Short-Term Blood Pressure Variability and Cognition in the Elderly: Mere Association or a Key Mechanism?

Many studies have been published on the consequences of high blood pressure (BP) on the brain and overall they show that, even without the occurrence of a stroke, hypertension in middle-age is associated with an increased risk of cognitive deterioration, cognitive decline over time, and ultimately dementia some decades later.1–3 The mechanisms underlying this relationship are not yet fully elucidated, but it is generally accepted that there is a mix of impaired autoregulation of cerebral blood flow, microvascular ischemia, and hypoperfusion. The cerebral circulation has several unique characteristics, including the absence of anastomoses in the vascular architecture of deep brain regions which renders them particularly vulnerable to chronic hypoperfusion4 as well as a particular susceptibility to pressure damage, since this is a torrential circulation with minimal vascular resistance. Both the mean and pulse pressures are therefore easily transmitted from the aorta to the small cerebral arteries, a phenomenon amplified by age and hypertension. Such age and hypertension-related cerebral small vessel disease would limit the autoregulation of cerebral blood flow and thus increase susceptibility to cerebral ischemia5 when BP is transiently low. BP variability may therefore be directly involved in transient cerebral ischemia that leads to cognitive deterioration, particularly in elderly hypertensive individuals. There are indeed several recent studies showing that long-term visit-to-visit variability was associated with cognitive deterioration6–9 and even dementia.10 However, the constraints related to the observation of visit-to-visit variability over a period of years render this marker essentially useless in terms of clinical practice, so other measures of BP variability are of interest. There have been few studies on the association between short-term BP variability measured by 24-hour ambulatory BP monitoring and cognition, and even these few they exhibit certain discrepancies.11–15 In this paper, Cho and colleagues analyzed the cross-sectional association between ambulatory BP variability and cognitive status in a sample of 232 elderly hypertensive individuals (mean age 77.7 years, 66% of them women) recruited over a short period of time in one center.16 The level of BP control (defined as 24-hour BP < 130/80 mm Hg) was high, with 87% of the participants having controlled hypertension and the sample mean 24-hour systolic BP (SBP) being 118.7 mm Hg. The authors analyzed the association between cognition assessed by the Japanese version of the Montreal Cognitive Assessment (MOCA) and SBP as well as the weighted SD of SBP, a marker of SBP variability taking into account the day-to-night BP changes. Results were statistically significant for BP variability, showing a lower cognitive score in the upper quartile of the weighted SD of SBP, mainly on executive and attention functions, while there was no relationship between cognition and the SBP quartiles. The results were also clinically significant, as in the highest quartile of the weighted SD of SBP, the mean MOCA score was approximately 3 points less than in the lowest quartiles. Should it be concluded that lowering BP is insufficient regarding cognition and that short-term BP variability must be taken into account? One interpretation of the study is indeed that achieving a high level of control of BP and a low mean SBP in elderly hypertensive individuals may be inadequate for preserving cognition, and that BP variability should be controlled as well in order to reach this aim. There is however another, more subtle and worrying interpretation of the results: the effect of short-term variability may have only been unmasked because the patients were too well treated and controlled. One unique feature of the current study is the homogeneity of the sample, since all of the participants were followed in a single center and by a single hypertension specialist who was able to achieve a high level of hypertension control (87 %). Important characteristics of the sample are therefore not only the low mean 24-hour SBP (118.7 mm Hg), but also the rather old age of the participants (mean age of approximately 78 years). It might be hypothesized that the variability of SBP in these elderly hypertensive patients with a low SBP directly translates into frequent episodes of transient cerebral ischemia and thus cognitive deterioration. This raises an important question: is it the combination of high BP variability and a low BP level that proves detrimental in some elderly individuals? If yes, in these individuals it might be important to try to reduce BP variability but not the BP level. Certain previous studies,11–13,15 but not all,14 have also shown that there is an association between 24-hour BP variability and cognition deterioration but, like the study of Cho and colleagues, they had a cross-sectional design. All these studies are therefore exposed to the risk of third-party factors explaining part of the association. Age, for example, is strongly associated with the risk of cognitive deterioration as it is with BP variability and the observed association between BP variability and cognition could then be partly or completely explained by aging-related factors that are not always measured, and adjusting for age may not be sufficient to account for these confounding factors. A reverse causation may also be involved: one major cause of cognitive deterioration in the population is dementia, which is frequent and known to develop insidiously over several decades before the onset of overt cognitive symptoms. The neuronal depopulation at work in Alzheimer’s disease could affect the deep structures of the brain, especially those regulating the vegetative nervous system and BP. The relationship would then be causal but inverted: it would be the dementing process that caused the BP variability and not the reverse. Where to go next? The merit of the study of Cho and colleagues as well as previous studies on the same topic is to point out 24-hour BP variability as a potential risk factor for cognitive deterioration in elderly individuals. It remains an important hypothesis and we need larger, less selected samples with longitudinal follow-up for both cognition and 24-hour BP variability, in order to further explore this association. Some vital questions should then be answered: what is the temporal pattern and does short-term BP variability precede cognitive decline? Does its effect on cognition depend on the baseline SBP level? Are there differences according to the individuals’ profiles (age, gender, vascular profile, etc.)? Future studies should also include a measure of vascular brain injury lesions determined by cerebral MRI. Vascular brain injury, which includes mainly white matter lesions and silent infarcts, has been shown to be very common in the elderly, and to be associated with cognitive deterioration and dementia.3,17 Further, hypertension is its most important modifiable risk factor.3 Vascular brain injury is therefore an important factor in the relationship of BP with cognitive deterioration and its assessment could help to answer some of the questions raised above. To start with, the plausibility of the association between 24-hour BP variability and cognitive deterioration would be singularly reinforced if it was shown that there is also an association between 24-hour BP variability and vascular brain injury. There is still a long way to go to confirm the association between 24-hour BP variability and cognitive deterioration, and to estimate its real impact in broader populations with longitudinal follow-up, including cerebral MRI. This may seem daunting, but we have to keep in mind that we have very limited options to try to limit the ongoing epidemic of dementia, and obtaining a better understanding of the relationship between BP and the brain is one of the best of them. The total number of people with dementia worldwide was estimated at 27 million in 2007 and is projected to reach 106 million in 2050.18 We know now that the vast majority of dementia cases, including Alzheimer’s disease, have a mix of vascular injury and neurodegenerative lesions.19 Reducing the vascular burden by a better control of high BP might help to reduce its prevalence.18 However, in elderly individuals, the situation may be more complex,3,20 and the SPRINT-MIND trial is expected to provide needed insights.21 Could a better control of short-term variability be an important factor above and beyond BP control? It is too early to raise real hope, but the study of Cho and colleagues confirms that it is certainly worth continuing research in this direction. Ambulatory BP monitoring is a relatively simple measure that could be performed in many centers on a regular basis, so if future studies confirm that it helps identify elderly individuals with a higher risk of dementia and to adjust their BP treatment, it would be a significant step forward in building our decision tree. Even if the real impact of short-term BP variability on cognition and on the risk of dementia remains speculative, we are so powerless in the face of the epidemic of dementia, with no known preventive or curative treatment, that no hypothesis should be neglected and no stone left unturned. The author declared no conflict of interest. Dr Tzourio has received investigator-initiated research funding from the French National Research Agency, the Fondation Leducq, and the Initiative of Excellence (IdEX) of Bordeaux University.