Study
Participantsa
Follow-up
Outcome
Main results
Kohler et al. (2014)
n = 1805, age 25–84 years
12 years
Verbal memory, executive function and processing speed (psychometric tests)
Baseline and incident HT (≥140/90 mmHg) associated with a faster decline in memory and processing speed
Gottesman et al. (2014)
n = 13,476, age 45–64 years
20 years
Verbal learning, short-term memory and executive function (as a composite)
HT (≥140/90 mmHg) at baseline was associated with a steeper cognitive decline later in life
Taylor et al. (2013)
n = 1484, age 40–67 years (multiethnic)
20 years
Global cognitive function, expressed as a composite
Low and high baseline DBP (quintiles) related to cognitive impairment
Ninomiya et al. (2011)
n = 534, age 65–79 years
32 years
AD and VaD (DSM-III and NINCDS-ADRDA
Greater mid-life BP (quartiles) associated with increased risk of VaD but not of AD
Whitmer et al. (2005)
n = 8845, mean age 42 years
30 years
Dementia (medical records)
Mid-life HT (≥140/90 mmHg) associated with a 20–40 % increased risk of dementia
Wu et al. (2003)
n = 301, age >65 years
15 years
AD (DSM-IV)
Mid-life severe HT (≥160/95 mmHg) was as a strong risk factor for AD later in life
Yamada et al. (2003)
n = 1774, age >65 years
25–30 years
AD and VaD (DSM-IV)
Higher SBP (continuous variable) associated with increased risk of VaD but not of AD
Kivipelto et al. (2001)
n = 1449 age 40–64 years
11–26 years
AD (DSM-IV and NINCDS-ADRDA)
High SBP (≥160 mmHg) in mid-life was a significant risk factor for AD later in life
Kilander et al. (2000)
n = 502, age 50 years (men only)
20 years
Global cognitive function, measured with 13 psychometric tests
Low DBP (≤70 mmHg) related to better performance in cognitive tests later in life
Launer et al. 2000
n = 3703, age 45–68 years (men only)
25–27 years
Dementia (DSM-III); AD (NINCDS-ADRDA)
Midlife severe HT (≥160/95 mmHg) increased risk of late-life dementia in untreated subjects
Kilander et al. (1998)
n = 999, age 50 years (men only)
20 years
Global cognitive function, measured by the MMSE
Positive association between mid-life DBP and prospective cognitive decline
Launer et al. (1995)
n = 3735, mean age 53 (men only)
25 years
Cognitive function, measured by the CASI
Midlife SBP predictor of poor cognitive function. No association with DBP
Elias et al. (1993)
n = 1702, age 55–88 years
14–20 years
Global cognitive function, expressed as a composite score
Inverse relation between BP (continuous variable) and cognition
The Framingham Study in the early 1990s was one of the first to show that blood pressure levels and chronicity of hypertension in individuals aged 55–88 years where inversely related to global cognitive performance and to specific measures of memory and attention, assessed 14 years after blood pressure examination (Elias et al. 1993). A similar adverse effect of midlife hypertension on late-life cognitive function was demonstrated by the Honolulu-Asia Aging Study (Launer et al. 1995). Although this program examined only men of Japanese-American origin, results demonstrated that the risk for poor cognitive function increased progressively with higher levels of SBP, which had been measured in the preceding two decades. In this report, there was not a significant association between midlife DBP and cognitive function. However, two additional Swedish studies with other large-scale men cohorts have found significant inverse correlations between DBP measured at age 50 and cognitive function 20 years later (Kilander et al. 1998; Kilander et al. 2000). Extending these observations, the multiethnic Southall and Brent study showed a significant U shaped-association between mid-life DBP (at age 40–67) and cognitive impairment in the subsequent two decades (Taylor et al. 2013). In other words, both low and high DBP were found to have adverse effects on cognitive function later in life. The U-shaped relationship between DBP and cognition was more prominent in the older participants (50–67 years) than in the younger subjects (40–49 years), and surprisingly pulse pressure showed little evidence of association after covariate adjustment (Taylor et al. 2013).
If hypertension has a negative effect on cognitive function, when do cognitive deficits appear after the onset of high blood pressure? In a recent prospective cohort (a Dutch population including men and women), the Masstrich Aging Study examined the cognitive trajectories of individuals with prevalent and incident hypertension (age 25–84) over a period of 12 years (Kohler et al. 2014). Interestingly, it was found that subjects who developed hypertension during the study duration exhibited a slow, steady decline in memory and processing speed within 6–12 years after blood pressure assessment, suggesting that the onset of hypertension may offer a window for therapeutic brain protection.
With respect to dementia, epidemiologic studies have also shown a significant adverse effect of elevated mid-life blood pressure on the risk of developing vascular dementia and Alzheimer’s disease in future years (Launer et al. 2000; Kivipelto et al. 2001; Wu et al. 2003; Yamada et al. 2003; Whitmer et al. 2005; Ninomiya et al. 2011). The Honolulu Heart Program, where a large cohort of Japanese-American men were followed over a period of 25 years, revealed that the risk of two of the most common subtypes of dementia (AD and VaD) was ~4 times higher (4.8 (95 % CI = 2.0–11.8)) for those individuals with untreated high SBP (≥160 mmHg) and high DBP (≥95 mmHg) (4.3 (95 % CI = 1.7–10.8) (Launer et al. 2000). Mid-life blood pressure (whether it was moderate or high) was not associated with dementia risk in men who received antihypertensive medication. This study further reported a trend for an association between low DBP (<80 mmHg) and increased risk of both types of dementia, although it was not significant.
In line with the results of the Honolulu Heart Program, Kivipelto and colleagues later demonstrated in a Finnish population (including men and women) that individuals with mid-life isolated systolic hypertension (SBP ≥160 mmHg) exhibited a ~ 2 fold risk (CI = 1.0–5.5) of developing Alzheimer’s disease over a follow-up period of 11–26 years (Kivipelto et al. 2001). This risk was enhanced in people who also had high serum cholesterol levels (≥6.5 mmol/L), and was not influenced by midlife DBP levels. The significant association between high SBP and dementia risk in the Honolulu Heart Program and in the Kivipelto cohort highlights the importance of monitoring and controlling isolated systolic hypertension. The lack of association between DBP and dementia in the study of Kivipelto and colleagues may be related to the fact that individuals with Alzheimer’s disease were more likely to have received treatment for hypertension, which could have corrected DBP but not SBP levels, as discussed by the authors. Also, some individuals were followed for less than 20 years and the population was smaller compared to that from the Honolulu Heart Program. Despite these differences, both studies highlight the relevance of high mid-life SBP as a risk factor for dementia.
2.2 Late-Life Hypertension, Cognitive Dysfunction and Dementia
Even though the association between midlife hypertension, cognitive dysfunction and dementia is generally well supported, the relationship between high blood pressure in late life (65+ years) and cognition is less consistent. A summary list of relevant reports is depicted in Table 2.
Table 2
Summary of studies investigating the association between late-life hypertension and risk of cognitive decline and dementia
Study | Participants | Follow-up | Outcome | Main results |
|---|---|---|---|---|
Ninomiya et al. (2011) | n = 668, age 65–79 years | 17 years | AD and VaD (DSM-III and NINCDS-ADRDA) | Significant association between late-life BP level (defined by JNC-7) and VaD but not AD |
Li et al. (2007) | n = 2356, age ≥65 years | 8 years | Dementia (NINCDS-ADRDA) | High SBP (≥160 mmHg) associated with increased risk of dementia (<70 years); risk declined with older age (70+) |
Waldstein et al. (2005) | n = 847, age 39–96 years | 11 years | Battery of six psychometric tests to assess attention, working and verbal memory, processing speed and executive function | At older ages, both high and low DBP were associated with poor performance on tests of executive function, confrontation naming and processing speed |
Solfrizzi et al. (2004) | n = 2963, age 65–84 years | 3.5 years | MCI (MMSE) and (DSM-III and NINCDS-ADRDA) | No significant effect of HT as a risk factor for MCI |
Tervo et al. (2004) | n = 747, age 60–76 years | 3 years | MCI (MMSE) | No effect of elevated blood pressure on conversion to MCI |
Kuo et al. (2004) | n = 70, mean age 72 years | Cross-sectional study | Verbal and visual memory, visuo-spatial skills and executive function | Greater SBP (quartiles) associated to impairment in executive function |
Hebert et al. (2004) | n = 4284, age ≥ 65 years | 3–6 years | Global cognition (composite of four cognitive tests) | No significant association between SBP or DBP and cognitive change |
Elias et al. (2003) | n = 1423, age 55–88 years | 4–6 years | Learning, memory, executive function and abstract reasoning | Positive association between HT (≥140/90 mmHg) and low cognitive performance only in men (not in women) |
Qiu et al. (2003) | n = 1270, age 75–101 years | 6 years | Dementia and AD (DSM-III) | Both high SBP (>180 mmHg) and low DBP (≤65 mmHg) associated with an increased risk of dementia and AD |
Verghese et al. (2003) | n =488, age ≥75 years | Median 6.7 years | Dementia (DSM-III) | High SBP (140–179 mmHg) and low DBP (<70 mmHg) influenced risk of developing AD |
Morris et al. (2001) | n = 378, age ≥65 years | 13 years | AD (NINCDS-ADRDA) | No association between high SBP (≥160 mmHg) and risk of AD |
Tzourio et al. (1999) | n = 1172, age 59–71 years | 4 years | Global cognitive function, measured with the MMSE | High BP (≥160/95 mmHg) associated with cognitive decline |
Kilander et al. 1998 | n =999, age 70 years (men only) | Cross-sectional study | Global cognitive function, measured with the MMSE | Greater DBP (quintiles) related to lower cognitive function |
Cacciatore et al. (1997) | n = 1106, age 65–95 years | Cross-sectional study | Global cognitive function, measured with the MMSE | Greater DBP (but not SBP) associated with increased risk of cognitive impairment |
Skoog et al. (1996) | n = 382, age 70 years | 15 years | Dementia (DSM-III) and AD (NINCDS-ADRDA) | Subjects who developed dementia at age 79–85 had higher SBP and DBP at age 70 than people who remained dementia free |
Several cross-sectional and longitudinal studies have indicated a significant adverse effect of late-life hypertension (defined as BP ≥140–160/90–95 mmHg) on global cognition (Kilander et al. 1998; Cacciatore et al. 1997; Elias et al. 2003; Tzourio et al. 1999) as well as on specific cognitive domains like executive function (Kuo et al. 2004). However, such associations between high BP and cognitive decline could not be demonstrated in other elderly populations (Hebert et al. 2004; Tervo et al. 2004; Solfrizzi et al. 2004; Gottesman et al. 2014). In addition, evidence from other study cohorts suggest either the inverse relation; that low blood pressure in older adults is related to poor global cognitive function, or even more complex U-shaped associations between late-life blood pressure and cognition (Launer et al. 1995; Guo et al. 1997b; Pandav et al. 2003; Waldstein et al. 2005). While high BP may be damaging to the brain by leading to atherosclerosis, white matter disease and altered neurovascular coupling (as reviewed later in this chapter), low BP may negatively affect cognition by leading to insufficient cerebral perfusion, rendering the brain more vulnerable to ischemic and neurodegenerative pathologies.
Similar discrepancies have been reported on the relationship between hypertension in old age and dementia (Qiu et al. 2005; Power et al. 2013). For example, a longitudinal study of 70 year-old residents of Gothenburg, Sweden, revealed a significant correlation between increased DBP (≥100 mmHg) at age 70 and the risk of developing dementia 15 years later (Skoog et al. 1996). In a study with a shorter follow-up (6 years), SBP and DBP were found to confer opposing risks. While subjects with very high SBP were at higher risk of dementia, very low DBP (≤65 mmHg), rather than high DBP, produced an adjusted relative risk of 1.7 (95 % CI = 1.1–2.4) for Alzheimer disease (Qiu et al. 2003). The association between low DBP on dementia risk was also found in other prospective cohorts (Verghese et al. 2003). The opposing trajectories of systolic and diastolic BP are in line with the concept that pulse pressure increases with older age, reflecting increased arterial stiffness; which could explain the positive associations between high SBP and low DBP with Alzheimer’s disease.
In contrast, Morris and colleagues found no significant relation between Alzheimer’s disease risk and high blood pressure measured 13 years before and 2 years after dementia diagnosis in subjects aged 65 or older (Morris et al. 2001). It should be noted that in this study only a small number of participants had high blood pressure at baseline, and that they tended to be older. Also, about 30 % of the population was receiving antihypertensive medication with diuretics or beta-blockers, although there was no association between antihypertensive use, Alzheimer’s disease and SBP measured between 4 and 13 years before diagnosis.
When examining the risk of high blood pressure on dementia across different age strata (65–74 years, 75–84 years and 85+), Li et al. found that only in the youngest age group there was a significant association between high SBP (>160 mmHg) and dementia (HR = 1.6, 95 % CI = 1.01–2.55) (Li et al. 2007). No such relationships were evident in the older participants, supporting the concept that the link between high blood pressure and dementia varies with age.
Since Alzheimer’s disease has a long incubation phase that lasts decades, it is possible that late-life hypertension is not related to a disease that had begun decades before, or alternatively, if hypertension is indeed a risk factor, there may not be enough time for the clinical expression of dementia to manifest when hypertension occurs in late life.
Interestingly, prospective longitudinal studies with very long durations (between 32 and 37 years of follow-up) have revealed an inverted U-shaped trajectory of blood pressure in people who developed incident dementia. In the Honolulu-Asia Aging Study, the development of Alzheimer’s disease or vascular dementia was related to a greater rise in SBP from midlife to late life followed by a decline in blood pressure in the years preceding the clinical diagnosis of dementia (Stewart et al. 2009).
Similar findings were presented by the Prospective Population Study of Women in Gothenburg, Sweden, a cohort that was followed for 37 years. In this study, the development of incident Alzheimer’s disease (between ages 79 and 85) was related to higher midlife SBP and DBP and a steeper increase of SBP between age 46 and 70, followed by declines in blood pressure in the years before dementia onset (between ages 75 and 85) (Joas et al. 2012). This decline could be explained by a disturbed control of BP with neurodegeneration and/or by the fact that hypotension itself promotes degenerative processes. Interestingly, the same study showed that such variations in SBP were more pronounced in demented subjects treated for hypertension, suggesting that these treatments in women who became demented did not prevent well the increase in blood pressure.
2.3 Post-stroke Dementia
It is estimated that more than 70 % of patients with ischemic or haemorrhagic stroke have a history of high blood pressure (Miller et al. 2014). Dementia is a frequent cause of disability after stroke, occurring in approximately 25–30 % stroke survivors, although prevalence rates may vary among study populations (Henon et al. 2006; Barba et al. 2000). In fact, having a stroke increases the risk of developing dementia by 3–5 times, irrespective of its type (vascular, Alzheimer’s or mixed), and this risk is highest within the first months after stroke (Leys et al. 2005).
Patients with post-stroke dementia have higher mortality rates and are often more impaired in functional daily activities (Henon et al. 2006). There is also a strong, linear association between stroke mortality and blood pressure (Palmer et al. 1992). Thus, it has been argued that among the modifiable risk factors, controlling blood pressure in patients with a prior stroke should result in a favourable reduction in the risk of developing dementia and cognitive impairment (Soros et al. 2013). Whether antihypertensive therapy is effective in preventing stroke-related cognitive decline and other forms of dementia will be discussed in Sect. 3 of this chapter.
2.4 How Does Hypertension Affect Cognitive Function?
Multiple mechanisms link hypertension to cognitive dysfunction and dementia (Fig. 1). High blood pressure affects the brain by altering the structure of the cerebral vasculature, by disrupting the mechanisms that regulate the cerebral circulation, as well as by contributing to the pathogenesis of Alzheimer’s disease (Faraco and Iadecola 2013; Iadecola and Davisson 2008; Girouard and Iadecola 2006; Skoog and Gustafson 2006). These effects will be briefly reviewed.
Fig. 1
The link between hypertension, cerebrovascular dysfunction and dementia. Persistent elevated blood pressure promotes the formation of atherosclerosis, leads to arterial smooth muscle hyperplasia and vascular remodelling, increasing arterial stiffening. The resulting increased pulsatility may promote reactive oxygen species production and inflammation in cerebral blood vessels and further lead to disruption in the blood brain barrier (BBB). Hypertension may also affect small cerebral arteries leading to microhemorrages and vessel wall necrosis, referred as lipohyalinosis. Besides affecting the structure of cerebral blood vessels, hypertension disrupts the mechanisms that regulate cerebral blood flow, such as neurovascular coupling (NVC) and cerebral autoregulation. These changes compromise the clearance of brain metabolites, such as amyloid-β and tau, favouring their accumulation. Taken together, the structural and functional alterations induced by hypertension lead to cerebrovascular dysfunction and render the brain more vulnerable to degenerative and ischemic disease
2.4.1 Structural Alterations in Cerebral Blood Vessels
Hypertension promotes the formation of atherosclerosis in large extracranial and intracranial arteries, leading to a reduced vascular lumen and hypoperfusion followed by vascular occlusion and ischemic injury, and consequently, to increased stroke risk (Hollander et al. 1993). Studies in rodents suggest that a moderate reduction in cerebral blood flow affects brain homeostasis. At declining flow rates, protein synthesis is inhibited first, at blood flow values approximately 80 % of the normal flow rate (Xie et al. 1989; Jacewicz et al. 1986), followed by alterations in glucose utilization and energy metabolism, the release of neurotransmitters (particularly of the inhibitory type), and finally by anoxic depolarization (at <20 % of normal flow rate) (Hossmann 1994). Chronic hypoperfusion has also been proposed as an early factor driving the neurodegenerative process of Alzheimer’s disease (Zlokovic 2011; de la Torre 2002).
Elevated blood pressure also induces occluding lesions in small cerebral arteries and arterioles supplying the white matter, known as lipohyalinosis. This refers to thickening of the vessel wall or in severe cases to vessel wall necrosis, which may lead to rupture (Lammie 2002). Thus, such morphological alterations facilitate the appearance of lacunes (infarcts of <20 mm in diameter) and microinfarcts (<1 mm in diameter), as well as microbleeds and large cerebral haemorrhages, leading to white matter damage, (Havlik et al. 2002). Notably, these vascular lesions have been associated to reduced cognitive function (Skoog et al. 1996; Kapasi and Schneider 2016; Wolf et al. 2000) and to an accelerated progression from mild cognitive impairment (MCI) to dementia (Clerici et al. 2012).
Sustained elevations in blood pressure may also lead to adaptive changes in cerebral blood vessels (i.e. vascular remodelling) to counteract the adverse effects of increased pulsatile stress induced by hypertension (Intengan and Schiffrin 2001). Vascular smooth muscle cells may grow in size or undergo a rearrangement that leads to reduced vessel lumen and capacity to dilate; and accumulation of extracellular matrix proteins may further lead to greater vessel wall thickness (Intengan and Schiffrin 2001; Baumbach and Heistad 1989; Heistad et al. 1990). Accumulation of collagen deposits and elastin fragmentation may also occur in large arteries, leading to reduced distensibility and vascular stiffening (Kaess et al. 2012). Arterial stiffness is often a neglected aspect of hypertension, which deserves further attention as it has been related to stroke, cognitive impairment and dementia (Pase et al. 2012; Hanon et al. 2005), although the pathogenic mechanisms involved continue to be elucidated (Sadekova et al. 2013).
2.4.2 Alterations in Cerebrovascular Function
Besides affecting the mechanical properties of cerebral blood vessels, hypertension interferes with the homeostatic mechanisms that control the regulation of cerebral blood flow. In experimental models, hypertension (induced by the infusion of angiotensin-II) impaired the endothelium-dependent relaxation of cerebral blood vessels as well as functional hyperemia, which is the increase in cerebral blood flow in response to neuronal activity (Kazama et al. 2004; Girouard et al. 2006). Alterations in neurovascular coupling are also supported by human studies. Patients with untreated hypertension have reduced evoked cerebral blood flow responses in the posterior parietal cortex when engaged in a memory task (Jennings et al. 2005).
Given that the integrity of endothelial cells is vital for the regulation of blood-brain barrier (BBB) permeability, the negative effects of hypertension on endothelial function inevitably impact on BBB maintenance (Abbott et al. 2010). Several lines of evidence point at oxidative stress, matrix metallo-protease activation and inflammation as the underlying mechanisms linking hypertension to BBB breakdown (Kahles et al. 2007; Yang and Rosenberg 2011).
In addition, hypertension alters cerebral autoregulation (Immink et al. 2004), which is the mechanism by which cerebral blood flow remains constant within a certain range of arterial pressures (60–150 mmHg). Taken together, the changes induced by hypertension compromise cerebral perfusion and render the brain more vulnerable to stroke and to neurodegenerative lesions. Interestingly, endothelium-dependent relaxation and functional hyperemia are altered in experimental models of Alzheimer’s disease (Niwa et al. 2000a, b; Park et al. 2004) as well as in humans suffering from this disorder (Hock et al. 1997; Rosengarten et al. 2006; Janik et al. 2016).
2.4.3 Alzheimer’s Disease Pathology
It is known that cerebrovascular lesions, such as those caused by hypertension, worsen cognitive performance and increase the likelihood of dementia development in individuals who also exhibit Alzheimer’s neuropathology (Snowdon et al. 1997; Esiri et al. 1999; Chi et al. 2013). Interestingly, histopathological analyses from the Honolulu-Asia Aging Study have demonstrated that people with elevated midlife blood pressure exhibited greater number of cortical and hippocampal amyloid plaques and neurofibrillary tangles, as well as reduced brain weight later in life (Petrovitch et al. 2000).
Likewise, Shah and colleagues recently reported that the risk of Alzheimer’s disease was higher in individuals with declining levels of plasma Aβ; an interaction which was enhanced by midlife blood pressure (Shah et al. 2012). In this study, reduced plasma Aβ was related to an increased likelihood of cerebral amyloid angiopathy (deposition of amyloid within the walls of cerebral blood vessels); suggesting that hypertension may interfere with the vascular clearance of Aβ leading to amyloid accumulation. As a case in point, experimental models of chronic and acute hypertension reproduce the enhanced deposition of Aβ in cerebral blood vessels and in the brain parenchyma, and also exhibit an altered permeability of the blood-brain barrier (Carnevale et al. 2012; Gentile et al. 2009; Faraco et al. 2016).
3 Treating Hypertension to Prevent Cognitive Decline and Dementia
Given the ample epidemiological and mechanistic evidence linking hypertension with lower cognitive abilities, a logical question emerges: does blood pressure control slow down cognitive decline and prevent dementia? A beneficial effect of antihypertensive medication on cognitive decline and dementia incidence is suggested by several observational studies with short/medium-term follow-ups (i.e. between 412 years) (Kohler et al. 2014; Tzourio et al. 1999; Qiu et al. 2003; Joas et al. 2012; Gelber et al. 2013; Khachaturian et al. 2006; Guo et al. 1999, 2001). Interestingly, Peila and colleagues suggested that the protective effect of antihypertensive therapy (after adjusting for age, education, APOE ε4 status, midlife and late-life BP) is proportional to its use: the longer the duration of treatment, the lower the risk of incident dementia and its subtypes (AD and VaD) (Peila et al. 2006). Despite these encouraging observations, the true effect of blood pressure lowering drugs on cognition can only be assessed by randomized placebo-controlled clinical trials. These are summarized in Table 3.
Table 3
Randomized controlled trials assessing the effect of antihypertensive drugs on cognitive decline and dementia
Trial | Participants | Follow-up | Intervention | Outcome | Main results |
|---|---|---|---|---|---|
SYST-EUR | n = 2902 with ISH, mean age 70 years | 4 years | CCB (nitrendipine) and/or ACEI (enalapril) and/or diuretic (HCTZ) vs. placebo | Dementia (AD, VaD and mixed) by DSM-III and MMSE | Significant dementia (AD and VaD) risk reduction by 55 % (95 % CI: 24–73 %) in treated group |
SHEP | n = 4376 with ISH, age >60 years | 4.5 years | Diuretic (chlorthalidone) with BB (atenolol) or reserpine vs. placebo | Dementia (Short-CARE test) | 14 % (95 % CI: −26 to 54 %) reduction in dementia (non significant) in treated group |
SCOPE | n = 4964 with ESH, age 70–89 years | 3.7 years | ARB (candesartan) with possible addition of diuretic (HCTZ) and/or open-label AH | Dementia and cognitive function (ICD-10 and MMSE) | Comparable incidence of dementia (all cause) and rate of cognitive decline between placebo and active treatment groups |
MRC | n = 2584 with ESH, age 65–74 years | 4.5 years | Diuretic (HCTZ) or BB (atenolol) vs. placebo | Change in cognitive function (measured by the PAL and TMT tests) | No significant effect of active treatment on change in cognitive function |
PROGRESS | n = 6105 with prior stroke or TIA, mean age 64 years | 3.9 years | ACEI (perindopril) with possible addition of diuretic (indapamide) vs. placebo | Dementia and cognitive decline (DSM-IV and MMSE) | Significant reduction in the risk of dementia with recurrent stroke by 34 % (95 % CI: 3–55 %) and of cognitive decline with recurrent stroke by 45% (95% CI: 21-61%) |
HOPE | n =9297 with vascular risk factors, age ≥ 55 years | 4.5 years | ACEI (ramirpil) vs. placebo | Stroke, TIA and cognitive function | Significant reduction in stroke-related cognitive decline by 41 % (95 % CI: 6–63 %) |
HYVET-COG | n = 3336 with ESH, age ≥ 55 years | 2.2 years | Diuretic (indapamide) with possible addition of ACEI (perindropil) vs. placebo | Cognitive decline and dementia (assessed by MMSE) | No significant difference in incident dementia rates between active treatment and placebo |
The Systolic Hypertension in Europe (SYST-EUR) trial was one of the first double-blind studies to show a significant reduction in the incidence of dementia (AD and VaD) due to antihypertensive treatment (Forette et al. 1998). This trial enrolled dementia-free subjects (with no prior history of stroke) who were 60+ years and had high SBP (160–219 mmHg). Active treatment consisted of nitrendipine, a calcium channel blocker (CCB), which could be combined with a diuretic (hydrochlorothiazide: HCTZ) and/or with enalapril, an angiotensin-converting enzyme (ACE) inhibitor. The goal of the study was to reach a threshold of SBP reduction of at least 150 mmHg. The SYST-EUR trial was stopped early after 2 years due to a significant reduction in stroke-related events (by ~40 % P < 0.001) in the treatment group. At this time point, nitrendipine also reduced the incidence of all cause dementia by 50 % (P = 0.05), including Alzheimer’s, vascular and mixed dementia cases. These findings were based on a total of 32 incident dementia cases (all forms). After the trial stopped, all participants (those previously treated and those in the placebo group) were invited to continue or begin treatment with the same BP-lowering regimen for another 2 years (Forette et al. 2002). Interestingly, this second phase showed that immediate antihypertensive therapy was more effective at reducing dementia risk compared to delayed treatment, supporting the observation from Peila et al. discussed above (Peila et al. 2006).
Despite comparable subject demographics to the SYST-EUR trial (participants aged 60+, SBP ranging from 160 to 219 mmHg) and a similarly reduced stroke risk (by 36 %), the Systolic Hypertension in the Elderly Program (SHEP) study did not find a significant protective effect of blood pressure lowering on dementia incidence, after a follow-up of 4.5 years (relative risk reduction; RRR: 14 %; 95 % CI: −26 to 54 %; P = 0.44) (SHEP Cooperative Research Group 1991). No distinction between AD and VaD dementia was made. It should be also noted that the antihypertensive regimen was different, consisting primarily of chlorthalidone (a diuretic). Thus, it is possible that for dementia prevention it may not be just about lowering blood pressure but also about the choice of antihypertensive drug. In particular CCBs of the dihydropyridine (DHP) type, as used in the SYST-EUR trial, could be beneficial in the context of Alzheimer’s disease given that sustained intracellular calcium elevations may promote Aβ production and tau phosphorylation (Paris et al. 2011; Green and LaFerla 2008; Yu et al. 2009). This suggests that DHP-CCBs may have additional protective effects on the brain other than decreasing BP.
In a subsequent report about the SHEP trial, it was observed that the non-significant cognitive and functional effects of the treatment might have been due to differential dropout between the groups; i.e. participants who missed the annual cognitive assessments tended to be older, to be in the placebo group and to have a higher occurrence of cardiovascular events. In other words, selective attrition may have biased the lack of significant differences between active treatment and placebo (Di Bari et al. 2001). Another important consideration is that due to ethical reasons, participants with high blood pressure in the placebo group also received open-label antihypertensive medication. These limitations put into question the observation that BP lowering does not affect dementia.
The Study on Cognition and Prognosis in the Elderly (SCOPE) was a prospective, double-blind trial designed to assess whether antihypertensive treatment with candesartan, an angiotensin-II type I (AT1) receptor blocker (ARB), was effective in reducing cardiovascular events, cognitive decline and dementia in elderly patients (70–89 years) with moderate high blood pressure (SBP 160–179 mmHg) (Lithell et al. 2003). Additional open-label antihypertensive drugs (i.e. HCTZ, diuretic, CCBs), were included as needed in both groups.
After 3.7 years of follow-up and after achieving considerable reductions in blood pressure and non-fatal stroke risk (by 27.8 %), no difference was noted in global cognition scores or dementia incidence between the two groups, assessed with the mini-mental state examination (MMSE) and ICD-10 criteria. Considering that by the end of the trial both candesartan and placebo groups had received additional AH medication and that both obtained significant BP reductions (from 166.0/90.3 to 145.2/79.9 in candesartan arm and 166.6/90/4 to 148.5/81.6 in placebo arm), it is possible that this could be masking any treatment benefit. The fact that a significant stroke risk reduction was seen with candesartan treatment (compared to placebo) could further highlight a protective cerebrovascular effect mediated through AT1 receptor antagonism. In addition, as the authors acknowledged, the rates of dementia incidence in the study cohort were lower (approximately 6.5 cases per 1000 patient years) compared to what would be expected for the age range (70–89 years) of the subjects, thus limiting the study’s power to detect a significant difference on the treatment. Notably, in a later report, SCOPE participants were segregated based on baseline cognitive function into high and low. This new analysis revealed that the incidence of dementia was higher in subjects with initially lower cognitive function, and importantly, that MMSE scores in the active treatment subgroup declined less than in the placebo group (Skoog et al. 2005).
Similarly, in a first report of the Medical Research Council (MRC) study, a subgroup of hypertensive subjects (n = 2584; age 65–74; SBP 160–209 mmHg, DBP < 115 mmHg) who received treatment with a diuretic (HCTZ + amiloride) or a beta-blocker (atenolol) were examined longitudinally with a neuropsychological testing battery, that included the paired associate learning and trail making tests, over a period of 4.5 years. These tests evaluate episodic memory and new learning as well as executive functions. The trial reported no difference in the rate of change of semantic memory and attention scores between the treated and placebo groups (Prince et al. 1996). When a subset of these patients (n = 387) were followed for an extended period of 9–12 years, it was found that poorer global cognition at follow-up was significantly associated to a smaller decline in SBP during the study period (Cervilla et al. 2000), suggesting that more extensive trial durations may be needed to detect significant differences in cognition.
The Perindopril Protection Against Recurrent Stroke Study (PROGRESS) trial evaluated the efficacy of monotherapy with perindopril (an ACE inhibitor), or in combination with indapamide (a thiazide-like diuretic), compared to other antihypertensive therapies in reducing the risk of cognitive decline and dementia in subjects with pre-existing stroke or transient ischemic attack (mean age 64 years). Contrary to the other studies discussed before, cognitive decline and dementia incidence (assessed by the MMSE and DSM-IV criteria) were primary outcomes in the PROGRESS trial analysis. After a mean follow-up of 3.9 years, the study showed a clear benefit of combination therapy in reducing the risk of post-stroke dementia and cognitive decline by 34 % (95 % CI: 3–55 %) and 45 % (95 % CI: 21–61 %) respectively, in individuals with recurrent stroke. Despite this positive observation, the effect was not significant in patients with dementia in the absence of repeated stroke events (Tzourio et al. 2003). In line with the PROGRESS study, the Heart Outcomes Prevention Evaluation (HOPE) trial revealed a significant 41 % reduction in stroke-related functional impairment (cognition, motor weakness, speech and swallowing) in patients with cardiovascular risk factors treated with an ACE inhibitor (ramipril) (Bosch et al. 2002). In the HOPE trial, an effect on dementia incidence was not evaluated.
A commonality of all previous studies was the testing of antihypertensive medication for cognitive protection in ‘young’ older adults (60–75 years). The double-blind, placebo-controlled Hypertension in the Very Elderly Trial (HYVET) included a cognitive assessment sub-study (HYVET-COG) to examine the benefit of treating very old hypertensive subjects (80+, SBP 160–200 mmHg; DBP < 110 mmHg), with no dementia at baseline (Peters et al. 2008). Participants received indapamide (a diuretic) with the possible addition of perindropil (ACE inhibitor) to reach a target blood pressure of <150/80 mmHg. The study had a short follow-up (mean 2.2 years) due to positive effects on stroke and total mortality reduction. At this time, although there was a risk reduction of 14 % in dementia incidence between active treatment and placebo, these differences were not statistically significant, likely due to the short follow-up and the low number of patients who had developed dementia.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
