Scientific investigations
A Comparison of Glycemic Control, Sleep, Fatigue, and Depression in Type 2
Diabetes with and without Restless Legs Syndrome
Norma G. Cuellar, D.S.N.1; Sarah J. Ratcliffe, Ph.D.2
1
University of Pennsylvania School of Nursing, Philadelphia, PA; 2Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania
School of Medicine, Philadelphia, PA
Objective: The purpose of this pilot study was to determine how sleep
quality, glycemic control, sleepiness, fatigue, and depression differ in
persons with type 2 diabetes with and without restless legs syndrome
(RLS).
Research Designs/Methods: The design was a descriptive, case-
control study of participants with type 2 diabetes with and without RLS
at the University of Pennsylvania, Rodebaugh Diabetes Center. Thirty-
nine participants (adults over 21 years of age who had been diag-
nosed with type 2 diabetes and had a HbA1c in the last 3 months) were
stratified based on RLS diagnostic criteria. Exclusion criteria included
severe hypoglycemia identified by seizures or coma related to hypogly-
cemia or known sleep disorder other than RLS. The primary outcome
of sleep was measured by self-report sleep quality (Pittsburgh Sleep
Quality Index) with secondary outcomes including HbA1c, sleepiness
(Epworth Sleepiness Scale), fatigue (Fatigue Severity Scale), and de-
pression (Center for Epidemiologic Studies-Depression Scale).
D iabetes is the fifth leading cause of death in America, af-
fecting 17 million people or 6.2% of the population. In
2002, the direct cost of diabetes was $91.8 billion dollars in
healthcare costs, with indirect costs of disability, work loss, and
premature mortality at $40.2 billion dollars.1 These alarming
statistics are due in part to the rising problem of obesity and
chronic diseases in our population, resulting in increased in-
sulin resistance and increased rates of type 2 diabetes, which
is most often diagnosed after the age of 40. Sleep disturbances
and sleep loss are implicated in insulin resistance, a precursor
to type 2 diabetes. Individuals who sleep more than 8 hours per
day or less than 7 hours per day are at modestly increased risk
of all-cause mortality, cardiovascular disease, and developing
symptomatic diabetes.2
Disclosure Statement
This was not an industry supported study. The authors have indicated no
financial conflicts of interest.
Submitted for publication August, 2007
Accepted for publication October, 2007
Address correspondence to: Norma G. Cuellar, Assistant Professor,
School of Nursing, University of Pennsylvania, 418 Curie Blvd, Philadel-
phia, PA 19104; Tel: (215) 898-1935; Fax: (215) 573-6464; E-mail: ncuel-
lar@nursing.upenn.edu
Journal of Clinical Sleep Medicine, Vol. 4, No. 1, 2008
Results: Participants with type 2 diabetes with RLS reported a signifi-
cant difference in quality of sleep (p = .001), sleep latency (p = .04),
sleep efficiency (p = .035), use of sleep medications (p < .001), and
daytime dysfunction (p = .005). In the total group, higher HbA1c levels
were positively correlated with sleepiness (p = .038). Global Pittsburgh
Sleep Quality Index scores were positively correlated with fatigue (r =
.58, p = .002) and depression (r = .74, p < .001). As well, fatigue and
sleepiness were positively correlated (r = .36, p = .04).
Conclusions: RLS is a significant sleep disorder that may have an
impact on diabetes management and health outcomes. More research
should be conducted on the impact of RLS in sleep to improve diabetic
management.
Keywords: type 2 diabetes, restless legs syndrome, sleep
Citation: Cuellar NG; Ratcliffe SJ. A comparison of glycemic control,
sleep, fatigue, and depression in type 2 diabetes with and without rest-
less legs syndrome. J Clin Sleep Med 2008;4(1):50-56.
One sleep disorder that may affect the management of diabe-
tes, yet is unrecognized as a significant contributor to diabetes
health outcomes, is restless legs syndrome (RLS) increasingly
seen in type 2 diabetes.3 However, the association between RLS
and diabetes has not been studied thoroughly, and the effects
of RLS on diabetes are unknown. In the results of the National
Sleep Foundation Sleep in America 2005 Poll, adults with dia-
betes were more likely to be at risk for RLS (p < 0.05). The
adults who were at risk for RLS were also at increased risk for
obstructive sleep apnea (OSA) and insomnia (p < 0.05), sleep
problems often reported in diabetes.2
Although the cause of RLS remains unknown, current hy-
potheses implicate abnormalities in central nervous system iron
and dopamine metabolism.4 RLS may affect up to 15% of the
population5,6 with debilitating symptoms of sleep deprivation,
discomfort, and fatigue that can interfere with family life and
occupational and social activities.7 The prevalence of RLS is
difficult to determine because it is often misdiagnosed or not
diagnosed due to the myriad of symptoms with which patients
may present. RLS symptoms worsen with age and may occur
daily, despite the use of pharmacologic agents that temporarily
treat the symptoms.5,8 A higher prevalence of women is seen
more frequently than men in the older population.3,8 Many in-
dividuals that are affected by the syndrome do not report it to
their healthcare providers, even when their symptoms are se-
50

vere because they do not realize that it can be treated or is a
problem worth mentioning.9
Secondary RLS is associated with comorbid disease, includ-
ing diabetes. In persons with RLS, 21% report having diabetes,8
a prevalence more than 3 times that of the general population;
however, there is no report found in the literature that provides
the incidence or prevalence of RLS in the diabetic population
in the US. RLS is significantly associated with self-reported di-
minished general health and poor mental health and correlates
with age, increasing body mass index, sedentary lifestyle, and
low exercise, all factors associated with type 2 diabetes.8
Studies specifically examining the symptomatology of RLS
in diabetes have not been done in the US. In Brazil, RLS was
found in 27% of patients with diabetes. In 45% of these pa-
tients, poor sleep quality was associated with age (p = 0.04),
peripheral neuropathy (p = 0.001), and RLS (p = 0.000). Lo-
gistic regression analysis revealed an association between
RLS and peripheral neuropathy (p = 0.001).10 In Italy, RLS
was diagnosed in 18% of patients with type 2 diabetes and
was independently associated with diabetes (p < .04). Poly-
neuropathy was the main risk factor for RLS but only partially
explained the increased prevalence.11 In another study recent-
ly conducted in the same country, RLS was present in 33 of
99 patients with neuropathy associated with diabetes mellitus,
impaired glucose tolerance, and impaired fasting glucose. The
patients with RLS were more commonly diagnosed with small
fiber sensory neuropathy and more often reported symptoms
of burning feet.12
It is unknown if the presence of RLS in patients with type 2
diabetes contributes to sleep disturbances that may affect gly-
cemic control, in turn, exacerbating the severity of symptoms
of RLS or contributing to diabetic complications. Poor diabetic
control with RLS may also impact the associated consequences
of diabetes, including sleep quality, sleepiness, fatigue, and de-
pression. Sleep disorders in diabetes have been reported pri-
marily as sleep-disordered breathing or OSA associated with
obesity, body mass index, diet, and exercise. The impact of
sleep loss related to other sleep disorders, besides OSA, should
be considered in patients with type 2 diabetes. Therefore, the
purpose of this pilot study was to determine how sleep qual-
ity, glycemic control, sleepiness, fatigue, and depression differ
in persons with type 2 diabetes with and without RLS. It was
hypothesized that participants with type 2 diabetes with RLS
would have worse self-reported sleep quality (the primary out-
comes measure of the study) as well as higher levels of HbA1c,
sleepiness, fatigue, and depression than those participants with
type 2 diabetes without RLS.
Methods
Research Design
The design of this pilot study was a descriptive, comparative,
case-control study. The target population was patients with type
2 diabetes with and without RLS. Thirty-nine participants were
stratified to 2 groups (with or without RLS). All participants
continued their standard care for diabetes and RLS, including
any pharmacologic interventions.
Journal of Clinical Sleep Medicine, Vol. 4, No. 1, 2008
RLS in type 2 DM
Participants
Participants (n = 39) with type 2 diabetes were recruited
from the Penn Rodebaugh Diabetes Center at the University of
Pennsylvania in Philadelphia between June 2004 and December
2005. One participant was excluded after data collection due to
the misreported diagnosis of type 1 diabetes mellitus. A second
participant was reassigned to the non-RLS group after reporting
only 3 out of the 4 criteria for RLS on the survey. Institutional
Review Board approval was received from the Office of Regu-
latory Affairs at the University of Pennsylvania.
Procedures
Patients who had been seen at the Diabetes Center in the pre-
vious year were sent letters asking for participation in a study
examining sleep in type 2 diabetics. The letters provided a brief
explanation of the study, and participants were asked to call
for more information. The letter stated that patients could be
included if they did not have sleep problems. Participants were
asked to contact the principal investigator if interested in par-
ticipating in the study.
When contact was made with the principal investigator, ver-
bal consent over the phone was given to provide the principal
investigator with information needed to screen the participant
and determine if the participant qualified for the study and had
symptoms of RLS. The principal investigator obtained verbal
consent over the phone to obtain healthcare information related
to inclusion and exclusion criteria. Inclusion criteria were men
and women older than 21 years of age who had been diagnosed
with type 2 diabetes and had a HbA1c in the last 3 months. Ex-
clusion criteria included persons with severe hypoglycemia, de-
fined as ever having seizures or coma related to hypoglycemia,
or a sleep disorder other than RLS (e.g., OSA, narcolepsy, other
sleep-disordered breathing).
The participants were screened for RLS based on the 4 diag-
nostic criteria developed in 1995, including (1) an urge to move
legs due to discomfort, (2) temporary relief with movement, (3)
worsening symptoms at rest, and (4) worsening symptoms at
night13 and were assigned to 1 of the 2 groups: with or without
RLS. At this time, they were asked if they would like to partici-
pate in the study.
While the participants were being enrolled in the study, re-
cords were collected based on meeting the 4 criteria and were
documented to estimate the prevalence of RLS in type 2 dia-
betes. If the participants agreed to be in the study, participants
were mailed the consent, demographic survey, and question-
naires with a 100% return rate. A $25.00 gift card from a local
pharmacy was mailed to participants when the surveys were
returned to the principal investigator in a stamped envelope
provided.
Measures
Demographic data were collected on age, sex, marital status,
ethnicity, employment status, education, socioeconomic status,
comorbid conditions, pharmacologic interventions, and non-
conventional treatments. The use of all pharmacologic medi-
cations prescribed, natural products, and vitamins with name,
51
NG Cuellar, SJ Ratcliffe

dose, and frequency were listed on the demographic data form.

All participants with type 2 diabetes
Objective measures were collected by self-report on HbA1c TOTAL assessed for eligibility
height, weight, and blood pressure. Information collected on n=121

the participants with RLS included family history and age of

onset of RLS. Common comorbid conditions associated with ALLOCATION: RLS symptoms No RLS
n=121 n=54 (45%) n=67 (55%)
RLS that were listed included iron deficiency, vitamin deficien-
cies, neuropathy, rheumatoid arthritis, Parkinson disease, em- Stratified by RLS
physema, hypothyroidism, kidney failure, fibromyalgia, hyper- EXCLUDED: n=51
criteria n=25 n=26
tension, hypotension, chronic fatigue syndrome, irritable bowel
Those who chose not
syndrome, and other. to participate
The primary outcome of the study was sleep quality mea-
sured by the Pittsburgh Sleep Quality Index (PSQI) examining
7 components (sleep quality, sleep latency, sleep duration, ha- EXCLUDED: n=31 n=11 n=20

bitual sleep efficiency, sleep disturbances, use of sleep medica- Those who did not Type 1 DM = 1* Type 1 DM = 3
meet inclusion OSA = 9 OSA = 13
tions, and daytime dysfunction) and a global score. With a total criteria Seizures = 1 Seizures = 2
of 19 questions, participants rate the components on a scale of Narcolepsy = 1
Death = 1
0 to 3 with a possible range from 0 to 21 and with higher scores
indicating worse sleep quality (> 6 = poor sleepers). Reliability n=18
One participant
of the instrument is .83, with strong correlates between sleep ANALYSIS: n=39
withdrawn during data
n=21
One participant
logs and PSQI (r = .81).14,15 Total number who analysis for having Type added from RLS
1 DM*
The Epworth Sleepiness Scale (ESS) is the gold standard for completed study
One participant moved
group**

subjectively measuring daytime sleepiness in adults. Partici- to non-RLS group**

pants are asked to rate the chances of falling asleep in 8 situa-

tions encountered in daily life on a scale of 0 to 3, with scores
ranging from 0 to 24 (> 10 indicates excessive daytime sleepi-
ness). Reliability of the instrument is .88, with validity corre-
lated with the mean sleep latency scale at -.30 (p < .0001).16,17
The Fatigue Severity Scale (FSS) is designed to rate state-
ments distinguishing fatigue from depression. With 9 items
rated on a scale of 1 to 7, possible scores range from 7 to 63,
with 63 representing the most fatigue. Reliability is reported at
.88, with strong convergent validity (≥ 0.6-0.7).18
The Center for Epidemiologic Studies Depression Scale
(CES-D) is a self-report measure of depressive symptomatolo-
gy developed for nonpsychiatric populations aged 18 and older.
The CES-D provides an index of cognitive, affective, and be-
havioral features of depression and the frequency of symptoms.
Participants rate each item on a scale of 0 to 3, with a pos-
sible score of 0 to 60 (higher scores being the most depressed).
A score of 16 or above indicates depressive symptomatology.
Every fourth item is reversed scoring. Reliability is reported at
.90, with a sensitivity of 77.8% and specificity of 84.7%.19, 20
Sample Size/Power Analysis
Sample-size calculations were based on the differences in
PSQI scores in primary and secondary RLS participants, since
sleep outcomes are a major complication of RLS and diabetes.
This study was powered to find a difference of 4.0 in the aver-
age PSQI score between groups, assuming 80% power, α of
0.5, and standard error of 4.5 (from previous studies) under a
Student t-test. A total of 40 participants was required for enroll-
ment with a 10% drop-out rate anticipated requiring a minimum
of 18 participants in each group.
Statistical Methods
The prevalence of RLS in the diabetic population was esti-
mated based upon the proportion of screened diabetic partici-
Journal of Clinical Sleep Medicine, Vol. 4, No. 1, 2008
Figure 1—CONSORT Flowchart. DM refers to diabetes mellitus;
RLS, restless legs syndrome.
pants found to have RLS. Descriptive statistics were used to
explore the demographic characteristics and outcome variables
in the case and control groups. Nonnormally distributed contin-
uous variables were transformed to normality, where possible.
Comparisons between groups were conducted using χ2, Fisher
exact, or t-tests. Additionally, Pearson correlations were used
to determine the strength of the relationships between HbA1c
and sleep outcomes. All analyses were conducted in SPSS 14.0
(SPSS, Inc., Chicago, Ill).
Results
One hundred twenty-one participants were screened for the
study from June 2004 to December 2005. The participants’
flowchart using the CONSORT statement can be seen in Figure
1. The estimate prevalence of RLS from this sample was 44.6%
(95% confidence interval [CI] = 38.3 - 50.4).
Of the 39 participants included in the analysis sample, the
average participant was female (56.4%), married or partnered
(59.0%), and Caucasian (46.2%) or African American (38.5%).
Participants with and without RLS were similar across all de-
mographic characteristics except ethnicity (Table 1). Cauca-
sians were 5 times more likely to have RLS in this sample (odds
ratio [OR] = 5.0, 95% CI = 1.3-19.6, p = 0.026). Comorbid
conditions reported in participants with and without RLS can
be seen in Figure 2. Participants reported that symptoms of RLS
developed a mean of 2 years after the diagnosis of type 2 dia-
betes was confirmed. Of the 18 participants with RLS, only 5
were being treated for RLS with gabapentin (4), pramipexole
(4), clonazepam (2), and temazepam (1).
Participants with type 2 diabetes with RLS were found to
have significantly worse sleep outcomes. Participants with RLS
had approximately twice the global PSQI score as those without
52
 RLS in type 2 DM
Table 1—Comparison of Demographic Characteristics Between Diabetic Participants with and without Restless Legs Syndrome

Characteristica Cases (n = 18) Controls (n = 21) Effect Sizeb 95% CI

Female 10 (55.6) 12 (57.1) OR = 0.94 0.26 – 3.34
Age, y 59.5 (± 11.6) 62.1 (± 10.8) D = -2.55 -9.96 – 4.86
Married or partnered 11 (61.1) 12 (60.0) OR = 1.05 0.28 – 3.86
Caucasian 12 (66.7) 6 (28.6) OR = 5.00c 1.28 – 19.61
Years in school 13.6 (± 2.5) 15.2 (± 4.7) D = -1.58 -4.00 – 0.84
Employment status
Full-time 2 (11.1) 6 (28.6) (ref)
Retired 7 (38.9) 6 (28.6) OR = 3.50 0.51 – 24.27
Other 9 (50.0) 9 (52.8) OR = 3.00 0.47 – 19.04
Socioeconomic status ($1,000)
0-20 5 (31.3) 8 (42.1) (ref)
20-40 5 (31.3) 4 (21.1) OR = 2.00 0.36 – 11.23
40-60 5 (31.3) 3 (15.8) OR = 2.67 0.43 – 16.39
60+ 1 (6.3) 4 (21.1) OR = 0.40 0.03 – 4.68
Cups of coffee/day
0 7 (38.9) 7 (33.3) (ref)
1 4 (22.2) 7 (33.3) OR = 0.57 0.11 – 2.87
2+ 7 (38.9) 7 (33.3) OR = 1.00 0.23 – 4.40
1+ Other caffeinated drinks/day 13 (72.2) 12 (57.1) OR = 1.95 0.51 – 7.49
Smokes 2 (11.1) 2 (9.5) OR = 1.19 0.15 – 9.41
Exercise / week, h 3.9 (± 3.3) 3.7 (± 2.8) D = 0.18 -1.82 – 2.18
body mass index, kg/m2 32.4 (± 7.1) 33.8 (± 8.8) D = -1.42 -6.66 – 3.82

Data are reported as frequencies (%) or mean (± SD).

a
Responses within a characteristic may not sum to 100% due to missing values.
b
Odds ratio (OR) for cases (with restless legs syndrome) versus controls (without restless legs syndrome), difference (D) for cases-controls.
CI refers to confidence interval.
c
p < 0.05

RLS (12.8 vs 6.7, p = 0.002). Further, in the individual PSQI
component scores, participants with type 2 diabetes with RLS
had significantly worse subjective sleep quality (p = 0.001),
sleep latency (p = 0.040), habitual sleep efficiency (p = 0.035),
and daytime dysfunction (p = 0.005) and took more sleep
medications (p < 0.001). For secondary outcomes, the only
significant difference between patients with and without RLS
was seen in the FSS scores. Participants with type 2 diabetes
with RLS had significantly higher FSS scores than participants
without RLS (3.8 vs. 2.6, p = 0.028). Comparison of sleep out-
comes between diabetic participants with and without RLS can
be found in Table 2.
The relationship between the sleep outcomes and HbA1c
were examined in the entire sample (Table 3). Significant
positive correlations were found between HbA1c and ESS
scores (r = 0.36, p = 0.037). Global PSQI scores had a strong
control. More importantly, the RLS participants reported that
their sleep quality was worse, that the time it took them to get to
sleep was longer, and how much time they did sleep was less than
that of the participants without RLS. The participants with RLS
took more medications to help them sleep, despite the fact that
they had worse sleep outcomes. These participants also reported
that the lack of sleep affected their daytime functioning and that
they had more fatigue than those without RLS, which could af-
fect their personal self care, including diet, exercise, and social
activities. It is also significant to note that less than one third of
the participants with RLS were being treated for RLS.
Our findings are in agreement with other studies that have
used objective sleep measures in patients with type 2 diabe-
16

positive correlation with both FSS (r = 0.58, p = 0.002) and 14

CES-D (r = 0.74, p < 0.001) scores. These significant correla- 12
tions persisted when RLS status was adjusted for via partial 10
correlations (not shown). Without RLS
8
With RLS
6
Discussion
4

To promote better management of type 2 diabetes, sleep dis- 2

orders must be addressed. Few studies have examined the sleep 0
Iron Def Neuropathy RA Thyroid Kidney HTN IBS
patters of persons with RLS and type 2 diabetes. This is the first
Figure 2—Report of comorbid conditions in participants with
study to look at outcomes of sleep in the US. This pilot shows type 2 diabetes with and without restless legs syndrome (RLS).
that participants with type 2 diabetes with RLS have significantly Iron def refers to iron deficiency; RA, rheumatoid arthritis; HTN,
worse sleep outcomes, reporting twice the global PSQI score as hypertension; IBS, irritable bowel syndrome.
those without RLS, with no significant difference on glycemic
Journal of Clinical Sleep Medicine, Vol. 4, No. 1, 2008 53
NG Cuellar, SJ Ratcliffe
Table 2—Comparison of Sleep Outcomes Between Diabetic Participants with and without Restless Legs Syndrome

Outcome Variable1 Cases (n = 18) Controls (n = 21) Effect Size2 95% CI or p value
Global PSQI score 12.9 (± 3.3) 6.7 (± 5.8) D = 6.11d 2.49 – 9.73
Sleep category
Good 0 (0.0) 8 (44.4) (ref)
Poor 17 (100.0) 10 (55.6) OR = 28.3d 1.48 – 542.99
Component scores
1. Subjective sleep quality 2.2 (± 0.8) 1.0 (± 1.0) Z = 3.22d 0.001
2. Sleep latency 1.9 (± 1.1) 1.0 (± 1.2) Z = 2.14c 0.040
3. Sleep duration 2.1 (± 0.9) 1.5 (± 1.4) Z = 1.07 0.308
4. Habitual sleep efficiency 1.9 (± 1.1) 1.1 (± 1.3) Z = 2.22c 0.035
5. Sleep disturbance 1.7 (± 0.6) 1.3 (± 0.8) Z = 1.33 0.229
6. Sleep medications 1.2 (± 1.6) 0.2 (± 0.7) Z = 3.91e < 0.001
7. Daytime dysfunction 1.6 (± 0.8) 0.7 (± 0.9) Z = 2.85d 0.005
ESS 10.3 (± 5.1) 8.4 (± 6.0) D = 1.94 -1.81 – 5.70
FSS 3.8 (± 1.2) 2.6 (± 1.6) D = 1.19c 0.20 – 2.18
CES-D 18.4 (± 9.1) 12.1 (± 11.6) D = 6.25 -0.47 – 12.96
HbA1c 7.1 (± 1.5) 7.4 (±1.5) D = -0.30 -1.32 – 0.72

Data are reported as frequencies (%) or mean (± SD). PSQI refers to Pittsburgh Sleep Quality Index; ESS, Epworth Sleepiness Scale; FSS,
Fatigue Severity Scale; CES-D, Center for Epidemiologic Studies-Depression Scale; CI, confidence interval.
a
Responses within a variable may not sum to 100% due to missing values.
b
Cases and controls compared via a t-test, Fisher exact test, and Mann-Whitney U tests (component scores). Odds ratio (OR) for cases (with
restless legs syndrome) versus controls (without restless legs syndrome), difference (D) for cases-controls, or z-values (Z) from nonparametric
test.
c
p < 0.05; dp < 0.01; ep < 0.001

tes and RLS who reported poor sleep outcomes related to in-
creased sleep latency. This interruption in sleep was based on
the symptom severity of RLS, which was not measured in this
study because all the participants did not have RLS. Nightly re-
duction of sleep of 2 to 4 hours affected sleep outcomes, includ-
ing daytime functioning and fatigue. These findings were also
consistent in our study, with significant differences reported in
fatigue between the 2 groups. The demographics of the study
are consistent with previous findings.21, 22
Overall, all (n = 39) the participants with type 2 diabetes
reported some sleep problems. Sleepiness and HbA1c were
positively correlated. As well, overall sleep quality was posi-
tively correlated with fatigue and depression. Few studies
have examined the effect of sleepiness on self-care of patients
with diabetes. It is uncertain if sleepiness is a result of the
symptoms of type 2 diabetes (like fatigue or nocturia), comor-
bid conditions that may be associated with type 2 diabetes,
polypharmacy, or obesity, which may contribute to inactivity,
to name a few possibilities. In studies of patients with type
2 diabetes, excessive daytime sleepiness has been correlated
with nocturia,23 with habitual snoring,24 and in RLS.10 Sleepi-
ness has also been correlated with fatigue and depression, of-
ten seen in type 2 diabetes.25 Although sleepiness or fatigue is
a symptom of diabetes, the impact of sleepiness or fatigue on
the management of type 2 diabetes has not been thoroughly
examined.
Several hypotheses may explain the findings for our study.
First, the misdiagnosis or lack of treatment for persons with
RLS is documented. Therefore, the worsening sleep loss may be
related to the management of RLS in persons with type 2 diabe-
tes who have not been evaluated for this sleep disorder. RLS is
treatable, and the symptoms that disrupt sleep can be managed
with carbidopa-levodopa, dopamine agonists, opioids, benzo-
diazepines, and anticonvulsants. Although promising break-
throughs are being made in the treatment of RLS, it remains
difficult to treat, as pharmacologic agents often stop working or
cause augmentation or rebound. As well, many other pharma-
cologic agents exacerbate symptoms of RLS. Since participants
in this study have many comorbid illnesses, the medications
used to treat the chronic illness may in fact contribute to the
development of RLS. For clinical practitioners, patients with
RLS should be reevaluated at regular intervals to determine if
their RLS symptoms remain under control or if the symptoms
are worsening. The RLS Severity Scale is 1 instrument that can
be used to measure symptom severity in these patients,26 with
the RLS Quality of Life Scale being a valid instrument to exam-
ine the effects of RLS in daily life.27 Secondly, sleep disruption
reported in the participants with RLS may also impact other
metabolic disturbances found in this group of participants that,
in turn, impact the negative symptoms of diabetes and RLS.
Because a high proportion of these participants had comorbid
conditions, all factors should be examined to determine their
contribution to sleep disturbance. Thirdly, it is unknown if the
negative health outcomes found in this study are the result of
Table 3—Correlations Between Sleep Outcomes and HbA1c in the
Sample of 39 Subjects with Diabetes
HbA1c PSQI ESS FSS CES-D
HbA1c 0.09 0.36a 0.06 0.22
PSQI Global Score 0.09 0.34 0.58b 0.74c
ESS 0.36a 0.34 0.36a 0.26
FSS 0.06 0.58b 0.36a 0.39a
CES-D 0.22 0.74c 0.26 0.39a
PSQI refers to Pittsburgh Sleep Quality Index; ESS, Epworth
Sleepiness Scale; FSS, Fatigue Severity Scale; CES-D, Center for
Epidemiologic Studies-Depression Scale.
a
p < 0.05; bp < 0.01; cp < 0.001

Journal of Clinical Sleep Medicine, Vol. 4, No. 1, 2008 54


RLS or other factors that are associated with type 2 diabetes,
such as depression, hypertension, or obesity. Because this is the
first study examining RLS in type 2 diabetes, more research
should be done on all of the variables that may impact health
outcomes in diabetes and RLS.
Limitations
The limitations of this study include the lack of objective
measures, including polysomnography, to evaluate sleep qual-
ity; the diagnostic criteria for RLS used for the study; and the
ability to generalize to the population due to the small sample
size, lack of clinical data, and the dependence on self-report for
other sleep diagnoses, including OSA, which is thought to be
common in type 2 diabetes.
The PSQI is commonly used to measure sleep quality in a
variety of sleep disorders. The use of polysomnography or ac-
tigraphy could be used to determine objective sleep measures.
The diagnostic criteria used to assess for symptoms of RLS have
been used in the majority of RLS studies, developed by the In-
ternational RLS Study Group.13 The use of the 4 criteria has been
reported to have a false-positive rate. A new criterion has been
developed by researchers at John Hopkins University, the Tele-
phone Diagnostic Interview, for diagnosing RLS that should be
used in studies that are screening for RLS.28 A large sample size
is needed to generalize these findings to the population.
Although this study is a pilot, it introduces healthcare pro-
viders to a possible cause of sleep disturbance that can impact
diabetic healthcare outcomes. Future studies should examine
the impact of RLS on sleep and health outcomes in the devel-
opment and management of type 2 diabetes. Poor sleep qual-
ity in patients with type 2 diabetes in association with periodic
limb movements as an independent cause for poor sleep quality
should be examined.
Summary
This is one of the first studies examining the impact of RLS
on outcomes in diabetes. RLS research has focused on primary
RLS, which does not include participants who have diabetes.
RLS has been found to be associated with a variety of medical
and psychiatric conditions and can be diagnosed in conjunction
with other sleep disorders. Patients with type 2 diabetes who
develop RLS, and their healthcare providers, should understand
that a variety of sleep disorders may be seen in diabetes and
that treatment options are available that can improve sleep and
thereby improve the management and health outcomes of dia-
betes and RLS. Healthcare providers must be able to identify
and manage RLS, which is prevalent in type 2 diabetes and has
been shown to be associated with increasing morbidity in this
population.
Acknowledgments
We gratefully acknowledge support for this study from the
American Association of Diabetic Educators and Sigma Theta
Tau International, Honorary Nursing Society. We want to ac-
knowledge support from Mark Shutta, MD, and Charles R.
Cantor, MD, for work as medical advisors on this project. Dr.
Journal of Clinical Sleep Medicine, Vol. 4, No. 1, 2008
RLS in type 2 DM
Schutta is the director of the PENN Rodebaugh Diabetes Cen-
ter and Clinical Assistant Professor of Medicine, Division of
Endocrinology, Diabetes & Metabolism, at the University of
Pennsylvania. Dr. Cantor is from the University of Pennsylva-
nia Health System, Department of Neurology, Philadelphia.
References
1. National Diabetes Statistics. National Institute of Health, 2005.
Available at: http://diabetes.niddk.nih.gov/dm/pubs/statistics/in-
dex.htm#12. Accessed: May 7, 2007.
2. Phillips B, Hening W, Britz P, Mannino D. Prevalence and cor-
relates of restless legs syndrome: results from the 2005 National
Sleep Foundation Poll. Chest 2006;129:76-80.
3. Garcia-Borreguero D, Egatz R, Winkelmann J, Berger K. Epide-
miology of restless legs syndrome: the current status. Sleep Med
Rev 2006;10:153-67.
4. Trenkwalder C, Paulus W. Why do restless legs occur at rest?—
pathophysiology of neuronal structures in RLS. Neurophysiology
of RLS (part 2). Clin Neurophysiol 2004;115:1975-88.
5. Ohayon MM, Roth T. Prevalence of restless legs syndrome and
periodic limb movement disorder in the general population. J
Psychosom Res 2002;53:547-54.
6. Allen RP, Earley CJ. Restless legs syndrome: a review of
clinical and pathophysiologic features. J Clin Neurophysiol
2001;18:128-47.
7. Rothdach AJ, Trenkwalder C, Haberstock J, Keil U, Berger K.
Prevalence and risk factors of RLS in an elderly population: the
MEMO Study. Neurology 2000;54:1064-8.
8. Phillips B, Young T, Finn L, Asher K, Hening WA, Purvis C. Epi-
demiology of restless legs symptoms in adults. Arch Intern Med
2000;160:2137-41.
9. Hening W, Walters AS, Allen RP, Montplaisir J, Myers A, Ferini-
Strambi L. Impact, diagnosis and treatment of restless legs syn-
drome (RLS) in a primary care population: the REST (RLS epi-
demiology, symptoms, and treatment) primary care study. Sleep
Med 2004;5:237-46.
10. Lopes LA, Lins CdMM, Adeodato VG, et al. Restless legs syn-
drome and quality of sleep in type 2 diabetes. Diabetes Care
2005;28:2633-6.
11. Merlino G, Fratticci L, Valente M, et al. Association of restless
legs syndrome in type 2 diabetes: a case-control study. Sleep
2007;30:866-71.
12. Gemignani F, Brindani F, Vitetta F, Marbini A, Calzetti S. Restless
legs syndrome in diabetic neuropathy: a frequent manifestation of
small fiber neuropathy. J Peripher Nerv Syst 2007;12:50-3.
13. Walters AS. Toward a better definition of the restless legs syn-
drome. The International Restless Legs Syndrome Study Group.
Mov Disord 1995;10:634-42.
14. Buysse DJ, Reynolds CF, 3rd, Monk TH, Berman SR, Kupfer DJ.
The Pittsburgh Sleep Quality Index: a new instrument for psychi-
atric practice and research. Psychiatry Res 1989;28:193-213.
15. Backhaus J, Junghanns K, Broocks A, Riemann D, Hohagen F.
Test-retest reliability and validity of the Pittsburgh Sleep Quality
Index in primary insomnia. J Psychosom Res 2002;53:737-40.
16. Johns MW. Reliability and factor analysis of the Epworth Sleepi-
ness Scale. Sleep 1992;15:376-81.
17. Johns MW. A new method for measuring daytime sleepiness: the
Epworth sleepiness scale. Sleep 1991;14:540-5.
18. Krupp LB, LaRocca NG, Muir-Nash J, Steinberg AD. The Fatigue
Severity Scale: application to patients with multiple sclerosis and
systemic lupus erythematosus. Arch Neurol 1989;46:1121-3.
19. Radloff L. The CES-D scale: A new self-report depression scale
for research in the general population. Appl Psychol Measure
1977;1:385-401.
55
NG Cuellar, SJ Ratcliffe

20. Orme JG, Reis J, Herz EJ. Factorial and discriminant validity
of the Center for Epidemiological Studies Depression (CES-D)
scale. J Clin Psychol 1986;42:28-33.
21. Mizuno S, Mihara T, Miyaoka T, Inagaki T, Horiguchi J. CSF
iron, ferritin and transferrin levels in restless legs syndrome. J
Sleep Res 2005;14:43-7.
22. Garcia-Borreguero D, Larrosa O, de la Llave Y, Granizo JJ, Allen
R. Correlation between rating scales and sleep laboratory mea-
surements in restless legs syndrome. Sleep Med 2004;5:561-5.
23. Chasens ER, Umlauf MG, Pillion DJ, Singh KP. Sleep apnea
symptoms, nocturia, and diabetes in African-American community
dwelling older adults. J Natl Black Nurses Assoc 2000;11:25-33.
24. Marchesini G, Pontiroli A, Salvioli G, et al. Snoring, hyperten-
sion and Type 2 diabetes in obesity. Protection by physical activ-
ity. J Endocrinol Invest 2004;27:150-7.
25. Bixler EO, Vgontzas AN, Lin HM, Calhoun SL, Vela-Bueno A,
Kales A. Excessive daytime sleepiness in a general population
sample: the role of sleep apnea, age, obesity, diabetes, and depres-
sion. J Clin Endocrinol Metabol 2005;90:4510-5.
26. Allen RP, Kushida CA, Atkinson MJ, Consortium RLSQ. Fac-
tor analysis of the International Restless Legs Syndrome Study
Group’s scale for restless legs severity. Sleep Med 2003;4:133-5.
27. Atkinson MJ, Allen RP, DuChane J, et al. Validation of the Rest-
less Legs Syndrome Quality of Life Instrument (RLS-QLI): find-
ings of a consortium of national experts and the RLS Foundation.
Qual Life Res 2004;13:679-93.
28. Hening WA, Allen RP, Thanner S, et al. The Johns Hopkins tele-
phone diagnostic interview for the restless legs syndrome: pre-
liminary investigation for validation in a multi-center patient and
control population. Sleep Med 2003;4:137-41.

Journal of Clinical Sleep Medicine, Vol. 4, No. 1, 2008 56