Journal of Diabetes and Its Complications 20 (2006) 216 – 223

Exercise training can modify the natural history of

diabetic peripheral neuropathy
Stefano Balduccia,b, Gianluca Iacobellisb,e,T, Leoluca Parisic, Nicolina Di Biasea,f,
Eugenio Calandrielloa,c, Frida Leonettib, Francesco Falluccad
a
Health Care Team, Metabolic Fitness Association, Monterotondo, Rome, Italy
b
Department of Clinical Sciences, Endocrinology, University of Rome, La Sapienza, Rome, Italy
c
Department of Clinical Neurology, La Sapienza University, Rome, Italy
d
Department of Clinical Sciences, La Sapienza University II Faculty (St. Andrea Hospital), Rome, Italy
e
Center for Human Nutrition, The University of Texas Southwestern Medical Center at Dallas, TX 75390-9052, USA
f
St. Pietro Hospital, FBF, Rome, Italy
Received 26 August 2004; received in revised form 25 February 2005; accepted 5 July 2005

Abstract

Background: Diabetes is the most important cause of peripheral neuropathy (DPN). No definitive treatment for DPN has been
established, and very few data on the role of exercise training on DPN have been reported. Aim of the study: We sought to examine the
effects of long-term exercise training on the development of DPN in both Types 1 and 2 diabetic patients. Participants and methods:
Seventy-eight diabetic patients without signs and symptoms of peripheral DPN were enrolled, randomized, and subdivided in two groups: 31
diabetic participants [15 f, 16 m; 49F15.5 years old; body mass index (BMI)=27.9F4.7], who performed a prescribed and supervised 4 h/
week brisk walking on a treadmill at 50% to 85% of the heart rate reserve (exercise group: EXE), and a control group of 47 diabetic
participants (CON; 24 f, 23 m; 52.9F13.4 years old; BMI=30.9F8.4). Vibration perception threshold (VPT), nerve distal latency (DL), nerve
conduction velocity (NCV), and nerve action potential amplitude (NAPA) in the lower limbs were measured. Results: We found significant
differences on # (delta) in NCV for both peroneal and sural motor nerve between the EXE and CON groups during the study period
( Pb.001, for both). The percentage of diabetic patients that developed motor neuropathy and sensory neuropathy during the 4 years of the
study was significantly higher in the CON than the EXE group (17% vs. 0.0%, Pb.05, and 29.8% vs. 6.45%, Pb.05, respectively). In
addition, the percentage of diabetic patients who developed increased VPT (25 V) during the study was significantly higher in the CON than
the EXE group (21.3% vs. 12.9%, Pb.05). Change on Hallux VPT from baseline to the end of the study was significantly different between
the EXE and CON groups ( Pb.05); no significant change in Malleolus VPT between the two groups occurred. Conclusions: This study
suggests, for the first time, that long-term aerobic exercise training can prevent the onset or modify the natural history of DPN.
D 2006 Elsevier Inc. All rights reserved.

Keywords: Diabetic peripheral neuropathy; Exercise training

1. Introduction Poncelet, 2003) and a leading cause of nontraumatic foot

Diabetic peripheral neuropathy (DPN) is a common
complication and quality-of-life damaging factor in diabe-
tic patients (Boulton, Malik, Arezzo, & Sosenko, 2004;
T Corresponding author. Fax: +1 214 648 7150.
E-mail address: gianluca.iacobellis@utsouthwestern.edu
(G. Iacobellis).
amputation (Boulton, 1998). Metabolic and vascular factors
seem to be involved in the pathogenesis of DPN. Distal
symmetric polyneuropathy is the most common form of
DPN, involving usually both small and large nerve fibers
(Boulton et al., 2004). Small nerve fiber neuropathies occur
early in the course of diabetes and frequently develop with
no objective signs or electrophysiologic evidence of nerve
damage (Vinik, Erbas, Stansberry, & Pittenger, 2001).

1056-8727/06/$ – see front matter D 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.jdiacomp.2005.07.005
 S. Balducci et al. / Journal of Diabetes and Its Complications 20 (2006) 216 – 223
Although no definitive treatment for DPN has been
established yet, several studies have shown that intensive
therapy and optimal glycemic control can significantly
reduce DPN (Azad et al., 1999; Dahl-Jorgensen et al.,
2.2. Exclusion criteria
1986; DCCT Research Group, 1995). Promising treatment
for DPN include metabolic treatments, autoimmune thera-
Patients carrying at least one of the following conditions
pies, and nerve growth factors (Vinik, 1999). Recent studies
were excluded from the study:
suggested that aerobic physical activity, alone or in
combination with resistance exercise, may be an effective
therapeutic modality for Type 2 diabetes (Castaneda et al.,
2002; Dunston et al., 2002; Maiorana, O’Driscoll, Goodman,
Taylor, & Green, 2002). The beneficial effects of a regular
exercise on glycemic control, insulin sensitivity, lipid
abnormalities, and hypertension in diabetic patients have
been previously described (Balducci, Leonetti, Di Mario,
& Fallucca, 2004; Goldhaber-Fiebert, Goldhaber-Fiebert,
Tristan, & Nathan, 2003).
Nevertheless, very few data on the effectiveness of
exercise treatment on DPN have been reported (Richardson,
Sandman, & Vela, 2001; Tesfaye, Harris, Wilson, & Ward,
1992). Prescribed and supervised long-term exercise pro-
grams may influence neuromuscular parameters in diabetic
patients, thereby inducing adaptive changes in the neuro-
muscular system in response to exercise training.
In this study, we sought to examine the effects of a long-
term exercise training on the development of peripheral
neuropathy in both Types 1 and 2 diabetic patients.
2. Study design
This was a 4-year prospective randomized intervention
study. Types 1 and 2 diabetic patients, from 500 consecutive
diabetic patients admitted in our Department, without signs
and symptoms of DPN and able to have a 1.6-km-distance
walk were enrolled in this study. Patients who did not accept
to perform the proposed exercise program were excluded
from the study. The enrolled participants were subsequently
randomized in two groups and followed up for each year of
the study. All patients had their usual diet during the study
period. The dietary regimen was based on the Mediterranean
diet. In addition, no significant changes in pharmacological
treatment for diabetes, lipid, and blood-pressure-lowering
drugs during the study were applied. This study was con-
ducted in accordance with the Declaration of Helsinki guide-
lines. Each participant gave an informed consent before the
study began.
2.1. Inclusion criteria
The following inclusion criteria were implemented:
! Type 2 and/or Type 1 diabetes
! No signs or symptoms of DPN
! Ability to walk a 1.6-km distance without assistance or
with a device. The ability to complete this walk-distance
217
was tested in the metabolic fitness center with heart
rate monitoring.
1. Physical and electrodiagnostic evidence of DPN,
according to the reported criteria (Feldman et al.,
1994).
Exclusion criteria based on physical examination. The
presence of DPN was evaluated with the Michigan
Neuropathy Screening Instrument (MNSI). MNSI con-
sisted of an inspection of the foot, examination of the
Achilles reflexes, and evaluation of the vibratory thresh-
old (VPT). Vibration perception was considered reduced
when the patients said that they did not feel any
vibration, while the operator perceives the vibrations at
the index of their own hand for another 10 s. If the final
score was of 2.5 or more (MNSI positivity), the patient
was excluded from the study.
Exclusion criteria based on electrodiagnostic exami-
nation:
(a) Sural response: absent or decreased amplitude
(b6 AV) with a normal or minimally prolonged
distal latency (DL; b3 ms) stimulating 14 cm
from the recording site posterior to the lateral
Malleolus. If the sural response was absent bila-
terally, the motor responses were not performed.
(b) Peroneal responses: absent or decreased in
amplitude (b 2 mV for peroneal), with a normal
DL (b 6.2 ms stimulating 9 cm from recording
sites over the extensor digitorum brevis).
2. A history or evidence on physical examination of
significant central nervous system dysfunction (i.e.,
hemiparesis, myelopathy, and cerebellar ataxia).
3. Significant musculoskeletal deformity [i.e., amputa-
tion, scoliosis, abnormality of range of motion
(ROM)] that would prevent participation (b908 of
humeral abduction, inability to grip, b108 of com-
bined ankle inversion/eversion).
4. Lower extremity arthritis or pain that limits exercise.
5. A history or clinical evidence of severe cardiovas-
cular diseases that limit or contraindicate the exercise.
6. A history or evidence on physical examination of
vestibular dysfunction.
7. A history of angina or angina-equivalent symptoms
(i.e., nausea, diaphoresis, and shortness of breath with
exercise).
8. Symptomatic postural hypotension defined as a fall in
blood pressure (i.e., N20 mm Hg for systolic or
N10 mm Hg for diastolic blood pressure) in response
to postural change, from supine to standing (Position
paper, 1996).
218 S. Balducci et al. / Journal of Diabetes and Its Complications 20 (2006) 216 – 223

Table 1 – 47 diabetic participants (24 f, 23 m; age =52.9F

Patients’ clinical characteristics at baseline and at the end of the
13.4 years; BMI=30.9F8.4; 14 Type 1 diabetes and
4-years study
33 Type 2 diabetes) with a sedentary lifestyle repre-
CON EXE
sented the control group (CON). The CON group did
Baseline 4 years Baseline 4 years not perform supervised physical activity for the entire
n 47 47 31 31 duration of the study.
Sex (F/M) 24/23 15/16
Type 1/Type 2 14/33 14/33 7/24 7/24
diabetes 2.4. Methods
Age (years) 52.9F13.4 49.0F15.5
Onset of 43.6F16.3 40.8F15.9 2.4.1. Neurophysiological measurements
diabetes (years) Vibration perception threshold (VPT), nerve DL, nerve
Mean 9.1F6.6 8.4F5.7
duration (years)
conduction velocity (NCV), and nerve action potential
Height (m) 1.64F0.1 1.65F0.1 amplitude (NAPA) in the lower limbs were measured by
Weight (kg) 73.4F11.3 73.2F12.2 77.2F13.7 75.6F13.1 experienced specialists in neurophysiology (L.P. and E.C.)
BMI (kg/m2) 27.2F3.6 27.3F4.5 27.9F4.7 27.5F4.4 using the same methods and instrumentation at baseline
Waist 87.4F13.2 91.2F10.5 98.0F11.8 92.3F13.1 and for each year throughout the 4 years of the study. The
circumference (cm)
FM (%) 30.9F8 30F9.4 29.5F9.8 29.5F10
measurements of mean peroneal motor nerve and sural
FFM (kg) 51.2F7.7 49.6F11.5 54.5F12.2 54.7F14.4 sensory nerve were performed using a Medelec MS
HbA1c (%) 7.95F1.6 8.09F1.4 8.26F 1.2 7.84F1.1 928 Neurostar (Oxford Instruments Medical, Old Woking,
Fasting 164F68 158F83 163F58 143F43 U.K.). NCV and NAPA were analyzed for the sensory
glucose (mg/dl) sural and motor peroneal nerves, and the reference values
Total 205F43 186F44 201F37 188F44
cholesterol (mg/dl)
used were based on a large number of volunteers (Falck
Tryglicerides (mg/dl) 129F77 130F94 154F71 137F79 et al., 1991; Liveson & Ma, 1992). For NCV, the lowest
Creatinine (mg/dl) 1.03F0.2 1.02F0.3 1.01F0.2 1.01F0.2 reference value is 40 m/s ( 2 S.D.) for all nerves, and for
Systolic blood 136F19 135F17 138F19 134F19 NAPA, the lowest reference value ( 2 S.D.) is 5 AV for
pressure (mm Hg) the sural nerve and 2 mV for the peroneal nerve. In our
Diastolic blood 82F10 80F10 87F9 82F10
pressure (mm Hg)
calculations, we used 15 m/s instead of 0 m/s for the
Microalbuminuria (n) 7 (15%) 7 (15%) 4 (13%) 4 (13%) patients with NCV registered as 0 m/s because, for this
Smoking (n) 12 (26%) 12 (26%) 10 (35%) 10 (35%) type of patient, the true value probably lies between the
Alcohol intake 12 (26%) 12 (26%) 11 (36%) 11 (36%) detection limit (approximately 30 m/s) and 0 m/s. If we
b250 ml/day (n) had used 0 m/s in our calculations, the differences
Alcohol intake 4 (9%) 4 (9%) 3 (10%) 3 (10%)
N250 ml/day (n)
between the groups would have been greater; thus, the
HR at rest 71.5F7.8 71.9F9.4 73.3F9.1 63.4F5.2 use of 15 m/s instead of 0 m/s reduces the possibility of
(beats min 1) overestimating the effect of exercise training (Larsen &
VO2 max 23.5F5.3 23.1F7.1 25.4F6.8 40.7F4.6 Jorgensen, 2003).
(ml kg 1 min 1)
Data are meanFS.D.

9. A history or evidence on physical examination of

plantar skin pressure ulcer.

2.3. Participants

Seventy-eight diabetic patients (39 f, 39 m; age=

51.3F14.3 years, duration of diabetes =8.79F6.2 years)
who met the inclusion criteria were enrolled and randomized
in two treatment groups:

– 31 diabetic participants [15 f, 16 m; age=49.0F

15.5 years; body mass index (BMI) =27.9F4.7; 7 Type
1 diabetes and 24 Type 2 diabetes] formed the exercise
group (EXE). They performed a prescribed and super-
vised 4-h/week (four sessions per week) brisk walking
on a treadmill (Technogym Gambettola, FC, Italy) at 50% Fig. 1. Change in NCV (DNCV) from baseline to the end of the study;
to 85% of the heart rate reserve estimated every month. **Pb.001.
 S. Balducci et al. / Journal of Diabetes and Its Complications 20 (2006) 216 – 223 219

Table 2
Neurological parameters at baseline and for each year of the study
CON EXE
Baseline 1 year 2 years 3 years 4 years Baseline 1 year 2 years 3 years 4 years
n 47 47 31 31

Peroneal motor nerve

DL (m/s) 4.41F0.64 4.36F0.67 4.45F0.65 4.47F0.60 4.40F0.61 4.38F0.83 4.15F0.57 4.23F0.52 4.3F0.5 4.34F0.49
NAPA (mV) 2.93F1.68 2.95F0.95 2.77F0.98 2.72F1.0 2.70F1.07 3.19F1.99 3.07F0.7 2.9F1.0 2.9F0.9 2.81F0.98
NCV (m/s) 46.6F3.2 47.1F3.1 47.1F3.2 46.3F5.8 46F5.36 47F3.27 48.1F3.52 48.3F2.48 48.9F1.9 48.8F2.24T
DNCV (m/s) 0.6F3.3 +1.8F2.7TT
Patients developed 0 (0.0%) 0 (0.0%) 2 (4.3%) 5 (10.6%) 8 (17.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) T
motor neuropathy

Sural sensory nerve

DL (m/s) 3.43F0.54 3.30F0.46 3.32F0.39 3.33F0.50 3.27F0.52 3.40F0.92 3.29F0.40 3.30F0.38 3.21F0.3 3.25F0.50
NAPA (AV) 20.3F5.02 20.8F4.05 19.9F3.9 19.7F3.9 19.4F4.66 21.5F5.24 21.4F4.4 21.5F5.1 20.6F5.5 21.7F5.44
NCV (m/s) 47.0F3.5 46.5F5.9 45.8F7.0 45.1F6.7 44.3F7.84T 47.1F4.01 47.4F5.05 47.1F3.97 47.9F2.9 47.5F3.18
DNCV (m/s) 2.7F2.8 + 0.4F3.3TT
Patients developed 0 (0.0%) 1 (2.1%) 4 (6.4%) 9 (19.1%) 14 (29.8%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 1 (3.2%) 2 (6.4%)T
sensory neuropathy

VPT
VPT Malleolus (mV) 18.6F8.5 18.9F7.1 19.6F7.2 20.5F8.9 21.4F9.60 17.9F6.20 18F6.6 18.4F6.21 18.0F7.1 17.9F6.90
VPT Hallux (mV) 14.2F6.7 14.9F6.2 16F7.4 16.2F9.1 16.7F9.23 14.1F4.00 14.1F5.1 14.0F4.6 13.6F4.3 13.6F7.00T
Patients with 0 (0%) 1 (2.1%) 2 (4.3%) 6 (12.8%) 10 (21.2%) 0 (0.0%) 0 (0.0%) 1 (3.2%) 3 (9.7%) 4 (12.9%)T
VPT altered
Data are meansFS.D.
T Pb.05.
TT Pb.001.

Large-fiber sensory nerve function was quantified by
VPT at the Hallux and Malleolus by means of a
Biothesiometer (Horwell, Nottingham, U.K.). We defined
reduced vibration detection to be a VPT test score of 25 V
(Abbott, Vileikyte, Williamson, Carrington, & Boulton,
1998; Young, Breddy, Veves, & Boulton, 1994).
Participants were tested in a quiet environment in closed
room with constant temperature. The Biothesiometer was
applied to the skin surface but held lightly so that almost
its full weight was on the foot to ensure that the pressure
was always the same during testing. The vibration
amplitude was increased until the participants could feel
vibration at the site that was being tested. The patients
were instructed not to respond to sensation perceived in
other parts of the foot or leg. If vibration was perceived in
the tested area, this was indicated by the patients by a
verbal response. The values of VPT were determined as the
mean of three measurements (Davis, Jones, Walsh, &
Byrne, 1997).
2.4.2. Anthropometric measurements
Weight (to the nearest 0.1 kg) and height (to the nearest
0.5 cm) were measured while the participants were fasting
and wearing only their undergarments. BMI was calculated
as body weight divided by height squared. Minimum waist
circumference (W, in centimeters; minimum circumference
between the lower rib margin and the iliac crest, midwaist)
was measured while the participants were standing with
their heels together.
2.4.3. Dual energy X-ray absorptiometry (DEXA)
measurements
Fat mass (FM, %) and free fat mass (FFM, kg) cal-
culations were performed using a whole body densitometer
(Hologic QDR 2000 plus, Hologic).
2.4.4. Exercise testing
All participants were started on a walking program on a
treadmill Technogym at baseline and for each year
throughout the 4 years of the study after instruction by a
qualified physical education instructor on suitable clothing,
shoes, and other precautions. The patients were fitted
with a heart rate feature strapped to the chest, for heart
rate monitoring. Maximal aerobic capacity (VO2 max)
was measured used a 1-mile track walk (Kline, Porcari, &
Rippe, 1987).
2.4.5. Blood pressure measurements
Blood pressure was measured using a standard manual
mercury sphygmomanometer for at least three measurements
in a seated position, after at least 15 min of rest, according to
the NIH guidelines.
2.4.6. Lifestyle habits
Information about smoking habits, use of medication,
and alcohol was obtained through questionnaires. The
Minnesota Leisure Time Physical Activity Questionnaire
was used to obtain information on the average frequency
220 S. Balducci et al. / Journal of Diabetes and Its Complications 20 (2006) 216 – 223
Fig. 2. Percentage of patients developing (A) motor neuropathy, (B) sensory
neuropathy, and (C) VPT z25 V during 4 years of follow-up in the (– –)
control and (—) exercise group.
and duration of participation for a specific list of physical
activities. Sedentary lifestyle was defined as performing no
vigorous activity and performing no light to moderate
activity during the past 1 week, 1 month, 3 months, and 1 year
preceding the survey.
2.4.7. Analytic procedures
Glycosylated hemoglobin (HbA1c) and microalbuminuria
were measured using monoclonal antibody method (DCA
2000+Analyzer, Milan, Italy). Overnight urine collections
for microalbuminuria detection were performed. Micro-
albuminuria was defined as urinary albumin excretion
between 20 and 200 Ag/min in two consecutive samples.
Plasma glucose was determined by the glucose oxidase
method [Autoanalyzer, Beckman Coulter, Fullerton, CA;
coefficient of variation (CV), 1.9F0.2%]. Plasma total
cholesterol (CV, 3.4F0.2%), triglycerides (CV, 3.1F
0.5%), and creatinine (CV, 1F0.4%) concentrations were
measured using enzymatic kits (Ortho-Clinical Diagnostic,
Milan, Italy).
2.5. Statistical analysis
All the results are expressed as meansFS.D. We
computed the changes over time in each group with paired
t test with the calculation of 95% confidence interval (CI).
Unpaired t test with the calculation of 95% CI was used to
estimate the difference between the delta in NCV for both
peroneal and sural motor nerves between the two treatment
groups during the study period. We used this t test for deltas
in NCV as equivalent to the test on the interaction term in
the repeated measures analysis of variance. Fisher’s Exact
Test was used to evaluate the percentage of number of
patients developing PDN in both groups. Two-tailed Pb.05
indicated statistical significance. Statistical analysis was
made by StatView, Version 4.5 (Abacus Concept, Berkeley,
CA, USA).
3. Results
We obtained an excellent compliance, and all patients
completed the study and attended N90% of the prescribed
program sessions. No adverse effects throughout the entire
study were observed.
3.1. Clinical parameters
The participants’ characteristics for both groups at base-
line and for each year of the study are summarized in Table 1.
No significant difference on baseline clinical parameters
between the EXE and CON groups were found (Table 1).
3.2. Neurophysiological parameters from baseline to the
end of the 4-year study
3.2.1. NCV parameters
We found significant differences on D (delta) in NCV for
both peroneal and sural motor nerve between the EXE and
CON groups during the study period ( Pb.001, 95%
CI= 0.53 to 2.26; Pb.001, 95% CI= 4.48 to 1.71,
respectively; Fig. 1).
Peroneal motor NCV significantly increased in the EXE
group ( Pb.05, 95% CI= 0.37 to 3.22) and insignificantly
decreased in the CON group. For sural sensory NCV, there
was no significant increase in the EXE group (95% CI=1.53
to 5.11), while a significant decrease in the CON group
( Pb.05, 95% CI=1.21 to 4.26) was found. No significant
difference in both peroneal and sural DL and NAPA
between the two groups was observed (Table 2).
3.2.2. PDN development during the 4-year study
The percentage of diabetic patients who developed
motor neuropathy during the 4 years of the study was
 S. Balducci et al. / Journal of Diabetes and Its Complications 20 (2006) 216 – 223
Fig. 3. Malleulus and Hallux VPT at baseline and after 4 years on the
control and exercise groups; *Pb.05.
significantly higher in the CON than the EXE group (17.0%
vs. 0.0%, Pb.05, OR= 0.07; 95% CI=0.003– 0.77). Sim-
ilarly, the percentage of diabetic patients that developed
sensory neuropathy during the 4 years of the study was
significantly higher in the CON than the EXE group (29.8%
vs. 6.5%, Pb.05, OR= 0.16; 95% CI=0.004 –1.3). In
addition, the percentage of diabetic patients who developed
increased VPT (25 V) during the study was statistically
significantly higher in the CON than the EXE group (21.3%
vs. 12.9%, Pb.05, OR=0.554; 95% CI= 0.04– 0.97; Fig. 2).
3.2.3. VPT parameters
Change on Hallux VPT from baseline to the end of the
study was significantly different between the EXE and CON
groups ( Pb.05); no significant change in Malleolus VPT
between the two groups occurred (Fig. 3).
No significant differences on NCV and VPT parameters
between Types 1 and 2 diabetic patients during the study
were observed.
4. Discussion
Our study shows, for the first time, that long-term
aerobic exercise training can modify the natural history of
peripheral diabetic neuropathy or even prevent its onset. In
fact, we found that a prescribed aerobic exercise regimen,
although of mild intensity, can positively influence and
modify both motor and sensory neuromuscular parameters
in diabetic patients.
The treatment of DPN has traditionally focused on
the control of hyperglycemia. The impact of an intensive
glycemic control on DPN has been widely evaluated
(Azad et al., 1999; Balducci et al., 2004; Castaneda et al.,
2002; Dahl-Jorgensen et al., 1986; DCCT Research Group,
1995; Dunston et al., 2002; Goldhaber-Fiebert et al., 2003;
Maiorana et al., 2002; Vinik, 1999), but with controversial
results. In fact, some studies showed substantial improve-
221
ments on peroneal NCV in Type 1 diabetic patients
(Muller-Felber et al., 1993; Muller-Felber et al., 1991;
Navarro, Sutherland, & Kennedy, 1997), whereas others
failed to demonstrate significant changes in the progression
of somatic or autonomic neuropathy (Azad et al., 1999;
Ruiz, 2004).
Our study can provide original information of potential
interest. First, this is the first long-term follow up study on
the effect of aerobic exercise training on the development of
DPN, because only short-term studies were previously
performed (Richardson et al., 2001). We evaluated the
progression of neurophysiological parameters for each year
of the 4-year study period. Second, we studied diabetic
patients without baseline signs or symptoms of DPN. This
study design allowed us to evaluate the DPN development
in two groups of diabetic patients with the same baseline
neurological condition. Third, we prescribed mild submax-
imal aerobic exercise training, as well as brisk walking, to
evaluate physical exercise easily reproducible and reliable
within the lifestyle of diabetic patients. Our data seem to
indicate the effectiveness of a submaximal and reliable
exercise on DPN. Fourth, no difference at baseline on age,
duration of diabetes, BMI, waist circumference, glycemic
control, smoking, alcohol intake, blood pressure, and
nephropathy between the study and control groups
occurred. We believe that this latter point could allow us
to evaluate the effect of exercise on DPN without the
confounding role of the most important risk factors. In fact,
it is well known that DPN is associated with poor glycemic
control, obesity, dyslipidemia, retinopathy, microalbuminu-
ria, and smoking (Boulton et al., 2004). The absence of
statistically significant changes on BMI, waist circum-
ference, and metabolic profile could suggest a possible
direct and local effect of exercise on peripheral nerves in
our diabetic patients. This partially unexpected finding
could be due to the mild intensity of the training and to the
fact that no modifications on participants’ dietary habitus
have been applied in this study. However, these data should
be interpreted with caution. Finally, this study included the
expertise of a highly specialized metabolic fitness assistant
who supervised on the technical aspects of the proposed
exercise and took good care of the patients. The presence of
this professional figure could explain the good compliance
of the patients.
Several exercise-induced vascular and metabolic
changes could be invoked to explain the effects of training
on DPN development. Human and experimental studies
suggest that short-term exercise stimulates endothelium-
dependent vasodilatation. Higher vascular endothelial
growth factor (VEGF) expression during short-term exer-
cise has been proposed to play a role in endoneurial blood
flow increase (Gustafsson, Puntschart, Kaijser, Jansson, &
Sundberg, 1999). Exercise may also improve abnormal
perfusion and plasma viscosity facilitating oxygen delivery
(Terjung, Mathien, Erney, & Ogilvie, 1988; Fuchsjager-
Mayrl et al., 2002). It is known that exercise training
222 S. Balducci et al. / Journal of Diabetes and Its Complications 20 (2006) 216 – 223
exposes the vessels to repeated episodes of hyperemia. The
elevated shear stress from the increased blood flow of
aerobic exercise augments vasodilatation over the long term
by increasing the vascular expression of nitric oxide
synthase and by enhancing the release of nitric oxide (NO;
Fukai et al., 2000; Maiorana, O’Driscoll, Taylor, & Green,
2003; McAllister, Hirai, & Musch, 1995; Niebauer & Cooke,
1996). The increase of NO synthesis or bioavailability may
be useful in preventing diabetes-induced changes in the
polyol pathway (Ramana, Chandra, Srivastava, Bhatnagar,
& Srivastava, 2003). Exercise-induced nerve function
changes could be also related to an improvement of Na/K
ATPase activity. In fact, training has been reported to
increase the concentration of Na/K ATPase in rat muscle
cells (Kjeldsen, Richter, Galbo, Lortie, & Clausen, 1986). In
addition, K (ATP) channel openers have been shown to
provide marked beneficial effects on nerve perfusion and
function in experimental diabetic neuropathy (Hohman,
Cotter, & Cameron, 2000).
In conclusion, this study suggests, for the first time, that
long-term prescribed and supervised aerobic exercise train-
ing may modify the natural history of DPN or even delay its
onset. Mild aerobic exercise training, like brisk walking,
could be an effective and reasonable treatment tool to
prevent the onset or modify the natural history of DPN.
5. Study limitations
Our data are promising, but further studies on a larger
population are necessary to confirm these findings. No
conclusions on the possible exercise-related mechanisms
inducing a delayed development of DPN can be drawn from
this study.
Acknowledgments
The authors thank Gianluca Balducci, Lorella Seniga-
gliesi, and the participants, for their contribution to these
studies. We also express our gratitude to Philippa Mungra in
the preparation of the manuscript.
References
Abbott, L., Vileikyte, S., Williamson, A. L., & Carrington, A. J.
(1998). Multicenter study of the incidence of and predictive risk
factors for diabetic neuropathic foot ulceration. Diabetes Care, 21,
1071 – 1075.
Azad, N., Emanuele, N. V., Abraira, C., Henderson, W. G., Colwell, J.,
Levin, S. R., Nuttall, F. Q., Comstock, J. P., Sawin, C. T., Silbert,
C., & Rubino, F. A. (1999). The effects of intensive glycemic control
on neuropathy in the VA cooperative study on type II diabetes
mellitus (VA CSDM). Journal of Diabetes and Its Complications, 13,
307 – 313.
Balducci, S., Leonetti, F., Di Mario, U., & Fallucca, F. (2004). Is a long-
term aerobic plus resistance training program feasible for and effective
on metabolic profiles in type 2 diabetic patients? Diabetes Care, 27,
841 – 842.
Boulton, A. J. (1998). Lowering the risk of neuropathy, foot ulcers and
amputations. Diabetic Medicine, 15 (Suppl. 4), S57 – S59.
Boulton, A. J. M., Malik, R. A., Arezzo, J. C., & Sosenko, J. M. (2004).
Diabetic somatic neuropathies. Diabetes Care, 27, 1458 – 1486.
Castaneda, C., Layne, J. E., Munoz-Orians, L., Gordon, P. L., Walsmith, J.,
Foldvari, M., Roubenoff, R., Tucker, K. L., & Nelson, M. E. (2002). A
randomized controlled trial of resistance exercise training to improve
glycemic control in older adults with type 2 diabetes. Diabetes Care,
25, 2335 – 2341.
Dahl-Jorgensen, K., Brinchmann-Hansen, O., Hanssen, K. F., Ganes, T.,
Kierulf, P., Smeland, E., Sandvik, L., & Aagenaes, O. (1986). Effect of
near normoglycaemia for two years on progression of early diabetic
retinopathy, nephropathy, and neuropathy: The Oslo study. British
Medical Journal, 8, 1195 – 1199.
Davis, E. A., Jones, T. W., Walsh, P., & Byrne, G. C. (1997). The use of the
biothesiometer to detect neuropathy in children and adolescents with
IDDM. Diabetes Care, 20, 1448 – 1453.
DCCT Research Group. (1995). The effect of intensive diabetes therapy on
the development and progression of neuropathy. Annals of Internal
Medicine, 122, 561 – 568.
Dunston, D. W., Daly, R. M., Owen, N., Jolley, D., De Courten, M., Shaw,
J., & Zimmet, P. (2002). High-intensity resistance training improves
glycemic control in older patients with type 2 diabetes. Diabetes Care,
25, 1729 – 1736.
Falck, B., Andreassen, S., Groth, T., Lang, H., Melander, M., Nurmi, A.,
Rosenfalck, A., Stalberg, E., & Suojanen, M. (1991). The development
of a multicenter database for reference values in clinical neuro-
physiology—Principles and examples. Computer Methods and Pro-
grams in Biomedicine, 34, 145 – 162.
Feldman, E. L., Stevens, M. J., Thomas, P. K., Brown, M. B., Canal, N., &
Greene, D. A. (1994). A practical two-step quantitative clinical and
electrophysiological assessment for the diagnosis and staging of
diabetic neuropathy. Diabetes Care, 17, 1281 – 1289.
Fuchsjager-Mayrl, G., Pleiner, J., Wiesinger, G. F., Sieder, A. E., Quittan,
M., Nuhr, M. J., Francesconi, C., Seit, H. P., Francesconi, M.,
Schmetterer, L., & Wolzt, M. (2002). Exercise training improves
vascular endothelial function in patients with type 1 diabetes. Diabetes
Care, 25, 1795 – 1801.
Fukai, T., Siegfried, M. R., Ushio-Fukai, M., Cheng, Y., Kojda, G., &
Harrison, D. G. (2000). Regulation of the vascular extracellular
superoxide dismutase by nitric oxide and exercise training. Journal of
Clinical Investigation, 105, 1631 – 1639.
Goldhaber-Fiebert, J. D., Goldhaber-Fiebert, S. N., Tristan, M. L., &
Nathan, D. M. (2003). Randomized controlled community-based
nutrition and exercise intervention improves glycemia and cardiovas-
cular risk factors in type 2 diabetic patients in rural Costa Rica. Diabetes
Care, 26, 24 – 29.
Gustafsson, T., Puntschart, A., Kaijser, L., Jansson, E., & Sundberg, C. J.
(1999). Exercise-induced expression of angiogenesis-related transcrip-
tion and growth factors in human skeletal muscle. American Journal of
Physiology, 276, H679 – H685.
Hohman, T. C., Cotter, M. A., & Cameron, N. E. (2000). ATP-sensitive
K(+) channel effects on nerve function, Na(+), K(+) ATPase, and
glutathione in diabetic rats. European Journal of Pharmacology, 2,
335 – 341.
Kjeldsen, K., Richter, E. A., Galbo, H., Lortie, G., & Clausen, T. (1986).
Training increases the concentration of [3H]ouabain-binding sites in rat
skeletal muscle. Biochimica Et Biophysica Acta, 11 (860), 708 – 712.
Kline, G. M., Porcari, J. P., & Rippe, J. M. (1987). Estimation of Vo2max
from a one-mile track walk, gender, age and body weight. Medicine and
Science in Sports and Exercise, 19, 253 – 259.
Larsen, J. R., Sjoholm, H., Hanssen, K. F., Sandvik, L., Berg, T. J., & Dahl-
Jorgensen, K. (2003). Optimal blood glucose control during 18 years
preserves peripheral nerve function in patients with 30 years’ duration
of type 1 diabetes. Diabetes Care, 26, 2400 – 2404.
 S. Balducci et al. / Journal of Diabetes and Its Complications 20 (2006) 216 – 223
Liveson, J. A., & Ma, D. M. (Eds.). (1992). Laboratory reference for
clinical neurophysiology (pp. 278 – 317). Philadelphia7 FA Davis.
Maiorana, A., O’Driscoll, G., Goodman, C., Taylor, R., & Green, D.
(2002). Combined aerobic and resistance exercise improves glycemic
control and fitness in type 2 diabetes. Diabetes Research and Clinical
Practice, 56, 115 – 123.
Maiorana, A., O’Driscoll, G., Taylor, R., & Green, D. (2003). Exercise and
the nitric oxide vasodilator system. Sports Medicine, 33, 1013 – 1035.
McAllister, R. M., Hirai, T., & Musch, T. I. (1995). Contribution of
endothelium-derived nitric oxide (EDNO) to the skeletal muscle blood
flow response to exercise. Medicine and Science in Sports and Exercise,
27, 1145 – 1151.
Muller-Felber, W., Landgraf, R., Scheuer, R., Wagner, S., Reimers, C. D.,
Nusser, J., Abendroth, D., Illner, W. D., & Land, W. (1993, Oct.).
Diabetic neuropathy 3 years after successful pancreas and kidney
transplantation. Diabetes, 42 (10), 1482 – 1486.
Muller-Felber, W., Landgraf, R., Wagner, S., Mair, N., Nusser, J.,
Landgraf-Leurs, M. M., Abendroth, A., Illner, W. D., & Land, W.
(1991). Follow-up study of sensory-motor polyneuropathy in type 1
(insulin-dependent) diabetic subjects after graft rejection. Diabetologia,
(Suppl. 1), S113 – S117.
Navarro, X., Sutherland, D. E., & Kennedy, W. R. (1997). Long-term
effects of pancreatic transplantation on diabetic neuropathy. Annals of
Neurology, 42, 727 – 736.
Niebauer, J., & Cooke, J. P. (1996). Cardiovascular effects of exercise: Role
of endothelial shear stress. Journal of the American College of
Cardiology, 28, 1652 – 1660.
Poncelet, A. N. (2003). Diabetic polyneuropathy. Risk factors, patterns of
presentation, diagnosis, and treatment. Geriatrics, 58, 16 – 18.
223
Position paper. (1996). The Consensus Committee of the American
Autonomic Society and the American Academy of Neurology: The def-
inition of orthostatic hypotension, pure autonomic failure, and multiple
system atrophy. Journal of Autonomic Nervous System, 58, 123 – 124.
Ramana, K. V., Chandra, D., Srivastava, S., Bhatnagar, A., & Srivastava,
S. K. (2003). Nitric oxide regulates the polyol pathway of glucose
metabolism in vascular smooth muscle cells. Federation of American
Societies for Experimental Biology, 17, 417 – 425.
Richardson, J. K., Sandman, D., & Vela, S. (2001). A focused exercise
regimen improves clinical measures of balance in patients with
peripheral neuropathy. Archives of Physical Medicine and Rehabilita-
tion, 82, 205 – 209.
Ruiz, J. (2004). Intensive care of type 2 diabetes: What does the STENO 2
study teach us? Revue Medicale De La Suisse Romande, 124, 125 – 127.
Terjung, R. L., Mathien, G. M., Erney, T. P., & Ogilvie, R. W. (1988).
Peripheral adaptations to low blood flow in muscle during exercise. The
American Journal of Cardiology, 62, 15E – 19E.
Tesfaye, S., Harris, N. D., Wilson, R. M., & Ward, J. D. (1992). Exercise-
induced conduction velocity increment: A marker of impaired peripheral
nerve blood flow in diabetic neuropathy. Diabetologia, 35, 155 – 159.
Vinik, A. I. (1999). Diabetic neuropathy: Pathogenesis and therapy. The
American Journal of Medicine, 107 (2B), 17S – 26S.
Vinik, A. I., Erbas, T., Stansberry, K. B., & Pittenger, G. L. (2001). Small
fiber neuropathy and neurovascular disturbances in diabetes mellitus.
Experimental and Clinical Endocrinology and Diabetes, 109 (Suppl. 2),
S451 – S473.
Young, M. J., Breddy, J. L., Veves, A., & Boulton, A. J. M. (1994). The
prediction of diabetic foot ulceration using vibration perception
thresholds: A prospective study. Diabetes Care, 17, 557 – 560.