Association of insomnia severity and comorbid medical and psychiatric disorders
in a health plan-based sample: Insomnia severity and comorbidities
Khaled Sarsoura,*, Charles M. Morinb, Kathleen Foleyc, Anupama Kalsekara, James K. Walshd
aGlobal Health Outcomes, Eli Lilly and Company, Lilly Research Laboratories, Indianapolis, IN 46285, USA
bDepartment of Psychology, Laval University, Que., Canada
cThomson Reuters, Ann Arbor, MI, USA
dSleep Medicine and Research Center, Chesterfield, MO, USA
a r t i c l ei n f o
Received 1 December 2008
Received in revised form 26 February 2009
Accepted 27 February 2009
Available online 1 May 2009
Chronic medical comorbidities
Insomnia severity index
Administrative health claims
a b s t r a c t
Background: Insomnia is commonly associated with one or more comorbid illnesses. Data on the rela-
tionship between insomnia severity and comorbid disorders are still limited, especially with regard to
the use of well-validated measures of insomnia severity.
Methods: A total of 2086 health plan enrollees, over-sampling for those with insomnia based on health
claims, completed a telephone survey between April and June of 2006. Participants were categorized
using four insomnia severity categories and compared on their administrative health claims’ psychiatric
and medical comorbidities.
Results: Controlling for age and gender, the odds ratio for having at least one psychiatric diagnosis was
5.04 (CI = 3.24–7.84) for severe insomnia, 2.63 (CI = 1.97–3.51) for moderate insomnia, and 1.7
(CI = 1.30–2.23) for subthreshold insomnia compared with those with no insomnia. Similarly, the odds
ratio for having treatment for at least one chronic disease was 2.83 (CI = 1.84–4.35) for severe insomnia,
2.34 (CI = 1.83–2.99) for moderate insomnia, and 1.55 (CI = 1.25–1.92) for subthreshold insomnia com-
pared with the no insomnia group.
Conclusions: Increasing insomnia severity is associated with increased chronic medical and psychiatric
illnesses. Further research is needed to better understand associations between insomnia severity and
individual psychiatric and chronic medical comorbidities.
? 2009 Elsevier B.V. All rights reserved.
Although insomnia may occur as a primary disorder, more com-
monly it is associated with one or more comorbid illnesses [1–3].
Insomnia has been found to be comorbid with neurological condi-
tions such as Alzheimer’s and Parkinson’s disease and medical con-
ditions such as chronic pain, cardiovascular disease, respiratory
symptoms, gastrointestinal diseases, heart disease, and hyperthy-
roidism [3–5]. Taylor et al.  reported that individuals with
insomnia are more likely to have multiple comorbid medical con-
ditions than those without insomnia. Additionally, the rate of
insomnia appears to be higher in people with chronic medical con-
ditions (for example, chronic pain, breathing problems, and heart
disease) than those without the same disorders .
In addition to comorbid medical conditions, insomnia has also
been associated with depression and other psychiatric disorders
including generalized anxiety disorders [6–12]. Though historically
thought to be a symptom of psychiatric disorders, data suggest that
insomnia sometimes precedes these disorders [1,5,12]. An NIH pa-
nel recently provided a consensus opinion that when insomnia oc-
curs concurrent with other conditions, it should be considered
comorbid rather than secondary . Secondary insomnia implies
that the insomnia is caused by the concurrent condition and should
respond to treatment of that condition. Comorbid insomnia, on the
other hand, recognizes that a cause-effect relationship has not
been established between insomnia and most other conditions.
The association between insomnia severity and comorbid med-
ical and psychiatric conditions has been examined in a number of
studies. Katz and McHorney  assessed the association between
insomnia severity and physician- and subject-reported chronic
medical conditions and depression. The presence or absence of
mild and severe insomnia was ascertained via self-administered
mail survey. Mild insomnia was defined by responses indicating
difficulty in initiating or maintaining sleep during the preceding
4 weeks ‘‘some” or ‘‘a good bit” of the time; severe insomnia was
defined by difficulty initiating or maintaining sleep ‘‘most” or ‘‘all”
of the time. The study found that mild and severe insomnia are
associated with greater odds of having self-reported or physi-
cian-diagnosed chronic medical and psychiatric conditions. More
1389-9457/$ - see front matter ? 2009 Elsevier B.V. All rights reserved.
* Corresponding author. Tel.: +1 317 2762641; fax: +1 317 433 2997.
E-mail address: email@example.com (K. Sarsour).
Sleep Medicine 11 (2010) 69–74
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/sleep
recently, in a review of insomnia prevalence and impact, Becker 
concludes that increased insomnia severity is associated with high-
er numbers of medical comorbidities. Ancoli-Israel , in a liter-
ature review on chronic insomnia, also found that insomnia, as a
feature of chronic disease, tends to be more severe and persistent
than insomnia that does not occur in the context of chronic illness.
Despite the above studies, data on the relationship between insom-
nia severity and comorbid disorders are still limited, especially
with regard to the use of well-validated measures of insomnia
This study combined subjects’ survey information with admin-
istrative health care claims data to examine the relationship be-
tween insomnia severity and treated chronic comorbid medical
conditions and diagnosed psychiatric illnesses. The Insomnia
Severity Index (ISI) was used to define categories of insomnia
2.1. Study design
The target population for this study consisted of members of a
health plan in the Midwest affiliated with Thomson Healthcare.
The ISI and treatment information were captured during inter-
viewer administered telephone surveys conducted between April
and June of 2006. Data on insomnia symptoms and daytime
impairment were also collected but are not included in this report.
Thomson Healthcare used subject health plan identification num-
bers to link telephone survey data to health plan claims data. For
each subject, the index date was calculated as the date of the sur-
vey minus one year (365 days). Claim-based comorbidities (a pri-
mary outcome of this study) were ascertained based on claims
data for 12 months prior to the survey date. The study protocol
was approved by the health plan’s medical director and submitted
to the health plan’s IRB by the medical director. The health plan’s
IRB is independent of Thomson Healthcare. A full review of the pro-
tocol, patient identification, informed consent and data linkage
processes was conducted and approved by the IRB. All subsequent
data analyses were conducted by Thomson Healthcare.
2.2. Study sample
Eligible subjects were between the ages of 18 and 80 years and
had been continuously enrolled in the participating health plan for
at least 24 months. Subjects were excluded from participation if
they had removed themselves from a contact list or had been diag-
nosed or treated for sleep disorders other than insomnia as indi-
cated by diagnostic code (ICD-9-CM diagnosis codes = 307.40
[non-organic sleep disorder unspecified], 307.47 [Other dysfunc-
tion of sleep stages or arousal from sleep], 780.50 [sleep distur-
bance unspecified], 780.51, 780.53, 780.57 [sleep apnea], 780.59
[other sleep disturbance], 89.17-89.18 [sleep studies]; Procedure
codes 93.90 and 94660 [sleep apnea], 95806–95807 [sleep
In order to ensure adequate variability of insomnia severity cat-
egories, Thomson Healthcare generated two groups of subjects
from the pool of eligible subjects (n = 150,000). Subjects in the
insomnia related claims group (n = 5000) were required to have
at least one inpatient or outpatient claim with a diagnosis of
insomnia (ICD-9-CM diagnosis codes = 307.41 [transient disorder
of initiating or maintaining sleep], 307.42 [persistent disorder of
initiating or maintaining sleep], 307.45 [phase shift disruption of
24-h sleep cycle], 307.49 [subjective insomnia complaint], 780.52
[other insomnia NOS], 780.55 [disruption of 24-h sleep cycle]) or
at least one pharmaceutical claim for a benzodiazepine, non-ben-
zodiazepine or low dose sedating antidepressant (commonly used
prescription medications for treating insomnia) between January 1,
2004 and December 31, 2005 (Table 1). Only two (2) individuals
were identified as having insomnia based on the use of a low dose
(n = 6000) were randomly selected and had no insomnia related
diagnosis or insomnia prescription claims.
Subjects were notified of their eligibility to participate in the
study via a written letter from their health plan and were given
the option of returning a postcard requesting their name to be
withdrawn from the study call list. Of the insomnia related claim
group a total of 3234 subjects were contacted. Of those, 1257 sub-
jects agreed to participate in the study and provided data to the
interviewer (a 39% response rate). Of the no insomnia related claim
group, a total of 3662 subjects were contacted. Of those, 1175 sub-
jects agreed to participate in the study and provided data to the
interviewer (a 32% response rate), bringing the total number of
study participants to 2432 health plan members (Fig. 1). The tele-
phone survey was conducted by Market Strategies, Inc. between
April and June of 2006. Remuneration was provided to subjects
who completed the telephone survey. Due to IRB regulations, no
demographic information was collected from subjects who refused
to participate in the study, and hence we were not able to test if
those who refused to participate or opted out of the study were
systematically different from the study population.
Since the study used a combination of healthcare claims and
self-reported survey data, claims-based and self-reported mea-
sures were defined for several key constructs of interest including
presence of comorbidity and insomnia severity levels.
2.3. Claim-based measures
2.3.1. Demographic characteristics
Health plan claims data were used to determine the demo-
graphic characteristics of subjects including age, sex, and geo-
graphic location (urban vs. rural).
2.3.2. Chronic medical conditions
The Chronic Disease Score (CDS) was used to assess each sub-
ject’s general level of medical conditions. The CDS is constructed
from claim-based numbers of therapeutic drug classes an individ-
ual is taking in 17 disease/condition categories in the year prior to
the survey. Depending on the number of therapeutic classes taken
in each category, a person was assigned a score from 1 up to 5. The
maximum score across all 17 categories was 35 [22,23]. The CDS
has high year to year stability (a correlation of 0.74 over time)
and has been shown to be related to physician ratings of physical
Prescription medications and doses considered insomnia related.
Drug classGeneric name Brand name Dose (mg)
TemazepamRestoril All doses
Halcion All doses
<50 mg DoxepinAdapin,
K. Sarsour et al./Sleep Medicine 11 (2010) 69–74
disease severity and to predict hospitalization and mortality [22–
25]. For analytic purposes, the CDS was converted into a binary
variable where a CDS score of 0 was coded as no treated chronic
medical conditions, while a score greater than 0 was coded as hav-
ing at least one treated chronic medical condition.
2.3.3. Psychiatric comorbidities
The presence or absence of psychiatric comorbidities was ascer-
tained based on the Psychiatric Index . The Psychiatric Index
consists of 12 psychiatric diagnostic groupings (PDGs) which coin-
cide with major diagnostic groups in the diagnosis-related groups
(DRG) system . Claims associated with at least one of the 12
PDGs were used to identify subjects with psychiatric diagnoses
in the 12 months prior to the survey date. For analytic purposes,
a binary variable was constructed where a score of 0 was consid-
ered no psychiatric diagnosis, while a score greater than 0 was con-
sidered to be having at least one psychiatric diagnosis.
2.4. Survey based measures
2.4.1. Insomnia severity Index
The Insomnia Severity Index (ISI) measures the subject’s per-
ceived severity of insomnia symptoms (falling asleep, staying
asleep, and waking early) and the degree of impact of the symp-
toms on daily functioning, quality of life, and the amount of worry
the subject feels from sleep problems. The scale includes a total of
7 items, each scored on a 0–4 scale, that are summed with a range
of 0–28. The score can then be divided into four categories of sever-
ity: no clinically significant insomnia (0–7), subthreshold insomnia
(8–14), clinical insomnia of moderate severity (15–21), and severe
clinical insomnia (22–28). Psychometric evaluation provides evi-
dence that the ISI measurements for difficulty falling asleep, stay-
ing asleep, and waking early correlate with sleep latency, wake
time after sleep onset (WASO), and early morning awakenings
based on sleep diary studies. Furthermore, the ISI has been vali-
dated against both polysomnographic and prospective sleep diary
measures and demonstrates convergence with clinical interview
2.4.2. Self-reported depression
Self-reported depression was ascertained using the Whooley
depression scale , a two question, validated case-finding tool
used to ascertain presence or absence of depression. The two ques-
tions are (1) ‘‘During the past month, have you often been bothered
by feeling down, depressed, or hopeless?” and (2) ‘‘During the past
month, have you often been bothered by little interest or pleasure
in doing things?”
Bivariate comparisons of demographic and comorbidity mea-
sures across the four ISI categories used ANOVA for continuous
variables and chi-square for categorical variables. Multivariate lo-
gistic regression analyses were used to examine the association be-
tween insomnia severity and claim-based presence or absence of
chronic medical and psychiatric treatment or diagnosis. Dependent
variables were claim-based presence or absence of chronic medical
and psychiatric illnesses. The primary independent variable of
interest was insomnia severity categories assessed with the no-
insomnia category as the reference group. Because psychiatric
Health Plan Members
(N ~ 150,000)
Claims Diagnosis of Insomnia
OR Receiving Rx for Insomnia
antidepressants in low dose;
Meet Exclusion Criteria (N =
5000) (complete ascertainment)
N = 3234
N = 1766
N = 1977
N = 1257
Randomly selected sample
that has no insomnia
claims and that meets
exclusion criteria, sample
is representative of health
plan members (N = 6000)
N = 3662
N = 2338
N = 1175
(N = 2432)
N = 1977
Fig. 1. Study sample flow diagram.
K. Sarsour et al./Sleep Medicine 11 (2010) 69–74
and medical conditions are often found to be comorbid, a second
logistic model was run to control for any confounding by concom-
itant comorbidities. Furthermore, a third model was run to control
for any confounding by depression. All regression analyses were
conducted while controlling for age and gender. All analyses were
conducted using SAS v9.1.
3.1. Characteristics of sample
Of the 2432 subjects who participated in the telephone survey,
2086 enrollees completed the survey, were successfully linked to
the claims data, and were included in the analyses. Subjects’ char-
acteristics are presented separately according to the 4 ISI insomnia
severity groups. Approximately 40.3% of respondents did not meet
criteria for having insomnia (ISI score < 8) and were therefore la-
beled as the ‘‘no insomnia” group. Of the remaining subjects
34.0%, 20.5%, and 5.2% met the criteria for subthreshold, moderate,
and severe insomnia, respectively. The mean age did not vary
(p = .5342) by insomnia severity, with the mean age ranging from
52.3 for the no insomnia group to 51.1 for the moderate insomnia
group. The percent of females in each group, however, differed
with just 61.8% of the no insomnia group being female compared
to 74.8% of the moderate group and 72.5% of the severe groups
(p < 0.0001). These data also show that individuals with moderate
and severe insomnia had higher CDS scores (p < 0.0001), such that
more than half of the individuals in those groups had at least one
chronic disease medication compared to less than a third of those
with no insomnia (p < 0.0001). The same pattern held for the psy-
chiatric index (p < 0.0001) and the presence of any psychiatric
diagnosis (p < 0.0001), with subjects with increasing severity of
insomnia being more likely to have comorbid psychiatric condi-
tions. Complete descriptive characteristics of the study population
are presented in Table 2.
3.2. Claim-based insomnia diagnosis and treatment by insomnia
Insomnia related diagnosis claims were far less common than
insomnia related prescription claims. Overall, just 5.2% of respon-
dents had an insomnia diagnosis, while 29.6% had at least one pre-
scription for an insomnia medication. Just 11.9% of the severe
insomnia group had a claim-based diagnosis versus 10.1%, 4.9%,
and 2.1% for the moderate, subthreshold, and no insomnia groups.
On the other hand, 54.1% of the severe insomnia group had a claim-
based prescription for insomnia versus 47.9%, 29.9%, and 16.8% in
the moderate, subthreshold, and no insomnia groups. Of the severe
insomnia group, 40.4% had no insomnia diagnosis or prescription
related claims versus 49.5% in the moderate group, 67.1% in the
subthreshold group, and 81.9% in the no insomnia group. Complete
details are presented in Table 3.
3.3. Claim-based comorbidities
Table 4 shows the results of multivariate analyses to estimate
the association between insomnia severity and study primary end-
points of psychiatric comorbidities and treated chronic medical
conditions. Controlling for age and gender, the odds of having at
least one psychiatric diagnosis was 5.04 times greater in those with
severe insomnia compared with those with no insomnia, and was
significantly elevated in the moderate and subthreshold groups as
well. To account for any confounding due to the coexistence of
chronic medical conditions and psychiatric comorbidities, we con-
trolled for the presence of chronic medical conditions to see how
the association between insomnia severity and psychiatric diagno-
sis will change. Controlling for age, gender and the presence of
chronic medical conditions, the odds of having at least one psychi-
atric diagnosis was 4.43 times greater in those with severe insom-
Descriptive characteristics of sample.
Mean age (SD)
Urban location (vs rural
Mean chronic disease score
Any chronic disease
Mean psychiatric index score
Any psychiatric diagnosis (%)*
*p < 0.0001.
+p < 0.01.
Insomnia diagnosis and prescription by severity group.
Claims based measures*
*p < 0.0001.
Associations between comorbidities’ measures and insomnia severity.
OR95% CI OR 95% CI
Chronic disease diagnosis
Self reported depression
OR, odds ratio; CI, confidence interval.
?P < 0.0001 for all models.
*Model I adjusted for age and gender.
±Model II/psychiatric diagnosis adjusted for age, gender, and chronic disease
diagnosis; Model II/chronic disease diagnosis adjusted for age, gender, and psy-
§Model III adjusted for age and gender.
K. Sarsour et al./Sleep Medicine 11 (2010) 69–74
nia compared with those with no insomnia, and remained signifi-
cantly elevated in the moderate and subthreshold groups as well.
The odds of having treatment for at least one chronic medical
disease were 2.83 times greater in those with severe insomnia
compared with the no insomnia group and were significantly ele-
vated in the moderate insomnia group but not in the subthreshold
group. To similarly account for any confounding by psychiatric
diagnosis, we controlled for presence or absence of psychiatric
diagnosis. When controlling for age, gender and the presence of
psychiatric diagnosis, the odds of having at least one treatment
for chronic disease were 2.36 times greater in those with severe
insomnia compared with those with no insomnia and remained
significantly elevated in the moderate group but not the sub-
3.4. Self-reported depression as an independent variable
To control for the degree to which depression accounts for the
reported association between insomnia severity and presence or
absence of chronic and psychiatric illnesses, we ran a third model
controlling for respondent self-reported depression. When control-
ling for depression, the odds of being treated for at least one
chronic medical disease increased slightly for each level of insom-
nia severity, with the odds being 2.67 times greater in those with
severe insomnia compared with those with no insomnia. The odds
of having a chronic disease treatment were 2.25 times higher in
those with moderate insomnia and 1.51 times higher in those with
subthreshold insomnia relative to those with no insomnia. Those
who reported depression had 1.10 times greater odds of having a
chronic disease treatment. However, this association was not sta-
Because claim-based psychiatric diagnoses necessarily contain
depression (Kendall Tau correlation coefficient between presence
or absence of claim-based psychiatric diagnosis and self-reported
depression was 0.212 [p < 0.0001]), when we controlled for self-re-
ported depression, the association between insomnia severity and
the presence of psychiatric diagnosis was diminished but remained
statistically significant. When controlling for self-reported depres-
sion, the odds ratio of having a psychiatric diagnosis was 3.18 and
1.89 for severe and moderate insomnia, respectively, compared
with those with no insomnia. When controlling for self-reported
depression, severe insomnia remained more strongly associated
with claim-based psychiatric diagnosis than claim-based presence
or absence of chronic disease treatments (odds ratio: 3.18 versus
2.67). However, this trend appeared to be reversed in the moderate
insomnia group (odds ratio: 1.89 versus 2.25).
This study quantified the association between insomnia sever-
ity and claim-based psychiatric and medical comorbidities. Our
findings extend previous research by demonstrating that insomnia
severity measured by a well-validated instrument is independently
associated with chronic medical conditions after controlling for
psychiatric illnesses and is independently associated with chronic
medical conditions after controlling for psychiatric comorbidities.
This is important because, unlike previous studies, which only
studied mild and severe insomnia (or which used only a dichoto-
mous classification – insomnia vs. no insomnia), this study demon-
strated that the observed trend holds in subthreshold, moderate,
and severe subgroups of insomnia. Even though the association be-
tween insomnia severity and presence of psychiatric diagnoses
was stronger than the association with presence of chronic disease
diagnosis, insomnia comorbid with chronic medical conditions re-
mains a significant burden that requires identification and
Psychiatric illnesses were found to be more significantly asso-
ciated with insomnia severity levels than chronic medical ill-
presence of psychiatric comorbidities even after adjusting for
depression ratings. However, the trend of association appears to
be inconsistent once we control for self-reported depression. Be-
cause of the overlapping confidence intervals, this finding may
not be substantiated. Other studies are required to estimate the
relative contributions of depression and other psychiatric ill-
nesses to insomnia severity.
This study found that among a random sample of health plan
members who had no insomnia related claims in the 12 months
preceding the survey administration, 46% fell within the ISI defini-
tion of subthreshold insomnia or greater (including 12% who fall
within the moderate insomnia category). Additionally, this study
found that 22.9% and 37.9% of the severe and moderate insomnia
groups had no insomnia related health claim, implying that insom-
nia is likely to be under-recognized in a health care setting despite
its severity. Given the associations between insomnia severity and
comorbid medical and psychiatric illnesses, these findings suggest
the need to screen and better manage insomnia in populations that
present with medical and psychiatric comorbidities.
This study found that 5.2% of the study sample had an insomnia
diagnosis claim. This reflects practice patterns where physicians
are less likely to submit a health insurance claim for an insomnia
diagnosis either because of their belief that insomnia will always
be secondary to the underlying primary comorbidity or that insur-
ance companies are less likely to reimburse health care providers
for an insomnia diagnosis. Hence health care providers tend to
underreport insomnia as a primary diagnosis and instead resort
to noting other comorbidities that are likely to be reimbursed. A
greater percentage of respondents were found to have an insomnia
prescription related claim (29.6%). Medications typically pre-
scribed for insomnia (Table 1) also have a whole host of other uses.
Patients appear to use multiple classes of medications in combina-
tion or at various time-points in the study period. Because many
subjects in this study were sampled based on their medication
usage (subsequently they were administered the insomnia severity
index to ascertain their insomnia status), it is conceivable that our
study sample contains a group of subjects who have no insomnia
and who have more comorbidities than is representative of the
health plan population. This will likely bias our findings. However,
this bias, if it exists, will cause our odds ratios to be underesti-
mated. Future studies should consider addressing this limitation
by conducting a large population-based survey.
This study found that 25% of the group with insomnia related
claims fell into the ISI category of no insomnia. This may be ex-
plained by a number of reasons. First, given the design of the study,
respondents who had an insomnia related claim in the 12 months
preceding the survey administration may have had their insomnia
resolved either due to using sleep promoting medications or
through other means. Furthermore, this may be due to coding or
other errors in the claims database. Given the problem of insomnia
misclassification in health claims research, future studies should
consider increased reliance on a mixed approach of linking health
claims to subject surveys as well as longitudinal follow-up studies.
Additionally, methodological studies are needed to examine if
administrative claim case finding strategies make good proxies
for actual in-person assessment.
Several limitations of research using claims databases have
been described including potential coding errors and enrollment
requirements, which apply to the current study. Because of privacy
considerations, claim-based data for non-respondents were not
available for comparison. Thus, we were unable to assess if those
who refused to participate in the survey introduced any selection
bias into our odds ratio estimates. Given the study design, a causal
K. Sarsour et al./Sleep Medicine 11 (2010) 69–74
direction of association between insomnia and comorbidities may Download full-text
not be ascertained.
In conclusion, to our knowledge, this is the first study to ascer-
tain insomnia severity using a validated tool and to link severity
levels with claim-based comorbidities. Clinical populations with
a chronic medical condition and/or psychiatric illnesses are more
likely to have comorbid insomnia. Further research is needed to
better understand associations between insomnia severity and
individual psychiatric and chronic medical comorbidities. More-
over, further research is needed to assess how better management
and treatment of insomnia in such populations may improve sub-
jects’ quality of life and chronic disease trajectory. Conversely, pop-
ulations with chronic insomnia should be assessed for comorbid
medical and psychiatric illnesses.
Disclosure of the presence or absence of financial support
Dr. Walsh reports that research support has been provided to
his institution by Pfizer, Merck & Co., Somaxon, Evotec, Actelion,
Vanda, Neurogen, Sanofi-Aventis, Ventus, and Jazz Pharmaceuti-
cals, and that he has provided consulting services to the following
companies: Pfizer, Sanofi-Aventis, Cephalon, Organon, Neurocrine,
Takeda America, Actelion, Sepracor, Jazz, Respironics, Transcept,
Neurogen, GlaxoSmithKline, Somaxon, Eli Lilly, Evotec, Merck,
Kingsdown, Vanda, and Somnus. Dr. Morin has received financial
support for consulting or serving on advisory boards from Sanofi-
Aventis, Sepracor, Takeda, Pfizer, Actelion, Roche, and Eli Lilly.
Khaled Sarsour and Anupama Kalsekar are employees of Eli Lilly
and Company. Kathleen Foley is an employee of Thomson Reuters
which conducted the study and analysis on behalf of Eli Lilly and
Company. This study was funded by Eli Lilly and Company.
Thank you to Drs. Joseph A. Johnston and David Van Brunt for
their thoughtful review of this manuscript, and many thanks to Ta-
cey Ann Boucher for her assistance preparing this manuscript.
 Becker PM. Insomnia: prevalence, impact, pathogensis, differential diagnosis,
and evaluation. Psychiatr Clin N Am 2006;29:855–70.
 Roth T. Challenges in the comorbid condition. Sleep Med 2007;8(Suppl.
 Taylor DJ, Mallory LJ, Lichstein KL, Durrence HH, Riedel BW, Bush AJ.
Comorbidityof chronic insomnia
 Ringdahl EN, Pereira SL, Delzell Jr JE. Treatment of primary insomnia. JABFP
 Thase ME. Correlates and consequences of chronic insomnia. Gen Hosp
 Ford DE, Kamerow DB. Epidemiologic study of sleep disturbances and
psychiatric disorders: an opportunity for prevention? J Am Med Assoc
 Breslau N, Roth T, Rosenthal T, Andreski P. Sleep disturbance and psychiatric
disorders: a longitudinal epidemiological study of young adults. Biol
 Lustberg L, Reynolds III CF. Depression and insomnia: questions of cause and
effect. Sleep Med Rev 2000;4:253–62.
 Bélanger L, Morin CM, Langlois F, Ladouceur R. Insomnia and generalized
anxiety disorder: effects of cognitive behavior therapy for gad on insomnia
symptoms. Anxiety Disord 2004;18:561–71.
 Roth T, Drake C. Evolution of insomnia: current status and future direction.
Sleep Med 2004;5(Suppl. 1):S23–30.
 Johnson EO, Roth T, Breslau N. The association of insomnia with anxiety
disorders and depression: exploration of the direction of risk. J Psychiatr Res
 Riemann D. Insomnia and comorbid psychiatric disorders. Sleep Med
 NIH State-of-the-Science Conference Statement on Manifestations and
Management of Chronic Insomnia in Adults. NIH Consens Sci Statements
2005; 22: 1–30.
 Katz DA, McHorney CA. Clinical correlates of insomnia in patients with chronic
illness. Arch Intern Med 1998;158:1099–107.
 Ancoli-Israel S. The impact and prevalence of chronic insomnia and other sleep
disturbances associated with chronic illness. Am J Manag Care 2006;12(8
 Bastien CH, Vallières A, Morin CM. Validation of the insomnia severity index as
an outcome measure for insomnia research. Sleep Med 2001;2:297–307.
 Mendelson WB, Roth T, Cassella J, Roehrs T, Walsh JK, Woods JH, et al. The
treatment of chronic insomnia: drug indications, chronic use and abuse
liability. Summary of a 2001 new clinical drug evaluation unit meeting
symposium. Sleep Med Rev 2004;8:7–17.
 Benca R. Diagnosis and treatment of chronic insomnia: a review. Psychiatr Serv
 Rodenbeck A, Cohrs S, Jordan W, et al. The sleep-improving effects of doxepin
are paralleled by a normalized plasma cortisol secretion in primary insomnia.
A placebo-controlled double-blind, randomized, cross-over study followed by
anopen treatment over3
 Hajak G, Rodenbeck A, Adler L, et al. Nocturnal melatonin secretion and sleep
after doxepin administration
 Hajak G, Rodenbeck A, Voderholzer U, et al. Doxepin in the treatment of
primary insomnia: a placebo-controlled, double-blind, polysomnographic
study. J Clin Psychiatry 2001;62(6):453–63.
 Von Korff M, Wagner EH, Saunders K. A chronic disease score from automated
pharmacy data. J Clin Epidemiol 1992;45:197–203.
 Clark DO, Von Korff M, Saunders K, Baluch WM, Simon GE. A chronic disease
score with empirically derived weights. Med Care 1995;33:783–95.
 Johnson RE, Hornbrook MC, Nichols GA. Replicating the chronic disease score
(CDS) from automated pharmacy data. J Clin Epidemiol 1994;47(10):1191–9.
 Putnam KG, Buist DS, Fishman P, Andrade SE, Boles M, Chase GA, et al. Chronic
diseasescore asa predictor
 Ashcraft ML, Fries BE, Nerenz DR, Falcon SP, Srivastava SV, Lee CZ, et al. A
psychiatric patient classification system. An alternative to diagnosis-related
groups. Med Care 1989;27(5):543–57.
 Whooley ML, Avins A, Miranda J, Browner W. Case-finding instruments for
depression, two questions are as good as many. J Gen Intern Med
in chronic primary insomnia.
K. Sarsour et al./Sleep Medicine 11 (2010) 69–74