Discussion
The present study was based on hospital-based, prospective cohort study aiming to assess BP parameters and clinical outcome in a Chinese population of patients with acute IS or TIA. The following are the main findings: (1) both SBPV and DBPV were associated with stroke recurrence; and (2) DBPV was a strong predictor of cardiovascular events, whereas there was no observed association between SBPV and cardiovascular events.
SBPV was more often investigated and reported in comparison with DBPV. We have noticed that previous studies had mainly focused on subjects with SBPV and stroke, while reports regarding the association between DBPV and stroke have yielded inconsistent results. Studies conducted in 8811 patients ≥55 years with diabetes (69% with a history of hypertension) from the ADVANCE (Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation) trial9 found no association between BPV and stroke. In a large cohort of 3 285 684 US veterans (63% with hypertension),10 greater SBPV was proven to be associated with a higher risk for stroke, while DBPV was not examined. Moreover, it has been shown in several previous studies that VVV in DBP had poor correlation with stroke.2 Both SBPV and DBPV showed associations with stroke recurrence in the analysis we conducted on patients with IS/TIA. Similar results were found at 1 year, although the trend was weaker than 3 months, since stroke recurrence happened more frequently during the acute phase of IS/TIA. The results of our study were, to some extent, consistent with a previous report from the Oxford Vascular Study.11 Moreover, we extended previous findings by assessment of BPV through day-to-day monitoring, enabling us to collect more reliable information.12
Another important finding of this study is that DBPV was not only associated with stroke, but was strongly associated with cardiovascular events, even after adjusting for confounding variables. In contrast, a negative association was observed between SBPV and cardiovascular events. BPV parameters may play different roles between cerebrovascular and cardiovascular events. In Asia, the association between SBP level and stroke is stronger than the association between SBP and ischaemic heart disease.13 Hata and colleagues14 showed that the coefficient of SBPV was slightly greater in patients with stroke than in controls,15 but not in patients with myocardial infarction. Consistent with the results of the present study, researchers from the Ohasama Study observed that the predictive power of SBPV level was only for stroke but not for cardiovascular disease.16 The results from the STABILITY trial showed that in patients with stable coronary heart disease, higher VVV of both SBP and DBP is a strong predictor of increased risk of cardiovascular events, but not stroke.4 These results suggest that DBPV should merit more attention in predicting the risk of cardiovascular events. With all the attention on target SBP, DBP is often relegated to a footnote. Our findings reported that episodic fluctuation in DBP might trigger cardiovascular events in a population with established stroke or TIA. The SPRINT trial suggested a benefit of SBP <120 mm Hg; however, optimal DBP treatment goal is still in question.6 The problem raised by the study is whether treatment of SBP that produces high DBPV might inadvertently increase the risk for cardiovascular events in a stroke population.
Several mechanisms may explain the prognostic impact of different patterns of variability in BP. It is widely recognised that SBP control is the most important factor for preventing cerebrovascular and cardiovascular diseases.17–20 However, both SBP and DBP contribute to cerebral blood flow (CBF). Rapid, large falls in BP could reduce CBF, leading to extension of cerebral infarctions.21 In ischaemic areas of the brain during stroke, CBF becomes dependent on mean arterial BP, which is equal to (SBP + 2 DBP)/3. Previous analyses have demonstrated that stroke was best predicted by mean arterial BP,22 23 compared with other BP parameters. The small penetrating end arteries, which supply the medial and basal portions of the brain and brainstem, seem to be particularly vulnerable to the adverse effects of high BP, in as much as these arteries arise directly from the main arterial trunks. Furthermore, greater BPV is associated with greater aortic stiffness and maladaptive carotid arterial remodelling.24 Endothelial dysfunction and subclinical inflammation have also been proposed as mechanisms underlying the association between BPV and outcomes.25–27 Additionally, the association between DBPV and cardiovascular events agrees with the key role of DBP in myocardial perfusion. As we have known, the heart is perfused during diastole. DBP plays a key role in myocardial perfusion. A marked increase in adverse cardiovascular outcomes was found for DBP values below 70 mm Hg according to results from an international cohort study.28 The association of high DBPV with cardiovascular events in this study is consistent with the hypothesis that the risk is mediated through coronary perfusion pressure.
The BOSS study had several advantages. For example, each patient enrolled was assigned a semiautomatic upper-arm BP monitor (HEM-4030; Omron), which made measurements of BP more standard. Additionally, BPV was assessed day to day in this study which is less affected by circadian rhythm, the patient’s status and the conditions of BP measurement. Repeated assessment of BP values by patients themselves or relatives may offer a solid base to assess BPV during the long-term follow-up period.29 There are also known and potential limitations associated with the current analysis. First, the possibility of residual confounding cannot be fully eliminated in an observational study, although several important potential confounders have been controlled in multivariable adjusted models. Second, the follow-up duration in this study is 1 year, which is relatively short compared with previous studies on cardiovascular disease. This may lead to lower cumulative occurrence of cardiovascular events, because the disease emerges over a long period of time determined by its pathological characteristics. In addition, patients receive repeat neuroimaging and electrocardiographic examination only if they had new symptoms or diagnostic signs during the study’s follow-up period, which led to the inability to identify covert strokes or equally subclinical cardiovascular events.