After the initial investigations by Perloff et al.,1 our knowledge of the prognostic significance of ambulatory blood pressure (ABP) measurements has considerably increased. There is currently little doubt that 24-hour ABP is a better predictor of mortality and morbidity than conventional or office blood pressure (BP). This has been shown in the general population, in elderly, in treated and untreated hypertensive patients, in older patients with systolic–diastolic or isolated systolic hypertension and in refractory hypertension.2 However, debate continues on the relative importance of daytime and night-time ABP, on the prognostic significance of the night–day BP ratio and of the dipping pattern; in the latter categorisation, reverse dippers have higher ABP during the night than during the day, and night-time ABP is lower by, respectively, 0–10%, 10–20% and more than 20% in non-dippers, dippers and extreme dippers.
Studies that reported on daytime and night-time ABP separately found that both BPs carried significant prognostic information in patients with hypertension.3–8 Whereas the prognostic value of daytime and night-time ABP was similar in two studies,3,5 others directly compared the prognostic value of the two BPs and found that night-time ABP was a significantly better predictor than daytime ABP.4,6,7 Also, the results on the nocturnal fall of BP or on the night–day BP ratio are not consistent in hypertension. Some studies have observed a significantly better prognosis in patients with greater decline in night-time ABP4,8 but this was not confirmed by others.5,9 In addition, Kario et al.10 reported in older hypertensives that the relative risk of stroke was greater in extreme dippers than in dippers and non-dippers.
To better appreciate the prognostic significance of daytime and night-time ABP, the night–day BP ratio and the dipping pattern in hypertension, the author and co-workers pooled the individual data of hypertensive patients from four prospective studies performed in Europe and co-ordinated in Belgium.4,5,11,12 Advantages of the meta-analytical technique, particularly when on individual patient data, are the increased statistical power and the greater precision of the estimates, and meta-analyses may resolve uncertainty when individual studies disagree. Disadvantages are potential differences in methodology in different studies and the potential heterogeneity of individual studies.13
Common features of the four studies included in the Belgian Ambulatory Blood Pressure Monitoring database14 are: the prospective design; follow-up for fatal and non-fatal events; and evaluation of events by blinded end-point committees, which used the same definitions of events. The different selection criteria of the various studies are a potential limitation. On the other hand, the database represents a wide and comprehensive spectrum of hypertensive patients with regard to age, gender, type of hypertension, antihypertensive treatment, and type of care (primary care and specialist care).
Hypertensive Patients With No History of Cardiovascular Disease
Daytime and Night-time Blood Pressure
The Belgian Ambulatory Blood Pressure Monitoring database contains 3,468 hypertensive patients with hypertension and without severe co-existing disease, debilitating illness, dementia, impairment of renal function or major cardiovascular disease at baseline. Their age averaged 61±13 years and 45% were men; 61% were under antihypertensive treatment at the time of ABP monitoring. Office BP averaged 159±20/91±12mmHg, daytime ABP 143±17/87±12mmHg, night-time ABP 130±18/75±12mmHg, and the night–day BP ratio 0.907±0.085/ 0.866±0.095. During a median follow-up of 6.57 years (range: 0.08–13.1), 324 patients died (145 from a cardiovascular cause: coronary heart disease [CHD]: 68; congestive heart failure [CHF]: 29; stroke: 31; other: 17) and 129 patients suffered a fatal or non-fatal CHD event (myocardial infarction [MI] or sudden death), and 113 a fatal or non-fatal stroke; altogether there were 272 first major cardiovascular disease events, that is cardiovascular death or non-fatal MI or stroke.14
Systolic daytime and night-time ABP significantly predicted all outcomes, independently from age, gender, smoking, total cholesterol, diabetes, antihypertensive treatment and office BP. Diastolic daytime and night-time ABP predicted CHD, stroke and the aggregate of cardiovascular disease events, but all-cause and cardiovascular mortality were only significantly and independently predicted by night-time ABP. Figure 1 summarises the hazard ratios (HRs) associated with one standard deviation increment in BP for systolic and diastolic BP when both daytime and night-time ABP were included in the models, together with the other covariates.
Systolic night-time ABP significantly predicted all-cause and cardiovascular death, CHD, stroke and the aggregate of major cardiovascular disease events. Diastolic night-time ABP independently predicted all-cause and cardiovascular mortality and major cardiovascular disease events. Neither systolic nor diastolic daytime ABP added prognostic precision to night-time ABP. Separate analyses in men and women, in treated and untreated patients and in older and younger patients according to median age (63 years) indicated that the results were roughly similar in these various subgroups.14
Apart from studies included in the meta-analysis,4,5,12 few other studies have investigated the prognostic significance of daytime and night-time ABP in hypertensive patients, with adjustment for office BP and other covariates. Daytime ABP significantly and independently predicted all-cause mortality,6,7,15 cardiovascular and stroke mortality6 in patients referred for ABP monitoring, all strokes in older hypertensive patients16 and an aggregate of cardiovascular events in refractory hypertension.17 Night-time ABP independently predicted all-cause,6,7 cardiovascular and stroke mortality6 but not total stroke.16 When daytime and night-time ABP were included in the same model for the prediction of death, night-time ABP was superior to daytime ABP for all-cause,6,7 cardiovascular, cardiac and stroke mortality.6 The author’s results confirm these observations for all-cause and cardiovascular mortality for both systolic and diastolic ABP. In addition, the author and co-workers observed that systolic night-time ABP was the better predictor for fatal and non-fatal cardiovascular events, CHD and stroke, and these findings were consistent in men and women, in young and old, and in treated and untreated patients for most of the outcomes.
There is no clear explanation why night-time ABP would be a better predictor of outcome than daytime ABP, but several factors could be involved. BP is more variable during the day than during the night due to physical and mental activity, so it is possible that intermittent BP measurements do not completely capture the true average daytime ABP. It is of note that Khattar et al.,3 who used continuous intra-arterial ABP recordings, reported that the prognostic value of daytime and night-time ABP for cardiovascular events was about similar, but the two BPs were not included in the same model.
Night-time ABP is likely to be more stable so that intermittent BP measurements may be more representative of the true night-time average BP. Moreover, BP during sleep is more closely related to basal BP, which has been shown to predict life expectancy better than casual BP.18 Finally, night-time ABP may be influenced by sleep apnoea in some patients, which is associated with a worse prognosis.19
Night–Day Blood Pressure Ratio and Dipping Pattern
In the Belgian Ambulatory Blood Pressure Monitoring database, the systolic night–day BP ratio significantly predicted all-cause and cardiovascular mortality, CHD, stroke and aggregates of cardiovascular events, with adjustment for age, gender, smoking, total cholesterol, diabetes and antihypertensive treatment. However, significance only persisted for all-cause mortality after additional adjustment for 24-hour ABP (HR: 1.13; 95% confidence interval (CI): 1.02–1.24). The diastolic night–day BP ratio predicted all-cause and cardiovascular mortality, both before and after adjustment for 24-hour ABP.14 In a subsequent analysis, in which the author and co-workers also included CHF, documented angina pectoris and transient ischaemic attack in the aggregate of cardiovascular disease events, amounting to a total of 473 first events, the systolic night–day BP ratio did add prognostic precision to 24-hour ABP, both for systolic ABP (HR: 1.12; 95% CI: 1.12–1.22) and for diastolic ABP (HR: 1.10; 95% CI: 1.01–1.21). Interactions with gender, age and treatment were not significant, indicating that results were similar in men and women, in younger and older patients, and in treated and untreated patients.
Finally, patients were classified in four categories according to the systolic night–day BP ratio as follows: reverse dippers if the ratio was >1.0, non-dippers if ≤1.0 and >0.9, dippers if ≤0.9 and >0.8, and extreme dippers if ≤0.8.10 The percentage of patients in these categories amounted to, respectively, 12%, 41%, 37% and 10%. The hazard ratios for all-cause mortality and the extended aggregate of cardiovascular disease events of reverse dippers, non-dippers and extreme dippers were assessed versus dippers, with adjustment for age, gender, smoking, total cholesterol, diabetes and antihypertensive treatment, and, in addition, for 24-hour ABP. The hazard ratios were never significant when non-dippers were compared with dippers.
The hazard ratio of reverse dippers versus dippers was significant for all cardiovascular events, both before (HR: 1.80; 95% CI: 1.37–2.37) and after adjustment for 24-hour ABP (HR: 1.51; 95% CI: 1.14–2.00); the hazard ratio for all-cause mortality was significant before (HR: 1.42; 95% CI: 1.03–1.94) but not after adjustment for 24-hour ABP (HR: 1.28; 95% CI: 0.92–1.76). All-cause mortality was lower in extreme dippers than in dippers, with similar hazard ratios before and after adjustment for 24-hour ABP (HR: 0.59; 95% CI: 0.39–0.88).
A number of other studies reported on the prognostic significance of the night–day BP ratio or the dipping pattern in hypertensive patients, but data on the four dipping categories are scarce. Ben-Dov et al.7 reported on patients referred for ABP monitoring and found that all-cause mortality was similar in extreme dippers and dippers in a model controlling for awake ABP and confounders; compared with all dippers, prognosis was significantly and progressively worse in non-dippers and reverse dippers. Kario et al.10 observed a J-shaped relationship between dipping pattern and incidence of stroke in older hypertensives and this relationship remained significant after controlling for 24-hour ABP and other covariates. Moreover, the incidence of stroke was significantly higher in extreme dippers than in dippers. The author and co-workers could not confirm this finding; the night–day BP ratio was not independently predictive of stroke and the incidence of cerebrovascular disease was not significantly different among the four dipping categories in the current analysis.
In general, the predictive power of the night–day BP ratio appears to be largely dependent on the worse prognosis of reverse dippers. Several mechanisms have been invoked to explain the higher night-time ABP and associated worse outcome: nocturnal autonomic dysfunction; disturbed baroreflex sensitivity; sleep apnoea; abnormal sodium handling; and nocturnal volume overload.14,20,21 It has also been suggested that the higher nocturnal ABP might be a marker of disease, leading to lower daytime ABP, or might result from intake of drugs to lower BP during the day. However, the author and co-workers excluded patients with a history of cardiovascular disease or other co-existing disease at baseline, and found similar results in treated and untreated patients, so that reverse causality is unlikely to explain the worse prognosis of reverse dippers in their study.
Hypertensive Patients with History of Cardiovascular Disease
Previous studies excluded patients with history of cardiovascular disease or included small numbers of patients with complicated hypertension. No study, however, has separately reported on the prognostic value of daytime and night-time ABP in hypertensive patients with cardiovascular disease, and none has included an interaction term in the Cox models to investigate whether the results differed according to the presence of cardiovascular disease. The Belgian Ambulatory Blood Pressure Monitoring database contains 302 patients with hypertension and cardiovascular disease at baseline,22 including CHD, cerebrovascular disease and CHF. Age of these patients averaged 69±9 years, 50% were men and 62% were under antihypertensive treatment at the time of ABP monitoring. Office BP averaged 161±20/86±12mmHg, daytime ABP 144±16/83±11mmHg and night-time ABP 132±18/72±12mmHg, and the night–day BP ratio 0.907±0.085/0.866±0.095. During median follow-up of 6.76 years, 91 patients died, 58 from a cardiovascular cause, and 84 suffered a first major cardiovascular event, including cardiovascular mortality (n=47), non-fatal MI (n=19) or stroke (n=18). Whereas daytime ABP did not predict any outcome, night-time systolic ABP predicted cardiovascular mortality (HR: 1.41; 95% CI: 1.06–1.87) and cardiovascular disease events (HR: 1.34; 95% CI: 1.06–1.69), and night-time diastolic ABP predicted cardiovascular events (HR: 1.38; 95% CI: 1.07–1.80), after adjustment for age, gender, smoking, total cholesterol, diabetes, antihypertensive treatment, history of CHD, cerebrovascular disease, CHF and office BP. As shown in Figure 2, systolic night-time ABP was a significant predictor of the three outcomes, when both daytime and night-time ABP were included in the models together with the confounding variables. Daytime ABP did not add prognostic precision to night-time ABP. Results were similar for diastolic ABP.22
In addition, the systolic night–day BP ratio was significant for the three outcomes before and after adjustment for 24-hour ABP, but this was not the case for the diastolic BP ratio. Finally, all-cause mortality and the incidence of cardiovascular events were significantly higher in reverse dippers than in dippers, both before and after adjustment for 24-hour ABP; similar results were observed for cardiovascular events when non-dippers were compared with dippers.22
Night-time ABP is, in general, a better predictor of mortality and cardiovascular disease events than daytime ABP in patients with hypertension. In addition, the night–day BP ratio significantly predicts outcome, with worse prognosis in reverse dippers and better prognosis in extreme dippers in comparison with dippers. The results are, in general, similar in men and women, in older and younger patients, in treated and untreated patients and in patients with and without history of cardiovascular disease at baseline.