The Journals of Gerontology Series B: Psychological Sciences and Social Sciences 63:P386-P389 (2008)
© 2008 The Gerontological Society of America
Exploring the Within-Person Coupling of Blood Pressure and Cognition in Elders
Alyssa A. Gamaldo,
Sarah R. Weatherbee and
Jason C. Allaire
Department of Psychology, North Carolina State University, Raleigh.
Address correspondence to Alyssa A. Gamaldo, Poe Hall, Box 7650, North Carolina State University, Raleigh, NC, 27695-7650. E-Mail: aagamald{at}ncsu.edu
 |
Abstract
|
|---|
In this study, we examined the relationship between within-person blood pressure and cognitive functioning. We conducted an analysis on 36 community-dwelling elderly individuals (age range = 60–87 years). Participants measured their blood pressure and completed cognitive tasks (i.e., the Rey Auditory Verbal Learning Task, the Letter Series test, and the Number Comparison test) twice a day over 60 consecutive days. We observed a significant interaction between within-person change in blood pressure and average blood pressure for the Letter Series test. Individuals with high blood pressure tended to perform poorly, particularly on occasions when their blood pressure level was above their personal average. These results demonstrate that the relationship between blood pressure and cognition at the between-person level and the relationship within each individual should be further explored simultaneously.
Key Words: Blood pressure • Cognition • Variability • Aging
High blood pressure (BP; i.e., hypertension) has been shown to be related to worse cognitive performance in older adults (Aleman, Muller, de Hann, & van der Schouw, 2005
; Insel, Palmer, Stroup-Benham, Markides, & Espino, 2005). However, some studies have reported that low BP (i.e., hypotension) also places individuals at risk for poor cognitive performance (Hestad, Kveberg, & Engedal, 2005; Paran, Anson, & Reuveni, 2003). In fact, high BP and extremely low BP have been shown to be related to deteriorating performance on cognitive tasks (André-Petersson, Hagberg, Janzon, & Steen, 2001; Guo, 1998). However, Paran and colleagues suggest that the quadratic association may not offer a better explanation for the BP–cognition relationship than a linear association.
Most studies examining the BP–cognition link have employed cross-sectional and traditional longitudinal designs. A number of researchers in the aging literature have begun to move away from such assessments in favor of "microlongitudinal" designs, which assess individuals on several occasions over short periods of time (Jones & Nesselroade, 1990; Nesselroade & Ford, 1985). Such a design is well suited to capture intraindividual variability (e.g., Hultsch & MacDonald, 2004; Nesselroade, 2004). Recent studies have found reliable and quantifiable intraindividual variability in cognition (e.g., Ram, Rabbitt, Stollery, & Nessleroade, 2005; Salthouse, Nesselroade, & Berish, 2006), stress (Neupert, Almeida, Mroczek, & Spiro, 2006), health, and daily activities (Ghisletta, Nesselroade, Featherman, & Rowe, 2002). These studies have observed that there are substantial fluctuations around a central mean, and that these fluctuations have important correlates.
Though previous research suggests that the relationship between BP and cognition might evolve gradually, there is some evidence that temporary within-person fluctuations in BP are related to variability in other constructs. Ong and Allaire (2005) observed that negative emotions on a particular occasion were associated with a concomitant but temporary increase in BP. Temporary increases in BP have also been linked to increases in stress (Carroll, Davey Smith, Sheffield, Shipley, & Marmot, 1995; Kamarck et al., 1997). Interestingly, temporary fluctuations in stress appear to be related to cognition, suggesting that on days of high stress, cognitive performance declines (Sliwinski, Smyth, Hofer, & Stawski, 2006). Thus, it is possible that a dynamic within-person association between BP and cognition may exist and suggests a cardiovascular reactivity to current environmental stressors. Although we did not assess stress in the current study, as a necessary first step, we wanted to explore the possible within-person association between BP and cognition.
We had two primary aims in the current study. First, we wanted to determine the amount of within-person and between-person variability found in BP. We expected that twice-daily assessments over the course of 2 months would reveal substantial intraindividual variability in the BP levels of elders. Second, we examined the within-person relationship between BP and cognitive functioning. We examined the coupling of BP and cognition and whether this coupling reflected a linear or quadratic relationship between BP and cognition.
|
METHODS
|
|---|
Participants
The sample consisted of 36 older adults, aged 60 to 87 years (M = 73, SD = 5.45; 26 women and 10 men), who were recruited from senior centers, church groups, and community organizations in the Detroit Metropolitan area. The average education of the participants was 16 years (SD = 2.98 years; range = 12–22 years); the average household income was $28,000 (SD = $6,000; range = $12,000–50,000+). There were 34 European Americans and 2 African Americans in the sample.
Measures
Our current analyses included the demographic variables of age, education, and gender. We assessed the cognitive abilities of memory, inductive reasoning, and perceptual speed by using psychometrically sound measures commonly used in the cognitive aging literature (see Allaire, 2001 and Allaire & Marsiske, 2005 for more detailed information about the cognitive measures). We assessed declarative memory by using the Rey Auditory Verbal Learning Task (Rey, 1941), which is a test of the ability to remember a list of semantically unrelated words (test–retest r = 62). We assessed inductive reasoning by using the Letter Series test (Thurstone, 1962), which required participants to identify the pattern in a series of letters and select the letter that would come next in the series (internal consistency, 
=.81). We assessed perceptual speed by using the Number Comparison test (Ekstrom, French, Harman, & Derman, 1976), which assesses how quickly participants can determine if two strings of numbers are the same (internal consistency, =.88). After the participants sat quietly for 5 minutes, we monitored their systolic blood pressure (SBP) while they were still in a seated position. We used the participants' nondominant arm, and we made the measurement with an Omron automatic digital BP monitor with IntelliSense (Omron, Schaumburg, IL).
Procedure
We asked the participants to complete daily workbooks twice a day, morning and evening, for 60 consecutive days (see Allaire & Marsiske, 2005). The daily workbooks included all the measures previously described. Prior to data collection, participants received instructions on how to self-administer all of the measures in the workbook, including information on how to use the automatic BP machine. After instructions were given, participants practiced self-administering the assessment under direct supervision. Participants received an instructional manual, which provided additional instructions, and pictorial demonstrations of the appropriate method for applying the BP cuff.
|
RESULTS
|
|---|
Brief Description of Multilevel Modeling
We ran three separate multilevel models to examine the association between BP and each cognitive measure. Each model represents a simultaneous regression analysis that is run concurrently at the within-person level (Level 1) and at the between-person level (Level 2). Following Raudenbush and Bryk (2002), the coupling parameter (Level 1 predictor) is each individual's average SBP across all measurement occasions deviated from his or her SBP on a particular occasion, as shown in the following equation. <--CO?1-->
where, β0 = intercept; occasional SBP reading when a participant's value on the predictors are both zero, u0j = within-person random component or error, and eij = between-person random component or error. Note that, after β0, the three sections of the equation show Level 1, Level 2, and the cross-level interaction.
A negative beta coefficient would suggest that the more an individual's SBP deviates above his or her average on a given occasion, the worse cognitive performance would be relative to the individual's average. The between-person (Level 2) parameter reflects the relationship between cognitive performance and an individual's mean SBP across the 120 occasions, so a negative beta would indicate that individuals with higher mean SBP perform worse on the cognitive measure across occasions.
We estimated the cross-level interaction to determine if the within-person relationship between SBP and cognitive performance differed, depending on mean level of SBP. We included the quadratic effect of occasion SBP as well as the quadratic effect of mean SBP in the model by squaring both the within- and the between-person SBP variables. We included the between-person variables of age, education, and gender as covariates.
We observed significant practice-related improvements for each of the cognitive variables (see Allaire & Marsiske, 2005). To control for practice effects, we included parameters representing linear and quadratic time trends in each model. Consequently, the results subsequently presented here have been adjusted for practice effects. Because the measurements were taken morning and evening, we included a time-of-day effect and the interaction between time of day and BP in each of the models. We did not observe any significant diurnal effects, and therefore we do not include them in the models reported.
Intraclass Correlations
We assessed the within- and between-person variability in BP by using the intraclass correlation (ICC), an intercept-only model. Results indicated that 29% of the total variance was within-person variance whereas 71% was between-person variance. When the model included linear and quadratic trends, 23% of the within-person variance remained. To ensure that there was sufficient within- and between-person variance in the dependent variables to conduct our hierarchical linear modeling analyses (Raudenbush & Bryk, 2002), we calculated intraclass correlations for each cognitive test. The results suggested that between 20% and 46% of the total variance was due to within-person variance and between 54% and 80% was due to between-person variance. When the model included the linear and quadratic practice effects, between 14% and 38% of the within-person variance remained; this serves as the baseline index of within-person variability.
Coupling of Blood Pressure and Cognitive Functioning
Occasion SBP and the quadratic effect of occasion SBP were both nonsignificant predictors of performance on all three cognitive measures (see Table 1). Age was a significant and negative predictor of the Letter Series test. Women performed significantly better than men did on the Rey Auditory Verbal Learning Task. Education, mean SBP, and quadratic mean SBP were not significantly associated with between-person differences in performance on any of the cognitive tests. We found a significant cross-level interaction between occasion SBP and mean SBP for the Letter Series test, indicating that the relationship between occasion SBP and Letter Series test performance at each occasion differed, depending on one's mean level SBP.
We estimated the coupling between occasion SBP and Letter Series performance at three levels of mean SBP (i.e., 1 SD above the mean SBP, mean SBP, and 1 SD below the mean SBP). A simple slopes analyses indicated that performance on the Letter Series test was significantly lower on high-SBP occasions (β = –.023, p <.05) only in participants who on average had high BP (i.e., SBP 
144.53). We found a similar, nonsignificant trend for individuals with average SBP (128.47; β = –.011, p >.05). We observed no significant association (β =.00, p >.05) for individuals with low SBP (SBP 
112.40). It appears that at a BP cutoff of around 130, individuals tend to perform significantly worse when their BP goes above their average level (see Table 2).
View this table:
[in this window]
[in a new window]
|
Table 2. Standardized Slopes Illustrating the Relationship Between BP Variability and Letter Series Performance as it Relates to Various Average BP Levels.
|
|
|
DISCUSSION
|
|---|
The current findings did not necessarily meet our expectations that there would be a dynamic within-person relationship between BP and cognition across all individuals, and they did not support previous literature suggesting that extremely high or low average BP levels alone were associated with worse cognitive performance, which is not surprising given the small sample size at the between-person level. However, our exploratory study did show an interesting finding that suggested that the within-person coupling was significantly moderated by between-person (i.e., average) BP; when individuals' average BP was above 129, which is considered prehypertensive, fluctuations in BP above their average level was significantly related to worse inductive reasoning performance. In fact, as illustrated in Table 2, the within-person coupling increases in magnitude as mean SBP increases. This significant within-person coupling relationship is not observed for individuals with average BP levels within a relatively normal range (i.e., SBP = 129–100).
The within-person association between BP and cognition may be a result of psychological stress, particularly in individuals who typically have prehypertension or hypertension. Given that fluctuations in stress are associated with fluctuations in cognition, individuals with at least prehypertension may be experiencing increased stress, which may impact their cognitive performance.
Although this study reveals some interesting findings, there are limitations that should be noted. First, we could not collect data on the use of antihypertensive medications. Because treated hypertension has been shown to be associated with better cognitive performance than untreated hypertension (Paran et al., 2003), it possible that not controlling for medicated hypertension resulted in the washing out of potential effects. Second, the majority of the participants in the current sample were well-educated, middle- to high-income Caucasian elders, which may have served to constrain our results. Third, the current study essentially has a correlational design. Thus, causality or the direction of the relationship between BP and cognition cannot be determined. Because each BP measurement occurred after the cognitive assessment, it is possible that metacognitive processes associated with poor cognitive performance could have subsequently elevated BP. Fourth, the current study examined a limited number of cognitive measures, so more significant findings may have been observed with a more multidimensional cognitive battery. Finally, given the unsupervised nature of the study, it is uncertain whether participants properly followed testing instructions. Though this is an important limitation, the frequency and duration of the testing occasions constitute a strength of the current study and could have only been achieved through an at-home self-administration.
The current study is unique because it is, perhaps, the first study to explore how cognition is associated with BP on a day-to-day basis. In addition, our study's findings warrant further investigation of the influence of between- and within-person differences in BP on various cognitive abilities in larger sample sizes. Future studies should also investigate whether the relationship between daily BP and cognitive performance is moderated by perceived stress.
|
Footnotes
|
|---|
Decision Editor: Elizabeth Stine-Morrow, PhD
Received for publication October 10, 2007.
Accepted for publication September 4, 2008.
|
References
|
|---|
- Allaire, J. C. (2001). Performance gains and fluctuations: The effects of cognitive practice and the role of intraindividual cognitive variability in older adults. Unpublished doctoral Dissertation, Wayne State University, Detroit, Michigan.
- Allaire, J. C., Marsiske, M. (2005). Intraindividual variability may not always indicate vulnerability in elders' cognitive performance. Psychology and Aging, 20,390-401.[Medline]
- Aleman, A., Muller, M., de Hann, E. H. F., van der Schouw, Y. T. (2005). Vascular risk factors and cognitive function in a sample of independently living men. Neurobiology of Aging, 26,485-490.[Medline]
- André-Petersson, L., Hagberg, B., Janzon, L., Steen, G. (2001). A comparison of cognitive ability in normotensive and hypertensive 68-year-old men: Results from population study "Men Born in 1914," in Malmö, Sweden. Experimental Aging Research, 27,319-340.[Medline]
- Carroll, D., Davey Smith, G., Sheffield D., Shipley, M. J., Marmot, M. G. (1995). Pressor reactions to psychological stress and prediction of future blood pressure: Data from the Whitehall II study. British Medical Journal, 310,771-775.[Abstract/Free Full Text]
- Ekstrom, R. B., French, J. W., Harman, H. H., Derman, D. (1976). Kit of factor-referenced cognitive tests. Princeton, NJ: Educational Testing Service.
- Ghisletta, P., Nesselroade, J. R., Featherman, D. L., Rowe, J. W. (2002). Structure and predictive power of intraindividual variability in health and activity measures. Swiss Journal of Psychology, 2,73-83.
- Guo, Z. (1998). Blood pressure and dementia in the very old. An epidemiologic study. Stockholm, Sweden: Karolinska Institute and Stockholm Gerontology Research Center.
- Hestad, K., Kveberg, B., Engedal, K. (2005). Low blood pressure is a better predictor of cognitive deficits than the apolipoprotein e4 allele in the oldest old. Acta Neurologica Scandinavica, 111,323-328.[Medline]
- Hultsch, D. E., MacDonald, S. W. S. (2004). Intraindividual variability in performance as a theoretical window into cognitive aging. In R. A. Dixon, L. Bäckman, & L. G. Nilsson (Eds.), New frontiers in cognitive aging (pp. 65–89). New York: Oxford University Press.
- Insel, K. C., Palmer, R. F., Stroup-Benham, C. A., Markides, K. S., Espino, D. V. (2005). Association between change in systolic blood pressure and cognitive decline among elderly Mexican Americans: Data from the Hispanic Established Population for Epidemiology study of the elderly. Experimental Aging Research, 31,35-54.[Medline]
- Jones, C. J., Nesselroade, R. J. (1990). Multivariate, replicated, single-subject designs and P-technique factor analysis: A selective review of intraindividual change studies. Experimental Aging Research, 16,171-183.[Medline]
- Kamarck, T. W., Everson, S. A., Kaplan, G. A., Manuck, S. B., Jennings, R., Salonen, R., et al. (1997). Exaggerated blood pressure responses during mental stress are associated with enhanced carotid artherosclerosis in middle-aged finish men. Circulation, 96,3842-3848.[Abstract/Free Full Text]
- Nesselroade, J. R. (2004). Intraindividual variability and short-term change. Gerontology, 50,44-47.[Medline]
- Nesselroade, J. R., Ford, D. H. (1985). P-technique comes of age: Multivariate, replicated, single subject designs for research on older adults. Research on Aging, 7,46-80.[Abstract]
- Neupert, S. D., Almeida, D. M., Mroczek, D. K., Spiro, A. (2006). Daily stressors and memory failures in a naturalistic setting: Findings from the VA Normative Aging Study. Psychology and Aging, 21,424-429.[Medline]
- Ong, A., Allaire, J. C. (2005). Cardiovascular intraindividual variability in later life: The influence of social connectedness and positive emotions. Psychology and Aging, 20,476-485.[Medline]
- Paran, E., Anson, O., Reuveni, H. (2003). Blood pressure and cognitive functioning among independent elderly. American Journal of Hypertension, 16,818-826.[Medline]
- Ram, N., Rabbitt, P., Stollery, B., Nesselroade, J. R. (2005). Cognitive performance inconsistency: Intraindividual change and variability. Psychology and Aging, 20,623-633.[Medline]
- Raudenbush, S. W., Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). Thousand Oaks, CA: Sage.
- Rey, A. (1941). Psychological examination of traumatic encephalopathy. Archives de Psychologie, 28,286-340.
- Salthouse, T., Nesselroade, J., Berish, D. E. (2006). Short-term variability in cognitive performance and the calibration of longitudinal change. Journal of Gerontology: Psychological Sciences, 61B,P144-P151.[Abstract/Free Full Text]
- Sliwinski, M. J., Symth, J. M., Hofer, S. M., Stawski, R. S. (2006). Intraindividual coupling of daily stress and cognition. Psychology and Aging, 21,545-557.[Medline]
- Thurstone, T. G. (1962). Primary mental ability for Grades 9–12 (Rev. ed.). Chicago, IL: Science Research Associates.
TOP
Abstract
Methods
