.9%. 1.2% had an absent right posterior tibial pulse, 12.5% had a weak and 86.2% had a normal right posterior tibial pulse. Similarly, 1.2% had an absent left posterior tibial pulse, 11.2% had weak and 87.5% had a normal left posterior tibial pulse.

Correlation between severity of neuropathy and disease/patient variables

Using multivariate regression, severity of diabetic neuropathy was found to have no correlation with gender, history of smoking, co-morbid illnesses, BMI, waist/hip ratio and random and fasting blood sugar readings as illustrated in Table 2.

Table 2 Variables with a lack of statistically significant correlation with severity of DPN
Full size table

However, a significant positive correlation was found between duration of diabetes and severity of diabetic neuropathy (r(75) = +0.699, p = 0.01). This finding is shown in the graph below (Fig. 1). Using a linear regression analysis, duration of diabetes was found to be significant predictor of severity of diabetic neuropathy (total MNSI score) (F(1, 71) 13.385 p = 0.003, r 2 = 0.231, adjusted r 2 = 0.489) and could account for 45.2% of total variance in the individual MNSI score by ANOVA.

Fig. 1
figure 1

Statistically significant correlation between duration of diabetes and severity of diabetic neuropathy (total MNSI score)

Full size image

Other variables correlating well with DPN severity were glycemic controls as expressed in terms of HbA1c (Fig. 2) and age (Fig. 3). For HbA1c, F(2,13) 11.145, p value 0.002, adjusted r 2 = 0.575, and for age F(3, 12)13.835, p value 0.000, adjusted r 2 = 0.720, denoting that duration of diabetes, glycemic controls and age can together account for 72% of total variance in MNSI scores (Table 3).

Fig. 2
figure 2

Statistically significant correlation between severity of diabetic neuropathy (total MNSI score) and HbA1c levels

Full size image
Fig. 3
figure 3

Statistically significant correlation between severity of diabetic neuropathy (total MNSI score) and age

Full size image
Table 3 Variables with a statistically significant correlation with severity of DPN
Full size table

We also found a statistically significant correlation between DNS scores and severity of diabetic neuropathy and noted that MNSI score on history was a significant predictor of MNSI score on physical examination and could account for 37.1% of the total variance in physical MNSI score (p = 0.00).

Lifestyle habits were found to correlate with symptomatology of diabetic neuropathy (MNSI score on history) but not with physical examination findings (MNSI score on physical).

Discussion

Sensorimotor diabetic peripheral neuropathy is a frequently encountered complication of diabetes. In clinical practice, DSMN (diabetic sensorimotor neuropathy) has historically been diagnosed using symptoms, signs and electrophysiological testing. Quantitative sensory testing and consideration of intra-epidermal nerve fiber density and corneal confocal microscopy [14] are some of the other useful diagnostic tests [15]. The requirement of specially trained and experienced personnel or the use of invasive testing (for nerve or skin biopsies) makes such testing inconvenient, and extensive use in population-based real-world practice has fallen out of favor. The MNSI is one such reliable and simple screening tool which bypasses these difficulties on patients’ and physicians’ part.

Diabetic neuropathy and its complications cause significant morbidity. Symptoms are distressing, refractory to treatment and loss of protective sensation which result in increasing risk for foot ulceration and lower limb amputation. Subclinical neuropathy at an asymptomatic stage can explain the lack of beneficial effect of more intensive treatment/type of treatment.

The study was conducted at diabetic clinic, KRL General Hospital, Islamabad, and CDLD (Centre for diabetes and Liver Diseases) which caters to patients of all socioeconomic strata, from Rawalpindi, Islamabad and its peripheries, to delineate the presence of neuropathy in patients with diabetes and describe its severity and correlates.

A total of 800 patients with diabetes were screened for neuropathy using a standardized tool—MNSI [16], and two trained clinicians. The prevalence of neuropathy in the cohort was found to be 11.25%. This is consistent with the previously published figures of 8% at the time of diagnosis increasing to 66% in those with increased disease duration [6, 7] although the cohort contained patients with both recent and relatively older diabetes diagnosis. Previously, prevalence as high as 59% has been reported [17].

The presenting symptoms in order of decreasing frequency were found to be numbness in legs/feet (63.6%), generalized body weakness (61.5%), pain, burning or electric current-like sensation in peripheries (51.2%) followed by muscle cramps which were found in 50.8%. This contrasts with the findings of a locally published study [18] which showed generalized pain as the most common symptom (8%) followed by numbness in 3%. 13.3% patients had previously been informed by their physicians as having diabetic neuropathy, and 11.8% had had prior access to health care. These observations show a stark contrast to those reported by George et al. [19] who found good knowledge score for diabetic neuropathy and foot care in 75% of patients reflecting a major need on the part of diabetologists and primary care physicians to make efforts to increase patient awareness about this illness. The MNSI score for symptomatology averaged at 4.72 (range 2–10). Comparison of this finding to previously known facts was not possible as the authors could not find any similar study in this regard.

On detailed physical assessment, the most common findings were dry skin/callus (38.7%) followed by deformities on either of the two feet (14.7%) and infected wound/ulcer in 10.7%. Loss of vibration perception which is considered the earliest manifestation of DPN was found in 6.8%, and vibration perception was reduced in 20.3% of the patients. Monofilament/fine touch was absent in 13.2% and reduced in 23.7%. Average MNSI on physical examination was 4.5. 7.2% had severely reduced quadriceps femoris strength, while 66.7% had moderately reduced muscle strength. 1.4% had severely reduced tibialis anterior strength, and 68.6% had moderately reduced muscle strength. Dorsalis pedis pulse was absent/weak in 14.8% of the patients, while 13.7% had a weak/absent posterior tibial pulse. There appears to be scarcity of literature studying the frequency of occurrence of these physical findings. The asymmetric nature of findings in 8.8% of the patients is comparable with 13% noted in a similar study [20].

As described earlier, impairment in glucostatic pathways and subsequent chronic hyperglycemia contributes significantly to the axonal damage that underlies the development of diabetic polyneuropathy. It is, therefore, sensible to imply a correlation between average or single blood glucose readings and the severity of diabetic polyneuropathy. Our study, however, fails to simulate the pattern shown in a previously published study which states glycemic control as a better predictor of severity of diabetic polyneuropathy than the duration of diabetes [21] by demonstrating a stronger correlation between the severity of diabetic neuropathy and duration of diabetes than with the glycemic controls. But, one study [22] has reported a failure of near normoglycemia in preventing the development of diabetic complications, hence indirectly supporting our observation. Also such an association between diabetic controls and development of diabetic neuropathy is said to be weaker in type 2 diabetics (the population included in this study) as compared to type 1 diabetics hence supporting our findings [23]. Discordance in impact of prior treatment may reflect deficiencies in susceptibility of small or large nerve fibers to glycemic exposure. In general, findings from studies of diabetic neuropathies have to be interpreted with caution given the broad range of diagnostic methods employed and lack of consistency in the criteria used to define neuropathy.

DCCT/EURODIAB [24] also reported that HbA1c was an important determinant of neuropathy incidence. DCCT/EURODIAB confirms that glycemic control is a significant and robust predictor of neuropathy. Current strategies for optimizing glucose control are insufficient to fully prevent or delay the development of neuropathic complications.

Conclusion

Duration of diabetes, rather than the level of control of diabetes may be a better predictor of development and severity of diabetic peripheral neuropathy. This fact undermines the importance of achieving near normoglycemia and emphasizes the need of primary prevention and risk factor control to delay the development of diabetes in order for our population (which is facing an increased risk for developing this disease) to avert this debilitating diabetic complication.