Hypertension has been linked to impaired cognitive/CNS function, and some of these changes may precede development of frank essential hypertension. The stress and fatigue of sleep deprivation may exacerbate these cognitive changes in young adults at risk. We hypothesize that individuals at risk for hypertension will show significant declines in cognitive function during a night of sleep deprivation. Fifty-one young adults were recruited for 28-hour total sleep deprivation studies. Hypertension risk was assessed by mildly elevated resting blood pressure and by family history of hypertension. A series of cognitive memory tasks was given at four test sessions across the sleep deprivation period. Although initially comparable in cognitive performance, persons at risk showed larger declines across the night for several indices of working memory, including code substitution, category, and order recall. These results suggest that cognitive/CNS changes may parallel or precede blood pressure dysregulation in the early stages of hypertension development. The role of CNS changes in the etiology of essential hypertension is discussed.
Hypertensive patients have impaired cognitive and CNS function, and some of these changes may precede development of frank essential hypertension. Subtle cognitive changes in younger persons at risk for hypertension may become more readily apparent during the systemic stress of sleep deprivation. We hypothesize that young adults at risk for hypertension will show significant declines in cognitive function during a night of sleep deprivation.
Cognitive decline in older persons with advanced hypertension is especially well documented and likely represents, at least in part, the damaging effects of sustained high blood pressure on CNS microvasculature, and, hence, brain function [
However, some cognitive changes may precede the development of frank clinical hypertension, suggesting a more complex association between the CNS and blood pressure in the development of essential hypertension. For example, decreased cognitive/CNS function has been recently found in middle aged and younger hypertensives [
Participants were fifty-one volunteers (28 males and 23 females) with an average age of 22.9 years. Young adult study participants were recruited from campus and the surrounding community with ages ranging from 19–32 years old. This sample included graduate and undergraduate students, university employees, and community citizens. Average body weight was 148 ± 25.7 lbs., and average body mass index was 22.6 ± 3.26. The participants completed questionnaires about personal and family medical history, sleep habits, and alcohol and tobacco use. Only persons reporting a regular diurnal sleep/wake cycle were selected for participation. The final study population was a healthy, normal sample without sleep disorders or significant cardiovascular, neurological, endocrine, or psychiatric disease. All subjects refrained from use of caffeine and tobacco and exercise throughout the study period. Informed consent was obtained from each subject before participation. All procedures were approved by the Clemson University Institutional Review Board.
The present investigation is part of a larger series of studies of sleep deprivation and sustained operations. Procedures have been described in detail by Pilcher and coworkers [
Participants also received a sleep log to provide information on sleep habits prior to the onset of the study. These sleep logs were to be completed each morning of the three days prior to the study. The sleep logs included questions inquiring about sleep quality, time going to bed, time getting out of bed, and napping throughout the day. Analysis of Actiwatch data confirmed accuracy of sleep logs to verify adherence to instructions. Participants reported at 9:30 AM on Day 1 and were transported to the residential sleep laboratory. Food and noncaffeinated beverages were provided throughout the study.
The participants completed a series of tasks and questionnaires and were given scheduled breaks and meals throughout the study period. Training on the cognitive tasks was completed in two periods on Day 1. Four testing sessions were scheduled at approximately 8:30 PM, 1:00 AM, 5:30 AM, and 10:00 AM across the sleep deprivation period. The study ended on Day 2 when participants were transported back to their residences and instructed to sleep before operating heavy equipment or driving.
Questionnaires were administered to obtain information on typical sleep patterns and individual and familial medical history for each participant. The Pittsburgh Sleep Quality Index (PSQI) was used to assess recent sleep quality. Its reliability and validity has been verified by Buysse and coworkers [
Resting blood pressure (BP) was measured upon arrival and departure at approximately 11 PM each day and at approximately 8:30 PM, 1:00 AM, 5:30 AM, and 10:00 AM over the study period. Electronic blood pressure measurements were taken using GE Dinamap Pro100 machines (Medical Solutions, Minneapolis, MN.). Dinamap performance was verified on a regular basis for zero offset, integral offset, and gain using a mercury manometer. All devices performed within manufacturer tolerances. Research assistants were trained on theory and application of blood pressure determination using both auscultatory and oscillometric techniques, including use of appropriate cuff sizes and other American Heart Association guidelines for blood pressure determination [
The participants also completed five subtests from the Automated Neuropsychological Assessment Metrics (ANAM). This measure is a battery of cognitive tests developed by the Office of Military Performance Assessment (OMPAT; Washington, DC). The ANAM tests have strong correlations with traditional measures of neuropsychological functioning, high test-retest reliability (typically = 0.80, 0.95 range), high differential stability, and a large database of studies providing construct validity [
Participants were presented with a set of symbols (
Participants were asked to recall the previous key of numbers and symbols that was given during the Code Substitution Practice task. They were presented with a symbol and number pair and indicated if the stimulus matched the previously memorized key. Participants were given 18 trials during each administration of the task. Each stimulus was displayed for 4000 milliseconds, and the time allowed for response was 4200 milliseconds before the next stimulus was displayed.
After completing two additional tasks (see ST6 and CPT below), participants were asked to recall the memorized key given during the Code Substitution Practice task after approximately 10 minutes had elapsed since it was presented. The participants were given 36 trials in the CDD. Each stimulus was displayed for 4000 milliseconds, and the time allowed for response was 4200 milliseconds before the next stimulus was displayed.
Participants were given a set of 6 letters to memorize. They could review the letters until pressing the space bar to continue the task. Single “probe” letters were then flashed on the screen. The participants then indicated if the stimulus was part of the memory set. The probe was displayed on the screen for 1400 ms. The maximum time to respond to each presented stimulus was 1500 milliseconds.
Participants completed a Category Recall (CatRecall) task presented on a computer to test memory. The CatRecall memory task was developed using E-prime at the University of Maryland’s Center for Advanced Study of Language and derived from previously formed category memory tasks [
Participants were presented with a “memory list,” which was a series of six words, one from each category. Each word in the memory list was displayed on the computer screen for 400 milliseconds. Participants were then presented with a probe category such as “music” from the 6 available category names. The probe category was displayed for 3000 ms. Participants were then asked to identify which word from the memory list was originally included in the category presented.
The order recall task presented participants with a memory list of 6 words, one at a time. Participants were asked to remember the 6 words in order, and when presented with a cue word from the memory list, to correctly choose the word from the memory list that followed the cue word.
Participants were presented single-digit numbers (1–9) one at a time. Once a second number appeared, they indicated whether that number was the same as the previously presented number (a 1-back task). Participants were presented with 179 numbers resulting in 178 trials. This task was divided into two sections: with feedback (1 minute) and without feedback (4 minutes). The probe was displayed on the screen for 1000 ms, and participants were given a maximum of 1500 milliseconds to respond.
Risk for subsequent development of hypertension was determined in two different ways, by resting blood pressure levels and by reported parental history of hypertension. Classification of risk by resting blood pressure levels was based on criteria outlined in the Seventh Report of the Joint National Committee on the Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC-7) [
Participants were classified with a positive family history of hypertension if one or both biological parents were identified as having been diagnosed with essential hypertension by a physician. Participants with no reported parental hypertension were classified as negative family history. Validity of self-reported parental hypertension has been consistently demonstrated through direct contact with parents and parents’ physicians [
All BP data were entered into Microsoft Excel and imported into SPSS (IBM, Armonk, NY, USA) for statistical analyses. Cognitive performance measures were initially analyzed in a
Means and standard errors for blood pressures and cognitive performance of all participants at the four test sessions are shown in Table
Means (± standard errors) for systolic and diastolic blood pressure and cognitive performance accuracy across a night of sleep deprivation with sustained cognitive work. All cognitive performance data are in percent correct.
8:30 PM | 1:00 AM | 5:30 AM | 10:00 AM | |
---|---|---|---|---|
Systolic blood pressure (mmHg) | 113.0 (1.64) | 112.6 (1.78) | 111.5 (1.72) | 112.0 (1.57) |
Diastolic blood pressure (mmHg) | 66.3 (1.13) | 67.2 (1.03) | 66.6 (1.22) | 66.5 (1.00) |
Category recall | 74.2 (2.05) | 75.1 (2.42) | 69.9 (2.58) | 72.2 (2.41) |
Code substitution learning | 96.6 (.40) | 94.9 (.66) | 91.4 (.87) | 92.7 (.90) |
Code substitution immediate recall | 93.3 (1.16) | 88.5 (2.18) | 83.2 (2.73) | 86.9 (1.92) |
Code substitution delayed recall | 91.5 (1.53) | 85.0 (2.37) | 80.2 (2.18) | 81.0 (2.31) |
Sternberg memory recall | 96.7 (.50) | 91.6 (1.20) | 90.1 (1.87) | 86.7 (1.99) |
Continuous performance test | 90.2 (1.15) | 83.9 (2.35) | 76.6 (2.32) | 75.3 (2.56) |
Using the JNC-7 criteria for classification of blood pressure, 19.6% of participants (10 of 51) were classified as prehypertensive. Average age was 21.9 years for the prehypertensive group and 23.0 years for the normal group. Using family history of hypertension as an index of risk, 23.5% of participants (12 of 51) reported a positive family history of essential hypertension in at least one biological parent. Average age was 23.0 years for the positive family history group and 22.7 years for the negative family history group. There were no significant group differences in age, weight, body mass index, or distribution of sex among risk groups (
Results for blood pressures across time and by risk groups have been reported elsewhere [
Repeated measures ANOVAs show significantly greater declines in cognitive performance of prehypertensives versus normotensives over the period of sleep deprivation. For example, Figure
The effect of sleep deprivation on immediate code substitution recall (% correct) in persons classified by JNC-7 criteria of prehypertensive or normotensive. The time by Risk Group interaction was significant [multivariate
The effect of sleep deprivation on delayed code substitution recall (% correct) in persons classified by JNC-7 criteria of prehypertensive or normotensive. The time by Risk Group interaction was significant [multivariate
Using family history as the risk variable, a significant Risk Group × Time interaction was observed for order recall [
The effect of sleep deprivation on the order recall task (% correct) in persons classified as positive or negative family history of hypertension. The time by Risk Group interaction was significant [multivariate
Throughout the night of sleep deprivation, blood pressure was determined prior to each cognitive testing session. Therefore, we also examined the correlations between sleep-deprivation-induced changes in blood pressure and cognitive function across the night. While most correlations were not statistically significant, we did observe a pattern of significant positive correlations (2-tailed probabilities) between blood pressure and corresponding cognitive function, especially at the 1:00 AM test session. For example, at the 1:00 AM test session, SBP was positively correlated with performance on simultaneous code substitution (
Sleep deprivation has been shown to impair functioning of the distributed thalamoprefrontal cortical networks subserving attention and higher order cognitive processes [
Interestingly, we also observed a decline in order recall performance in persons with a positive family history of hypertension. Because there were no initial differences in resting blood pressure between positive and negative family history groups, this finding suggests that even mild elevations in resting pressure are not necessary for expression of the association between risk for hypertension and cognitive performance decline. This suggests that CNS changes may occur before significant blood pressure dysregulation.
Sleep deprivation and/or disruption can influence the sympathetic nervous system and blood pressure acutely [
Interestingly, sleep-deprivation-induced changes in acute blood pressure and cognition across the night showed a trend for positive correlations, especially at the 1:00 AM testing. This suggests that individuals with the worst cognitive performance at this time of night also showed signs of decreased arousal as indexed by lower acute blood pressure at the time of testing. This may reflect the relationship between cortical and autonomic arousal and is consistent with reports of decreases in blood pressure and attentiveness with increasing fatigue [
Overall, the cognitive decline in persons with mild elevations in resting blood pressure (JNC-7 prehypertensive) is consistent with results of Ditto et al. [
Changes in cognitive function in younger persons without a history of significant and sustained high blood pressure may provide insight into the early etiological mechanisms in the developmental pathophysiology of hypertension. It is possible that even modest, subclinical resting blood pressure elevations could directly engender subtle CNS functional damage [
Notwithstanding the above, there are other possible links between blood pressure and CNS function in otherwise healthy young adults at risk. For example, CNS changes could be involved intimately with the progressive blood pressure dysregulation that eventuates in frank essential hypertension later in life. Several pathways could influence this process. Firstly, changes in CNS function and blood pressure control could be related indirectly with each other, but correlated with a heretofore unknown, underlying mechanism affecting both through distinct, but noninteractive pathways. Secondly, and even more intriguingly, CNS changes could contribute directly to dysfunction of autonomic and neuroendocrine systems involved in regulation of blood pressure. Under the latter scenario, changes in higher CNS function could cascade distally via central control of sympathoadrenomedullary, hypothalamic pituitary adrenocortical (HPA), and perhaps other neuroendocrine axes [
In support of this notion, an accumulating body of evidence points to a broad spectrum of subtle CNS changes in otherwise healthy persons with either mildly elevated blood pressure or normal resting blood pressure with other risk factors such as a positive history of hypertension. For example, prior work in our lab has shown abnormalities in opioidergic inhibition of both the HPA and the sympathoadrenomedullary axes in persons at risk for hypertension [
The notion of CNS changes preceding blood pressure dysregulation is consistent with a recent report of reduced cry response to painful vitamin K injections in newborn infants with hypertensive grandparents [
Several limitations should be considered in interpretation of the present results. First, the 28-hour sleep deprivation methodology requires significant costs, as well as intensive time burden on participants. Thus, modest samples sizes are typical of studies where sleep deprivation is experimentally manipulated. Although a larger sample size would have been preferable, the present study nevertheless produced several significant results in line with hypothesized outcomes. This suggests that the current sample size has sufficient statistical power to expose reliable links between brain function and hypertension risk. Moreover, the experimental manipulation of sleep deprivation in the current study avoids the many potential covariates that confound correlational epidemiological studies of chronic sleep loss. Nevertheless, risk group differences in aerobic fitness or other related variables may have affected these results, despite similarities in body weight and BMI. Interestingly, only two subjects were classified as both prehypertensive and positive family history. This finding could reflect the relatively young age of the present sample. For example, the categorization by family history in our sample could favor a later developing blood pressure rise. An older study sample might show more persons with a family history of hypertension in the prehypertensive blood pressure range. Regardless of methodological limitations, the present results suggest that healthy young adults at risk for hypertension show decreased memory performance during a single night of sleep deprivation.
The overall findings of this study indicate that young adults at risk for hypertension show decreases in higher cognitive function during sleep deprivation. The sleep deprivation-induced cognitive declines in persons at enhanced risk are significantly greater than those seen in their low-risk counterparts. These risk-related cognitive declines are seen primarily in tasks that require sustained attention and working memory and emerge only after significant sleep deprivation. These results suggest that some cognitive effects in hypertensive patients may reflect CNS changes that occur prior to development of significant and sustained high blood pressure. Taken together with other findings, these results suggest that CNS changes may parallel, precede, or even contribute to blood pressure dysregulation in the early stages of the development of hypertension. Better understanding of the potential CNS origins of essential hypertension could lead to new strategies for treatment and prevention of this costly and widespread disease.
The authors gratefully acknowledge support from the Center for Advanced Study of Language, University of Maryland, College Park, MD, USA. They also thank their research staff, graduate students, and undergraduate students at Clemson University for their assistance in data gathering for the study and data management. These individuals include Jesse Allen, J. Adam Beeco, Tracy Lindquist, Bridget McConnell, Eric McKibben, Joe Mulvihill, Kristina O’Connell, Tyler Pierce, Ginger Pottorff, Susan Redmond, Hilary Schoon, Colin Shive, Sarah Spainhour, Rebekkah Wills, Cassandra Wright, Mark Zajack, Crystal Burnette, and Skye Gillispie.