Nocturnal blood pressure dipping, blood pressure variability, and cognitive function in early and middle‐aged adults

Abstract

Higher nighttime blood pressure (BP), less BP dipping, and higher BP variability have been linked with worse cognitive function in the elderly. The goal of this study is to explore whether this relationship already exists in early and middle adulthood. We further examined whether ethnic differences between African Americans and European Americans in BP parameters can explain ethnic differences in cognitive function. 24‐h ambulatory BP monitoring and cognitive function were obtained from 390 participants (average age: 37.2 years with a range of 25–50; 54.9% African Americans; 63.6% females). We observed that higher nighttime BP, decreased dipping, and higher variability were significantly associated with lower scores on the Picture Sequence Memory Test. Significant negative associations between variability and overall composite scores were also observed. No significant associations between average 24‐h or daytime BP and cognitive function were observed. Ethnic differences in nighttime diastolic pressures and dipping can explain 6.81% to 10.8% of the ethnicity difference in the score of the Picture Sequence Memory Test (ps < .05). This study suggests that the associations of nighttime BP, dipping, and variability with cognitive function already exist in young and middle‐aged adults. Ethnic differences in nighttime BP and dipping can at least partially explain ethnic differences in cognitive function. The stronger association of these parameters with cognitive function than daytime or average BP in this age range raises the importance of using ambulatory BP monitoring for more precise detection of abnormal BP patterns in young adulthood.

1. INTRODUCTION

As technology and medicine advance, our population becomes older and stays older longer than at any other time in human history. With this aging population comes a new set of challenges to improve health and wellness in the final decades of life. Cognitive decline has a particularly large impact on older populations, and its incidence is continually increasing. Studies estimate that 1 in 45 Americans will have Alzheimer's disease by 2050. ¹ Cognitive decline and dementia are especially impactful, not only due to the lack of disease‐modifying treatments, but also their immense negative influence over performing activities of daily living. Thus, there is growing interest and research in identifying risk factors for early prevention.

Hypertension has been recognized as a modifiable risk factor for cognitive impairment and dementia. Ambulatory blood pressure (BP) monitoring is increasingly employed to diagnose hypertension. Collecting data throughout an entire day allows accurate calculation of mean BP, which has been found to be a better predictor of cerebrovascular disease and cognitive impairment than readings in the clinic. ² In addition to mean BP, ambulatory BP monitoring allows for the determination of BP variability and diurnal variation. Increased variability has been associated with impaired cognitive function and cognitive decline in older adults. ³ , ⁴ In addition, variability has been associated with cerebral white matter hyperintensities. These lesions, thought to indicate small vessel cerebrovascular disease, are strongly associated with cognitive impairment. ⁵ Diurnal variation is another important feature of BP regulation. During sleep, BP should drop by about 10%−20% of its daytime value. ⁶ However, the pressures of some people remain constant or even increase overnight, which is termed “non‐dipping.” The non‐dipping phenomenon, or higher nighttime BP, is not only a stronger predictor of target organ damage and cardiovascular events compared with average BP levels, but is also associated with cognitive dysfunction. ⁷ , ⁸ , ⁹

The majority of studies that have assessed relationships between BP and cognitive function, apart from the Coronary Artery Risk Development in Young Adults (CARDIA) Study, ¹⁰ have been conducted in older adults. It has been suggested that impaired cognition could be a clinically silent disorder that starts during midlife or even earlier. Therefore, it is important to understand whether the association between higher BP levels, higher variability, or blunted nocturnal dipping and cognitive function already exists in early and middle adulthood. The findings will provide important guidelines for early detection and prevention. In this study, we aim to evaluate whether BP parameters obtained from 24‐h ambulatory BP recording are already associated with cognitive function in early and middle‐aged adults. Moreover, it is of note that African Americans (AAs) bear a disproportionate burden of Alzheimer's disease and related disorders compared to European Americans (EAs), with estimates suggesting that they may have more than a 2‐fold increased risk. ¹¹ Previous studies, including ours, have also documented the ethnic difference in BP parameters, with AAs having higher BP levels, higher variability, and lower nocturnal dipping since childhood. ¹² , ¹³ Our sample included roughly equal numbers of both ethnic groups. Therefore, we further aimed to examine whether the ethnic difference in BP parameters can at least partially explain the ethnic difference in cognitive functions. The findings may help reduce ethnic health disparities.

2. METHODS

2.1. Participants

This study included 390 participants (average age: 37.2 years with a range of 25–50; 54.9% AAs; 63.6% females) who were enrolled in either the Georgia Stress and Heart Study (160 participants) or the Georgia Cardiovascular Twin Study (230 participants from 75 twin pairs and 80 singletons). All participants had valid nighttime BP measurements and cognitive measurements that were obtained during a visit conducted between May 2019 and April 2022. Criteria for classifying participant ethnicity, as well as the study design and selection criteria, have previously been described. ¹⁴ , ¹⁵ Of the 390 participants, 46 were taking anti‐hypertensive medications. Sensitivity tests were conducted by excluding these participants from the analysis. Anthropometric measurements were obtained during a laboratory visit, including height on a wall‐mounted stadiometer and weight on a digital scale. Body mass index (BMI) was calculated as weight/height² (kg/m²). Participants self‐reported their smoking status; participants were considered smokers if they had smoked at least five cigarettes in the past 30 days.

The institutional review board at the Medical College of Georgia gave approval for the studies. Each participant gave written consent prior to any involvement, in accordance with institutional guidelines.

2.2. Ambulatory BP monitoring

Participants underwent ambulatory BP monitoring over a 24‐h period. The cuff was fitted to each person's non‐dominant arm (model 90217; SpaceLabs, Redmond, WA). Measurements were recorded every 20 min from 8 a.m. to 10 p.m. and every 30 min from 10 p.m. to 8 a.m. The transitional periods between 6 a.m. to 8 a.m. and 10 p.m. to midnight were excluded from analysis. An adequate number of recordings was defined as ≥14 recordings during the day (from 8 a.m. to 10 p.m.) and ≥6 recordings during the night (from midnight to 6 a.m.). Acceptable readings were defined according to the following criteria: pulse pressure ≥ 20 mmHg; pulse pressure ≤ 140 mmHg; heart rate ≥ 40 beats/min; and heart rate ≤ 180 beats/min.

Dipping was defined by mean arterial pressure (MAP) and calculated as (Daytime MAP—Nighttime MAP)/Daytime MAP. BP variability was indexed by the 24‐h Average Real Variability (ARV) in BP which was calculated using the following algorithm: $ARV=\frac{1}{N-1}{\sum }_{t=1}^{N-1}|B{P}_{k+1}-B{P}_k|$, where N denotes the number of valid BP measurements in the data corresponding to a given subject. In comparison with the other variability parameters such as the standard deviation of ambulatory BP recordings, 24‐h ARV further considers the order in which the BP measurements are obtained. ²

2.3. Cognitive function testing

Cognitive function was assessed using the National Institute of Health (NIH) Toolbox: Cognition Battery. All cognitive function testing was performed on an iPad by a trained examiner proctoring the test during the participant's laboratory visit. The NIH Toolbox was chosen for its accessibility, broad application, low cost, and high test‐retest reliability. ¹⁶ We used five tests from the Toolbox: The Flanker Inhibitory Control and Attention Test (assessing attention and executive function), Picture Sequence Memory Test (assessing episodic memory), List Sorting Working Memory Test (assessing working memory), Dimensional Change Card Sort Test (assessing executive function), and Pattern Comparison Processing Speed Test (assessing processing speed). ¹⁷ The Cognition Fluid Composite Score was calculated from the above five tests and was used as an index of overall cognitive function.

2.4. Statistical analysis

All analyses were performed using STATA software. Generalized estimating equations (GEE) were used to test whether ethnic differences exist in BP parameters and cognitive function with age, sex, education level, BMI, and smoking status as covariates. GEE is a multiple regression technique that allows for non‐independence of twin or family data yielding unbiased standard errors and p‐values. GEE was also used to test the association between BP parameters and cognitive function with age, sex, ethnicity, education level, BMI, and smoking as covariates. An interaction term between ethnicity and BP parameters was also included in the model to check whether the association between BP parameters and cognitive function was dependent on ethnicity. We further tested whether the ethnic differences in BP parameters can mediate the ethnic differences in cognitive function using the Sobel test. Mediation was considered to be present when the ethnic differences in cognitive function decreased on the addition of BP parameters to the model and the mediation test gave p < .05. Covariates in all models included age, sex, education level, BMI, and smoking. Due to the non‐independence of the twin data, we used bootstrapping to test for mediation significance.

3. RESULTS

General characteristics of the participants are listed by ethnicity in Table 1. AA participants had significantly higher BMIs and lower education levels compared with EA participants. There were also more female participants among AAs. After adjustment of age, sex, education level, BMI, and current smoking status, AA participants had significantly higher BP levels, higher variability, and less nighttime dipping, except for systolic variability. Every cognitive test score, as well as the composite score, was also significantly lower in AA participants.

TABLE 1.

General characteristics of the participants (n = 390).

	European American	African American	p
N	176	214
Females, %	57.4%	68.7%	.021
Age, years	35.9 ± 6.0	36.9 ± 5.4	NS
Education, years	15.2 ± 2.0	14.8 ± 2.1	.037
BMI, kg/m2	29.0 ± 6.4	31.7 ± 8.0	<.001
Current smokers, %	23.9%	24.3%	NS
24‐h Systolic BP, mmHg a	112.0 ± 11.9	116.6 ± 12.4	<.001
24‐h Diastolic BP, mmHg a	71.9 ± 7.9	74.9 ± 9.0	.001
Nighttime Systolic BP, mmHg	102.6 ± 13.9	108.6 ± 14.1	<.001
Nighttime Diastolic BP, mmHg	63.1 ± 9.2	66.9 ± 10.7	.001
Daytime Systolic BP, mmHg a	117.2 ± 12.1	120.7 ± 13.0	<.001
Daytime Diastolic BP, mmHg a	76.6 ± 8.3	79.0 ± 9.5	.004
Systolic BP variability a	8.10 ± 1.67	8.51 ± 1.69	NS
Diastolic BP variability a	7.29 ± 1.63	7.80 ± 1.50	.046
Dipping a	0.15 ± 0.07	0.12 ± 0.08	.006
Flanker Inhibitory Control and Attention Test	101.0 ± 7.6	95.1 ± 9.9	<0.001
Pattern Comparison Processing Speed Test	110.4 ± 17.2	98.6 ± 18.6	<.001
Picture Sequence Memory Test	110.5 ± 14.7	102.9 ± 14.4	<.001
Dimensional Change Card Sort Test	108.0 ± 8.2	104.1 ± 9.0	<.001
List Sorting Working Memory Test	112.6 ± 11.7	103.7 ± 12.2	<.001
Cognition Fluid Composite	110.5 ± 10.9	100.0 ± 11.6	<.001

Note: For all the BP parameters and cognitive function measurements, age, sex, education levels, BMI, and current smoking status were included as covariates.

Abbreviations: BMI, body mass index; BP, blood pressure.

^a The sample size for this variable is 173 for European American participants and 207 for African American participants.

Associations between BP parameters and the Picture Sequence Memory Test are presented in Table 2, with statistically significant results bolded. Higher nighttime BP parameters, including systolic and diastolic BP, decreased dipping, and higher variability were significantly associated with lower scores on the Picture Sequence Memory Test. Table S1 presents the associations between BP parameters and the other four cognitive tests, as well as the overall composite scores. Significant negative associations were also identified between systolic variability and the scores from the Pattern Comparison Processing Speed Test, as well as the scores from the Dimensional Change Card Sort Test. This was consistent with the significant associations between systolic and diastolic variability and the overall composite scores (Table S1). No significant association between 24‐h average BP levels or daytime BP and cognitive function was observed (Table 2 and Table S1). No significant interaction between ethnicity and BP parameters on cognitive function was identified, indicating the observed associations exist in both ethnic groups. These findings were essentially unchanged when participants on anti‐hypertensive medication were excluded from the analysis.

TABLE 2.

The association between BP parameters and Picture Sequence Memory Test.

	β	p
24‐h Systolic BP	−.10	.100
24‐h Diastolic BP	−.16	.072
Nighttime Systolic BP	−.13	.014
Nighttime Diastolic BP	−.21	.004
Daytime Systolic BP	−.06	.362
Daytime Diastolic BP	−.07	.389
Dipping	2.65 a	.004
Systolic BP variability	−1.14	.008
Diastolic BP variability	−1.15	.016

Note: Age, sex, ethnicity, education levels, BMI, and current smoking status were included as covariates.

Abbreviation: BP, blood pressure.

^a Per 10% of changes in dipping, the changes in the score of the picture sequence memory test.

We further tested whether the ethnic differences in nighttime BP parameters, dipping, and variability can explain the ethnic differences in the cognitive score of the Picture Sequence Memory Test and the overall composite score. The ethnic differences in the score of the Picture Sequence Memory decreased when adding nighttime diastolic pressures (Figure 1A) and dipping (Figure 1B) to the model, and the Sobel tests were significant (all ps < .05), indicating the presence of mediation effects. Specifically, nighttime diastolic pressure and dipping are able to explain 6.81%–10.8%, respectively, of the ethnicity difference in the cognitive score of the Picture Sequence Memory Test. No ethnic difference in nighttime systolic BP or diastolic variability was observed that could mediate the ethnic difference in the score from the Picture Sequence Memory test or the overall composite score.

FIGURE 1.

Mediation models illustrating BP parameters mediating the effect of ethnicity on the Picture Sequence Memory Test. Age, sex, education levels, BMI, and current smoking status were included as covariates. (A) Nighttime diastolic BP, (B) BP dipping. * Per 10% of changes in dipping, the changes in the score of the picture sequence memory test.

4. DISCUSSION

In this study, we observed that, in early and middle‐aged adults, higher nighttime BP, blunted dipping, and higher variability were already associated with worse episodic memory (Picture Sequence Memory Test). We also found an association between variability and overall fluid cognitive function.

The association of high nighttime BP and blunted or reversed dipping with abnormal cognitive function (i.e., cognitive impairment or dementia) has been well documented in the elderly. ¹⁷ , ¹⁸ The CARDIA study, however, has been the only study reported so far that explored this topic in young adults (mean age: 30.2 years) and observed that less nighttime dipping and higher nighttime diastolic BP were associated with worse executive function twenty years later. ¹⁰ Findings of the present investigation are consistent with those of the CARDIA study. Furthermore, our study extends the findings of CARDIA in at least two important ways. First, the CARDIA study did not have cognition measurements during the young adult period. Therefore, it was not clear whether the observed associations already existed during this age range. Present findings indicate that in early adulthood (mean age: 37 years), higher nighttime BP and blunted dipping were already associated with worse cognitive function. Second, the present investigation included roughly equal numbers of AAs and EAs, enabling us to explore whether the well‐documented ethnic differences in nighttime BP and dipping can explain ethnic differences in cognitive function. The present findings demonstrate that the ethnic difference in nighttime diastolic pressures and dipping can, at least partially, explain the ethnicity difference in the cognitive score of the Picture Sequence Memory Test. Regarding another aspect, both the CARDIA study and the present study observed a stronger association of nighttime BP, rather than daytime BP, with cognitive function. Hypertension‐induced mechanisms of brain damage have been described in detail elsewhere. ¹⁹ Accordingly, it could be reasonably hypothesized that individuals whose BP does not drop during the nighttime, according to the physiological circadian rhythm, and remains substantially higher throughout the whole 24‐h period, are more susceptible to the detrimental effects of hypertension in the brain. Future studies are certainly warranted to test this hypothesis.

Accumulating evidence has also demonstrated that increased variability may contribute to a higher risk of cognitive impairment and dementia, independent of BP levels. ²⁰ , ²¹ Most of these studies have been conducted in elderly individuals or have been conducted using long‐term visit‐to‐visit variability. The present study is the first study to examine whether variability calculated from 24‐h ambulatory BP recording (i.e., short‐term variability) is associated with cognitive function in apparently healthy young and middle‐aged adults. Significant associations between higher variability and worse episodic memory and overall fluid cognitive function were observed and suggest that short‐term variability could be a risk factor for cognitive impairment in early and middle adulthood. The exact mechanism by which increased variability may impair cognitive function is unknown, although a few potential explanations have been proposed. For example, variability has been linked with increased white matter lesions, white matter abnormalities, and lower brain and hippocampal volume. ⁵ , ²² Increased variability secondary to increased arterial stiffness can lead to under‐perfusion of the brain and subsequent damage. ⁷ In addition, an association has also been observed between increased variability and neurofibrillary tangle pathology in an autopsy study. ²³ Current findings demonstrate that variability, compared to average BP levels, is more strongly associated with cognitive function in early and middle adulthood, which emphasizes the importance of further understanding the underlying mechanisms which contribute to these relationships.

In this study, we focused on fluid cognitive function because fluid abilities are generally more sensitive to neurobiological integrity and age decline compared with crystallized abilities. ²⁴ Most of the significant findings observed in the present study were from the Picture Sequence Memory Test, which is designed to test episodic memory. In the context of this specific cognitive test, episodic memory also requires strong short‐term and working memory, other components of fluid intelligence. Episodic memory is also generally more susceptible to age‐related decline, due to its fragile network in the brain. ²⁴ Present findings indicate that the episodic component of fluid cognition could be the first or most prominent factor affected by higher nighttime BP, blunted dipping, or higher variability. Previous studies indicate that elevated BP is associated with later impairment of attention, executive function, motor speed, and information processing, all components required for a participant to perform well on the Picture Sequence Memory Test. ²⁵

4.1. Experimental considerations

There are three experimental considerations within the present study that need to be recognized. First, this is a cross‐sectional study with cognitive testing only performed one time. Therefore, we cannot test the effect of 24‐h ambulatory BP parameters on cognitive decline over time. The Georgia Stress and Heart Study cohort and twin cohort are longitudinal cohorts and therefore, we will be able to answer this question as the participants come back for subsequent visits. Second, the present study only included EA and AA participants. Accordingly, findings cannot be generalized to any other racial or ethnic groups. Future studies including more racial and ethnic groups are warranted. Third, because this study is cross‐sectional, it is impossible to claim the causality of this BP‐cognition relationship. Additionally, there could be other factors, which were not addressed in this study, that could be the common contributing factors to both abnormal BP patterns and worse cognitive functions such as alcohol consumption, dietary habits, occupation, or living environment. It is important for future studies to incorporate these potential contributing factors.

5. CONCLUSION

In summary, the present study indicates that the associations between higher nighttime BP, blunted dipping, and higher variability with worse cognitive function already exist in young and middle‐aged adults. In addition, ethnic differences in nighttime BP and dipping can, at least partially, explain the ethnic difference in cognitive function. The stronger association among these BP parameters with cognitive function in this age range, compared to just daytime BP or average BP levels, emphasizes the importance of using ambulatory or nighttime BP monitoring for earlier and more precise detection of abnormal BP patterns in young adulthood.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

PATIENT CONSENT STATEMENT

Each participant gave written consent prior to any involvement, in accordance with the institutional guidelines.

Supporting information

Supporting Information

Massengale K, Barnes VA, Williams C, et al. Nocturnal blood pressure dipping, blood pressure variability, and cognitive function in early and middle‐aged adults. J Clin Hypertens. 2024;26:235–240. 10.1111/jch.14764

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.