Archives of Medical Research 40 (2009) 180e185

ORIGINAL ARTICLE
Diabetic Polyneuropathy May Increase the Handicap Related
to Vestibular Disease
Catalina Aranda,a Anabel Meza,a Raymundo Rodrı́guez,a Marı́a Teresa Mantilla,a and
Kathrine Jáuregui-Renaudb
a
Hospital Regional 72, Instituto Mexicano del Seguro Social, Estado de México, Mexico
b
Unidad de Investigación Médica en Otoneurologı́a, CMN Siglo XXI, Instituto Mexicano del Seguro Social, México, D.F., Mexico
Received for publication November 28, 2008; accepted December 26, 2008 (ARCMED-D-08-00541).

Background and Aims. We undertook this study to assess the influence of diabetic
peripheral neuropathy on self-reported disability and postural control during quiet stance
of patients with peripheral vestibular disease, before and after a standardized program of
vestibular rehabilitation (Cawthorne & Cooksey exercises).
Methods. Twenty patients with peripheral vestibular disease participated in the study
(mean age 56  7.8 years), 10 with and 10 without peripheral neuropathy (age matched).
The Dizziness Handicap Inventory and static posturography (eyes open/closed and firm/
soft surface) were evaluated prior to rehabilitation and at week 7 of follow-up.
Results. Compared to patients without neuropathy, patients with neuropathy had more
time elapsed since the diabetes was diagnosed, higher glycemia and HbAc level and high-
er composite scores on the Dizziness Handicap Inventory, but similar results on static
posturography. After rehabilitation, although scores on the Dizziness Handicap Inventory
decreased in the two groups, the difference between them persisted. In patients with
neuropathy, static posturography showed improvement of postural control only with
the eyes closed and soft surface, whereas in patients without neuropathy the postural
control improved during all sensory conditions (eyes open/closed and firm/soft surface).
Conclusions. In diabetic patients with peripheral vestibular disease, peripheral neurop-
athy contributes to self-reported disability and may interfere with complete balance
recovery. Ó 2009 IMSS. Published by Elsevier Inc.
Key Words: Diabetes mellitus, Peripheral neuropathy, Vestibular disease.

Introduction
Postural control is dependent upon integration of signals
from the vestibular, visual and somatosensory systems to
generate the motor responses that maintain upright position
and adjust to destabilizing forces (1). Under altered sensory
conditions, the central nervous system reweighs each
sensory input, enhancing the influence of those senses
providing accurate information and suppressing the con-
flicting or inaccurate input (2). In general practice, the most
frequent cause of balance disorders is vestibular disease.
Address reprint requests to: Dr. Kathrine Jáuregui Renaud, Unidad de
Investigación Médica en Otoneurologı́a, P.B. Edificio C-Salud en el Traba-
jo, Centro Médico Nacional Siglo XXI, Av. Cuauhtemoc 330 Colonia Doc-
tores, C.P. 06720, México D.F.; E-mail: kathrine.jauregui@imss.gob.mx
During passive stance, regions of the feet that are more
anterior receive the most mechanical pressure in supporting
the body against gravity (3). This provides a mapping of body
orientation to the upright. Although plantar sensation is of
moderate importance for the maintenance of normal standing
balance, the impact of reduced plantar sensitivity on postural
control is expected to increase with the loss of additional
sensory modalities such as the concomitant proprioceptive
deficits commonly associated with peripheral neuropathies
(4). Furthermore, in patients with peripheral polyneuropathy,
evidence suggests that visual and vestibular inputs may not be
sufficient to compensate for impaired somatosensation (5,6).
The most frequent known cause of peripheral
neuropathy is diabetes mellitus; symptomatic degrees of
polyneuropathy may occur in |15% of patients with

0188-4409/09 $esee front matter. Copyright Ó 2009 IMSS. Published by Elsevier Inc.
doi: 10.1016/j.arcmed.2009.02.011
 Diabetic Neuropathy and Vestibular Disease
insulin-dependent diabetes mellitus and 13% of patients
with non-insulin-dependent diabetes mellitus (7,8). Dia-
betic neuropathy is a long-term complication. The mean
fasting blood or plasma glucose concentrations during long
term follow-up are higher in patients in whom polyneurop-
athy develops than in those in whom polyneuropathy does
not develop (9).
Compared with diabetic patients without neuropathy, dia-
betic patients with neuropathy report 15 times more injuries
during gait (10). Evidence has shown that diabetic distal
neuropathy is related to postural instability (11,12). Even with
vision, postural mechanisms at ankle level are impaired during
quiet stance (13), and the postural control deficit is greatest
when visual or vestibular input is absent or degraded (11). In
these studies, posturography has been useful to disclose failure
of postural control related to diabetic neuropathy.
The increasing prevalence of diabetes mellitus entails
that the co-occurrence of vestibular disease and diabetic
peripheral neuropathy may increase. This study was de-
signed to assess the influence of diabetic peripheral neurop-
athy on self-reported disability and postural control of
patients with peripheral vestibular disease before and after
a standardized program of vestibular rehabilitation (Caw-
thorne & Cooksey exercises).
Subjects and Methods
Participants
Twenty patients with peripheral vestibular disease and type
2 diabetes mellitus (T2DM) accepted to participate in the
study as follows: Group 1e10 patients with T2DM without
diabetic peripheral neuropathy, all women (mean age
54.9  7.5 years old). The mean time elapsed since diabetes
was diagnosed was 5.9  3.4 years, and the evolution time
of the balance symptoms was from 2 months to 3 years
(median 3 months). Five patients had a history of medically
controlled high blood pressure (50%). Group 2 was
comprised of 10 patients with T2DM and diabetic periph-
eral neuropathy, seven women and three men (57.2  8.3
years old). The mean time elapsed since diabetes was diag-
nosed was 13  10.3 years, and the evolution time of the
balance symptoms was from 2 months to 10 years (median
18 months). Three patients had a history of high blood pres-
sure under medical control (30%).
All patients were selected at a neurotology outpatient
clinic and were invited to participate when there was no
clinical evidence of middle ear disease, musculoskeletal
compromise, neurological disease other than diabetic
neuropathy, rheumatic disease or a history of alcohol abuse.
Patients with benign paroxysmal positional vertigo or Me-
nièrés disease were not included in the study. The study
protocol was evaluated and approved by the local research
committee, and informed consent was obtained from all the
subjects prior to participation.
181
Procedures
Patients reported their balance symptoms using a standard-
ized questionnaire (14), which includes the symptoms
shown in Figure 1. Because symptom duration was long
in the majority of patients, an accurate etiological diagnosis
was not possible. None of the participants had spontaneous
nystagmus either in the light or in darkness but had
abnormal results on a 30 C and 44 C caloric test. In this
study, unilateral hypofunction was defined as an asymmetry
of $ 20% between right and left responses to 30 C and
44 C caloric stimuli, and bilateral hypofunction was
defined as absent responses to both 30 C and 44 C caloric
stimuli. Caloric test showed unilateral hypofunction in all
but one patient without neuropathy who had bilateral hypo-
function. Directional preponderance was not evident in any
case.
After evaluation by a physician specialized in internal
medicine who administered the Michigan diabetic
neuropathy score (15), a blood sample was obtained to
assess HbAc, glucose, cholesterol and triglycerides. Dia-
betic retinopathy was investigated by an ophthalmologist,
and a physical medicine and rehabilitation physician per-
formed nerve conduction studies using standardized tech-
niques with temperature control and fixed distances.
Nerve conduction velocity was evaluated on the median,
ulnar, tibial, peroneal and sural nerves (Spirit, Nicolet,
Madison, WI).
At day 1 (before starting the rehabilitation program) and
at week 7 of follow-up, an evaluation was performed using
the Dizziness Handicap Inventory (16) and static posturog-
raphy (Posturolab 40/16, Medicapteurs, Cedex, France).
Cawthorne & Cooksey Rehabilitation Program
This program is based on a series of exercises of increasing
complexity, which include movements of the head, tasks
requiring coordination of the eyes with the head, total body
movements and balance tasks (17,18). All patients were
given written instructions with diagrams describing the
exercises. After explaining the premise of the program,
the level of complexity to start the program for each patient
was evaluated. All patients were instructed to carry out the
exercises for at least 10 min twice daily and to keep prac-
ticing the same group of exercises for as long as the vertigo
persisted and progress to the next level whenever they were
able to tolerate the exercises. Patients were also instructed
to use a visual focal point whenever dizziness started. Each
week the same physician evaluated treatment compliance
by patient self-report.
None of the patients underwent any other type of treat-
ment for their vestibular disorder while they participated
in the study. All patients reported regular practice of the
exercises, and all made progress on the exercise perfor-
mance during follow-up.
182 Aranda et al./ Archives of Medical Research 40 (2009) 180e185

Vertigo

Dizziness

Frequent falls

Frequent stumbles

Instability when looking at moving objects

Instability when changing posture

Instability when moving the head rapidly

Instability when walking in the dark

Instability when walking on uneven surfaces

Neuropathy 0% 20% 40% 60% 80% 100%

No neuropathy

Figure 1. Frequency of balance symptoms reported by 20 patients with type 2 diabetes mellitus and chronic peripheral vestibular disease, 10 without and 10
with peripheral neuropathy (no significant difference between groups, t-test for proportions).

Dizziness Handicap Inventory Results

Twenty five questions are subgrouped into three content Compared to patients without neuropathy, as expected,
domains representing functional, emotional, and physical patients with neuropathy had more time elapsed since dia-
aspects of dizziness and unsteadiness. A ‘‘yes’’ response betes was diagnosed, higher glucose and HbAc levels, longer
is scored 4 points, ‘‘sometimes’’ is scored 2 points, and evolution time of balance symptoms, and higher scores on the
a ‘‘no’’ response is scored 0 points. Thus, the composite Michigan diabetic neuropathy score ( p !0.05) (Table 1).
score ranges from 0 to 100, whereas lower scores are closer Patients with neuropathy showed slower nerve conduction
to normal (16). velocity for all nerves tested (upper and lower limbs) than
patients without neuropathy (t-test, p !0.01); sensory/motor
Static Posturography polyneuropathy was identified in all but one patient who
showed mainly a motor component. Additionally, diabetic
Body sway during 25.6 sec of quiet upright stance was re-
retinopathy was identified in three patients with neuropathy
corded with a force platform (Medicapteurs). Subjects were
and none of the patients without neuropathy.
asked to stand upright and barefoot on the platform with
arms at their sides and remaining as still as possible.
Recordings were made under the following conditions: Table 1. Clinical characteristics of 20 patients with type 2 diabetes
mellitus (T2DM)*
standing with eyes open or closed on either a hard or a soft
surface made with a layer of foam rubber (5-cm thick, Variable No neuropathy Neuropathy p**
density 2.5 pfc).
Mean  SD Mean  SD
BMI 30.6  4.3 28.2  4.8 e
Statistical Analysis Glucose (mg/dl) 138  16 185  52 0.01
HbAc (%) 6.4  0.4 7.9  1.3 0.002
Statistical analysis was performed after comparing data Cholesterol (mg/dl) 200  38 225  38 e
distribution with the normal distribution (Kolmogorov- Triglycerides (mg/dl) 186  76 224  103 e
Smirnov test). According to data distribution and scale of Michigan diabetic 10.8  6.8 24.1  8 0.0005
measurement the following tests were used: Wilcoxon test, neuropathy score
Mann-Whitney U test, t-test for means, t-test for propor- SD, standard deviation; BMI, body mass index.
tions and analysis of covariance (CSS Statsoft, Tulsa, *Ten without and 10 with peripheral neuropathy.
OK); p values !0.05 were considered significant. **t-test.
 Diabetic Neuropathy and Vestibular Disease
Balance symptoms reported by each group are shown in
Figure 1. Frequency and score of symptoms were similar
for the two groups. Median score was 6 (Q1eQ3: 5e8.5)
for patients without neuropathy vs. median score of 7.5
(Q1eQ3: 7e8.75) for patients with neuropathy (Mann-
Whitney U test, p O0.05).
At the first evaluation, the Dizziness Handicap Inven-
tory showed higher composite scores for patients with
neuropathy than for those without it. Median score was
67 (Q1eQ3: 45e81) vs. median score of 37 (Q1eQ3:
18e49), respectively (Table 2) (Mann-Whitney U test,
p !0.01). At this time, static posturography showed
similar results for the two groups on the length and area
of oscillation and the Romberg coefficient during all
conditions (Table 3) (Mann-Whitney U test, p O0.05).
Although the anterior/posterior and mediolateral position
of the center of pressure of the two groups was also similar
during recordings with eyes open on a hard surface, the
variance of the anterior/posterior position of the center
of pressure was less in patients without neuropathy than
in those with neuropathy. Median values were 9.3 mm
(Q1eQ3: 8.7e12.8) vs. 20.3 mm (Q1eQ3: 17.2e40.1)
(Mann-Whitney U test, p O0.02). Additionally, for all
patients, covariance analysis including group, age, time
elapsed since diabetes was diagnosed and evolution time
of the balance symptoms showed that the duration of both
diabetes and balance symptoms were consistently corre-
lated with the length of oscillation either with the eyes
open or closed during the hard surface conditions
(adjusted whole model R2 5 0.5, p 5 0.002), whereas
b values were positive for the time elapsed since diabetes
was diagnosed but negative for the time of evolution of the
balance symptoms. During the soft surface conditions, the
length of oscillation was correlated just with the time
elapsed since diabetes was diagnosed (adjusted whole
model R2 5 0.33, p 5 0.02).
After 7 weeks, within each group the composite scores and
the scores of the three content domains decreased (Wilcoxon
test, p #0.01) (Table 2); a decrease of the composite score of
at least 18 points was observed in six patients from each
group. However, the difference between groups on the Dizzi-
ness Handicap Inventory persisted. At this time, in patients
Table 2. Total score on the Dizziness Handicap Inventory before and after vestibular rehabilitation of 20 patients with T2DM*
No neuropathy
Median Q1eQ3 Median Q1eQ3
Domain Day 1 Week 8
Physical 10 (4e16) 4 (4e10)
Emotional 6 (4e26) 6 (2e11)
Functional 8 (6e17) 4 (2e11)
Total score 38 (18e49) 14 (9e30)
*Ten without and 10 with peripheral neuropathy.
**Wilcoxon test.
183
with neuropathy static posturography showed a decrease of
the area of oscillation only when the eyes were closed while
standing on a soft surface (Table 3), whereas in patients
without neuropathy posturography showed a decrease of
the oscillation during the four sensory conditions (Table 3).
Additionally, only in patients without neuropathy, a decrease
of the Romberg coefficient during the soft surface conditions
was observed. For the whole group, the correlation between
the length of oscillation and the duration of both diabetes
and balance symptoms was similar to the correlation
observed at the first evaluation.
Discussion
Study results show that a combined deficit of vestibular and
somatosensory input may preclude adjustments to postural
control and increase the handicap during activities of daily
life. Although the balance symptoms reported by each
group were similar compared to patients without neurop-
athy, in vestibular patients with peripheral neuropathy the
postural control improved less and the disability was
always higher.
Though diabetes per se has no effect on postural stability
(11), diabetic neuropathy may compromise postural control
(11,12). Evidence has shown that diabetic neuropathy leads
to a decrease in rapidly available ankle strength (19) and
ankle movement perception (20), which are very important
for anterior/posterior balance during side-by-side upright
stance (21). The larger variance of the anterior/posterior
position of the center of pressure observed on patients with
neuropathy during recordings with the eyes open on a hard
surface suggests the use of additional muscular activity in
order to keep the center of pressure within the limits of
stability. The positive correlation between the length of
oscillation and the time elapsed since the diabetes was diag-
nosed can be explained by the fact that diabetic neuropathy
is a long-term complication, whereas the negative correla-
tion with the evolution time of balance symptoms may indi-
cate that the patients learned to exploit anticipatory postural
strategies.
Although at the beginning of the study static posturogra-
phy showed similar results for the two groups, after vestibular
Neuropathy
Median Q1eQ3 Median Q1eQ3
p** Day 1 Week 8 P**
0.005 20 (13e24) 12 (9e18) 0.01
0.01 20 (11e28) 10 (7e20) 0.01
0.01 25 (18e32) 14 (9e19) 0.02
0.007 67 (45e81) 39 (36e45) 0.02
184 Aranda et al./ Archives of Medical Research 40 (2009) 180e185

Table 3. Static posturography results before and after vestibular rehabilitation of 20 patients with T2DM*

No neuropathy Neuropathy

Median Q1eQ3 Median Q1eQ3 Median Q1eQ3 Median Q1eQ3

Condition Day 1 Week 8 p** #0.05 Day 1 Week 8 p** #0.05

Hard surface
Eyes open
Length (mm) 375 (343-403) 296 (239e352) — 409 (398e462) 354 (239e410) —
Area (mm2) 110 (86e170) 78 (53e136) 0.009 159 (86e288) 109 (67e193) —
Eyes closed
Length (mm) 635 (537e684) 534 (281e579) 0.01 493 (332e664) 489 (305e636) —
Area (mm2) 206 (135e365) 127 (97e283) — 213 (131e263) 141 (135e323) —
Romberg coefficient 176 (138e256) 132 (111e171) — 156 (119e162) 158 (131e189) —
Soft surface
Eyes open
Length (mm) 488 (440e725) 434 (364e485) 0.01 608 (440e754) 550 (319e732) —
Area (mm2) 269 (179e476) 174 (117e301) — 359 (204e583) 269 (158e572) —
Eyes closed
Length (mm) 952 (670e1525) 645 (519e857) 0.02 960 (509e1168) 830 (455e870) —
Area (mm2) 461 (271e950) 453 (181e671) — 873 (302e1123) 454 (251e731) 0.03
Romberg coefficient 316 (197e460) 203 (165e233) 0.02 191 (118e270) 167 (146e271) —

*Ten without and 10 with peripheral neuropathy.

**Wilcoxon test.

rehabilitation patients with peripheral neuropathy showed
improvement only when the eyes were closed and the surface
was soft, when vestibular information would be more essen-
tial to use. This finding may be attributed to a better use of
vestibular signals with no change in the combined use of
somatosensory, visual and vestibular information. In
contrast, patients who had adequate somatosensory informa-
tion available showed an oscillation decrease during the four
sensory conditions. Better use of the vestibular signals
combined with an enhanced ability to interpret somatosen-
sory and visual input may have improved their static postural
control, either when the surface was hard or soft.
The results suggest that in patients with peripheral
vestibular disease and diabetic peripheral neuropathy,
a standardized program of vestibular rehabilitation may
improve the vestibular dysfunction with no effect on the
additional deficit related to peripheral neuropathy. Further
studies are needed to design the best approach to rehabili-
tate balance in patients with combined vestibular and
somatosensory deficits.
Acknowledgments
This work was supported by Fondo para el Fomento de la Inves-
tigación Médica del Instituto Mexicano del Seguro Social.
References
1. Nashner LM, Shupert CL, Horak FB, Black FO. Organization of
posture controls: an analysis of sensory and mechanical constraints.
Prog Brain Res 1989;80:411e418.
2. Diener HC, Horak FB, Nashner L. Influence of stimulus parameters on
human postural responses. J Neurophysiol 1988;59:1888e1905.
3. Kavounoudias A, Roll R, Roll JP. Foot sole and ankle muscle inputs
contribute jointly to human erect posture regulation. J Physiol May
2001;532(Pt 3):869e878.
4. Meyer PF, Oddsson LI, De Luca CJ. The role of plantar cutaneous
sensation in unperturbed stance. Exp Brain Res 2004;156:505e512.
5. Jáuregui-Renaud K, Kovacsovics B, Vrethem M, Ödkvist LM,
Ledin T. Dynamic posturography and randomized perturbed posturog-
raphy in the follow-up of patients with polyneuropathy. Arch Med Res
1998;29:39e44.
6. Giacomini PG, Bruno E, Monticone G, Di Girolamo S, Magrini A,
Parisi L, et al. Postural rearrangement in IDDM patients with periph-
eral neuropathy. Diabetes Care 1996;19:372e374.
7. Diabetes. Facts sheet/312. Geneva: World Health Organization;
September 2006.
8. Dyck PJ, Kratz KM, Karnes JL, et al. The prevalence by staged
severity of various types of diabetic neuropathy, retinopathy, and
neuropathy in a population based cohort: the Rochester Diabetic
Neuropathy Study. Neurology 1993;43:817e824.
9. Partanen J, Niskanen L, Lehtinen J, Mervaala E, Siitonen O, Uusitupa M.
Natural history of peripheral neuropathy in patients with non-insulin-
dependent diabetes mellitus. N Engl J Med 1995;333:89e94.
10. Cavanagh PR, Derr JA, Ulbrecht JS, Maser RE, Orchard TJ. Problems
with gait and posture in neuropathic patients with insulin-dependent
diabetes mellitus. Diabet Med 1992;9:469e474.
11. Simoneau GG, Ulbrecht JS, Derr JA, Becker MB, Cavanagh PR.
Postural instability in patients with diabetic sensory neuropathy. Dia-
betes Care 1994;17:1411e1421.
12. Oppenheim U, Kohen-Raz R, Alex D, Kohen Raz A, Azarya M. Postural
characteristics of diabetic neuropathy. Diabetes Care 1999;22:328e332.
13. Lafond D, Corriveau H, Prince F. Postural control mechanisms during
quiet standing in patients with diabetic sensory neuropathy. Diabetes
Care 2004;27:173e178.
14. Jáuregui-Renaud K, Gutierrez MA, Viveros RL, Villanueva PL. Sı́nto-
mas de inestabilidad corporal y enfermedad vestibular. Rev Med IMSS
2003;41:373e378.
 Diabetic Neuropathy and Vestibular Disease
15. Feldman EL, Stevens MJ, Thomas PK, Brown MB, Canal N. A prac-
tical two-step quantitative clinical and electrophysiological assessment
for the diagnosis and staging of diabetic neuropathy. Diabetes Care
1994;17(11):1281e1289.
16. Jacobson GP, Newman CW. The development of the Dizziness Handicap
Inventory. Arch Otolaryngol Head Neck Surg 1990;116:424e427.
17. Cawthorne T. The physiological basis for head exercises. J Chart Soc
Physiother 1944;29:106e107.
18. Cooksey FS. Physical medicine. Practitioner 1945;155:300e305.
185
19. Gutierrez EM, Helber MD, Dealva D, Ashton-Miller JA,
Richardson JK. Mild diabetic neuropathy affects ankle motor function
Clin Biomech (Bristol, Avon) 2001;16:522e528.
20. Simoneau GG, Derr JA, Ulbrecht JS, Becker MB, Cavanagh PR. Dia-
betic sensory neuropathy effect on ankle joint movement perception.
Arch Phys Med Rehabil 1996;77:453e460.
21. Winter DA, Prince F, Frank JS, Powell C, Zabjek KF. Unified theory
regarding A/P and M/L balance in quiet stance. J Neurophysiol 1966;
75:2334e2343.