Abstract
The relationship between cardiovascular disease and diabetic peripheral neuropathy is mainly sustained by data retrieved from cross-sectional studies focused on cardiovascular risk factors. We aimed to assess the presence of cardiovascular disease as a risk factor for developing diabetic peripheral neuropathy in a type 2 diabetes mellitus population.
A 10-year prospective, primary care, multicentre study in a randomly selected cohort. Cardiovascular disease presence included stroke, coronary artery disease and/or peripheral ischaemia. Diabetic peripheral neuropathy diagnosis was based on clinical neurological examination as well as the neuropathy symptoms score and nerve conduction studies.
Three hundred and ten (N=310) patients were initially recruited. Two-hundred and sixty seven (N=267) patients were included in the study. Diabetic peripheral neuropathy cumulative incidence was 18.3% (95% confidence intervals 14.1–23.4; N=49). Diabetic peripheral neuropathy development was significantly more frequent in participants presenting with cardiovascular disease at baseline (P=0.01). In the final logistic regression analysis, the presence of cardiovascular disease remained associated with an increased risk for diabetic peripheral neuropathy (odds ratio 2.32, 95% confidence intervals 1.03–5.22) in addition to diabetes duration and low density lipoprotein-cholesterol levels.
In our series, type 2 diabetes mellitus patients with cardiovascular disease at baseline present with an increased risk of developing diabetic peripheral neuropathy at 10 years of follow-up. Our results suggest that measures aimed at the prevention, control and treatment of cardiovascular disease can also help prevent diabetic peripheral neuropathy development.
Introduction
Diabetic peripheral neuropathy (DPN) is the most commonly reported long-term diabetic complication, affecting up to 20–40% of type 2 diabetic (T2DM) patients. Additionally, DPN is a major contributory factor in patients affected by diabetic foot ulceration and is responsible for up to 50–75% of non-traumatic foot amputations.1–3 Furthermore, DPN is closely related to other micro-angiopathic complications as well as with peripheral vascular disease.4 Thus, taking into account its prevalence, its socioeconomic burden, the impact this has on quality of life and on associated anxiety and depression, screening and appropriate treatment for DPN are of paramount importance.2,5,6 The slow and silent evolution of DPN is a known hazard since, once established, it is irreversible. Thus, it is necessary to deploy significant efforts in clinical research focused on prevention and early detection of DPN or its risk factors, particularly in primary care settings where preventive measures have proven efficiency and efficacy in patient care as well as in the control of modifiable risk factors.7
Despite the fact that cardiovascular risk control has widely increased worldwide, cardiovascular disease (CVD) remains the main cause for morbidity and mortality in T2DM patients.8 About 70% of mortality in diabetic patients is secondary to CVD events.9–11 Diabetes is associated with a marked increase (2–4 times) of coronary heart disease.9–12 Cardiac autonomic neuropathy, in addition to other cardiovascular risk factors, is responsible for a great number of cardiovascular deaths in TD2M patients according to the ACCORD trial.13 Furthermore, cardiac autonomic neuropathy in the presence of DPN appears as the highest predictor of CVD in that same study. The relationship between the different forms of neuropathy and CVD has been studied erratically. While the relationship between cardiac autonomic neuropathy and sudden death from CVD is demonstrated,14 the evidence on the relationship between DPN and CVD has focused primarily on identifying its relationship with cardiovascular risk factors,15 especially by cross-sectional studies.16 Conversely, a recent publication has reported previous DPN being associated with a higher risk of presenting a first cardiovascular event.17
To the best of our knowledge, there are no prospective studies focusing on the relationship between CVD presence and DPN development in T2DM patients. Because of this, we decided to set forth the current study whose main objective was to analyse the presence of CVD as a risk factor for DPN development in a cohort of participants with T2DM after 10 years of clinical follow-up.
Materials and method
Design
A 10-year (2002–2012) follow-up multicentre study was performed in a primary care setting including all state primary care centres (N=7) in North Catalonia, Spain. The region is a semi-urban area with a population of 92,912 inhabitants and whose main medical facilities providing primary care are state run.
Sample
An initial random sample of T2DM subjects enrolled in the North Catalonia Diabetes Study18 was selected by computer software among our population. Diabetes mellitus was diagnosed according to the World Health Organization guidelines.19 Briefly, a fasting blood glucose >125 mg/dl on two separate occasions (after a minimum of 8 hours fasting), blood glucose ≥200 mg/dl 2 hours after a 75 g of glucose OGTT (Oral Glucose Tolerance Test) and/or a random blood glucose ≥200 mg/dl.
Inclusion criteria were as follows: T2DM patients aged between 30 and 70 years.
Exclusion criteria were as follows: DPN presence; neuropathies of other aetiology; alcohol ingestion ≥60 g/day in women and ≥80 g/day in men; previous foot ulcers; refused consent. Figure 1 outlines the selection process undergone by the random sample cohort to become the sample initially included for the follow-up.
Selected sample and sample initially included in the observational prospective study (2002–2012). Values are numbers or percentages as appear indicated. DPN: diabetic peripheral neuropathy
Variables
Outcome
The main (dependent) variable considered was the appearance of DPN.
Patients were studied by a team of 28 health care providers including 16 nurse practitioners and 12 general practitioners. Study personnel underwent an intensive 24-hour training session.
Training included both theoretical group sessions (25%) and practice sessions (75%). Training was provided by nurse practitioners and general practitioners, coordinated by the principal investigator (J.J.-C.) of the projects FIS 01/0846, FIS PI 040181, ETS PI0690324 and FIS PI07/0340 (funded by the Health Research Institute Carlos III of Spain). At the end of the training, each one of the trainees had to pass a personal evaluation as well as an individual real practical test.
All examinations in each patient were performed by the same health care provider to control for inter-rate reliability. This same team had already been involved in previous DPN studies.18,20,21
DPN diagnosis was based on clinical neurological examination as well as the reduced Neuropathy Symptoms Score described previously20 and nerve conduction studies in borderline subjects (i.e. those with uncertain diagnosis).
In the clinical neurological evaluation, DPN was established by the presence of bilateral signs and symptoms.
Neuropathy signs
Vibration perception thresholds evaluated by a quantitative tuning fork and Neurothesiometer; muscle strength; Achilles tendon reflex; tactile sensitivity by Semmes-Weinstein Monofilament 5.07/10 g; superficial pain sensation and temperature perception).
Neuropathy symptoms
Patients were questioned about the presence or absence and possible nocturnal exacerbation of muscular cramps, numbness, abnormal hot or cold sensations, tingling sensations, burning pain and irritation from bed clothes in the lower legs and feet and whether moving about could reduce the symptoms.
Complete procedures of signs and symptoms determination were performed as described previously.20
Two or more significant signs (vibration perception thresholds, tactile sensitivity by Semmes-Weinstein Monofilament and Achilles reflex) or one sign plus two symptoms (sensitivity alterations, pain perception and decrease in muscle strength (≥3 points in the reduced Neuropathy Symptoms score) indicated DPN. A high Neuropathy Symptoms score (7–9 points) was considered indicative of DPN only in conjunction with the alteration of at least one major sign. Furthermore, DPN diagnosis is a dichotomy (yes/no) using the aforementioned combination of signs and symptoms, in agreement with previous reports.20
Nerve conduction studies
Nerve conduction studies were performed by the Neurophysiology Department staff at the Josep Trueta University Hospital (Girona/Spain), in the median and ulnar nerves for upper extremities and in the sural, peroneal and tibial nerves for lower extremities (sensory and motor nerve conduction velocities, by Synergy T-EPEMG/EP), as described previously.20
Main exposure (at baseline)
CVD was defined as either a history of physician-diagnosed CVD (e.g. previous myocardial infarction, angina, coronary artery disease bypass-grafting, stroke, peripheral vascular disease) or ischaemic changes detected on a 12-lead electrocardiogram. Peripheral ischaemia was assessed by dorsal and rear tibial pedal foot pulses, sporadic intermittent claudication, bypass and medical history records.
Covariates
The main independent variables considered at baseline were gender, age and duration of T2DM (in years). Other independent variables considered at baseline were the degree of metabolic control (HbA1c) and lipid profiles.
The same team performed regular screenings to identify new cases of DPN as well as changes in the other baseline variables during the ensuing 10 years of follow-up (2002–2012). Diagnostic data were retrieved from case records and nerve conduction studies. Clinical information was obtained through patient interviews, clinical examinations, health care records review as well as administrative data review.
Metabolic control was estimated by HbA1c levels (JSCC/JDS; mmol/mol) (liquid chromatography technique, Menarini, Firenze, Italy). Biochemical parameters were analysed in a single clinical laboratory (ICS, Girona, Spain). Lipid profiles were evaluated by serum plasma levels (mg/dL) of total cholesterol (enzymatic method), high-density lipoprotein (HDL) cholesterol (enzymatic method after very low-density lipoprotein (LDL) and LDL precipitation), triglycerides (enzymatic hydrolysis and glycerol measurement) and LDL cholesterol (mg/dL) (by Friedewald’s formula).
Statistical analyses
Proportions between groups were compared with the Chi-square test. Continuous variables were expressed as mean (SD) or median (25th–75th percentile) and were compared by Student’s t test, analysis of variance or U-Mann–Whitney test, as appropriate. Overall cumulative incidence (95% confidence intervals (CI)) of DPN was calculated.
Covariates that were of known clinical interest or that were identified as significant (p<0.05) in the bivariate analyses were selected for multivariate analyses, which were performed by employing logistic regression models for analysing the association between the CVD at baseline and the appearance of DPN (at the end of follow-up). We fitted all models in a customized way by means of the ENTER method. Results were expressed as odds ratios with the corresponding robust 95% CI. All analyses were performed with the STATA v.12 statistical package.
Ethical considerations
All participants gave written informed consent to participate and the study was approved by the Primary Care Ethics Committee of IDIAP Jordi Gol, expedient codes: P00/005, P04/17 and P07/01. The study was carried out in accordance with the principles of the Declaration of Helsinki as revised in 2000 and 2008.
Results
Table 1 provides the details of our observational prospective study in four groups: the initial sample who met the inclusion criteria; those who finished the study; those who moved (drop-outs); those who died during follow-up.
Demographic description of the variables at baseline
| Initial selected | Dead | Moved | Included | |
|---|---|---|---|---|
| N=310 | (n=30) | (n=13) | (n=267) | |
| Women | 38.4 (33.1–43.9) | 40.0 | 61.5 | 37.1 |
| Agea | 57.5 (8.0) | 62.3 (5.6) | 52.5 (8.2) | 57.2 (8.0) |
| History of CVD | 17.0 (12.7–20.9) | 30 | – | 15.3 |
| Duration DM (years)b | 5 (2–8) | 4.5 (3–11) | 0 (0–3) | 5 (3–9) |
| Initial selected | Dead | Moved | Included | |
|---|---|---|---|---|
| N=310 | (n=30) | (n=13) | (n=267) | |
| Women | 38.4 (33.1–43.9) | 40.0 | 61.5 | 37.1 |
| Agea | 57.5 (8.0) | 62.3 (5.6) | 52.5 (8.2) | 57.2 (8.0) |
| History of CVD | 17.0 (12.7–20.9) | 30 | – | 15.3 |
| Duration DM (years)b | 5 (2–8) | 4.5 (3–11) | 0 (0–3) | 5 (3–9) |
CVD: cardiovascular disease (previous myocardial infarction, angina, coronary artery disease, bypass-grafting, stroke, peripheral vascular disease, or ischaemic changes detected on a 12-lead electrocardiogram); DM: diabetes mellitus.
Mean (SD).
Median (P25–P75).
Values are percentages (95% CI) unless stated otherwise.
The percentage of women was lower than men, as is the case in our general population.
Demographic description of the variables at baseline
| Initial selected | Dead | Moved | Included | |
|---|---|---|---|---|
| N=310 | (n=30) | (n=13) | (n=267) | |
| Women | 38.4 (33.1–43.9) | 40.0 | 61.5 | 37.1 |
| Agea | 57.5 (8.0) | 62.3 (5.6) | 52.5 (8.2) | 57.2 (8.0) |
| History of CVD | 17.0 (12.7–20.9) | 30 | – | 15.3 |
| Duration DM (years)b | 5 (2–8) | 4.5 (3–11) | 0 (0–3) | 5 (3–9) |
| Initial selected | Dead | Moved | Included | |
|---|---|---|---|---|
| N=310 | (n=30) | (n=13) | (n=267) | |
| Women | 38.4 (33.1–43.9) | 40.0 | 61.5 | 37.1 |
| Agea | 57.5 (8.0) | 62.3 (5.6) | 52.5 (8.2) | 57.2 (8.0) |
| History of CVD | 17.0 (12.7–20.9) | 30 | – | 15.3 |
| Duration DM (years)b | 5 (2–8) | 4.5 (3–11) | 0 (0–3) | 5 (3–9) |
CVD: cardiovascular disease (previous myocardial infarction, angina, coronary artery disease, bypass-grafting, stroke, peripheral vascular disease, or ischaemic changes detected on a 12-lead electrocardiogram); DM: diabetes mellitus.
Mean (SD).
Median (P25–P75).
Values are percentages (95% CI) unless stated otherwise.
The percentage of women was lower than men, as is the case in our general population.
Briefly, of the 310 patients initially selected, 267 finished the follow-up and were included in the study.
There were no significant differences in age (p=0.664) and gender (p=0.747) between patients in the initially selected group and the remaining cohort included in the study (data not shown). As shown, the majority of patients were males in their late fifties whose T2DM median duration was 5 years where CVD affected 17% of them.
Forty nine DPN events were diagnosed during follow-up accounting for an overall DPN cumulative incidence of 18.3% (95% CI 14.1–23.4). Table 2 depicts the relationships between DPN development and the main variables at baseline. These newly diagnosed DPN cases were more frequent in those participants presenting CVD at baseline (p=0.01), were older (p=0.03) and had greater median duration years of T2DM (p=0.03).
Baseline variables association with total DPN events at the end of follow-up (n=267)
| Variables | Diabetic polyneuropathy | p-value | |
|---|---|---|---|
| Yes | No | ||
| Women | 44.0 | 36.7 | 0.33 |
| Age (years)a | 59.5 (6.9) | 56.9 (8.2) | 0.03 |
| History of CVD | 30.0 | 14.7 | 0.01 |
| Duration DM (years)b | 7 (4–10) | 5 (2–8) | 0.03 |
| Variables | Diabetic polyneuropathy | p-value | |
|---|---|---|---|
| Yes | No | ||
| Women | 44.0 | 36.7 | 0.33 |
| Age (years)a | 59.5 (6.9) | 56.9 (8.2) | 0.03 |
| History of CVD | 30.0 | 14.7 | 0.01 |
| Duration DM (years)b | 7 (4–10) | 5 (2–8) | 0.03 |
DPN: diabetic peripheral neuropathy; CVD: cardiovascular disease; DM: diabetes mellitus.
Mean (SD).
Median.
Values are percentages unless stated otherwise.
Baseline variables association with total DPN events at the end of follow-up (n=267)
| Variables | Diabetic polyneuropathy | p-value | |
|---|---|---|---|
| Yes | No | ||
| Women | 44.0 | 36.7 | 0.33 |
| Age (years)a | 59.5 (6.9) | 56.9 (8.2) | 0.03 |
| History of CVD | 30.0 | 14.7 | 0.01 |
| Duration DM (years)b | 7 (4–10) | 5 (2–8) | 0.03 |
| Variables | Diabetic polyneuropathy | p-value | |
|---|---|---|---|
| Yes | No | ||
| Women | 44.0 | 36.7 | 0.33 |
| Age (years)a | 59.5 (6.9) | 56.9 (8.2) | 0.03 |
| History of CVD | 30.0 | 14.7 | 0.01 |
| Duration DM (years)b | 7 (4–10) | 5 (2–8) | 0.03 |
DPN: diabetic peripheral neuropathy; CVD: cardiovascular disease; DM: diabetes mellitus.
Mean (SD).
Median.
Values are percentages unless stated otherwise.
Those who developed DPN during follow-up disclosed higher total cholesterol (p=0.014) and LDL-cholesterol levels (p=0.013). Conversely, HDL-cholesterol (p=0.295), HbA1c (p=0.946) and triglycerides values (p=0.221) showed no differences. In the final logistic regression analysis, the presence of CVD remained associated with an increased risk of DPN (odds ratio (OR) 2.32 (95% CI 1.03-5.22)] as well as T2DM’s years of duration and LDL-cholesterol levels (Table 3). Gender, body mass index and age did not show a significant association with and increased risk of DPN.
Multivariate final logistic regression model at the end of follow-up
| Factors (at baseline) | OR (95%) | p-value |
|---|---|---|
| CVD | 2.32 (1.03–5.22) | 0.04 |
| Age | 1.03 (0.98–1.07) | 0.22 |
| Women | 1.43 (0.71–2.88) | 0.31 |
| Years of duration | 1.05 (1.00–1.10) | 0.04 |
| LDL-cholesterol | 0.98 (0.97–0.99) | 0.01 |
| Factors (at baseline) | OR (95%) | p-value |
|---|---|---|
| CVD | 2.32 (1.03–5.22) | 0.04 |
| Age | 1.03 (0.98–1.07) | 0.22 |
| Women | 1.43 (0.71–2.88) | 0.31 |
| Years of duration | 1.05 (1.00–1.10) | 0.04 |
| LDL-cholesterol | 0.98 (0.97–0.99) | 0.01 |
CVD: cardiovascular disease.
Reference categories are no previous history of CVD, men.
Age, years of type 2 diabetes mellitus duration and LDL-cholesterol is in continuous form.
Multivariate final logistic regression model at the end of follow-up
| Factors (at baseline) | OR (95%) | p-value |
|---|---|---|
| CVD | 2.32 (1.03–5.22) | 0.04 |
| Age | 1.03 (0.98–1.07) | 0.22 |
| Women | 1.43 (0.71–2.88) | 0.31 |
| Years of duration | 1.05 (1.00–1.10) | 0.04 |
| LDL-cholesterol | 0.98 (0.97–0.99) | 0.01 |
| Factors (at baseline) | OR (95%) | p-value |
|---|---|---|
| CVD | 2.32 (1.03–5.22) | 0.04 |
| Age | 1.03 (0.98–1.07) | 0.22 |
| Women | 1.43 (0.71–2.88) | 0.31 |
| Years of duration | 1.05 (1.00–1.10) | 0.04 |
| LDL-cholesterol | 0.98 (0.97–0.99) | 0.01 |
CVD: cardiovascular disease.
Reference categories are no previous history of CVD, men.
Age, years of type 2 diabetes mellitus duration and LDL-cholesterol is in continuous form.
Discussion
The current study discloses a significant relationship between the presence of CVD at baseline and the appearance of DPN during a 10-year follow-up in our cohort of patients with T2DM (Table 2). Furthermore, this relationship maintains its significance after adjustment for gender, age, time of T2DM duration and LDL-cholesterol levels at baseline.
The cumulative incidence of CVD (18.3%) is somewhat lower than in other reports performed in the same geographical setting21 where it was 21.9%. Age (57.5 vs. 59.6 years) and time of T2DM duration (5.0 vs. 8.5 years) can be accounted for the observed differences.
The lack of evidence sustaining a longitudinal association between CVD and DPN in patients with T2DM prompted the setting of this study. Our findings stand as a novelty since, to the best of our knowledge, no prospective study has yet disclosed an association between CVD and DPN in T2DM patients. Conversely, a recent prospective study with 30 months of follow-up has specifically addressed the opposite hypothesis (i.e. does DPN diagnosis induce CVD in T2DM?). Noteworthy, the authors report previous DPN being associated with a higher risk for presenting a first cardiovascular event.17
To the best of our knowledge there are, aside from the current paper, insufficient available data to sustain proof of longitudinal associations between CVD presence and DPN development in T2DM. Hence, direct and straight comparisons will have to wait for the time being. Interestingly, some prospective works have proven a relationship between cardiovascular risk factors and distal symmetrical DPN development in type 115,22,23 and in type 2 diabetes mellitus16,24,25. Noteworthy, CVD at baseline was associated with double risk of neuropathy independently of cardiovascular risk factors.22 These results must be interpreted with caution since type 1 and type 2 DPNs differ in their characteristics.26
The association between pre-existing CVD and DPN development can be at least partially understood when the proven evidence on the relationship between T2DM and cardiovascular risk factors are taken into account.16,24,25 Furthermore, it is likely, although not proven in our cohort, that T2DM patients with CVD at baseline, who developed DPN during follow-up, already presented with asymptomatic cardiac autonomic neuropathy, the most overlooked of all serious complications of diabetes.27 In our series, the baseline incidence of cardiac autonomic neuropathy might well have increased during follow-up, along with age and duration of diabetes, in agreement with previous reports.28 Cardiac autonomic neuropathy may have additionally co-segregated with distal symmetric polyneuropathy, micro-angiopathy and macro-angiopathy in our series, also in agreement with previous reports.28
It would have been interesting to evaluate the cardiovascular autonomic function, but although cardiac autonomic neuropathy and DPN share common aspects in the affectation of long fibres, the DPN primarily affects small fibres29 and, unfortunately, there is no consensus on the best measures for assessing cardiovascular autonomic neuropathy.30 Additionally, accurate and homogeneous assessment of pre-clinical forms of cardiac autonomic neuropathy by cardiovascular autonomic function is a difficult task to perform in a primary care setting such as ours.
On the other hand, there is a relationship between the presence of cardiovascular risk factors and the incidence of chronic idiopathic distal symmetric neuropathy in elderly patients without diabetes.31 Thus, if the development of neuropathy in non-diabetic individuals is associated with potentially modifiable cardiovascular risk factors, the inflammatory/atherogenic background of T2DM patients32 might play a role in this relationship.
In the final logistic regression analysis (Table 3), T2DM’s years of duration remained associated with an increased risk of DPN in agreement with previous results.20,33,34 Moreover, the LDL-cholesterol levels remained associated with an increased risk of DPN, in agreement with previous works.34–36 Several underlying mechanisms have been identified: free fatty acids have been shown to directly damage Schwann cells in vitro37 and promote inflammatory cytokine release from adipocytes and macrophages.38 Plasma LDL-cholesterol particles can be modified by glycation and/or oxidation. The latter can bind to extracellular receptors, triggering signalling cascades that activate nicotinamide adenine dinucleotide phosphate-oxidase and lead to oxidative stress.39
Additionally, cholesterol can be oxidized to oxysterols, which have been shown to cause neuronal apoptosis.40,41 Additionally, significant independent associations have been reported in T2DM patients among lipid fractions and risk for micro-angiopathy.22,42
The overall DPN cumulative development was, as expected, higher than that at 3 years (18.3% vs. 10.44%).21
Our results must be interpreted with due caution due to several study limitations, as outlined below.
First, the different clinical forms of CVD (myocardial infarction, angor pectoris, stroke, peripheral vascular disease, etc.) were not considered individually; basically, due to the concurrence in their presentation (i.e. atherosclerosis is a systemic disease). Otherwise, our study would have required a much larger population size; nevertheless, each one of them has been related to DPN presentation in different cross-sectional studies.16
Second, several continuous variables such as HbA1c and cholesterol sub-fractions were only considered at baseline, but subsequent determinations were not included in our analysis due to their timely variability and the influence of different type of medications used for T2DM metabolic43,44 and lipid control45 as simultaneous treatment, which also kept evolving during the study (i.e. a significant percentage of our cohort followed either isolated or combined treatment for dyslipidaemia with statins and/or fibrates). Moreover, high blood pressure treatment also included isolated or combined treatment (angiotensin-converting-enzyme inhibitors, angiotensin II receptor antagonists and/or diuretics). Last but not least, the use and dosages of these drugs kept evolving during follow-up.
The lack of assessment of cardiovascular autonomic function cannot be considered as a weakness since it was not the purpose of the study.
Furthermore, we have not been able to establish a precise timing in the disclosure of new cases of DPN in all the patients; nevertheless, the latter does not affect the final results, taking into account the overall cumulative incidence.
On the other hand, the strengths of our study include its design (longitudinal with long follow-up), its well-defined and homogeneous population and the strict application of well recognized international diagnostic criteria. Hence, DPN diagnosis was based on the San Antonio Consensus Statement Criteria46 and, particularly, the definitions no. 3 and no. 4 of the Toronto Diabetic Neuropathy Expert Group47 for research studies.
In conclusion, our paper brings a novel and pioneering approach to DPN with the novelty of the disclosure of CVD as a risk factor for DPN development at 10 years of follow-up, in addition to T2DM’s years of duration and LDL-cholesterol levels. Thus, CVD might well stand as an early marker of DPN in patients with T2DM. Our results suggest that measures aimed at the prevention, control and treatment of CVD, can also help prevent DPN development.
Conflict of interest
None declared.
Funding
This work was supported by grants from the Fondo de Investigación Sanitaria SPAIN: FIS 01/0846, FIS 04/0181 and FIS PI070340, Instituto de Salud Carlos III of Spain, Plan Nacional I+D+i 2004-2007, 2008–2010 and FEDER.
Cardiovascular disease increases the risk of developing diabetic peripheral neuropathy
Cardiovascular disease control and treatment can contribute to diabetic peripheral neuropathy prevention
References
Author notes
These authors contributed equally to this work
Comments