Sleep and future cardiovascular risk: prospective analysis from the English Longitudinal Study of Ageing

Highlights

• Short sleep duration and disturbance increase the risk of incident hypertension.

• The link between sleep duration and hypertension is stronger in middle than old age.

• Sleep disturbance predicts inflammation in women but not in men.

Abstract

Study objectives

Hypertension and inflammation may contribute to the increased risk of cardiovascular disease in individuals with suboptimal sleep, but large prospective studies are lacking. This study tested whether sleep duration and disturbance were predictive of incident hypertension and inflammation four years later.

Methods

Participants were men and women aged 50 years and older from the English Longitudinal Study of Ageing. Sleep was assessed by self-report, incident hypertension (N = 3068) was defined by clinical examination and C-reactive protein and fibrinogen (N = 3768) were measures of inflammation.

Results

Both men (odds ratio, OR:1.73, confidence interval, C.I. 1.08–2.76) and women (OR: 1.44, C.I. 1.00–2.07) reporting short sleep at baseline had increased odds of incident hypertension 4 years later, after adjustment for covariates. Age-stratified analyses revealed that short sleep was predictive of incident hypertension in men (OR: 2.27, C.I. 1.01–5.11) and women (OR: 2.10, C.I. 1.08–4.09) younger than 60 years but not in older people. Disturbed sleep also predicted incident hypertension in men (OR: 1.20, C.I. 1.02–1.41). In women, disturbed sleep was associated with elevated C-reactive protein (B = 0.030, C.I. 0.00–0.06) and fibrinogen (B = 0.030, C.I. 0.01–0.05) at follow-up controlling for baseline inflammation and other covariates. Sleep duration was unrelated to inflammatory markers in either sex.

Conclusions

This study of older men and women adds to growing evidence that aberrant sleep patterns may increase the risk of cardiovascular outcomes through its adverse impact on blood pressure and inflammation.

Introduction

Disturbed sleep is common in developed countries including the UK [1], [2]. Insufficient (typically <5 or <6 h) and too long (typically >8, >9 or >10 h) sleep hours are prospectively associated with higher risk of all-cause and cardiovascular disease (CVD) and CVD mortality [3], [4], [5]. Suboptimal and disturbed sleep are also linked to major cardiovascular risk factors including adverse metabolic outcomes [6], hypertension [7], and obesity [8].

Experimental sleep studies have reported that acute sleep deprivation is associated with an increase in inflammatory markers such as C-reactive protein and interleukin-6 (IL-6) [9], [10]. This has led to the hypothesis that low-grade inflammation, a well-established cardiovascular risk factor [11], may be one biological mechanism through which aberrant sleep patterns increase the risk of future cardiovascular outcomes. Indeed, using data from the English Longitudinal Study of Ageing (ELSA) we have reported that in men, but not in women, long (>8 h) and disturbed sleep were both associated with raised concentrations of C-reactive protein and fibrinogen (another marker of inflammation implicated in the development of CVD) [12]. These findings are in line with previous studies [13], although some authors have found these relationships only in women [14], [15], [16], [17].

Sleep and inflammatory markers have been largely explored in cross-sectional studies and the findings of prospective investigations [13], [15], [17], [18] are inconsistent. Our previous investigation [12] was cross-sectional and therefore shed no light on the temporal precedence between sleep measures and inflammatory markers. Although abnormal sleep patterns may lead to poor health, for instance, through low-grade inflammation or metabolic dysfunction, poor health is also likely to impair sleep [19], [20]. Therefore, the first aim of our study was to test prospective associations between self-reported sleep and inflammatory markers. We hypothesised that both short and long as well as disturbed sleep would be associated with higher levels of C-reactive protein and fibrinogen at a 4-year follow-up.

Experimental studies suggest that sleep curtailment leads to increases in blood pressure [21]. This is thought to be mediated by increased sympathetic nervous activity and catecholamine concentrations [21], [22], [23], among other processes. The prevalence of hypertension is higher in sleep-disordered populations, in particular in those diagnosed with insomnia and sleep apnoea [24], [25]. However, in population-based samples the evidence relating sleep measures with hypertension is conflicting. For example, in the Whitehall II and the Western New York Health studies, short sleep was (cross-sectionally) associated with hypertension only in women [26], [27]. Analysis of the first National Health and Nutrition Examination Survey (NHANES I) revealed that short sleep was prospectively linked to hypertension in respondents younger than 60 years of age, but not in their older counterparts [28]. These age-dependent associations between sleep hours and hypertension were subsequently corroborated by a study in South Korea [29]. Relatedly, short sleep duration was unrelated to hypertension in samples of older Dutch and Spanish men and women [30], [31]. However, in contrast with the NHANES I [28] and Korean studies [29], sleep hours were not predictive of hypertension in middle-aged women from the Study of Women's Health Across the Nation (SWAN) [32], or in young (mean age 36 years) Spanish adults [33]. More recently, objective short sleep duration and lower sleep efficiency were prospectively associated with an increase in diastolic and systolic blood pressure, but not with incident hypertension in early middle-aged respondents from the Coronary Artery Risk Development in Young Adults (CARDIA) study [34].

Taken together, the literature relating the relationship between sleep and hypertension is conflicting, which, at least in some studies, seems to be moderated by sex and age [7]. Given the high prevalence of both aberrant sleep patterns [2] and hypertension [35] in industrialised countries, further studies are warranted on this import health issue. In the present study, we hypothesised that both short sleep and disturbed sleep would be associated with incident hypertension four years later.

Gender differences in associations between sleep measures, inflammatory markers and hypertension have been reported in the literature [7], [14], [36]; so the analyses presented here were stratified by gender.