Short sleep duration and the risk of hypertension: snoozing away high blood pressure?

Abstract

First noted over half a century ago [1], the epidemiologic association between daily short sleep duration (SSD) and prevalent hypertension (HTN) has been demonstrated in multiple cross-sectional studies, using self-reported as well as objectively assessed sleep duration [2–4]. As the natural next step towards proving causality, prospective studies have generally found that SSD significantly increases the risk of developing HTN [5–7]. However, the evidence is by no means unanimous, with significant inconsistencies among both cross-sectional and longitudinal studies. Consolidating extant prospective data, Li et al. [8] present a dose-response meta-analysis of nine prospective studies examining this association, in the current issue of Journal of Human Hypertension, and found that sleep deprivation increases the risk of developing HTN in a dose-response manner, with a relative risk of ≈10% and ≈20–30% for 6 h and ≤ 5 h of daily sleep, respectively, when compared with 7 h. No association was found between long sleep duration and incident HTN. Previous meta-analyses have reported similar findings both in terms of presence and strength of this association [9–11]. Subgroup analyses in these meta-analyses offer important clues as to factors underlying the heterogeneity observed among published studies. Wang et al. found that SSD predicted HTN only in those < 65 years old and in females, and the observed heterogeneity among studies disappeared within age and gender subgroups [9]. Meng et al. found that duration of follow up, and objective versus self-reported sleep duration and HTN assessment were important sources of heterogeneity [10]. Additionally, recent investigations using polysomnographically assessed sleep duration and sleep quality questionnaires suggest that both quality and quantity of sleep contribute to the causal relationship with HTN [4, 12]. Notwithstanding these caveats, overall evidence now strongly suggests a significant, though somewhat nuanced, link between sleep deprivation and HTN. Supporting the extant epidemiologic evidence, experimental sleep deprivation in healthy adults blunted the normal 10–20% nocturnal blood pressure (BP) “dipping”, and elevated the daytime ambulatory BP [13, 14]. Mechanistically, acute and short-term experimental sleep deprivation is associated with increased sympathetic and decreased parasympathetic cardiovascular modulation [15, 16], increased catecholamine levels and venous endothelial dysfunction [15, 17], and increased arterial stiffness [18]. Chronic insomniacs were found to have increased urinary free cortisol and catecholamine levels, as well as alterations in the hypothalamic-pituitary-adrenal axis regulation, indicating an increased activity of both limbs of the human stress response [19, 20]. Additionally, experimental and epidemiologic studies indicate wide-ranging effects of disturbed or reduced sleep on metabolic and endocrine homeostasis, including increased insulin resistance, glucose intolerance, and incident diabetes, as well as appetite regulation and caloric intake with resultant increases in adiposity and obesity, all well-known features of metabolic syndrome(MetSyn) [21–24]. Not surprisingly, and as further proof of the clinical impact of these cardiometabolic perturbations, SSD and insomnia symptoms have been found to predict incident coronary heart disease, heart failure, stroke, and cardiovascular and all-cause mortality [22, 25, 26]. Obviously, incontrovertible proof of causality and ability to make solid therapeutic recomendations requires demonstration of reversibility, i.e., improvement in BP control and/or clinical events, and perhaps other measurable metabolic parameters, with prolongation of sleep in a * Harsh Goel [email protected]