Vitamin D deficiency is associated with painful diabetic neuropathy

The aetiology of painful diabetic neuropathy is unclear. We have evaluated vitamin D levels in diabetic patients with and without painful neuropathy.


| BACKGROUND
The prevalence of painful diabetic neuropathy is 21% and painful symptoms are more prevalent in patients with type 2 diabetes, females and South Asians. [1][2][3][4] Painful diabetic neuropathy is characterized by symmetrical lower limb paraesthesiae, dysaesthesiae, lancinating pains and allodynia with nocturnal exacerbation 5 and is associated with significant sleep disturbance and reduced quality of life. 6 Observational studies in people with diabetes demonstrate a significant association between vitamin D deficiency with paraesthesiae and numbness, 7 the severity of DPN using the neuropathy symptom score, neurological deficits and electrophysiology, 8 impaired nerve conduction velocity 9 and reduced parasympathetic function. 10 In a recent study there was an inverse U-shaped association between serum vitamin D levels and the E/I ratio, 30/15 ratio and three heart rate variability indices. 11 Furthermore, a recent systematic review and meta-analysis showed a significant association between vitamin D deficiency and the development of DPN, particularly in Asian patients. 12 In another study, vitamin D deficiency was related to diabetic painful neuropathy in Greek but not Bangladeshi patients. 13 Shillo et al 14  In relation to a mechanistic link between vitamin D and neuropathic pain, nociceptive calcitonin gene-related peptide (CGRP)-positive neurones have a distinct vitamin D phenotype with hormonally regulated ligand and receptor levels. 15 Vitamin D deficiency results in increased numbers of axons containing CGRP, and in culture, vitamin D receptor (VDR) expression is increased in growth cones, and sprouting appears to be regulated by VDR-mediated rapid response signalling pathways. 16 Nerve growth factor (NGF) is depleted in experimental diabetes 17 and a preservation of NGF expression was shown in sciatic nerves of diabetic animals treated with a vitamin D analogue (CB1093). Similarly, tacalcitol, active vitamin D3, induces NGF production in human epidermal keratinocytes. 18 Treatment with vitamin D3 has been shown to limit demyelination in a cuprizone experimental model of demyelination 19 and in a spinal cord compression model, it has been shown to induce axonal regeneration. 20 In a prospective study of 51 patients with type 2 diabetes and painful neuropathy treated with 2000 IU of cholecalciferol daily for 3 months, there was a 50% decrease in pain scores as measured by the Visual Analogue Score (VAS). More recently in a placebo-controlled study of 112 patients with type 2 diabetes randomized to 50 000 IU of cholecalciferol once weekly for 8 weeks, there was a significant increase in 25OHD and an improvement in the neuropathy symptom score, but no change in NDS or neurophysiology. 21 In a study of people with diabetes given 400 IU of vitamin D daily for 12 weeks, there was a significant improvement in pain, numbness and weakness. 22 We have also shown a significant improvement in neuropathic symptoms 23,24 and quality of life in patients with painful diabetic neuropathy treated with 600 000 IU of vitamin D.
In the present study we have evaluated the levels of 25

| Corneal confocal microscopy
Patients underwent examination with the Heidelberg retina tomography III in vivo corneal confocal microscope employing our established methodology for image acquisition. 26,27 Several scans of the entire depth of the central cornea were recorded using the section mode, which enables manual acquisition and storage of single images of all corneal layers. This provides en face two-dimensional images with a lateral resolution of 2 mm/pixel and final image size of 400 × 400 pixels of the sub-basal nerve plexus of the cornea. Five images per patient from the centre of the cornea were selected and examined in a masked and randomized fashion. 28 Three corneal nerve parameters were quantified: (a) CNFD, the total number of major nerves per square millimetre of corneal tissue (no mm 2 ); (b) corneal nerve branch density (CNBD), the number of branches emanating from all major nerve trunks per square millimetre of corneal tissue (no mm 2 ) and (c) corneal nerve fibre length (CNFL), the total length of all nerve fibres and branches (mm/mm 2 ) within the area of corneal tissue.

| Skin biopsy and immunohistochemistry
Intra-epidermal nerve fibre density (IENFD) was assessed in a subcohort of participants (Controls n = 10, DPN n = 11, and PDN n = 9) who agreed to undergo a 3-mm punch skin biopsy from the dorsum of the foot, 2 cm proximal to the second metatarsal head, after local anaesthesia (1% lidocaine). The biopsy specimen was immediately fixed in PBS-buffered 4% paraformaldehyde and after 18 to 24 hours rinsed in tris-buffered saline and soaked in 33% sucrose (2-4 hours) for cryoprotection. It was then embedded in optimal cutting temperature-embedding compound, rapidly frozen in liquid nitrogen, and cut into 50-μm sections using a cryostat (model OTF; Bright Instruments, Huntington, UK). Four floating sections per subject were subjected to melanin bleaching (0.25% KMnO 4 for 15 minutes followed by 5% oxalic acid for 3 minutes), a 4-hour protein block with a tris-buffered saline solution of 5% normal swine serum, 0.5% powdered milk and 1% Triton X-100, and overnight incubation with 1:1200 Biogenesis polyclonal rabbit anti-human PGP9.5 antibody (Serotec, Oxford, UK). D were added. Following a further incubation step, the magnetic particles were 'captured' using a magnet. After a washing step and addition of trigger reagents, the light emitted by the acridinium label was inversely proportional to the concentration of 25(OH)D in the original sample. The concentration of 25(OH)D was calculated automatically using a 4-point logistic curve. The cross reactivity for vitamin D 2 (of the assay) as per manufacturers assertion was 100% (relative to vitamin D 3 ) and the assay has excellent correlation to existing globally recognized assays, in combination with good sensitivity and precision. 30 The reportable range of the assay was 5-140 ng/mL. Inter-and intra-assay variation of the in-house control was 5.6% and 9.7%, respectively. Vitamin D deficiency (<20 ng/ mL) and insufficiency (<30 ng/mL) were defined according to the Institute of Medicine (IOM) of the National Academies. 31

| Statistical analysis
Statistical analyses were undertaken on StatsDirect (StatsDirect, Cheshire, UK). All values are presented as mean ± SD. ANOVA method or a non-parametric counterpart Kruskal-Wallis was used to assess differences between groups depending on normality of the data. Overall, the P value was maintained at .05 for multiple comparison tests (Bonferroni adjustment or Dwass-Steel-Chritchlow-Fligner pairwise comparison).
Unpaired t test or Mann-Whitney U test were used for analysis for DPN vs PDN for the duration of diabetes and IENFD. Chi-squared analyses were used to assess frequencies of gender, ethnicity and aetiology of diabetes. Odds Ratios for painful symptoms were calculated by further delineating DPN and PDN groups based on the cut offs for vitamin D deficiency (<20 ng/mL) and insufficiency (<30 ng/mL).

| Demographics, metabolic and anthropometric assessment
The participant demographics and metabolic and anthropometric measurements in people with diabetes and control subjects are summarized in Table 1. There were no significant differences in age, gender, BMI, duration or type of diabetes and all subjects were of white European origin. HbA 1c (P < .0001) was significantly higher in those with diabetes compared with control subjects with no difference between patients with PDN and DPN. The total cholesterol was significantly lower in participants with DPN (P = .003) and PDN (P = .02) compared with control subjects, due to greater statin use. HDL, triglycerides, systolic and diastolic blood pressure were comparable between diabetes groups (DPN and PDN) and control subjects. The estimated glomerular filtration rate was comparable between diabetes groups and control participants, but the albumin-creatinine ratio (ACR) was higher in the DPN (P = .009) and PDN (P = .002) groups compared with control subjects with a higher ACR in PDN compared with DPN.
Serum vitamin B12 levels were comparable between groups.

| Symptoms and deficits
The NSP was significantly higher in PDN compared with DPN (P < .0005) and control subjects (P < .0001) ( Table 2)

| Quantitative sensory tests
VPT, CST and WST did not differ between diabetes groups and control subjects or between patients with DPN and PDN (Table 3).

T A B L E 1
Participant demographics and metabolic parameters in control subjects and patients with DPN and PDN, with statistically significant differences between groups

| Electrophysiology
Peroneal nerve conduction velocity and amplitude were significantly lower in DPN (P = .004, P = .003, respectively) and PDN (P = .0008, P = .001, respectively) compared with control subjects, but there was no difference between patients with DPN and PDN (Table 3). Sural nerve conduction velocity and amplitude were significantly lower in DPN (P = .02, P = .007, respectively) and PDN (P = .005, P = .04, respectively) compared with control subjects, but there was no significant difference between patients with DPN and PDN.

| Autonomic function, IENFD and CCM
HRV-DB did not differ between diabetes groups and control subjects or between patients with DPN and PDN (

T A B L E 3
Small and large fibre tests of nerve structure and function in control subjects and diabetic patients with DPN and PDN, with statistically significant differences between groups

| DISCUSSION
Painful diabetic neuropathy is an extremely disabling condition, which may be present in at least one-fifth of people with diabetes. 4 The treatment of this condition is unsatisfactory with a 'good' response to conventional medication rated at between 30% to 50% pain relief. 32 Available drugs are often moderately effective and their use is limited by side effects. Furthermore, recent studies of novel drugs in the treatment of painful diabetic neuropathy have failed to show efficacy. 33,34 There is an urgent need to explore new mechanisms and treatments for diabetic painful neuropathy.
The aetiology of painful diabetic neuropathy is not clear. Central changes comprising of increased thalamic vascularity, 35 Aβ fibre sprouting into lamina II of the dorsal horn and reduced inhibition via descending inhibitory pathways 5 together with axonal atrophy in peripheral nerves have been demonstrated in patients with painful diabetic neuropathy. 36 Painful diabetic neuropathy has also been associated with autonomic dysfunction. 37 Previously, we have shown that the LDI flare, a measure of small fibre function, is abnormal in patients with painful diabetic neuropathy, whereas conventional quantitative sensory testing and dermal nerve fibre density did not differ from those with painless diabetic neuropathy. 38 In the present study, we carefully phenotyped diabetic patients into those with painful and painless diabetic neuropathy and undertook detailed assessment of large and small fibre neuropathy. There was no difference for electrophysiology, quantitative sensory testing and autonomic function between painful and painless neuropathy.
Previously, we have shown a greater reduction in both intraepidermal nerve and corneal nerve fibre length in painful diabetic neuropathy 39 and a detailed immunophenotyping study has shown increased axonal growth (higher GAP43/PGP) and axonal swellings, positive for tropomyosin-receptor-kinase A and substance P in patients with painful compared with painless neuropathy. 40 In the present study, there was no significant difference in autonomic function, IENFD or corneal nerve morphology between painful and painless neuropathy. Indeed, we have recently shown that corneal nerve length at the inferior whorl as opposed to more central corneal nerve parameters differ between patients with and without painful diabetic neuropathy. 41 Given that painful neuropathic symptoms vary in their severity over time, particularly with nocturnal exacerbation, it is difficult to reconcile these 'hard wired' changes with the fluctuating symptoms.
Therefore, changes in sodium channel distribution and expression, altered peripheral blood flow and glycaemic flux have also been implicated in painful diabetic neuropathy. 5 Given the potential link between vitamin D and pain, together with the high prevalence of vitamin D deficiency in diabetic populations, 42 we have explored the link with painful diabetic neuropathy. Previous studies have shown a relationship between vitamin D deficiency and diabetic neuropathy, 7,8 but did not specifically assess the relationship to painful diabetic neuropathy. In the present study, we show a markedly increased risk of painful diabetic neuropathy particularly in patients with vitamin D deficiency but also in those with insufficiency. These data are in keeping with those of a recently published study. 14 A large epidemiological study has shown a higher prevalence of painful diabetic neuropathy in South Asians compared with Europeans. 4 In our previous study, we showed that 55% of South Asians were severely vitamin D deficient with a 25(OH)D < 10 ng/mL. 42 The cohort of subjects evaluated in this study was exclusively white European thus minimising ethnicity as a confounding factor. However, a large population-based study (n = 1461) from China indicated a low vitamin D concentration to be a risk factor for diabetic neuropathy in older adults (≥65 years) (P < .05). 43 There was no such relationship in the youth or middleaged groups suggesting an age-related differential effect. 43 Although vitamin D has been used to treat pain in rheumatological conditions, 44 a Cochrane review concluded that there was poor evidence for the efficacy of vitamin D in the treatment of chronic pain. 45  Treatment with vitamin D produced a dramatic improvement in the symptoms of painful diabetic neuropathy in a type 1 patient, refractory to a range of standard therapies. 48 We have also shown that treatment with high-dose vitamin D results in a significant reduction in neuropathic symptoms 23