Heart Health Meets Brain Health: The Role of Blood Pressure and Hypertension

Abstract

The prevalence and financial burden of Alzheimer’s disease and related dementias is rising and the lack of successful pharmacologic treatments necessitates the identification of risk factors and pathways to delay the onset of cognitive decline and reduce dementia risk. Blood pressure (BP) management and control has been identified as one promising area where data are inconclusive. This narrative review provides a summary of selected contemporary interventional and observational data on the associations between BP, BP control, and hypertension on brain health and discusses the methodological challenges and future considerations for this critical research area. Given the long prodromal phase of dementia and data supporting the importance of BP control earlier in the life course, longer-term interventional and observational studies in racially/ethnically diverse populations, with novel imaging and fluid-based biomarkers of neurodegeneration and vascular cognitive impairment to understand the pathophysiology, are needed to advance the field of BP and brain health.

Introduction

In the absence of effective pharmacological treatments to slow Alzheimer’s disease and related dementias (ADRD) progression, research over the last decade has focused on identifying risk factors and mechanisms that contribute to cognitive impairment and dementia. A particular priority has been to identify modifiable factors that can prevent or delay progression of cognitive decline and dementia. Evidence accumulates from observational studies that cardiovascular and cerebrovascular disease and contributing modifiable risk factors such as hypertension,^1,2 which in 2010 affected an estimated 1·4 billion people worldwide,^3,4 are associated with an increased risk of cognitive decline and dementia. Similar to dementia,^5–7 hypertension also disproportionally impacts minority populations, with Blacks shouldering the highest burden (~40%) compared to other racial/ethnic groups.⁸ This review focuses on hypertension and blood pressure (BP) lowering and control, which was identified by the recently updated Lancet Commission on Dementia Prevention, Intervention and Care to be one of 12 potentially modifiable risk factors that account for ~40% of worldwide dementias.⁹

The recent National Academies of Sciences, Engineering, and Medicine (NASEM) report on Preventing Cognitive Decline and Dementia: A Way Forward identified BP management as one of three promising interventions to prevent cognitive decline and dementia, but where data are encouraging but still inconclusive. The NASEM report aimed to synthesize data on the growing number of studies designed to prevent cognitive decline and dementia. The conclusions were partly based on a review – conducted specifically for the committee – by the Agency for Healthcare Research and Quality (AHRQ). The NASEM report concluded there was insufficient evidence on effective interventions for cognitive decline and dementia. However, the committee did identify three promising classes of intervention: cognitive training, BP management in people with hypertension, and increased physical activity. There is consistently strong evidence from observational data that BP control and lowering has beneficial effects, including reducing adverse cognitive outcomes. However, the report also concluded that data from randomized controlled trials (RCTs) do not offer strong support for the use of BP management in patients with hypertension to slow the progression of cognitive decline or prevent MCI and dementia.¹⁰ In both RCTs and observational studies, it was recommended that study design, sampling and selection, exposure/outcome ascertainment, and other sources of bias be taken into account when interpreting study results. Since publication of this report in 2017, additional studies have tested the effectiveness of BP management interventions. Therefore, our goal is to provide a narrative review of the recent evidence - both from RCTs and observational studies - on the role of BP, BP lowering, and hypertension control on the reduction of cognitive decline and delay of dementia, and discuss novel considerations for future research in this area.

Randomized controlled trials of BP lowering on prevention of cognitive decline and dementia

A key unanswered question is whether lowering BP reduces the risk of dementia or cognitive impairment and if it slows decline in cognitive function. This question has been formally tested through RCTs of BP lowering treatments and target levels, but with inconclusive results. In a recent meta-analysis of RCTs published through 2019, the authors were unable to identify an optimal BP range for dementia prevention.¹¹ Null results may be attributable to limited sample sizes, short follow-up (range: 2·2–4·3 years^12–14), or the inclusion of older participants who may not experience a cognitive benefit from BP lowering. Furthermore, dementia and/or cognitive decline was not a primary outcome in these trials, leading to concerns about type two error and limited power.

Since publication of the NASEM report in 2017, additional RCTs have augmented these extant findings (Table 1). The recent Heart Outcomes Prevention Evaluation-3 (HOPE-3) trial tested whether specific types of antihypertensive medications delay cognitive decline. HOPE-3 was a double-blind, randomized, 2×2 factorial design placebo-controlled trial of candesartan/hydrochlorothiazide, rosuvastatin, or their combination. Participants (n=1,626 from 21 countries) had a mean age of 74 years with at least one cardiovascular risk factor, and were followed for a median of 5·7 years.¹⁵ No significant benefit of antihypertensive treatment versus placebo on domain-specific cognitive decline was observed.¹⁵ This was despite a mean group difference in BP of 6·0 mmHg favoring the combination candesartan and hydrochlorothiazide therapy. A second RCT, the Intensive versus Standard Ambulatory Blood Pressure Control on Cerebrovascular Outcomes in Older People [INFINITY], uniquely examined the impact of BP control through two levels of targeted 24-hour average systolic blood pressure (SBP), based on ambulatory BP monitoring, on cognitive function over three years.¹⁶ Participants (n=199) with visible white matter lesions (mean age: 80·5 years) enrolled in a prospective open-label trial with blinded endpoints and were randomized to intensive versus standard (24-hour mean SBP: ≤130 mm Hg and ≤145 mm Hg, respectively) treatment with antihypertensive therapies. No differences in cognitive function were observed, but there was less progression of vascular lesions in the intensive treatment arm; white matter hyperintensity volume increases were smaller between baseline and follow-up. Despite no significant effect on cognitive function, this result is promising because it shows that intensive BP lowering may have intermediate effects on accumulation of white matter disease, which is a strong risk factor for cognitive decline¹⁷ and dementia.¹⁸

Table 1.

Contemporary randomized controlled trials of blood pressure lowering with cognition, cognitive decline, dementia, and subclinical MRI markers published between 2017–2020 (reverse chronological order).

Author	Study	Location	Sample	Treatment Randomization	Outcome	Median Follow-up Time	Summary of Main Results
Rapp et al. 202022	Systolic Blood Pressure Intervention Trial (SPRINT)-MIND	U.S. (including Puerto Rico)	2,921 participants; mean age: 68·4 years 37% women 30% black Inclusion criteria: participants with hypertension but without diabetes, dementia, or history of stroke	Blood pressure target 1:1 random assignment to SBP<120 mmHg (intensive treatment) versus <140 mmHg (standard treatment)	Primary: Composite memory and processing speed domain scores (individual test scores examined in secondary analyses) Exploratory: Composite language, executive function and global cognition domain scores	4·1 years	No significant difference in annual decline in memory composite score between intensive and standard treatment groups (between group difference= −0·004 (95% CI: −0·012, 0·004). Processing speed declined more in the intensive treatment group compared to standard treatment (between group difference= −0·010, 95% CI: −0·017, −0·002)
Williamson et al. 201920	SPRINT-MIND	U.S. (including Puerto Rico)	9,361 participants; mean age: 67·9 years; 35·6% women; 30% black Inclusion criteria: participants with hypertension but without diabetes, dementia, or history of stroke	Blood pressure target 1:1 random assignment to SBP<120 mmHg (intensive treatment) versus <140 mmHg (standard treatment)	Primary: Adjudicated probable dementia Secondary: Adjudicated MCI and composite MCI and dementia	5·11 years	No difference in incidence of probable dementia between intensive treatment group and standard treatment group (HR=0·83, 95% CI: 0·67, 1·04). Intensive blood pressure control was associated with a reduced risk of MCI (HR=0·81, 95% CI: 0·69, 0·95) and reduced risk of combined MCI and dementia outcome (HR=0·85, 95% CI: 0·74, 0·97)
Bryan et al. 201921	SPRINT-MIND	U.S. (including Puerto Rico)	670 participants (449 with repeat MRI scan); mean age: 67·3 years; 40·4% women 32·5% black Inclusion criteria: participants with hypertension but without diabetes, dementia, or history of stroke	Blood Pressure Target 1:1 random assignment to SBP<120 mmHg (intensive treatment) versus <140 mmHg (standard treatment)	Primary: Change in total white matter lesion volume from baseline Secondary: Change in total brain volume	3·4 years	Compared to standard treatment, intensive treatment was associated with a smaller increase in cerebral white matter lesion volume and a greater decrease in total brain volume
White et al. 201916	Intensive Versus Standard Ambulatory Blood Pressure Lowering to Prevent Functional Decline in the Elderly (INFINITY)	U.S.	199 participants; mean age: 80·5 years; 54% women; 6·5% black Inclusion criteria: diagnosis of systolic hypertension based on clinic and ambulatory blood pressure measurements, and had visible white matter hyperintensity lesions on MRI (typically ≥0·5% lesion volume corrected for intracranial cavity volume size). Excluded participants with uncontrolled diabetes, history of stroke or dementia, and clinically limited gait	Blood Pressure Target Randomization to a 24-hour mean SBP of ≤130 mmHg (intensive treatment) versus ≤145 mmHg (standard treatment) with antihypertensive therapies	Primary: White matter hyperintensity volume Secondary: Change in executive function	3 years	Smaller % changes in white matter hyperintensity volumes in intensive treatment (0·29%) compared to standard treatment (0·48%). No differences in executive function between the treatment groups
Bosch et al. 201915	Heart Outcomes Prevention Evaluation-2 (HOPE-3)	21 countries	2,361 participants; mean age: 74 years; 59% women Inclusion criteria: men ≥ 55 years of age and women ≥ 65 years of age with at least 1 cardiovascular risk factor OR women ≥ 60 years with 2 additional risk factors	Type of Antihypertensive medication 2×2 factorial design Randomization to candesartan (16 mg) plus hydrochlorothiazide (12·5mg) or placebo and to rosuvastatin (10 mg) or placebo	Primary: Change in Digit Symbol Substitution Test Secondary: Change in modified Montreal Cognitive Assessment and Trail Making Test, Part B	5·7 years	Treatment with candesartan plus hydrochlorothiazide, rosuvastatin, or their combination was not significantly associated with cognitive decline

Abbreviations: MCI=mild cognitive impairment, HR=hazard ratio, CI= confidence interval, MRI=magnetic resonance imaging, SBP=systolic blood pressure, CASI=Cognitive Abilities Screening Instrument

Some of the most encouraging recent evidence comes from the Systolic Blood Pressure Intervention Trial (SPRINT).¹⁹ Originally designed to test if a lower SBP goal (<120 mmHg) would reduce cardiovascular events in persons without diabetes or preexisting stroke, SPRINT-MIND (SPRINT Memory and Cognition in Decreased Hypertension) aimed to test the effect of intensive BP control (<120 mmHg), compared to a standard treatment goal (<140 mmHg), on the rate of probable dementia and MCI. SPRINT was successful at achieving a large spread between intensive and standard treatment groups (13·3 mmHg).²⁰ Despite suggesting a protective effect, there was no significant reduction in the incidence of probable dementia among participants in the intensive versus standard treatment groups over a median follow-up of 5·11 years (hazard ratio [HR]: 0·83, 95% confidence interval (CI): 0·67, 1·04),²⁰ but there was a reduced incidence of MCI (HR: 0·81, 95% CI: 0·69, 0·95) and a combined MCI/dementia (HR: 0·85, 95% CI: 0·74, 0·97) outcome. A substudy of SPRINT-MIND was conducted in 2,921 participants who underwent magnetic resonance imaging. Similar to INFINITY, SPRINT-MIND showed that intensive BP lowering, compared to standard treatment, was associated with smaller increases in white matter lesions, but also a small reduction in total brain volume.²¹

To date, few RCTs have examined treatment effects of BP lowering on domain-specific cognition. Another substudy of SPRINT-MIND evaluated the intervention effects on domain-specific cognitive function derived from multiple validated neuropsychological tests: memory and processing speed, were the primary domains of interest,²² with exploratory cognitive domains including language, executive function, and global cognition. The study found no significant differences in memory, language, executive function, or global cognition declines between the intensive and standard treatment groups.²² Greater declines in processing speed, over a median follow-up of 4·1 years, were observed in the intensive treatment group (−0·025 standard deviation (SD), 95% CI: −0·030, −0·019) compared to the standard treatment group (−0·015 SD, 95% CI: −0·021, −0·009). These differences were nominal and determined to not be clinically meaningful,²² particularly when compared to typical age-related cognitive decline which occurs at a rate of 0.04–0.05 SD units per year (~0.5 SD/decade) among older adults.^23–25 Overall, results from SPRINT-MIND for cognition are consistent with those documented in HOPE-3¹⁵ and INFINITY.¹⁶

Despite the null findings related to BP lowering benefits for cognitive function and probable dementia, the SPRINT-MIND Study has advanced the field beyond prior trials. First, the length of treatment (~3·3 years) and follow-up (~5 years) has been the longest among RCTs to date. Second, a lack of “state-of-the-art” criteria for cognitive assessment, including the use of at least four cognitive tests measuring multiple cognitive domains (at a minimum of two), was a notable weakness in many prior published trials.²⁶ Both INFINITY and SPRINT-MIND included a comprehensive battery of neuropsychological tests which meets this “state-of-the-art” criteria. However, an important and novel strength of SPRINT-MIND is that it is the only contemporary RCT of BP lowering to include expert adjudication of dementia and MCI, as compared to the commonly used Mini Mental State Examination (MMSE) test²⁷ or Montreal Cognitive Assessment (MOCA).²⁸ Neither the MMSE nor MOCA captures functional status, and although useful in detecting suspected cognitive impairment, the MMSE does not provide information on impairments in different cognitive domains. The MMSE also has well-documented ceiling effects²⁹ potentially rendering the test insensitive for detecting early or subtle cognitive impairments. Furthermore, the test is not sensitive to underlying brain pathology or normal fluctuations in test performance. Lastly, SPRINT-MIND tested much lower BP targets than prior studies. This informs how low of a target BP is beneficial for brain health in the absence of adverse sequelae and whether an even lower BP target adds benefit beyond current guidelines for BP thresholds for general cardiovascular benefit. By treating to a much lower target, SPRINT (and SPRINT-MIND) was able to demonstrate a BP spread between groups rarely seen in other studies of BP management, which may have contributed to the study’s overall success and findings for MCI.

Methodological Challenges in Randomized Controlled Trials

Although trial results showing a reduced risk of cognitive impairment and smaller progression of white matter disease with BP lowering have been encouraging, there are several considerations when interpreting results.¹⁰ The first is length of follow-up. Contemporary RCTs have exceeded the follow-up of past trials; however, given the latency of dementia, and the mounting evidence suggesting that it takes decades to develop clinically manifest symptoms,³⁰ longer trial follow-up is needed. In the absence of longer follow-up, important clinical endpoints for targeted interventions may be missed. Loss-to-follow-up and crossover between treatment and placebo groups also continue to be limitations.³¹

The characteristics of the study population in RCTs is important with most trials recruiting participants of older ages, and in some instances, although non-demented, participants have some level of cognitive impairment. Older individuals, whose underlying pathology is potentially too advanced, may not therefore benefit from BP lowering. In fact, there is evidence to suggest that intensive BP lowering at older ages may pose benefits, but also harms,^32,33 and should be considered carefully in the context of each patient’s age, health status, and comorbidities.³⁴ Furthermore, only some of the blood pressure trials have included diverse ethno-racial groups. SPRINT had ~30% Blacks, but this representation was lower in the INFINTY (~6·5%) and HOPE-3 (~2%) trials. HOPE-3 was the only blood pressure trial to include a large Hispanic (~27%) sample, which similar to Blacks have a higher prevalence of dementia compared to non-Hispanic Whites.³⁵ Despite improvements in the incorporation of racially/ethnically diverse populations in blood pressure trials, there continues to be narrow representation across all racial/ethnic groups. This has limited the ability to adequately understand the sources of race/ethnic disparities in dementia and suggests a significant need for future studies to continue to make efforts to include individuals from racial, ethnic, and socioeconomically diverse backgrounds.^36,37

Lastly, ascertainment methods for cognition and dementia are inconsistent. A comprehensive neuropsychological battery is not always administered, with most studies including only a single test for cognition or a dementia screening tool. SPRINT-MIND was the only contemporary study to incorporate a formal adjudication process, with multiple stages of testing, for probable dementia and MCI. These key methodological challenges highlight the continued need to complement data from RCTs with data from observational studies, particularly longitudinal population-based data, that can, in part, address these critical limitations.

Observational data on BP levels/control and hypertension for prevention of cognitive decline and dementia

Observational studies are desirable to characterize the prospective associations of BP levels and control on cognitive outcomes across the life course. Cross-sectional data suggest that higher BP levels are associated with lower cognitive performance and cognitive impairment,¹ but these associations are likely confounded by demographic and socioeconomic factors. The availability of more recent longitudinal data have extended these cross-sectional findings.

Association of Blood Pressure and Hypertension and Cognition and Cognitive Decline

Longitudinal studies of cognitive change were previously limited by short follow-up - often <5 years - and frequently did not span the mid- to late-life transition period, a critical time in the life course in which subclinical changes in the brain are most likely to occur. More recently, several large population-based U.S. and international cohort studies have found greater rates of decline in global and domain-specific cognition associated with midlife hypertension or BP levels (Table 2). Midlife hypertension was associated with global cognitive decline during the mid- to late-life transition over eight years in the Reasons for Geographic and Racial Differences in Stroke (REGARDS) Study³⁸ and over 20 years in the Atherosclerosis Risk in Communities (ARIC) Study.³⁹ In the Coronary Artery Risk Development in Young Adults (CARDIA) study, midlife hypertension was related to accelerated cognitive decline over five years during midlife.⁴⁰ This was the first study to show that midlife elevated BP or hypertension has measurable impacts on cognitive change that appear as early as in midlife. Vascular risk factors may be as important when measured in young adulthood as in mid- or late-life. Children (mean age: 10·8 years) from the Cardiovascular Risk in Young Finns Study were followed from childhood; measured elevated SBP in childhood, adolescence, and young adulthood were all associated with lower visual and episodic memory and visuospatial learning in midlife.⁴¹ These data shed light on the potential importance of targeting BP levels even earlier than midlife.

Table 2.

Contemporary prospective observational studies of baseline and changes in blood pressure/blood pressure control with cognition, cognitive decline, and dementia published between 2014–2020 (reverse chronological order).

Author	Study	Location	Sample	Exposure	Outcome	Follow-up Time	Summary of Main Results
Studies of Cognitive Function and Cognitive Decline
Levine et al. 202066	BP-Cog (ARIC, CARDIA, CHS, Framingham Offspring, NOMAS)	U.S.	34,349 black and white participants; mean age: 59·8 years; 55% women; 19·9% black	Midlife Time-dependent cumulative mean of all SBP measurements before cognitive measurement	Global cognitive performance, memory (learning and delayed recall/recognition, and executive function (complex and/or speeded cognitive functions)	Median: 12·4 years	Cumulative mean SBP associated with faster declines in global cognition, memory, and executive function
Hestad et al. 202048	Tromso Study	Norway	4,465 participants; mean age: 60·3 years; 55% women	Late-Life Single continuous measure of SBP	Cognitive function assessed with 3 tests: Mini Mental State Examination (MMSE), Digit Symbol Test, and the Twelve-Word Test	8 years	Higher SBP associated with lower cognitive performance in men. No interaction by antihypertensive medication
Yaffe et al. 202040	Coronary Artery Risk Developing in Young Adults (CARDIA) Study	U.S.	2,675 black and white participants; mean age: 52 years; 57% women; 45% black	Midlife Baseline HTN	Accelerated cognitive decline on a composite cognitive score (defined as race-specific decline ≥ 1·5 SD	5 years	Midlife HTN associated with accelerated composite cognitive decline over 5 years
Lane et al. 201976	Insight 46 (1946 British Birth Cohort)	U.K.	499 dementia-free participants; mean age: 70·7 years; 49% women	Late-Life	Exploratory outcomes of domain-specific cognitive function: Wechsler Logical memory and Face-Name Associative Memory Exam (episodic memory), WAIS Digit Symbol Substitution Test (processing speed), Mini-Mental State Examination (global cognition)	1–33 years	Neither SBP or DBP at any age or blood pressure change was associated with cognitive function
Levine et al. 201938	Reasons for Geographic and Racial Differences in Stroke (REGARDS)	U.S.	22,164 black and white participants; mean age: 642 years; 56% women; 38% black	Midlife Single continuous measure of SBP, DBP, pulse pressure, mean arterial pressure	Primary: Change in global cognition assessed using the Six-item Screener Secondary: Change in new learning, verbal memory, executive function	Median: 8·1 years	High SBP, low DBP, and high pulse pressure associated with faster decline in global, but not domain-specific, cognition. Blacks and men had faster rates of decline.
Yano et al. 201864	Atherosclerosis Risk in Communities (ARIC) Study	U.S.	11,408 black and white participants; Mean age: 54·3 years; 56% women; 21% black	Midlife Changes in BP estimated as the average real variability in SBP	Baseline and change in global cognition z score over 14 years	∼15 years	SBP variability associated with lower global cognitive function, but not cognitive decline. Higher SBP and DBP at baseline associated with global cognitive decline
Rovio et al. 201741	Cardiovascular Risk in Young Finns Study	Finland	2,026 children mean age: 10·8 years; 54·5% women	Early adulthood Area under the curve (AUC) for SBP to estimate long-term burden of BP	Principal component analysis of cognitive battery (included tests of executive function, visual and episodic memory, visuospatial associative learning, spatial working memory, reaction time, attention)	∼40 years	Elevated childhood, adolescence, and young adulthood SBP associated with lower visual and episodic memory and visuospatial learning in midlife
Hajjar et al. 201749	Health Aging and Body Composition Study	U.S.	1,657 cognitively intact older adults treated for HTN mean age: 73·7 years; 54·8% women; 47·3% black	Late-life Time-dependent SBP at each visit was classified as 120 mmHg or lower (SPRINT target), 121 to 139 mmHg (JNC-7 target), 140 to 149 mmHg (JNC-8 target), or 150 mmHg or higher	10-year cognitive trajectories in Modified Mini-Mental State Examination (3MSE) and Digit Symbol Substitution Test (DSST)	∼10 years	Participants with 150mmHg or higher SBP: greater declines in 3MSE and DSST compared to lowest declines in participants with SBP of 120 mmHg or lower
Goldstein et al. 201750	NIH-NIA Alzheimer’s Disease Centers (34 centers)	U.S.	1,029 participants 60 years and older with normal cognition at baseline; 68·7% women; 67·4% white	Late-Life Group 1: always normotensive (SBP <140 mmHg and DBP <90 mmHg) and taking antihypertensive medications Group 2: SBP 140–149 mmHg on 3 of 4 visits Group 3: SBP ≥ 150 mmHg on 3 of 4 visits	Changes in domain z-scores of attention, executive functioning, language, and memory Overall cognitive status based on the Mini-Mental State Exam (MMSE)	∼5 years	Compared to normotensive and SBP 140–149, SBP ≥ 150 mmHg had poorer cognitive status based on MMSE and had higher risk of conversion to MCI Participants with SBP 140–149 had poorer performance in executive functioning and attention
Studies of Dementia
Walker et al. 201962	ARIC	U.S.	4,761 dementia-free black and white participants; mean age at dementia diagnosis: 75 years; 59% women; 21% black	Midlife to Late-life Mid- to late-life BP patterns based on standard HTN definitions (BP value and/or use of antihypertensive medication): Group 1: midlife normotension and late-life normotension; Group 2: midlife normotension and late-life HTN; Group 3: midlife and late-life HTN; Group 4: midlife normotension and late-life hypotension; and Group 5: midlife HTN and late-life hypotension	Primary: Dementia based on expert adjudication and with additional information from informant interviews, telephone cognitive assessments, hospital discharge and death certificate codes Secondary: Mild cognitive impairment based on expert adjudication	Mean: 24 years	Compared to normotensive, midlife and late-life HTN and midlife HTN and late-life hypotension associated with higher dementia risk Sustained midlife HTN associated with higher dementia risk
Ma et al. 201957	The Rotterdam Study	Netherlands	5,273 dementia-free participants; mean age; 67·6 years 58·1% women	Midlife Within individual SBP variation calculated as the absolute difference in SBP divided by the mean SBP over two sequential visits (|difference|/mean), with a median 4·2 years apart	Dementia ascertainment based on consensus panel and electronic medical record linkage	Median: 14·6 years	Larger SBP variation associated with increased dementia risk Elevated dementia risk observed with large rise or fall in SBP
Gregson et al. 201956	UK Clinical Practice Research Database	U.K.	2,593,629 participants; median age: 54 years; 54·1% women	Midlife Continuous BP and time since BP measurement	Routine clinical record review of physician diagnosed dementia	Median: 8·2 years	For each 10 mmHg higher SBP, lower dementia risk BP not associated with AD BP associated with vascular dementia
Murray et al. 201860	Ibadan Dementia Project	U.S. Africa	1,236 black participants; mean age: 75·8 years; 73·3% women	Late-life Averaged BP and antihypertensive medication use	Dementia screening based on in-home clinic evaluation; laboratory, inpatient and outpatient encounters; coded diagnoses and procedures; or medication records	∼24 years	Treated vs. untreated HTN associated with reduced dementia risk Suboptimal treatment of BP not associated with dementia
Abell et al. 201855	Whitehall II Cohort	U.K.	8,639 participants; mean age: 50 years; 32·5% women	Midlife and late-life Continuous BP and HTN	Dementia ascertained from electronic medical records	∼30 years	SBP ≥ 130 mmHg at age 50, but not ages 60 and 70, associated with increased dementia risk Longer exposure to HTN associated with increased dementia risk
Peng et al. 201754	Health Improvement Network in the UK	U.K.	265,897 participants free of vascular dementia; Stratified by age cohorts: 60–65 years (53.3% women) 70–75 years (57% women)	Midlife and Late-life Continuous BP categorized into 4 groups: Normal, pre-HTN, Stage 1 HTN, Stage 2 HTN	Vascular dementia identified from recorded clinical diagnoses	Mean: 8·84 years	Higher midlife BP s associated with risk of developing vascular dementia in untreated No association between BP and vascular dementia among the treated group Late-life BP associated with dementia in treated group only
McGrath et al. 201753	Framingham Offspring Study	U.S.	1,440 participants; mean age: 69·6 years 52·6% women	Midlife and Late-life Midlife and late-life systolic HTN was defined as SBP ≥140 mmHg and diastolic HTN as DBP ≥90 mmHg Persistent HTN was defined as SBP≥140 mmHg and/or DBP ≥90 mmHg during both midlife and late-life Low late-life BP was defined as SBP <100 mmHg and DBP <70 mmHg	Dementia based on expert adjudication	Mean: 8 years	Midlife systolic HTN and persistence of systolic HTN (HR=1·96) into late-life associated with elevated dementia risk. In non-hypertensives/pre-hypertensives in midlife, a steep decline in BP associated with increased dementia risk
Gottesman et al. 201752	ARIC	U.S.	15,744 black and white participants; mean age: 57·4 years 57% women; 34·9% black	Midlife Presence of HTN (SBP ≥140 mmHg, DBP ≥90 mmHg, or use of antihypertensive medications) Presence of pre-HTN (not meeting HTN criteria SBP ≥120 mmHg, and DBP ≥80 mmHg)	Dementia based on expert adjudication and with additional information from informant interviews, telephone cognitive assessments, hospital discharge and death certificate codes	∼25 years	Midlife HTN and pre-HTN associated with a higher dementia incidence
Gabin et al. 201751	Health and Memory Study of Nord-Trøndelag Health Study (the HUNT Study)	Norway	24,638 participants; mean age: ∼57·9 years 54·8% women	Midlife Continuous SBP	Dementia ascertainment based on WHO algorithms and ICD-10 manual, and used retrospective electronic hospital records, clinical examinations and standard interviews	Mean: 17·6 years	Higher SBP associated with lower all-cause, mixed Alzheimer/vascular, and AD dementia regardless of treatment use Higher SBP associated with higher AD risk in treated participants <60 years
Corrada et al. 201759	90+ Study (Leisure World Cohort Study (LWCS))	U.S.	599 dementia-free participants; mean age: 93·2 years; 71% women; 1% black	Late-life Participant report of physician-diagnosed HTN and year of first diagnosis Age of onset categorized as: no HTN, <70 years, 70–79 years, 80–89 years, 90 years or older	Baseline cognitive status determined based on neurological examiners applying DSM-IV criteria and information from performance on the MMSE and 3MS Informants contacted when needed	Mean: 2.8 years	HTN developed between 80–89 years or 90 years or older had a lower dementia risk

Abbreviations: HTN=hypertension, HR=hazard ratio, CI= confidence interval, SD=standard deviation, MRI=magnetic resonance imaging, BP=blood pressure, SBP=systolic blood pressure, MMSE=Mini-Mental State Exam, 3MS=Modified Mini-Mental State Examination, WHO=World Health Organization, NINDS=National Institute of Neurological Disorders and Stroke, DSM-IV=Diagnostic and Statistical Manual 4^th Edition, ICD-9=International Classification of Diseases

Age-dependent associations of BP on cognition have been noted.¹ In a combined analysis of the Rotterdam Study and the Leiden 85-Plus Study, higher BP was associated with worse cognition among adults <75 years, but was associated with better cognition among adults >75 years.⁴² Because these are observational data, these findings may in part be explained by bias (survival bias, comorbidities associated with lower BP, or reverse causality (dementia leading to reduction in BP over time)). For example, survival bias is of particular concern in studies of cognitive aging since older participants with more comorbidities and poorer health, and consequently those at highest risk for poor cognitive function or dementia, will be less likely to survive to be examined in longitudinal studies.^43,44 This results in a biased overrepresentation of healthier adults with likely better cognitive performance and lower risk of dementia in longitudinal studies. Survival biases that are unaccounted for in analyses can therefore lead to a result showing better cognition with high BP and worse cognition with low BP. Because of these inherent risks of bias, investigations of late-life BP and BP control with late-life cognitive function have often been inconclusive⁴⁵ and as indicated by the Rotterdam and Leiden 85-Plus studies, may risk being interpreted as a protective effect of hypertension on cognitive function.⁴² Prior data from the Honolulu Asia Aging Study⁴⁶ and the Baltimore Longitudinal Study of Aging Study⁴⁷ also found lower cognitive performance among older participants with lower BP, but may suffer from similar sources of bias. However, other contemporary U.S. and international studies of late-life BP have replicated the inverse associations observed with midlife BP. A higher SBP in late-life was associated with lower cognitive performance in the Tromso Study.⁴⁸ Higher time-dependent late-life BP in the Health Aging and Body Composition Study, also showed a greater decrease in global cognitive function and executive function for participants with a SBP ≥150 mmHg or higher compared to ≤120 mmHg.⁴⁹ This finding was similarly replicated in a U.S.-based study of data aggregated across 34 of the NIH/NIA Alzheimer’s disease centers.⁵⁰

Association of Blood Pressure and Hypertension and Risk of Dementia

Several contemporary studies have examined the relationship between midlife and late-life hypertension and BP on dementia risk (Table 2), with associations stronger in midlife compared to late-life. Age-dependent associations have also been documented for dementia risk. Prior evidence suggests that midlife elevated BP or hypertension are associated with a higher risk of dementia.^51–57 The data on late-life measures of BP and hypertension remains inconsistent^{54,55,58–60} and potentially subject to issues of reverse causality due to lower levels of BP with severe brain atrophy later in life, and selection bias with the healthiest of individuals living long enough to participate in a study.⁶¹ The 90+ study was the first to report on the association between age of onset of hypertension with dementia risk among the oldest-old, or individuals 90 years or older.⁵⁹ Among 599 dementia-free participants, participants who developed hypertension between 80–89 years of age had a lower risk of dementia compared to participants without hypertension.⁵⁹ This risk was even lower for participants who developed hypertension at age 90 years or older. Lower BP in late-life, however, may reflect underlying ischemic heart disease, heart failure, frailty or dementia, all of which could cause associations with worse cognitive function. The associations may also be dependent on an individual’s history of hypertension, as documented in the ARIC⁶² and Age, Gene/Environment Susceptibility (AGES)-Reykjavik studies,⁶³ further emphasizing the importance of a life-course approach to this study question. It has been hypothesized that individuals with a history of hypertension may already have extensive end organ damage requiring a higher BP to sustain cerebral blood flow perfusion, particularly in the periventricular areas of the brain. However, there is as yet no evidence to support this hypothesis. An alternative explanation may be that spontaneously lowering BP in later life in a patient with a history of hypertension reflects a failure to maintain BP levels, e.g. due to heart failure or frailty. There continues to be speculation over the exact mechanisms relating low BP in older adults to adverse cognitive outcomes, but additional replication of the hypothesized U-shaped association is needed, taking into consideration population (e.g. age, race/ethnicity) subgroups and comorbidities. Furthermore, elevated blood pressure, even in older age, remains an important risk factor for stroke and cardiovascular disease, and its treatment is known to reduce these sequelae. Considering the impacts of stroke and cardiovascular disease (acute myocardial infarction, heart failure, and resultant arrhythmias) on cognition as well, there may be indirect consequences on cognition of hypertension, beyond those discussed in this review.

Time-varying Blood Pressure and Hypertension Assessments and Risk of Cognitive Decline and Dementia

Longitudinal studies are uniquely positioned to examine time-varying BP-levels and hypertension changes in relation to cognitive outcomes. In an analysis of 11,408 participants from the ARIC study, greater visit-to-visit BP variability during the mid- to late-life transition was associated with lower cognitive function – but not cognitive decline - over ten years.⁶⁴ These findings were replicated in the CARDIA study for BP variability estimated over 25 years during midlife with a single measurement of cognitive function.⁶⁵ In a pooled analysis of five independently reported U.S. population-based studies (ARIC, CARDIA, Cardiovascular Health Study, Framingham Offspring Study, and Northern Manhattan Study),⁶⁶ higher time-dependent cumulative mean SBP was associated with faster rates of decline in global cognition, memory, and executive function. Unique to this pooled analysis was the ability to have sufficient power to examine BP-cognition differences by race subgroups. Levine et al. determined that the observed differences between Blacks and Whites in cognitive decline was no longer evident after adjusting for cumulative BP level, suggesting that racial disparities in cognitive change are likely in part explained by differences in BP and its control.⁶⁶

Time-varying measures of BP and hypertension have also been associated with dementia risk. In a sample of 5,273 dementia-free participants from the Rotterdam Study, large variation in SBP was associated with increased dementia risk over a median of 4·2 years.⁵⁷ A higher risk of dementia was also evident with both large rises (HR= 3·31, 95% CI: 2·11, 5·18) and large falls (HR=2.20, 95% CI: 1·33, 3·63) in SBP over this same time period. Hypertension chronicity, estimated from BP patterns during the mid- to late-life transition, was associated with dementia risk in the ARIC Study.⁶² Compared to normotensives, sustained hypertension in midlife (HR=1·41, 95% CI: 1·17, 1·71) and hypertension in both mid- and late-life (HR=1·49, 95% CI: 1·06, 2·08) was associated with elevated dementia risk over 24 years. However, participants who had hypertension in midlife, but hypotension in late-life, had an even higher risk of dementia compared to normotensives (HR=1·62, 95% CI: 1·11, 2·37).⁶² Data from the Framingham Offspring Study corroborate these findings in ARIC related to persistent hypertension in mid and late-life, but also showed that among non-hypertensive participants in mid-life, a steep decline in blood pressure during the mid- to late-life transition was associated with a 2-fold higher risk of dementia.⁵³ These findings may be attributable to declines in blood pressure as a function of comorbidities (e.g. heart failure) or may be due to the impacts of vascular aging and resulting decline in vascular compliance and elasticity leading to impaired cerebral autoregulation and hypoperfusion.⁶⁷ These studies support the continued need for a life course approach to characterizing BP effects on brain and cognitive health to shed light on the optimal period of implementation and intensity of future BP interventions.

Class of Anti-hypertensive Medication Use and Risk of Cognitive Decline and Dementia

Trials to date, including SPRINT-MIND, were not designed to test the effect of type of antihypertensive medication and the data from observational studies are limited. A recent harmonized meta-analysis of six prospective community-based studies examined five antihypertensive medication classes (angiotensin-converting enzyme inhibitors, angiotensin II receptor antagonist, beta blockers, calcium channel blockers, and diuretics) and found that although any antihypertensive medication use, compared to none, was associated with a lower risk of dementia (HR=0·88, 95% CI: 0·79, 0·98), there was no evidence that one medication class was more effective over another.⁶⁸ A second meta-analysis included over 27 studies and 50,000 participants and found no relationship between anti-hypertensive medication class and incident cognitive decline and dementia.⁶⁹ Diuretics did show a benefit in some analyses, but the associations were inconsistent across follow-up, comparator group, and outcome.⁶⁹ Differences by race/ethnic subgroups were not explored in either study. Duration, intensity, and adherence to antihypertensive medication use was also not examined. Reports have shown discontinuation of antihypertensive medication use⁷⁰ and low adherence⁷¹ to be associated with a higher dementia risk. This finding may be particularly relevant in underrepresented minority groups who often have less access to adequate care for BP management and other conditions⁷², emphasizing the importance of including these subgroups in studies to inform clinical guidelines on the use of antihypertensive medications for prevention of dementia.

Methodological Challenges in Observational Studies

As with RCTs, observational studies are subject to their own limitations. Although they overcome issues of study design and length, other sources of bias become a concern. Longer follow-up can lead to informative attrition and selection biases, particularly when cognitively impaired participants are less likely to return for study visits. Established analytic methods, including inverse probability of attrition weighting, multiple imputation, and shared parameter models, may address this bias, but are rarely used in longitudinal analyses of BP and cognition.^39,62,64 As is the case with RCTs, the heterogeneity in exposure and outcome ascertainment (Table 2) poses challenges when synthesizing data. In some studies, dementia was ascertained via a formal adjudication process, while in others determined through dementia screening measures such as the MMSE and MOCA which are often insensitive, electronic medical record review, or physician diagnosis of dementia.

Despite recent improvements, there is a continued need for observational studies to include diverse study populations to adequately identify and address race/ethnic disparities in dementia. Recent data from the Health and Retirement Study showed differential effects of hypertension across racial/ethnic groups in dementia risk,⁷³ suggesting that high BP may partially explain the observed racial and ethnic disparities in dementia. Greater efforts should be focused on the recruitment and retention of racially/ethnically diverse samples in both RCTs as well as observational studies.⁷⁴ With underrepresented minorities shouldering the greatest burden of both hypertension and dementia, this is an urgent methodological issue that requires immediate action.

In summary, observational studies, despite their limitations, lend themselves to examining a life course approach to BP and brain health, with an opportunity to answer many critical questions, including should age-specific thresholds for BP be considered; at what age should treatment be initiated; how aggressive should treatment be to optimize brain health in late-life and reduce adverse cognitive outcomes; and how do age-specific BP thresholds and treatment effects differ across population subgroups and comorbidities?

Mechanisms linking blood pressure levels/control and hypertension to cognitive outcomes

Although the pathophysiology underlying BP and the brain is beyond the scope of this review, both animal and human studies provide sufficient evidence to support two biologically plausible pathways (Figure 1), one focused on cerebrovascular pathologies and disease and the other AD pathology. Evidence from observational studies show that elevated BP results in increased inflammation and oxidative stress, functional and structural vascular changes, and vascular dysregulation leading to several intermediate pathways to dementia, including, cerebral small vessel disease and stroke.^63,75–77 BP may also lead to neurodegeneration, specifically a reduction in brain volumes.^63,76–78 However, whether BP and BP changes relate to neuropathological hallmarks of AD, including beta amyloid (Aβ)^76,79–83 and tau^84,85 burden has been more inconclusive with a need for more compelling data to support this mechanistic pathway.

Figure 1.

Proposed mechanisms linking blood pressure and hypertension to cognitive decline and dementia.

To advance this field, conventional and novel imaging measures of cerebral small vessel disease (e.g. arteriolosclerosis, cerebrovascular reactivity, skeletonized mean diffusivity, optical coherence tomography angiography⁸⁶) and novel fluid-based biomarkers of Aβ, tau, neurodegeneration, and vascular cognitive impairment (e.g. markers of endothelial injury and inflammation, neurofilament light being used in the MarkVCID Consortium⁸⁷), will be critical to continue to understand the mechanisms and sequence of vascular and neuropathological changes prior to dementia that are a result of long-term BP levels and BP control. Specifically, when these biomarkers, which are cost-effective and non-invasive, are further validated, they can be included in future studies to improve our understanding of potential pathways and impacts of BP lowering in the prodromal phase of dementia where primary and secondary prevention efforts are critical.

Hypertension Treatment Guidelines and Population-Level Public Health Initiatives for Promoting Brain Health

Public health recommendations for anti-hypertensive treatment guidelines for general and cardiovascular health benefit exist, but specific BP treatment guidelines to optimize brain health are not in place. Despite the methodological limitations of observational and interventional studies stated in this review, and no clear consensus on the optimal BP targets for brain health, initiatives to promote cognitive health and reduce the risk of ADRD through vascular risk factor control, specifically, continue to remain an important opportunity for primary prevention of ADRD. Several ongoing U.S.-based and international initiatives have been developed to ensure that accumulating evidence-base from the research community is distributed to the broader public health community (Panel 2). Global public health initiatives have ranged from recommended tools to monitor and manage hypertension (HOPE Asia), public health educational campaigns (Mind your Risks®, AARP Global Council on Brain Health), provider-level guidelines to improve detection and management of hypertension (Health Matters), and multimodal approaches to improve global heart health (Global Hearts Initiative, Million Hearts®). As we continue to refine our understanding of BP and brain health from observational and interventional studies, the existing infrastructure of these initiatives will be essential to quickly and effectively disseminate clinically relevant findings to the general population.

Panel 2:

Public Health Initiatives for Disseminating Research on Promoting Cognitive Health and Reducing the Risk of Alzheimer’s disease and related dementias

Public Health Initiative	Location	Description	Access for additional information
Hypertension, brain cardiovascular, and renal Outcome Prevention and Evidence in Asia (HOPE Asia)	Asia	Group of experts from 12 countries/regions across Asia, part of the World Hypertension League, focused on the management of hypertension in Asia. The goal of the HOPE Asia Network is to improvement the management of hypertension and organ protection toward achieving “zero” cardiovascular events in Asia. This work has culminated in publications and consensus recommendation in 3 key areas for hypertension in the region: Asian characteristics, home blood pressure monitoring, and ambulatory blood pressure monitoring.	Kario K. The HOPE Asia Network activity for “zero” cardiovascular events in Asia: Overview 2020. J Clin Hypertens (Greenwich) 2020; 22(3): 321–30. Turana Y, Tengkawan J, Chia YC, et al. Hypertension and Dementia: A comprehensive review from the HOPE Asia Network. J Clin Hypertens (Greenwich) 2019; 21(8): 1091–8.
Health Matters	U.K.	A professional resource that outlines how providers and commissioners can reduce the population average blood pressure through improved prevention, detection, and management	https://www.gov.uk/government/publications/health-matters-combating-high-blood-pressure/health-matters-combating-high-blood-pressure
Million Hearts®	U.S.	National initiative by the Centers for Disease Control and Prevention (CDC) and the Centers for Medicare & Medicaid Services (CMS) to prevent 1 million heart attacks and strokes within 5 years. It focuses on implementation of a small set of evidence-based priorities and targets (including through hypertension control, cholesterol management, physical activity, and tobacco use) that can improve cardiovascular health	https://millionhearts.hhs.gov/
Mind Your Risks®	U.S.	Public health campaign by the National Institute of Neurological Disorders and Stroke (NINDS) that educates people with high blood pressure about the importance of controlling blood pressure in midlife to reduce the risk of stroke and possible dementia in late- life	https://www.mindyourrisks.nih.gov/
Centers for Disease Control (CDC) Healthy Brain Initiative: A Road Map for State and Local Public Health	U.S.	Creates and supports partnerships, collects and reports data, increases awareness of brain health, supports populations with a high burden of Alzheimer’s disease and related dementias, and promotes the use of its Road Map series which seeks to ensure health equity, collaboration across multiple sectors, and leverage resources for sustained effect	https://www.cdc.gov/aging/healthybrain/index.htm
AARP Global Council on Brain Health	U.S.	Committee of experts summarize evidence-based information in reports and fact sheets to the public on how individuals can maintain or improve brain health	https://www.aarp.org/health/brain-health/global-council-on-brain-health/
Alzheimer’s Association: Risk Reduction Messaging for Health Education, Accelerate Risk Reduction and Promote Brain Health	U.S.	Resource for the public community aggregating information on current briefs/reports, results from population-based research through reviews and papers, and additional featured resources across the local, state, and federal level	https://www.alz.org/professionals/public-health/core-areas/brain-health-risk-reduction
Global Hearts Initiative	International	Partnership between the World Health Organization (WHO) and the CDC. The initiative includes 5 focal areas identified as high-impact, evidence-based interventions that will have the greatest impact at improving global heart health: (1) MPOWER: tobacco control (2) ACTIVE: increasing physical activity (3) SHAKE: salt reduction (4) REPLACE: eliminate trans fat (5) HEARTS: management of cardiovascular disease in primary care	https://www.who.int/cardiovascular_diseases/global-hearts/en/
Alzheimer Disease International (ADI)	International	International federation of Alzheimer and dementia associations around the world in conjunction with the WHO. Based on the WHO global action plan on the public health response to dementia, ADI works across countries to implement and fund dementia action plans and monitor their effectiveness. Publisher of the World Alzheimer Report	https://www.alzint.org/

Conclusions and Future Directions

The growing burden of dementia, accompanied by the lack of successful pharmacological treatments, necessitates the identification of modifiable factors that can be readily intervened upon to prevent cognitive decline and dementia. BP lowering was determined to be one area of focus with a potential for success at preventing cognitive decline and dementia. However, the observational and RCT data still have not reconciled the optimal timing of BP lowering, BP target levels, and population subgroups (underrepresented minorities at higher risk for hypertension and dementia) expected to benefit most from BP lowering.

Despite these unanswered questions and in lieu of pharmacological therapies, lifestyle and environmental factors, including BP lowering, continue to be the most promising avenues to address the increasing burden of dementia. Moreover, some data suggests that multimodal interventions of lifestyle factors, that would include BP lowering, may be even more successful at reducing cognitive decline and dementia risk.⁸⁸ The Finnish Geriatric Intervention Study to Prevent Cognitive Impairment and Disability (FINGER) trial⁸⁹ included a multimodal intervention on lifestyle and dietary modification and vascular risk factor monitoring, specifically BP. FINGER found less cognitive decline among individuals who received this combined intervention, as compared to standard care,⁸⁹ although reported differences were of small clinical magnitude. In comparison, other multimodal interventions, such as MAPT⁹⁰ and Pre-DIVA,⁹¹ have shown no benefit on reducing cognitive decline or dementia risk, suggesting the effectiveness of multimodal interventions still requires further investigation. The World-Wide FINGERS Network will continue to test similar trial designs in a range of international sites.⁹²

In conclusion, although results from some RCTs and observational studies on the benefits of BP control and lowering are encouraging, they continue to be inconclusive. The observational evidence suggests the optimal timing of intervention is likely in midlife, which is difficult to assess in RCTs due to the length of follow-up needed. In addition, identifying the optimal level of BP is an added challenge beyond simply banning a particular treatment or product that may cause harm. Further inquiry is needed to address several methodological considerations, including, the lack of diverse study populations, population-subgroup specific optimal BP thresholds for cognitive benefit, timing of optimal BP control, duration of study follow-up, incorporation of novel intermediate imaging and fluid-based biomarkers, and use of validated cognitive function and dementia ascertainment methods. Given the long prodromal phase of dementia and data supporting the importance of BP control earlier in the life course, longer-term interventional and observational studies in racially/ethnically diverse populations, with novel imaging and fluid-based biomarkers of neurodegeneration and vascular cognitive impairment to explain underlying pathology, are needed to advance the field of BP and brain health.

Panel 1: Search Strategy and Selection Criteria.

We sought to provide a comprehensive narrative review of recent data on blood pressure/hypertension, blood pressure control and lowering on cognition and dementia from human studies only since the publication of the 2017 National Academies of Sciences, Engineering, and Medicine report on Preventing Cognitive Decline and Dementia: A Way Forward report. Our literature search focused on articles from January 1, 2017 to November 1, 2020. PubMed was searched without language or subject restrictions. Controlled vocabulary (MeSH and emtree), free text, and keywords were used in the search strategy. Our search involved several combinations of initial keywords, including: (‘blood pressure’ OR ‘hypertension’ OR ‘blood pressure control’) AND (‘cognition’ OR ‘dementia’ OR ‘cognitive impairment’) AND (‘epidemiology’ OR ‘clinical trial’ OR ‘observational study’). Observational studies and clinical trials that report on key clinical outcomes (i.e. cognition and cognitive decline, mild cognitive impairment, and Alzheimer’s disease and related dementias) were included. We also thoroughly reviewed references from published manuscripts and reviews to identify additional studies. This narrative review is not a systematic review and meta-analysis and may not be all-inclusive of all published articles. The final list included references selected to illustrate the objectives of this review, specifically, a narrative review on contemporary clinical trials and observational studies reporting original results on the associations between blood pressure and cognition/dementia outcomes.