Hypertension and Cognitive Decline: Impact on Executive Function

Abstract

Hypertension is a risk factor for mild cognitive deterioration and vascular dementia. Cognitive deterioration attributable to normal aging was distinguished from cognitive deterioration related to hypertension by means of neuropsychologic tests. Sixty hypertensive patients, aged 65‐80 years, were compared with 30 normotensive individuals. Patients with a history of stroke and/or transient ischemic attacks, diabetes, atrial fibrillation, hypercholesterolemia, or bypass surgery, and those diagnosed with dementia, depression, or anxiety, were excluded. Neither gender differences, duration of hypertension (10.2±8.2 years), nor prescribed antihypertensive drug treatment had an influence on study results. Immediate recall was impaired in both groups. The hypertensive patients evinced impairment in all tests vs. the normotensive subjects. Mean deferred recall scores ± SD were 5.68±2.6 vs. 7.13±2A; p<0.01. Deficits in attention speed and executive function, as measured by nonperformance on the Trail Making Test Part B, were present in 46% of hypertensive patients vs. 13% of normotensive patients (p<0.005), with more errors made by the hypertensive patients (1.15±1.54 vs. 0.46±0.9; p<0.02). Scores on the Stroop Color and Word Test also revealed deficits in the hypertensive patients (24.7±7.6 vs. 32±10.7; p<0.005). Compared with the control group, the hypertensive participants revealed more deficits in skills involving delayed recall and prefrontal‐region skills. The relevant neuropsychologic tests were sufficiently sensitive and proved easy to use in clinical practice.

A total of 9.7% of the population in Argentina is older than 65 years of age, with seniors over 81 years old accounting for 1.8% of the population. ¹ Consistent with trends in other countries, Argentinean projections estimate a 9% increase in the elderly every 5 years. ² These demographic changes present challenges. Hypertension, with a prevalence of 60%, ³ is one of them, and dementia, with its exponential ratio to age—its incidence amounting to 6% at the age of 70 and as much as 40% in seniors over 90 ⁴ —is another.

Dementia is the primary cause of behavioral deterioration worldwide. This disabling syndrome is characterized by behavior disorders and by a progressive and irreversible decline in higher intellectual functions. The functional and anatomic changes produced by high blood pressure (BP) on the brain's vascular system constitute a modifiable risk for vascular dementia and may play a significant role in Alzheimer's disease. ⁵ , ⁶

Several studies have established a relationship between hypertension and cerebrovascular disease (CBVD) on the one hand (Systolic Hypertension in the Elderly Program [SHEP], ⁷ Perindopril Protection Against Recurrent Stroke [PROGRESS] ⁸ ), and cognitive deterioration (Kilander et al., ⁹ Study on Cognition and Prognosis in the Elderly [SCOPE], ¹⁰ the Feedback of Outcomes to Users and Staff [FOCUS] study ¹¹ ) and vascular dementia (Systolic Hypertension‐Europe [Syst‐Eur] Trial ¹² ) on the other.

At present, there is no clear evidence that cognitive deterioration of a vascular etiology is a prelude to vascular dementia in hypertensive patients. It may, nonetheless, be a cause, a contribution, or a concurrent factor. ¹³ , ¹⁴ Cognitive deterioration is a manifold syndrome involving memory as well as other cognitive domains. Cognitive deterioration may increase the risk of developing dementia by a factor of 5.7. ¹⁵ Cognitive decline, though less severe than dementia, is part and parcel of aging, and may be difficult to differentiate from the impairment resulting from a subcortical vascular lesion. Subjective memory loss is the first symptom of cognitive decline and the most frequent complaint in patients over 70, but only 20% of cases may be attributable to CBVD. Anxiety and/or depression may account for 40% of cases. ¹⁶

In recent years, neuropsychology has recognized the impact of vascular pathology on cognitive function, as well as the importance of early detection and prevention. The purpose of the present study was to identify the effect of vascular pathology—and more specifically, hypertension—on cognition, and to differentiate this effect from the deterioration produced by a normal aging process.

Participants

Out of a total of 520 consecutive patients of both sexes who were treated in a cardiology office, individuals aged 65‐80 with hypertension were selected. All patients invited to participate accepted the offer. Exclusion criteria included previous stroke and/or transitory ischemic attacks (determined by history and through clinical examination); diabetes mellitus (according to the standards of the American Diabetes Association ¹⁷ ); hypercholes‐terolemia (defined as the use of cholesterol‐lowering drugs, low‐density lipoprotein >160 mg/dL, or non‐high‐density lipoprotein >190 mg/dL); atrial fibrillation; coronary bypass surgery; or a diagnosis of anxiety, depression, or dementia (with Folstein's Mini Mental State Examination [MMSE] [Psychological Assessment Resources, Inc., Lutz, FL] score according to Diagnostic and Statistical Manual of Mental Disorders IV ¹⁸ lower than 24). A group of 60 patients was compared with 30 normotensive patients (BP <139/89 mm Hg). Patients were considered to be hypertensive if their BP was > 140/90 mm Hg ¹⁹ at office visit and/or on three occasions throughout their clinical history, or if they were taking specific antihypertensive medication. Participants underwent a baseline evaluation consisting of the following: clinical history, risk factors, anthropometric indices, physical examination, 12‐lead resting electrocardiogram, and baseline BP readings following a standardized technique (seated, after a 5‐minute rest, and with an average of three measurements) using a mercury sphygmomanometer. Benzodiazepine use was discontinued 72 hours before examination.

Neurocognitive Tests

A group of previously trained cardiologists administered a neuropsychologic test battery to the participants. A blind interpretation of such tests was then performed by specialized neuropsychologists. These were the only professionals in the study who were blinded. The battery of tests included the following: 1) the MMSE ²⁰ with a standard cut point of 24 points; 2) the Trail Making Test, Parts A and B, in written and oral versions, ²¹ , ²² used to assess mental flexibility, speed of attention, executive function (the ability to plan a task, perform it systematically, and adapt it upon receiving new information), and potential brain damage. Written tests lasting over 20 seconds (Part A) or 30 seconds (Part B) were considered abnormal, as were oral tests taking more than 60 seconds or containing more than three errors; 3) the New York University Paragraph Test was used to assess short‐ and long‐term memory (immediate and deferred recall), ²³ with a cut point of five points selected to ensure greater predictive accuracy; 4) the Stroop Color and Word Test ²⁴ (Stoelting, Wood Dale, IL) was used to assess inhibition and potential brain damage; 5) the alternating series test and the test of drawing copy were used to evaluate executive function; and 6) the Hospital Anxiety‐Depression Scale ²⁵ was used to assess whether anxiety and/or depression—two conditions that alter cognitive results—were present. Scores of eight or over were defined as significant, whereas scores of 11 or over were defined as very significant. All cognitive tests have been validated and are characterized by their high specificity (between 91% and 97%) and sensitivity (between 73% and 89%). All participants provided informed consent before the examination.

Design and Statistics

All statistical tests were performed with a significance level <0.05. Continuous variables are described as mean ± SD. Parametric analysis of variance was used for analysis of variation, and the nonparametric Kruskall‐Wallis test was employed for those cases where distribution was non‐gaussian or where scores were involved. Percentage comparisons were made with contingency tables using Yates' correction.

RESULTS

Table I illustrates certain demographic characteristics of the normotensive and hypertensive groups. Educational level and average age (71.5±4.2 years vs. 72.5±4.2 years) were similar in both groups. Hypertension had been present for 10.2±8.2 (range, 5‐30) years.

Table I.

Characteristics of the Normotensive and Hypertensive Groups

Variable	Normotensive (n=30)	Hypertensive (n=60)	p Value
Age (yr)*	71.5 ± 4.2	72.5 ± 4.2	NS
Blood pressure (mm Hg)*
Systolic	121±11.09	141.4±17.8**	<0.001
Diastolic	72.8±10.9	80.8±9.02**	<0.001
Pulse pressure	48.1±10.2	60.5±16.6	<0.001
Female (n [%])	17 (56.6)	39 (65)	NS
Educational level (n [%])
Primary school	16 (53.3)	35 (58.3)	NS
High school	11 (36.6)	20 (33.3)	NS
College	3 (10)	5 (8.3)	NS
Body mass index >30 kg/m2 (n [%])	3 (10)	10 (16.6)	NS
Smoking status (n [%])
Nonsmoker	25 (83.3)	50 (83.3)	NS
Smoker	5 (16.6)	10 (16.6)
Cardiac disease (n [%])
Coronary artery disease	4 (13.3)	7 (11.6)	NS
Extrasystolic arrhythmia	9 (30)	12 (20)	NS
Benzodiazepine use (n [%])	11 (36.6)	27 (45)	NS
NS=nonsignificant; *mean ± SD; **all patients in the hypertensive group were taking antihypertensive medication, resulting in the relatively low blood pressure values for the group.

Table II summarizes the average values ± SD from the neurocognitive test battery. No significant differences (NS) were observed between groups in terms of MMSE scores (28.2±1.6 points vs. 28.01±1.5 points), the alternating series test, and the test of drawing copy. Both groups revealed some impairment in short‐term memory (5.30±1.8 points vs. 4.56±2.04 points; p=NS), but in the hypertensive participants there was poorer performance in long‐term memory, speed of attention, executive function, and inhibition. The Hospital Anxiety and Depression scale ruled out both disorders in all study participants (2.6±1.2 points vs. 3.0±1.3 points; p=NS).

Table II.

Results of the Neurocognitive Test Battery

Test/Examination	Normotensive (n = 30)	Hypertensive (n = 60)	p Value
Mini‐Mental State*	28.2±1.6	28.01±1.5	NS
Hospital Anxiety and Depression*	2.6±1.2	3±1.3	NS
Immediate recall*	5.30±1.8	4.56±2.04	NS
Delayed recall*	7.13±2.8	5.68±2.6	<0.01
Trail Making Part A nonexecution (n [%])	0 (0)	39 (65)	<0.005
Trail Making Part B nonexecution
Oral (n [%])	4 (13.3)	28 (46)	<0.005**
Written (n [%])	2 (6)	22 (36)	<0.01
No. of errors*	0.46±0.9	1.15±1.54	<0.02
Stroop Color and Word*	32.7±10.7	24.7±7.6	<0.005
NS=nonsignificant; *scores are mean ± SD; **Yates correction

Table III compares neuropsychologic performance of the hypertensive patients whose BP values met therapeutic objectives vs. hypertensive patients whose BP values did not achieve therapeutic objectives (controlled, 124.3±6/78.9±5.6 mm Hg vs. noncontrolled, 152±14/82±10.5 mm Hg). All hypertensive participants were receiving some form of treatment, with 37 (61.6%) taking one drug, 21 (35%) taking two drugs, and two (3.3%) taking three drugs. Sixty percent of participants were receiving angiotensin‐converting enzyme inhibitors or angiotensin I receptor blockers, 33% were taking β blockers, 15% were being treated with calcium channel blockers (CCBs), and 15% were on diuretics. No differences were observed between cognitive performance and BP values, number of drugs, or types of antihypertensive medications. No drug‐related trends were observed in the medication subgroups. Upon dividing the group of hypertensive participants into those aged up to 70 years of age and those aged 71 or older, no differences in BP were observed (140.2±17/80.8±7 mm Hg vs. 142.1±18/80.8±0.5 mm Hg; p=NS), but the older group did evince more failure to perform on Part B of the Trail Making Test (1 [4%] vs. 10 [27%]; p<0.03).

Table III.

Cognitive Assessment Between Hypertensive Controlled vs. Noncontrolled Subjects

	Ccontrolled (n = 23)	Noncontrolled (n = 37)	p Value
Systolic blood pressure (mm Hg)*	124.3±6.0	152.0±14.0	<0.001
Diastolic blood pressure (mm Hg)*	78.9±5.6	82.0±10.5	NS
Pulse pressure (mm Hg)*	45.4±7.3	70.0±13.5	<0.001
Immediate recall*	4.7±2.1	4.4±1.9	NS
Delayed recall*	5.6±3.0	5.6±2.3	NS
Trail Making Part A nonexecution (n [%])	1 (4.3)	0 (0)	NS
Trail Making Part B nonexecution (n [%])
Oral	14 (60.8)	14 (37)	NS
Written	7 (30.4)	9 (24.3)	NS
No. of errors*	1.34±1.4	1.02±1.59	NS
Stroop Color and Words*	24.8±8.5	26.8±7.7	NS
NS=nonsignificant; *mean ± SD

DISCUSSION

The connection between hypertension and cognitive decline is not well understood. Although neurocognitive tests can reveal neuropsychologic alterations in hypertensive patients, it is premature to establish a link between brain dysfunction and hypertension. Although progression toward vascular dementia (between 5% and 10% of all dementia disorders) is more frequent in hypertensive individuals, it occurs even more rapidly when BP control is inadequate. ²⁶ CBVD is the second major cause of mortality in the world and the second most frequent etiology of vascular dementia. CBVD is characterized by anomalies in cerebral perfusion (self‐regulation), ²⁷ greater metabolic impairment, white matter lesions (ischemic demyelination), ²⁸ and/or atrophy, which may amount to subclinical damage over the years and thus contribute to the development of specific cognitive deficiencies. Therefore, it is difficult to establish a difference between cognitive deterioration as a result of brain aging and the prodromic decline of a dementia disorder. In this context, vascular dementia may be the consequence of CBVD and may take either one of two shapes: multiple cerebral microinfarctions, and/or large vessel thromboembolism (multi‐infarction dementia or stroke), and is characterized by amnesia, aphasia, and apraxia, which may take different forms depending on the locus of the insult(s)—cortical, subcortical, or mixed. ²⁹

Against this background, we decided to analyze hypertensive and normotensive participants who were comparable in terms of age, educational level, and use of psychotropics—all factors known to influence cognitive response.

Of all cognitive functions, memory is the first to decline. “Forgetfulness” is a memory disorder typical of subcortical damage; it affects between 30% and 45% of the population, is characterized by the inability to remember, and is defined, in the absence of dementia, by a memory functionality score of one SD less than normal. The immediate and deferred recall tests, which were used with a low cut point, demonstrated that both groups had suffered some impairment in working memory (short‐term memory), but the deterioration of deferred recall (long‐term memory) was higher among hypertensive participants (Table II).

The studies whose hypotheses inversely relate hypertension to cognitive function are still debatable. The Framingham Study, ³⁰ which administered a neuropsychologic test to 2123 participants (aged 55‐89), concluded that it was not possible to establish a causative connection between hypertension‐related subclinical vascular disease and cognitive deficit because it was difficult to interpret BP changes, antihypertensive treatments, or white matter lesions in the light of cross‐sectional studies. During longitudinal follow‐up (10 years), cognitive performance declined in 88% of 1702 participants who did not receive antihypertensive treatments. ³¹ , ³² Other studies reflected similar results: the Epidemiology of Vascular Aging (EVA) Study Group ³³ noted, during a 4‐year follow‐up period, that untreated hypertensive patients were at a higher risk of cognitive decline than those taking antihypertensive drugs (risk ratio, 4.3 vs. 1.9). A study conducted in Sweden by Skoog et al. ³⁴ related the development of dementia to hypertension over a 15‐year follow‐up period. Kilander et al. ⁹ studied 999 men over 70 years of age during a 20‐year period and established a connection between cognitive deterioration and hypertension, insulin resistance, and diabetes. It is necessary, nonetheless, to consider other pathogenic mechanisms that might explain the differences illustrated by these results, such as patient selection, length of follow‐up periods, neurocognitive tests used, and questionable adherence to prescribed treatments.

Our results have led us to conclude that hypertension as a disease has a negative impact on cognitive performance, and that this negative impact may not be based on BP values alone. Cognitive deterioration was present both in the patients who met treatment objectives and in those who did not (Table III).

The prevalence of anxiety and depression is high in this age group (50%), and both conditions impair cognitive performance in both normal and pathologic circumstances. In selecting our participants, besides considering mood disorders as exclusion criteria, we considered medical conditions that are known to have a direct and negative influence on cognitive function; i.e., patients who have undergone heart surgery with extracorporeal circulation, ³⁵ and individuals with a history of diabetes mellitus, ³⁶ , ³⁷ hypercholesterolemia, ³⁸ , ³⁹ atrial fibrillation, ⁴⁰ , ⁴¹ or transient cerebral ischemia or stroke. ⁴²

Some studies have reported that antihypertensive medication (in particular, β blockers and centrally acting drugs) can be a cause of memory deficit or slowdown. ⁴³ , ⁴⁴ Although SHEP ⁷ revealed a lower risk of stroke in patients treated with diuretics and β blockers, it did not note any cognitive improvement. On the other hand, Syst‐Eur ¹² reported that a treatment program based on nitrendipine, a CCB, reduced cognitive decline (dementia) by 50%. This effect might be attributed to strict BP control, or to the use of CCBs. Calcium ions that accumulate intraneuronally have been associated with several neuropsychiatric diseases—both severe (stroke) and chronic (dementia, major depression). ⁴⁵

Hanson et al. ⁴⁶ contend that BP control with this family of drugs might have neuroprotective effects, thus preventing cognitive decline. On the basis of our results, we cannot demonstrate whether BP or certain drug types or dosages have an influence on cognitive performance.

Finally, the most striking observation in our study has been the impairment of cognitive function in the brain's prefrontal areas. This is the most developed region, and is responsible for processing the most complex intellectual tasks, concentrating perceptive, volitional, cognitive, and emotional aspects. When the brain begins to age, the frontal‐cortical region becomes reduced by 20% in seniors between 65 and 80 years of age as a result of atrophy and a decrease in neuronal population. The deafferentation of the frontal lobes as a result of lesions in specific circuits (prefrontal, dorsolateral, orbitofrontal, and anterior cingulate) that connect the subcortex with this region brings about “executive dysfunction,” which is characterized by impaired performance of mental operations, psychomotor retardation, attention deficit (distract‐ibility), visual and spatial alterations, difficulty in planning and starting an activity, apathy, and loss of inhibition. ⁴⁷ Such afflictions may be present without affecting daily activities and are not detected in primary care. The impairment of prefrontal regions implies a loss of voluntary attention control, which then becomes dependent on external stimuli. Thus immediate (short‐term) memory and time‐dependent memory are impacted, and the patient may display emotional instability.

The Trail Making Test, particularly in its Part B, studies cognitive flexibility. It is a highly sensitive test which helps detect diffuse brain damage and frontal lobe dysfunction regardless of the damaged hemisphere. It does not show any gender differences, its scores are not altered (according to some authors) by the participant's educational level, and it is used to determine early neurocognitive decline. Nearly 50% of the hypertensive participants in our study were unable to complete the test, and they experienced more errors than their normotensive counterparts. Younger hypertensive patients experienced fewer problems than older hypertensive patients.

Similar results were obtained when inhibition was measured through the Stroop Color and Word Test. Although our protocol did not include any correlation between neurocognitive involvement and brain imaging, the poor performance of hypertensive vs. normotensive patients may be attributed to diffuse prefrontal cortex lesions. ⁴⁸ , ⁴⁹ Even though the connection between the factors is not entirely accepted in the literature, these may be present in senior patients ⁵⁰ without hypertension and may even remain undetected by current diagnostic methods.

CONCLUSIONS

There is no clear evidence that mild cognitive deterioration in hypertensive patients is a prelude to vascular dementia. Many experts consider that cognitive decline is a functional and potentially reversible, rather than a structural and progressive, condition. Thus, several studies advocate intensive BP treatment as a way of modifying a vascular risk factor involved in the development of cognitive decline. The above‐mentioned multiplicity of causes, along with the extended periods during which brain damage is subclinical, makes it difficult to establish a connection between hypertension and cognitive decline. Our observations have shown that in hypertension, cognitive impairment was present in areas such as attention, memory, and executive function, which involve prefrontal regions. The use of a neuropsychologic test battery of easy administration (Trail Making Test Part B) might help detect on an uncomplicated and early basis whether patients are suffering from cognitive decline, and thus prevent structural brain damage. Nevertheless, additional studies involving a greater number of patients will be necessary to identify the potential causes of brain damage progression.