Heart to Heart: Exploring Heart Rate Variability in Insomnia Patient Subtypes

Abstract

Insomnia is one of the most common complaints in medical practice and the sleep disorder of highest prevalence. At least 10% of the worldwide population has chronic insomnia, which has been associated with a range of negative health outcomes. Within the clinical setting, patient subtypes have been defined according to symptomology. More recently insomnia researchers have proposed phenotypes based on total sleep time during overnight polysomnography (PSG). Short-sleeping insomnia patients are purported to be a biologically severe phenotype at higher risk of cardiovascular morbidity, poor mental health, and obesity (compared to healthy controls). Heart rate variability (HRV) is an objective marker that provides insight into autonomic nervous system dynamics. The overarching aim of my research was to explore a large clinical sample of patients with Insomnia Disorder to determine whether differences in HRV exist during sleep in empirically-derived insomnia patient subtypes. The aim of the work presented within Chapter 2 was to identify all previous insomnia-HRV research to determine if HRV was impaired in adult patients with insomnia, and whether treatments altered HRV. A systematic review of five web databases located 22 relevant articles; 17 case-control studies and 5 interventions studies. Results were difficult to synthesise due to incomparable methodology and reporting. There was a high risk of bias in the majority of studies. It was concluded that although HRV impairment in insomnia may be a widely-accepted concept, it is not supported by research nor has it been determined if it varies after treatment or according to patient subtype. The aim of the first empirical study of the thesis (Chapter 3) was to objectively-derive insomnia patient subtypes and evaluate their physiological signals (HRV and electroencephalography [EEG]) during sleep onset. Patients (n = 96) with clinically-diagnosed Insomnia Disorder underwent overnight PSG to determine sleep metrics for cluster analysis using Ward’s method: Total Sleep Time (TST), Wake After Sleep Onset (WASO) and Sleep Onset Latency (SOL). Electrocardiogram (ECG) from the PSG was extracted in the 10 minutes before and after sleep onset. After R-wave detection, the ECG was visually checked and manually corrected as required. Six time and frequency-domain HRV measures were analyzed; heart rate (HR), standard deviation of all N-N intervals (SDNN), root mean square of successive R-R intervals (RMSSD), percentage of successive R-R intervals that differ by > 50 ms (PNN50), high frequency (HF), and low frequency (LF)/HF ratio. Cluster analysis derived two solutions; one comprising two subtypes and another with three subtypes. The two cluster solution consisted of insomnia with short-sleep duration (I-SSD: n = 43) and insomnia with normal objective sleep duration (I-NSD: n = 53). At sleep onset, between-group HRV analysis revealed reduced parasympathetic activity (PNN50 and RMSSD) in the short-sleeping subtype. This was not mirrored by significant increases in HR and/or the LF/HF ratio. These findings suggested that reduced parasympathetic activity during sleep onset might contribute to poor cardiometabolic health outcomes previously reported in short-sleeping insomnia patients. The final component of this thesis was a case-control study (Chapter 4) which examined whether HRV measures differed between insomnia subtypes across the nocturnal period. It was hypothesized that short-sleeping insomnia patients would have impaired HRV compared to normal-sleep duration insomnia patients, consistent with differences observed at sleep onset (Chapter 3). Insomnia patients underwent overnight PSG, which provided sleep metrics for cluster analysis and ECG for HRV analysis. ECG was visually checked for accurate R-wave detection, and manually corrected as required. HRV analysis was performed from lights-off to lights-on (and separately by sleep/wake stage) using time and frequency-domain measures. Differences in HRV measures (HR, SDNN, RMSSD, LF, HF, LF/HF) were tested between the subtypes using General Linear Models controlling for age as a core confounder. Short-sleeping insomnia patients (I-SSD: n = 34; 45.5 ± 10.5 years) and normal-sleep duration insomnia patients (I-NSD: n = 41; 37.6 ± 10.9 years) were included in the HRV analysis. There were no statistically significant nocturnal HRV differences between subtypes after controlling for age. As such, short-sleeping insomnia patients did not have statistically significant reductions in HRV measures representative of parasympathetic activity.«br /» In summary, there was a lack of persistent nocturnal HRV disparities (between empirically-derived insomnia patient subtypes) that extended beyond sleep onset in this large clinical sample of patients with Insomnia Disorder. The central tenet of 24-hour hyperarousal amongst short-sleep duration insomnia patients cannot be supported by the combined findings of these two empirical studies. Post-hoc calculations revealed larger sample sizes would be required to determine a small to medium effect size difference in nocturnal HRV between insomnia patient subtypes. Until this time, the directional relationship between insomnia, heart rate variability, hyperarousal and cardiovascular disease remains unclear in the heterogeneous insomnia population.