Graduation Year


Document Type




Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Epidemiology and Biostatistics

Major Professor

Skai Schwartz, Ph.D.

Committee Member

Janice Zgibor, R.Ph., Ph.D., CPH, FACE

Committee Member

Lakshminarayan Rajaram, Ph.D.

Committee Member

Aurora Sanchez-Anguiano, MD, Ph.D., CPH

Committee Member

Maria Parr, MD


administrative claims, case-control, epidemiology, healthcare, sleep disorders


Background and Significance: Narcolepsy is a chronic neurological disorder. These patients experience various psychiatric and physical comorbid diseases and mortality at an increased rate compared to the general population. Additionally, patients with narcolepsy experience approximately a doubling of various annual healthcare related facility visits, transactions, and costs comparatively. Narcolepsy with cataplexy is generally believed to be more prevalent than narcolepsy without cataplexy. However, incidence and prevalence estimates of narcolepsy (with or without cataplexy) vary widely with few large epidemiological studies conducted worldwide and none in the U.S evaluating these proportions in both children and adults utilizing a large health care claims database.

One of the main mechanisms underlying narcolepsy, the destruction of hypocretin neurons, is not clear. Two of the more noted hypotheses for this pathology are autoimmune and infection based triggers in allele carrier patients. These have been highlighted since narcolepsy diagnoses increased following the late 2000s influenza vaccinations, especially across Europe. Specific influenza and streptococcal infections have also been considered. Large U.S. healthcare claims database investigations of the association between specific infections and development of narcolepsy were not found in the published scientific literature. Our goals were to enhance the knowledge regarding the epidemiology and possible infection triggers of narcolepsy. The information gained may aid in the overall understanding of the condition, the possible vulnerable populations, and lead to hypotheses regarding which subpopulations research should be focused upon and those triggers that may be avoided or reduced in exposure.

Methods: The Truven Health MarketScan Commercial Dissertation Database (THMCDD) was used to estimate prevalence and incidence of narcolepsy, with and without cataplexy, by age groups, gender, and region among patients under age 66 with continuous enrollment for years 2008-2010. THMCDD contains health claims information for over 18 million people. Prevalence was expressed as cases/100,000 persons. Average annual incidence (using varying criteria for latency between the diagnostic tests, polysomnograph coupled with MSLT, and the diagnosis) was expressed as new cases/100,000 persons/year. Subsequently, we conducted a case-control study to assess the differences in respiratory infections between patients with incident narcolepsy diagnosis and controls. Continuously enrolled patients under age 66 were included. Cases of narcolepsy occurring from July 1, 2009 through December 31, 2010 were included based on two diagnosis criteria (using varying criteria for latency between diagnosis and the diagnostic tests). Non-narcolepsy controls were frequency matched on look-back time by assigning an index date equal to a case diagnosis date. Occurrence of prior respiratory infections was compared between cases and controls based on narcolepsy criteria and four different time periods pre-index date. Infections were grouped into 9 types based on pathogen and clinical manifestation.

Results: From 2008 through 2010, there were 8,444,517 continuously enrolled patients and 6,703 diagnosed with narcolepsy (prevalence overall:79.4/100,000; without cataplexy:65.4/100,000; with cataplexy: 14.0/100,000). Based on the 3 definitions of incidence, overall average annual incidence was 7.67, 7.13, and 4.87/100,000 persons/year. Incidence for narcolepsy without cataplexy was generally several times higher than narcolepsy with cataplexy. Prevalence and incidence were approximately 50% greater for females compared to males across most age groups. Prevalence was highest among the 21-30 age group, with incidence highest among enrollees in their early 20s and late teens. Regionally, the North Central U.S. had the highest prevalence and incidence, while the West was the lowest. For the case-control study, Adjusted odds ratio (aOR) increases were statistically significant for Group 5 (acute respiratory infections), Group 8 (other pneumonias, bronchopneumonia, etc.) and Group 9 (influenzas) across various time periods pre-index date and for both narcolepsy criteria. Overall, the most significant aORs were for acute respiratory infections during the 3 to 15 months pre-index date for both narcolepsy diagnosis criteria (aOR=1.73, 95% 1.52 to 1.98 and aOR=1.83, 95% CI 1.57 to 2.19). The aORs for acute respiratory infections were approximately 50% greater among females than males.

Conclusion: We observed higher prevalence and incidence of narcolepsy compared to most previous studies. Females were associated with approximately 50% increased proportions compared to males. We also found that the greatest prevalence and incidence of narcolepsy occurred in patients in their early 20s, and those residing in the North Central region of the U.S. Perhaps most striking was the observation of much greater proportions of narcolepsy without cataplexy compared to narcolepsy with cataplexy. In the case-control assessment, we found increased occurrences of acute respiratory infections, pneumonias, and influenza prior to incident narcolepsy diagnosis, compared to controls. Generally, these rates appeared higher for females than males and occurred for both narcolepsy diagnosis criteria. Additionally, these associations were observed in the infection assessment periods 3 to 15 months and 6 to 18 months prior to incident narcolepsy diagnosis. Increased awareness and early notification among healthcare providers for signs and symptoms of narcolepsy is critical in helping this population of patients manage this burdensome condition. Also, the identification of potential narcolepsy triggers by certain infections may aid in the understanding of the disease. These findings may have implications in the understanding of mechanisms and causation of other acute onset neurological disorders. Our observations of consistently increased risk of incident narcolepsy related to recent previous viral respiratory infections and the inconsistent results for bacterial infections require additional study to confirm these findings.

Included in

Epidemiology Commons