Harvard Catalyst Profiles

Excessive daytime sleepiness (EDS) affects 10-20% of the population and is associated with increased risks of cardiovascular diseases (CVD) and mortality. Emerging data suggest reduction in EDS as novel intervention targets for improving CVD. However, findings are limited by self-reported data and heterogeneity. There is a need to dissect and understand the underlying drivers of EDS subtypes, and to determine whether there are subtypes causally related to CVD and potentially modifiable. Our recent work identified two primary subtypes of EDS sleep propensity (SP; characterized by objectively measured long sleep duration, high efficiency, and less fragmentation) and sleep fragmentation (SF; short sleep duration and low efficiency). Each of them is common in the population, associated with different genetic backgrounds and adverse cardiovascular outcomes. We hypothesize that SP is a novel sleep phenotype that reflects a property of the need of staying asleep; dissecting EDS into SP and SF subtypes will facilitate identification of genetic, behavioral and physiological mechanisms for EDS, and improve understanding of pleiotropic or causal associations with CVD risk. In order to test these hypotheses, we will leverage macro- and micro- sleep architecture measurements, genomics, and other -omics data in population-based cohorts. We will address the following specific aims: 1) To identify social behavior risk factors, comorbidities and neurophysiological characteristics (assessed by actigraphy and electroencephalography) for SP and SF, and refine classification of EDS subtypes if needed; 2) To identify genetic variants and molecular pathways associated with EDS subtypes and generate robust polygenetic risk score for risk stratification; 3) To systematically evaluate the causal or non-causal associations between EDS subtypes and CVD traits; 4) To identify the modification effect of each EDS subtype on genetic susceptibility of CVD using gene-environment interaction analyses. This work will advance our understanding of the heterogeneity of EDS, reveal biological mechanisms and pathways linking to CVD, and provide information that will guide clinical and public health interventions as well as provide directions for future laboratory research. The trainee will have the opportunity to lead one study of this project to investigate the neurophysiological/genetics background of EDS. He/she will use Python and R programming, often on a Linux-based computer cluster. He/she may also learn how to apply the most modern genetics statistical software to quantify genetic associations, model gene by environmental interactions, and generate polygenic risk scores.

Circadian misalignment and disruption increase CVD risk; however, most evidence comes from lab-controlled experiments. While previous epidemiological studies have used self-reported chronotype (i.e., early bird vs. night owl) as a proxy for circadian preference, this measure does not accurately capture circadian misalignment. Furthermore, there have been no systematic investigations in the general population, and research on variations in effects across age, sex, and race/ethnicity groups remains limited. This project will utilize accelerometry-derived circadian metrics and self-reported chronotype data to develop a novel metric for circadian misalignment, by comparing circadian preference with actual activity patterns. We will investigate the prospective associations of circadian misalignment with CVD risk and mortality in a longitudinal cohort. We will compare the effects across age, sex, race/ethnicity, and chronotype groups.