Orthostatic Hypotension and Cognitive Function: Cross-sectional Results From the ELSA-Brasil Study

Orthostatic Hypotension and Cognitive Function: Cross-sectional Results From the ELSA-Brasil Study Abstract Background The association between orthostatic hypotension (OH) and cognitive impairment is controversial, and most studies have investigated older white adults from Western Europe and the United States. Therefore, we investigated the association between OH and cognitive performance in a large and racially diverse sample of adults using cross-sectional data from the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil). Methods OH was defined when systolic blood pressure decreased 20 mmHg and/or diastolic blood pressure decreased 10 mmHg from supine to standing position. We investigated the association between OH and composite global cognition, memory, verbal fluency (VF), and Trail Making Test z-scores, using multiple linear regression models. We also investigated the association of orthostatic hypertension and systolic blood pressure/diastolic blood pressure changes with cognitive performance, as well as the interaction between OH and compensatory heart rate after postural change on cognitive performance. Results We evaluated 12,826 participants (mean age = 51.5 ± 9.0 years, 46% male, 53% white). Participants with OH (4% of the sample) had poorer z-scores for VF (β = −0.108, 95% confidence interval = −0.189; −0.025, p = .01) than participants without OH. Orthostatic hypertension was also associated with worse performance on the VF test (β = −0.080, 95% confidence interval = −0.157; −0.003, p = .04). Systolic blood pressure orthostatic change had a nonlinear association with VF. The interaction terms between OH and compensatory increase in heart rate for the Trail Making Test z-score (p = .09) was borderline significant, suggesting that participants who lack compensatory heart rate after postural change might have poorer performance. Conclusion OH and orthostatic hypertension were associated with poorer performance on the VF test in participants from Brazil. Hypotension, Orthostatic, Hypertension, Cognition, Dementia Orthostatic hypotension (OH) occurs when blood pressure drops substantially as a result of postural change from supine to upright position. It is classically defined as a drop of 20 mmHg or more in systolic blood pressure (SBP) and/or a drop of 10 mmHg or more in diastolic pressure (DBP) after 3 minutes of standing (1). OH is more frequent among older adults and those with functional limitations (2,3). In a recent meta-analysis of the association of OH with different outcomes, OH was associated with increased risk of cardiovascular events and overall mortality, whereas the evidence on cognitive outcomes was less clear (4). In the baseline analysis of the Irish Longitudinal Study of Aging (TILDA), participants with OH had poorer global cognitive performance, particularly among women and when hypertension was present in the supine position (5,6). However, OH was not associated with cognitive decline in the follow-up of the TILDA and in the ARIC (Atherosclerosis Risk in Communities) studies (7,8). Moreover, most of the evidence showing the association between OH and cognitive impairment comes from the United States and Western Europe and from older participants (2,6,7), with limited data from low-middle income countries and middle-aged participants (9,10). Therefore, the goal in this study was to investigate the cross-sectional association between OH and cognitive function in a large sample of participants aged 35–74 years from the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil). Material and Methods Participants The ELSA-Brasil is a longitudinal study of 15,105 employees from public institutions in six Brazilian cities (São Paulo, Belo Horizonte, Porto Alegre, Salvador, Rio de Janeiro, and Vitória). Participants were aged 35–74 years old and did not have dementia at enrollment (2008–2010). The study was approved by the local institutional review boards, and all the participants signed an informed consent form at the time of enrollment. A detailed description of the study can be found elsewhere (11,12). For this analysis, we excluded participants with a self-reported medical diagnosis of stroke to minimize the effect of previous cerebrovascular disease on cognitive function. We also excluded participants who regularly used some medications (antipsychotic medications, antiparkinsonian agents, and anticonvulsants) that indicated the presence of active neurologic or psychiatric diseases that could directly affect cognitive performance. Finally, we excluded participants with incomplete data for the cognitive tests or OH. Cognitive Function Evaluation Cognitive function was assessed using the following tests: (a) the Consortium to Establish a Registry for Alzheimer’s Disease Word List Memory Test (CERAD-WLMT) (13); (b) the Phonemic Verbal Fluency Test (PVFT) (14); and (c) the Trail Making Test version B (TMT) (15). Trained examiners administered the tests in a fixed order during a single session, in a quiet room with good lighting and low levels of noise or other distractions. We used the Brazilian version of the CERAD-WLMT (16). This test comprises immediate word recall, delayed word recall, and word recognition and evaluates the memory domain (13). Participants were asked to read and learn 10 words after three exposures; the sum of the number of words recalled in each of these attempts was the score in the immediate word recall. After a 5-minute filled delay, the participants were given 60 seconds to record the words. The delayed word recall score was equal to the number of recalled words. Finally, 20 words were presented and the participants had to recognize the 10 words that were presented previously. In the PVFT, participants were asked to generate as many words as possible that start with the letter F to evaluate language and executive ability. The score on this test was the total number of generated words (17). Finally, the TMT is a test of executive function, processing speed, and visual-spatial organization. The participants were instructed to draw lines connecting letters and numbers in an order that alternated between increasing numeric values and alphabetic order. The test score was the total time taken to complete the task, in seconds. To facilitate comparisons across cognitive tests, we generated z-scores by subtracting the participant’s score in each test from the sample mean score and dividing the difference by the standard deviation (SD) of the sample (18,19). Therefore, a z-score of −1 describes a cognitive performance that is 1 SD below the mean sample score. We combined the three scores in the CERAD-WLMT (ie, immediate recall, delayed recall, and recognition) by averaging the z-scores for each test and then standardizing this averaged score. We used the same process to generate z-scores for the PVFT and TMT. We inverted the TMT z-score signal because higher scores indicate poorer performance in the original TMT, whereas higher scores on the other tests are related to better performance. Finally, we averaged and standardized the z-scores of all tests as a measure of composite global cognitive score. Orthostatic Hypotension and Blood Pressure Assessment We followed the procedures of the ARIC study to assess OH (7,20). Blood pressure and heart rate (HR) were measured in the right arm using an automatic oscillometric device (Omron HEM 705CPINT) after a rest period of 5 minutes in a seated position, in a quiet, temperature-controlled room (20°C–24°C). Three measurements were taken at 1-minute intervals, and the mean of the two latest measurements was considered as the seated blood pressure. Blood pressure after postural change was measured in another moment, after evaluation of the ankle–brachial index. Blood pressure and HR were measured in the right arm using the automatic oscillometric device after 20 minutes of rest. The participant was then asked to stand up, and both measurements were reassessed at 2, 3, and 5 minutes after standing. OH was defined by a decrease of at least 20 mmHg in SBP and/or a decrease of at least 10 mmHg in DBP for any of the three measurements after standing. On the other hand, orthostatic hypertension was defined by an increase of at least 20 mmHg in SBP and/or an increase of at least 10 mmHg in DBP for any of the three measurements after standing (21). Changes in orthostatic SBP and DBP were calculated by the mean of the differences between each of the three standing position blood pressure measurements and the supine position measurement. The mean change in SBP measurement between these positions was then standardized. For that, the SBP change z-score was calculated by subtracting the participant’s SBP difference in response to postural change from the sample mean SBP change and dividing this difference by the sample SD for SBP change. The same calculation was performed for the mean difference in DBP. Therefore, an SBP change z-score of −1 would describe an SBP change that is 1 SD below the sample mean for SBP change. Moreover, we calculated the maximum HR response after postural change (ie, maximum difference between any of the three standing measurements and the supine measurement). Cognitive evaluation and OH assessment were performed within a 3-hour interval on the same day. Other Measurements We collected information on sociodemographic variables: age, sex, race, and education. Race and level of education were self-reported. Race was classified in white, black, brown (admixed race of white and black), or other race (ie, Asian and indigenous), and education was classified in high school or less versus college or more. Diabetes was self-reported and categorized into three groups: no history of diabetes, history of diabetes ≤5 years (median time of diabetes diagnosis in our sample), and history of diabetes >5 years. The following clinical information was self-reported: cardiovascular disease (myocardial infarction, myocardial revascularization, or heart failure), use of lipid-lowering drugs, alcohol use, smoking, and physical leisure activity. Alcohol use and smoking were categorized in never, former, and current user. Leisure activity was categorized into three groups: no physical activity, ≤150 min/wk, and >150 min/wk). Weight and height were measured in participants with light clothing and without shoes, and body mass index were calculated in kilogram per square meter. Low-density lipoprotein cholesterol was determined by the enzymatic colorimetric model. The diagnosis of depression was based on the International Classification of Diseases (ICD-10) criteria (22) and used the participant’s response to the Clinical Interview Schedule Revised (23). The use of antidepressants was reported by the participant. Thyroid stimulating hormone levels (TSH) and free thyroxine (FT4) were determined using a third generation immunoenzymatic assay (Siemens, Deerfield, IL). Thyroid function was classified as follows: (a) normal: TSH from 0.4 to 4 mIU/L and FT4 levels from 0.8 to 1.9 ng/dL; (b) hypothyroidism: TSH > 4.0 mIU/L and FT4 < 0.8 ng/dL; and (c) hyperthyroidism: TSH < 0.4 mIU/dL and FT4 > 1.9 ng/dL. We also investigated the use of drugs that could cause OH among participant’s self-reported medication: antihypertensive drugs, vasodilators, muscle relaxants, anticholinergics, sedatives, and alpha-blockers (8,24). Statistical Analysis We compared participants with and without OH using the unpaired t test for continuous variables and chi-square test for categorical variables. The independent variable was the OH status. We investigated the association between the dependent and independent variables using linear regression models. In these models, the dependent variables were the z scores for composite global, memory, verbal fluency, and TMT. We presented the beta coefficients for the crude model, as well as for the hierarchical linear regression models adjusted for (a) age and sex; (b) age, sex, race, and education; (c) age, sex, race, education, seated SBP, use of drugs that could cause OH, diabetes, low-density lipoprotein cholesterol, use of lipid-lowering drugs, cardiovascular disease, body mass index, alcohol use, smoking, physical activity, depression, use of antidepressants, and thyroid function. We also tested for the association between orthostatic hypertension and cognitive performance. We determined the association of cognitive performance with changes in orthostatic SBP and DBP (ie, mean change in SBP and DBP from supine to standing position in 2, 3, and 5 minutes) continuously by SD increase, using multiple linear regression models adjusted for the same set of possible confounders cited above. We tested for nonlinear associations of postural changes in SBP and DBP z-scores with cognitive performance by adding a quadratic terms for SBP/DBP change z-scores. Model selection between the linear and the quadratic models was based on the Akaike information criterion. Alternatively, we also investigated the association of cognitive performance with the absolute values of SBP and DBP change (per 10 mmHg) at 2, 3, and 5 minutes after standing. In addition, we tested whether the association between OH and cognitive test performance was modified by the HR response after postural change, using an interaction term between the OH and HR response in multiple linear models adjusted for the same set of confounders. To facilitate interpretation, HR response after postural change was used as a binary variable, using the median in the sample as the cutoff (≤14 bpm vs. >14 bpm). Moreover, we tested whether age (<65 vs. ≥65 years), sex, race (white vs. nonwhite, excluding Asian and other races), hypertension, diabetes, cardiovascular disease, depression, and drugs that may cause OH (25,26) would change the association between OH and cognitive performance by including an interaction term between OH and these variables in multiple linear regression models adjusted for the set of variables described above. As sensitivity analysis, we investigated the association of cognitive performance with early OH (ie, evaluated at 2 minutes after postural change), classical OH (ie, evaluated at 3 minutes after postural change), and delayed OH (ie, evaluated at 5 minutes after postural change), using the hierarchical linear models described above. Finally, we investigated the clinical meaning of our findings by testing the association between OH and cognitive impairment, defined as scores on cognitive tests 1 SD below the sample mean (27). We used hierarchical logistic regression models adjusted for the same set of possible confounders. We used Stata 13.0 software (StataCorp, College Station, TX) to perform the statistical analyses. The alpha level was set at .05, and tests were two tailed. Results Of the 15,105 eligible participants, 12,826 fulfilled the study criteria and 2,279 were excluded from this study (Supplementary Figure 1). The mean age of the sample was 51.5 ± 9.0 years, 46% were male, 53% were white, and 56% had college education or more. Hypertension was present in 34% of the sample, OH in 4%, orthostatic hypertension in 5%. OH was diagnosed in 1.7% of participants at 2 minutes of standing, 2.8% at 3 minutes, and 2.4% at 5 minutes. The mean SBP was 120.6 ± 16.6 mmHg, and mean DBP was 56.6 ± 14.4 mmHg (Supplementary Table 1). Participants recorded an average of 7.0 ± 2.0 words in the delayed word recall, 12.8 ± 4.4 words on the verbal fluency test, and completed the TMT in 121.5 ± 83.7 seconds. Participants with OH were older, more likely to be female, and had less education than participants without OH. The diagnosis of hypertension was more frequent among participants with OH, as well as the fact that the SBP and DBP measured in the seated position were higher in the OH group compared with those without OH. In addition, use of lipid-lowering drugs, cardiovascular disease, and depression were also more frequent among participants with OH (Table 1). Table 1. Baseline Characteristics of the Study Population by Orthostatic Hypotension Status (n = 12,826) Variable Without OH With OH p n = 12,293 n = 533 Immediate word recallc, mean (SD)a 21.3 (3.8) 20.8 (4.1) .0007 Delayed word recallc, mean (SD)a 7.0 (2.0) 6.8 (2.0) .001 Word recognitionc, mean (SD)a 9.6 (0.8) 9.5 (0.9) .08 Verbal fluencyc, mean (SD)a 12.8 (4.4) 11.9 (4.2) <.0001 Trail making test (s), mean (SD)a 120.9 (83.0) 135.1 (97.8) .001 Age (y), mean (SD)a 51.4 (8.9) 54.5 (9.3) <.0001 Male, %b 45.7 41.3 .04 Race, %b .12  White 53.6 51.2  Black 15.0 18.6  Brown 27.9 27.4  Other 3.5 2.8 College education or more, %b 56.1 48.8 .0009 Hypertension, %b 33.4 44.5 <.0001 Seated SBP (mmHg), mean (SD)a, d 120.4 (16.7) 124.8 (20.6) <.0001 Seated DBP (mmHg), mean (SD)a, d 76.0 (10.6) 77.3 (12.3) .02 SBP change(mmHg), mean (SD)a, e −4.3 (7.6) 15.0 (9.8) <.0001 DBP change(mmHg), mean (SD)a, e −7.7 (5.2) 3.7 (7.8) <.0001 SBP change z-score, mean (SD)a, e −0.1 (0.9) 2.1 (1.1) <.0001 DBP change z-score, mean (SD)a, e −0.1 (0.9) 1.9 (1.3) <.0001 Use of drugs that can cause OH, %b 31.3 41.5 <.0001 Diabetes, %b .16  No 91.2 88.7  Yes, ≤5 y 5.0 6.4  Yes, >5 y 3.8 4.9 LDL cholesterol (mg/dL), mean (SD)a 130.8 (34.7) 131.3 (36.3) .72 Use of lipid-lowering drugs, %b 12.2 16.7 .002 Cardiovascular diseasef, n (%)b 2.3 6.2 <.0001 Body mass index (kg/m2), mean (SD)a 26.9 (4.7) 26.9 (4.8) .94 Smoking, %b .52  Never 57.7 58.2  Former 29.6 30.8  Current 12.7 11.1 Alcohol use, %b .13  Never 9.8 12.2  Former 18.6 19.7  Current 71.6 68.1 Physically active, %b .06  No 61.9 66.8  Yes, <150 min/wk 12.8 12.2  Yes, ≥150 min/wk 25.3 21.0 Depression, %b 3.7 6.0 .006 Use of antidepressants, %b 4.9 5.6 .45 Thyroid function status, %b .12  Normal 85.9 83.3  Clinical or subclinical hypothyroidism 12.2 13.9  Clinical or subclinical hyperthyroidism 1.8 2.8 Variable Without OH With OH p n = 12,293 n = 533 Immediate word recallc, mean (SD)a 21.3 (3.8) 20.8 (4.1) .0007 Delayed word recallc, mean (SD)a 7.0 (2.0) 6.8 (2.0) .001 Word recognitionc, mean (SD)a 9.6 (0.8) 9.5 (0.9) .08 Verbal fluencyc, mean (SD)a 12.8 (4.4) 11.9 (4.2) <.0001 Trail making test (s), mean (SD)a 120.9 (83.0) 135.1 (97.8) .001 Age (y), mean (SD)a 51.4 (8.9) 54.5 (9.3) <.0001 Male, %b 45.7 41.3 .04 Race, %b .12  White 53.6 51.2  Black 15.0 18.6  Brown 27.9 27.4  Other 3.5 2.8 College education or more, %b 56.1 48.8 .0009 Hypertension, %b 33.4 44.5 <.0001 Seated SBP (mmHg), mean (SD)a, d 120.4 (16.7) 124.8 (20.6) <.0001 Seated DBP (mmHg), mean (SD)a, d 76.0 (10.6) 77.3 (12.3) .02 SBP change(mmHg), mean (SD)a, e −4.3 (7.6) 15.0 (9.8) <.0001 DBP change(mmHg), mean (SD)a, e −7.7 (5.2) 3.7 (7.8) <.0001 SBP change z-score, mean (SD)a, e −0.1 (0.9) 2.1 (1.1) <.0001 DBP change z-score, mean (SD)a, e −0.1 (0.9) 1.9 (1.3) <.0001 Use of drugs that can cause OH, %b 31.3 41.5 <.0001 Diabetes, %b .16  No 91.2 88.7  Yes, ≤5 y 5.0 6.4  Yes, >5 y 3.8 4.9 LDL cholesterol (mg/dL), mean (SD)a 130.8 (34.7) 131.3 (36.3) .72 Use of lipid-lowering drugs, %b 12.2 16.7 .002 Cardiovascular diseasef, n (%)b 2.3 6.2 <.0001 Body mass index (kg/m2), mean (SD)a 26.9 (4.7) 26.9 (4.8) .94 Smoking, %b .52  Never 57.7 58.2  Former 29.6 30.8  Current 12.7 11.1 Alcohol use, %b .13  Never 9.8 12.2  Former 18.6 19.7  Current 71.6 68.1 Physically active, %b .06  No 61.9 66.8  Yes, <150 min/wk 12.8 12.2  Yes, ≥150 min/wk 25.3 21.0 Depression, %b 3.7 6.0 .006 Use of antidepressants, %b 4.9 5.6 .45 Thyroid function status, %b .12  Normal 85.9 83.3  Clinical or subclinical hypothyroidism 12.2 13.9  Clinical or subclinical hyperthyroidism 1.8 2.8 Note: DBP = diastolic blood pressure; LDL = low-density lipoprotein; OH = orthostatic hypotension; SBP = systolic blood pressure. aInpaired t test. bChi-square test. cNumber of recorded words. dMean of the two last measurements of blood pressure in the seated position. eMean of the differences between each of the three blood pressure measurements in the standing position and the one in the supine position. fMyocardial infarction, myocardial revascularization, or heart failure. View Large Table 1. Baseline Characteristics of the Study Population by Orthostatic Hypotension Status (n = 12,826) Variable Without OH With OH p n = 12,293 n = 533 Immediate word recallc, mean (SD)a 21.3 (3.8) 20.8 (4.1) .0007 Delayed word recallc, mean (SD)a 7.0 (2.0) 6.8 (2.0) .001 Word recognitionc, mean (SD)a 9.6 (0.8) 9.5 (0.9) .08 Verbal fluencyc, mean (SD)a 12.8 (4.4) 11.9 (4.2) <.0001 Trail making test (s), mean (SD)a 120.9 (83.0) 135.1 (97.8) .001 Age (y), mean (SD)a 51.4 (8.9) 54.5 (9.3) <.0001 Male, %b 45.7 41.3 .04 Race, %b .12  White 53.6 51.2  Black 15.0 18.6  Brown 27.9 27.4  Other 3.5 2.8 College education or more, %b 56.1 48.8 .0009 Hypertension, %b 33.4 44.5 <.0001 Seated SBP (mmHg), mean (SD)a, d 120.4 (16.7) 124.8 (20.6) <.0001 Seated DBP (mmHg), mean (SD)a, d 76.0 (10.6) 77.3 (12.3) .02 SBP change(mmHg), mean (SD)a, e −4.3 (7.6) 15.0 (9.8) <.0001 DBP change(mmHg), mean (SD)a, e −7.7 (5.2) 3.7 (7.8) <.0001 SBP change z-score, mean (SD)a, e −0.1 (0.9) 2.1 (1.1) <.0001 DBP change z-score, mean (SD)a, e −0.1 (0.9) 1.9 (1.3) <.0001 Use of drugs that can cause OH, %b 31.3 41.5 <.0001 Diabetes, %b .16  No 91.2 88.7  Yes, ≤5 y 5.0 6.4  Yes, >5 y 3.8 4.9 LDL cholesterol (mg/dL), mean (SD)a 130.8 (34.7) 131.3 (36.3) .72 Use of lipid-lowering drugs, %b 12.2 16.7 .002 Cardiovascular diseasef, n (%)b 2.3 6.2 <.0001 Body mass index (kg/m2), mean (SD)a 26.9 (4.7) 26.9 (4.8) .94 Smoking, %b .52  Never 57.7 58.2  Former 29.6 30.8  Current 12.7 11.1 Alcohol use, %b .13  Never 9.8 12.2  Former 18.6 19.7  Current 71.6 68.1 Physically active, %b .06  No 61.9 66.8  Yes, <150 min/wk 12.8 12.2  Yes, ≥150 min/wk 25.3 21.0 Depression, %b 3.7 6.0 .006 Use of antidepressants, %b 4.9 5.6 .45 Thyroid function status, %b .12  Normal 85.9 83.3  Clinical or subclinical hypothyroidism 12.2 13.9  Clinical or subclinical hyperthyroidism 1.8 2.8 Variable Without OH With OH p n = 12,293 n = 533 Immediate word recallc, mean (SD)a 21.3 (3.8) 20.8 (4.1) .0007 Delayed word recallc, mean (SD)a 7.0 (2.0) 6.8 (2.0) .001 Word recognitionc, mean (SD)a 9.6 (0.8) 9.5 (0.9) .08 Verbal fluencyc, mean (SD)a 12.8 (4.4) 11.9 (4.2) <.0001 Trail making test (s), mean (SD)a 120.9 (83.0) 135.1 (97.8) .001 Age (y), mean (SD)a 51.4 (8.9) 54.5 (9.3) <.0001 Male, %b 45.7 41.3 .04 Race, %b .12  White 53.6 51.2  Black 15.0 18.6  Brown 27.9 27.4  Other 3.5 2.8 College education or more, %b 56.1 48.8 .0009 Hypertension, %b 33.4 44.5 <.0001 Seated SBP (mmHg), mean (SD)a, d 120.4 (16.7) 124.8 (20.6) <.0001 Seated DBP (mmHg), mean (SD)a, d 76.0 (10.6) 77.3 (12.3) .02 SBP change(mmHg), mean (SD)a, e −4.3 (7.6) 15.0 (9.8) <.0001 DBP change(mmHg), mean (SD)a, e −7.7 (5.2) 3.7 (7.8) <.0001 SBP change z-score, mean (SD)a, e −0.1 (0.9) 2.1 (1.1) <.0001 DBP change z-score, mean (SD)a, e −0.1 (0.9) 1.9 (1.3) <.0001 Use of drugs that can cause OH, %b 31.3 41.5 <.0001 Diabetes, %b .16  No 91.2 88.7  Yes, ≤5 y 5.0 6.4  Yes, >5 y 3.8 4.9 LDL cholesterol (mg/dL), mean (SD)a 130.8 (34.7) 131.3 (36.3) .72 Use of lipid-lowering drugs, %b 12.2 16.7 .002 Cardiovascular diseasef, n (%)b 2.3 6.2 <.0001 Body mass index (kg/m2), mean (SD)a 26.9 (4.7) 26.9 (4.8) .94 Smoking, %b .52  Never 57.7 58.2  Former 29.6 30.8  Current 12.7 11.1 Alcohol use, %b .13  Never 9.8 12.2  Former 18.6 19.7  Current 71.6 68.1 Physically active, %b .06  No 61.9 66.8  Yes, <150 min/wk 12.8 12.2  Yes, ≥150 min/wk 25.3 21.0 Depression, %b 3.7 6.0 .006 Use of antidepressants, %b 4.9 5.6 .45 Thyroid function status, %b .12  Normal 85.9 83.3  Clinical or subclinical hypothyroidism 12.2 13.9  Clinical or subclinical hyperthyroidism 1.8 2.8 Note: DBP = diastolic blood pressure; LDL = low-density lipoprotein; OH = orthostatic hypotension; SBP = systolic blood pressure. aInpaired t test. bChi-square test. cNumber of recorded words. dMean of the two last measurements of blood pressure in the seated position. eMean of the differences between each of the three blood pressure measurements in the standing position and the one in the supine position. fMyocardial infarction, myocardial revascularization, or heart failure. View Large Although OH and performance on all the cognitive tests were associated in simple linear models, only the PVFT z-score (adjusted β = −0.107, 95% confidence interval = −0.189; −0.025, p = .01) was associated with OH in multivariable analyses (Table 2). The associations of OH with cognitive impairment in all tests were not significant (Supplementary Table 2). In sensitivity analysis, we investigated whether cognitive performance would be associated with early, classical, or delayed OH. We found borderline associations of PVFT with classical OH (adjusted β = −0.081, 95% confidence interval = −0.179; 0.018, p = .11) and delayed OH (adjusted β = −0.097, 95% confidence interval = −0.204; 0.010, p = .08) in fully adjusted models (Supplementary Table 3). Similarly, orthostatic hypertension was associated with worse performance in all tests in simple linear models, but only the association with the PVFT z-score remained significant in the fully adjusted model (adjusted β = −0.080, 95% confidence interval = −0.157; −0.003, p = .04; Table 3). Table 2. Association Between Orthostatic Hypotension and Cognitive Performance (n = 12,826) Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.239 (−0.326; −0.152) <.0001 −0.159 (−0.242; −0.076) <.0001 −0.083 (−0.156; −0.010) .02 −0.071 (−0.143; 0.001) .05 Memory −0.150 (−0.237; −0.063) .001 −0.093 (−0.176; −0.010) .03 −0.048 (−0.127; 0.032) .24 −0.050 (−0.129; 0.030) .22 Verbal fluency −0.205 (−0.292; −0.118) <.0001 −0.169 (−0.255; −0.0.82) <.0001 −0.119 (−0.201; −0.037) .004 −0.107 (−0.189; −0.025) .01 Trail making −0.169 (−0.256; −0.083) <.0001 −0.087 (−0.171; −0.002) .04 −0.015 (−0.091; 0.060) .69 0.0001 (−0.075; 0.075) 1.00 Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.239 (−0.326; −0.152) <.0001 −0.159 (−0.242; −0.076) <.0001 −0.083 (−0.156; −0.010) .02 −0.071 (−0.143; 0.001) .05 Memory −0.150 (−0.237; −0.063) .001 −0.093 (−0.176; −0.010) .03 −0.048 (−0.127; 0.032) .24 −0.050 (−0.129; 0.030) .22 Verbal fluency −0.205 (−0.292; −0.118) <.0001 −0.169 (−0.255; −0.0.82) <.0001 −0.119 (−0.201; −0.037) .004 −0.107 (−0.189; −0.025) .01 Trail making −0.169 (−0.256; −0.083) <.0001 −0.087 (−0.171; −0.002) .04 −0.015 (−0.091; 0.060) .69 0.0001 (−0.075; 0.075) 1.00 Note: CI = confidence interval; LDL = low-density lipoprotein; OH = orthostatic hypotension. Model 1: linear regression model, adjusted for age and sex. Model 2: Model 1 + race and education. Model 3: Model 2 + seated systolic blood pressure, use of drugs that can cause OH, diabetes, LDL cholesterol, use of lipid-lowering drugs, cardiovascular disease, body mass index, alcohol use, smoking, physical activity, depression, antidepressant use, and thyroid function. View Large Table 2. Association Between Orthostatic Hypotension and Cognitive Performance (n = 12,826) Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.239 (−0.326; −0.152) <.0001 −0.159 (−0.242; −0.076) <.0001 −0.083 (−0.156; −0.010) .02 −0.071 (−0.143; 0.001) .05 Memory −0.150 (−0.237; −0.063) .001 −0.093 (−0.176; −0.010) .03 −0.048 (−0.127; 0.032) .24 −0.050 (−0.129; 0.030) .22 Verbal fluency −0.205 (−0.292; −0.118) <.0001 −0.169 (−0.255; −0.0.82) <.0001 −0.119 (−0.201; −0.037) .004 −0.107 (−0.189; −0.025) .01 Trail making −0.169 (−0.256; −0.083) <.0001 −0.087 (−0.171; −0.002) .04 −0.015 (−0.091; 0.060) .69 0.0001 (−0.075; 0.075) 1.00 Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.239 (−0.326; −0.152) <.0001 −0.159 (−0.242; −0.076) <.0001 −0.083 (−0.156; −0.010) .02 −0.071 (−0.143; 0.001) .05 Memory −0.150 (−0.237; −0.063) .001 −0.093 (−0.176; −0.010) .03 −0.048 (−0.127; 0.032) .24 −0.050 (−0.129; 0.030) .22 Verbal fluency −0.205 (−0.292; −0.118) <.0001 −0.169 (−0.255; −0.0.82) <.0001 −0.119 (−0.201; −0.037) .004 −0.107 (−0.189; −0.025) .01 Trail making −0.169 (−0.256; −0.083) <.0001 −0.087 (−0.171; −0.002) .04 −0.015 (−0.091; 0.060) .69 0.0001 (−0.075; 0.075) 1.00 Note: CI = confidence interval; LDL = low-density lipoprotein; OH = orthostatic hypotension. Model 1: linear regression model, adjusted for age and sex. Model 2: Model 1 + race and education. Model 3: Model 2 + seated systolic blood pressure, use of drugs that can cause OH, diabetes, LDL cholesterol, use of lipid-lowering drugs, cardiovascular disease, body mass index, alcohol use, smoking, physical activity, depression, antidepressant use, and thyroid function. View Large Table 3. Association Between Orthostatic Hypertension and Cognitive Performance (n = 12,826) Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.197 (−0.279; −0.116) <.0001 −0.133 (−0.211; −0.055) .001 −0.066 (−0.134; 0.003) .06 −0.057 (−0.125; 0.011) .10 Memory −0.133 (−0.215; −0.051) .001 −0.091 (−0.169; −0.013) .02 −0.051 (−0.126; 0.023) .18 −0.055 (−0.130; 0.020) .15 Verbal fluency −0.162 (−0.244; −0.081) <.0001 −0.133 (−0.214; −0.051) .001 −0.089 (−0.166; −0.012) .02 −0.080 (−0.157; −0.003) .04 Trail making −0.137 (−0.219; −0.056) .001 −0.068 (−0.148; 0.011) .09 −0.004 (−0.075; 0.067) .91 0.009 (−0.061; 0.80) .80 Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.197 (−0.279; −0.116) <.0001 −0.133 (−0.211; −0.055) .001 −0.066 (−0.134; 0.003) .06 −0.057 (−0.125; 0.011) .10 Memory −0.133 (−0.215; −0.051) .001 −0.091 (−0.169; −0.013) .02 −0.051 (−0.126; 0.023) .18 −0.055 (−0.130; 0.020) .15 Verbal fluency −0.162 (−0.244; −0.081) <.0001 −0.133 (−0.214; −0.051) .001 −0.089 (−0.166; −0.012) .02 −0.080 (−0.157; −0.003) .04 Trail making −0.137 (−0.219; −0.056) .001 −0.068 (−0.148; 0.011) .09 −0.004 (−0.075; 0.067) .91 0.009 (−0.061; 0.80) .80 Note: CI = confidence interval; LDL = low-density lipoprotein; OH = orthostatic hypotension. Model 1: linear regression model, adjusted for age and sex. Model 2: Model 1 + race and education. Model 3: Model 2 + seated systolic blood pressure, use of drugs that can cause OH, diabetes, LDL cholesterol, use of lipid-lowering drugs, cardiovascular disease, body mass index, alcohol use, smoking, physical activity, depression, antidepressant use, and thyroid function. View Large Table 3. Association Between Orthostatic Hypertension and Cognitive Performance (n = 12,826) Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.197 (−0.279; −0.116) <.0001 −0.133 (−0.211; −0.055) .001 −0.066 (−0.134; 0.003) .06 −0.057 (−0.125; 0.011) .10 Memory −0.133 (−0.215; −0.051) .001 −0.091 (−0.169; −0.013) .02 −0.051 (−0.126; 0.023) .18 −0.055 (−0.130; 0.020) .15 Verbal fluency −0.162 (−0.244; −0.081) <.0001 −0.133 (−0.214; −0.051) .001 −0.089 (−0.166; −0.012) .02 −0.080 (−0.157; −0.003) .04 Trail making −0.137 (−0.219; −0.056) .001 −0.068 (−0.148; 0.011) .09 −0.004 (−0.075; 0.067) .91 0.009 (−0.061; 0.80) .80 Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.197 (−0.279; −0.116) <.0001 −0.133 (−0.211; −0.055) .001 −0.066 (−0.134; 0.003) .06 −0.057 (−0.125; 0.011) .10 Memory −0.133 (−0.215; −0.051) .001 −0.091 (−0.169; −0.013) .02 −0.051 (−0.126; 0.023) .18 −0.055 (−0.130; 0.020) .15 Verbal fluency −0.162 (−0.244; −0.081) <.0001 −0.133 (−0.214; −0.051) .001 −0.089 (−0.166; −0.012) .02 −0.080 (−0.157; −0.003) .04 Trail making −0.137 (−0.219; −0.056) .001 −0.068 (−0.148; 0.011) .09 −0.004 (−0.075; 0.067) .91 0.009 (−0.061; 0.80) .80 Note: CI = confidence interval; LDL = low-density lipoprotein; OH = orthostatic hypotension. Model 1: linear regression model, adjusted for age and sex. Model 2: Model 1 + race and education. Model 3: Model 2 + seated systolic blood pressure, use of drugs that can cause OH, diabetes, LDL cholesterol, use of lipid-lowering drugs, cardiovascular disease, body mass index, alcohol use, smoking, physical activity, depression, antidepressant use, and thyroid function. View Large We tested for quadratic associations between SBP/DBP standardized orthostatic changes and cognitive performance. SBP change z-score was nonlinearly associated with the PVFT z-score. Both decrease and increase in SBP after standing were associated with worse performance on the PVFT (p = .03; Figure 1). SBP and DBP change z-scores were not associated with other cognitive tests (p > .05 for all testes). When we considered the absolute values of SBP and DBP change at 2, 3, and 5 minutes, we found and association of SBP change after 2 minutes of standing with verbal fluency z-score (Supplementary Table 4), and an association of SBP and DBP change after 3 minutes with memory z-scores (Supplementary Tables 4 and 5). Figure 1. View largeDownload slide Association between phonemic verbal fluency test z-score and systolic blood pressure (SBP) standardized orthostatic change. SBP change was calculated as the mean of the SBP difference of each SBP measurements after 2, 3, and 5 min of standing and the supine position measurement. Regression model included a quadratic term for SBP change (p = .03), and it was adjusted for age, sex, race, education, use of drugs that can cause orthostatic hypotension, diabetes, LDL-cholesterol, use of lipid-lowering drugs, cardiovascular disease, body mass index, alcohol use, smoking, physical activity, depression, antidepressant use, and thyroid function. Figure 1. View largeDownload slide Association between phonemic verbal fluency test z-score and systolic blood pressure (SBP) standardized orthostatic change. SBP change was calculated as the mean of the SBP difference of each SBP measurements after 2, 3, and 5 min of standing and the supine position measurement. Regression model included a quadratic term for SBP change (p = .03), and it was adjusted for age, sex, race, education, use of drugs that can cause orthostatic hypotension, diabetes, LDL-cholesterol, use of lipid-lowering drugs, cardiovascular disease, body mass index, alcohol use, smoking, physical activity, depression, antidepressant use, and thyroid function. We observed a borderline interaction term between OH and HR change and the TMT z-score (p value for interaction = .09), suggesting that participants who did not have a compensatory increase in HR had poorer performance on this test (Figure 2). We did not find evidence of effect modification of age, sex, race, hypertension, diabetes, cardiovascular disease, depression, and drugs that may cause OH on the association between OH and cognition (Supplementary Table 6). Figure 2. View largeDownload slide Predicted z-scores for (A) composite global cognition, (B) memory, (C) verbal fluency, and (D) trail making tests, considering an interaction term between the presence of orthostatic hypotension and heart rate (HR) response after postural change. The sample median of HR response was used as the cutoff to dichotomize the HR variable (HR ≤ 14 bpm and HR > 14 bpm). p Values are for the interaction terms. Figure 2. View largeDownload slide Predicted z-scores for (A) composite global cognition, (B) memory, (C) verbal fluency, and (D) trail making tests, considering an interaction term between the presence of orthostatic hypotension and heart rate (HR) response after postural change. The sample median of HR response was used as the cutoff to dichotomize the HR variable (HR ≤ 14 bpm and HR > 14 bpm). p Values are for the interaction terms. Discussion In the first large study with a racially diverse sample from a low-middle income country (mean age = 51 years, 53% white), participants with OH and orthostatic hypertension had lower PVFT scores compared with participants without orthostatic changes in blood pressure. In addition, we found a nonlinear association between orthostatic SBP standardized change and verbal fluency performance. The interaction terms between OH and compensatory increase in HR on the TMT was borderline significant, suggesting that participants who lack compensatory HR after postural changing might have poorer performance on an executive function test. The association between OH and cognitive impairment is controversial. The cross-sectional results from TILDA (mean participant age = 63 years) showed that OH was associated with poorer performance on global cognition and in the memory domain in women (6), and orthostatic hypertension was associated with poorer performance on global cognition and executive function (5). We found a similar association between orthostatic hypertension and poor performance on the PVFT, which is also a test related to executive function. However, OH was not associated with cognitive decline after two years of follow-up in TILDA (8). Similarly, OH did not predict cognitive impairment when adjusted for demographics and cardiovascular risk factors in the ARIC study after 3 and 10 years of follow-up (7). In a convenience sample of 184 frail older adults with history of falls (mean age = 80 years), OH was not related to cognitive impairment, as evaluated by the Mini–Mental State Examination (MMSE) and the Cambridge Examination for Mental Disorders of the Elderly (CAMCOG) (28). On the other hand, in some longitudinal studies, OH was associated with dementia and mild cognitive impairment (9,10,25,26). Among 6,204 participants in the Rotterdam study (mean age = 69), OH and SBP variability were associated with an increased risk of dementia after 15 years of follow-up. Similar to our findings for borderline interaction between compensatory HR and OH, participants who did not have a compensatory increase in HR after standing had an increased risk of dementia in the Rotterdam study (25). The lack of HR increase in the presence of OH may lead to more severe cerebral hypoperfusion due to a further decrease in the cardiac output. In the Malmo Preventive Project, 18,240 Swedish participants (mean age = 45) were followed for an average of 23 years. DBP decrease on standing at baseline predicted a 22% higher risk of dementia, whereas SBP and OH at baseline did not predict dementia (10). Similarly, OH predicted the incidence of mild cognitive impairment in a Swedish cohort of 1,480 participants (mean age = 68) (26). In another cross-sectional study with 495 older patients (mean age = 76), OH was related to poorer performance on global cognitive function and to cognitive status and was more prevalent among patients with mild cognitive impairment, Alzheimer’s disease, and vascular dementia (29). We found consistent results for the association of worse performance on the PVFT with OH and orthostatic hypertension. The U-shaped association with SBP standardized change reinforces these findings, and it has been previously described in a small sample of individuals with asymptomatic OH (30). The PVFT evaluates executive function, which is related to frontal lobe function (31). Executive dysfunction is commonly seen in vascular cognitive impairment, which may be the result of hypoperfusion related to orthostatic changes in blood pressure. Indeed, the association between OH and cognitive impairment can result from three possible mechanisms (32). First, OH may lead to episodes of cerebral hypoperfusion, causing permanent cognitive deterioration (33). Second, cognitive impairment may be a transient symptom of OH with no permanent effects. The presence of OH was associated with poorer performance on global and executive function tests in small samples of patients with neurogenic OH and autonomic disorders who underwent cognitive evaluation during postural changes, supporting this hypothesis (34,35). Finally, OH and cognitive impairment may be the result of a common neuropathological process, which affects brain areas related to cognition and cardiovascular autonomic control (eg, hypothalamus, brain stem, telencephalic structures) (36). Most evidence about the association between OH and cognitive impairment is based on studies with older participants with high prevalence of OH and related diseases (5,6). On the other hand, ELSA-Brasil participants were mostly middle aged and had low prevalence of OH and cardiovascular disease (5,6). Our study has several strengths; it was the first to show evidence about the association between OH and cognitive impairment in a large sample of most middle-aged Brazilian participants. Our sample size was comparable to large studies on this topic, such as the ARIC (n = 12,702) and the Malmo Preventive Project (n = 18,240) (7,10). This large sample size allowed us to investigate interactions between OH and certain variables on cognitive performance; however, we did not find evidence of effect modification by any of these variables in our sample. Furthermore, we included most middle-aged participants; studies with middle-aged participants are necessary to determine risk factors for cognitive impairment, and to develop preventive strategies against dementia, because neuropathological lesions may start decades before the clinical onset of dementia (37). Moreover, we presented evidence in a racially diverse sample (11), whereas most studies included mainly whites from Western European and American cohorts (7,10,25). We were able to adjust our analyses for race, and also test for the interaction between race and OH on cognitive performance, which has not been investigated previously, even in other racially diverse cohort (7). The association of OH with memory and trail making scores lost significance after adjustment for race and education, suggesting that these socioeconomic variables were important confounders. Finally, we used standardized neuropsychological tests, not only screening tools, like the MMSE (9). However, our study should be considered in light of its limitations. This is a cross-sectional study, which limits any considerations about the effect of OH on cognitive function. In addition, although we excluded participants with previous history of stroke, we did not have neuroimaging exams that could explore a possible mediating effect of cerebrovascular lesions in the association between OH and cognitive performance. Although OH and cognition were not evaluated simultaneously in the ELSA-Brasil study as in previous small studies, they were evaluated within a short interval on the same day (34,35). In addition, OH is classically evaluated within 3 minutes of standing (1), but our definition of OH also included delayed orthostatic blood pressure response at 5 minutes. When we included a sensitivity analysis on the association of cognitive performance with early, classical, and delayed OH, only borderline association was found with classical and delayed OH. In addition, ELSA-Brasil is an occupational cohort with higher educational attainment than the Brazilian population, which could be related to lower prevalence of OH and cognitive impairment, limiting the study external validity. However, it is important to note that enrollment in the ELSA-Brasil study was based on occupation (ie, one third of participants were unskilled, one third were technicians, and one third were professionals/faculties), which brings a diverse socioeconomic background to our sample. Finally, the association between OH and poorer performance on PVFT was based on small effects and may have limited clinical meaning, as the association of OH with cognitive impairment was not significant. These findings were expected because we examined a large sample of mostly middle-aged individuals who were free of dementia at baseline. Conclusion OH was weakly associated with poorer performance on verbal fluency scores in a large sample of mostly middle-aged Brazilians. Further longitudinal studies with this racially diverse population will be important to confirm whether OH is associated with cognitive decline and dementia. Supplementary Material Supplementary data is available at The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences online. Funding This work was supported by the Brazilian Ministry of Health and CNPq (grant numbers 01060010.00RS, 01060212.00BA, 01060300.00ES, 01060278.00MG, 01060115.00SP, and 01060071.00RJ). Conflict of Interest None reported. References 1. Freeman R , Wieling W , Axelrod FB , et al. Consensus statement on the definition of orthostatic hypotension, neurally mediated syncope and the postural tachycardia syndrome . Clin Auton Res . 2011 ; 21 : 69 – 72 . doi: 10.1007/s10286-011-0119-5 Google Scholar CrossRef Search ADS PubMed 2. Viramo P , Luukinen H , Koski K , Laippala P , Sulkava R , Kivelä SL . Orthostatic hypotension and cognitive decline in older people . J Am Geriatr Soc . 1999 ; 47 : 600 – 604 . Google Scholar CrossRef Search ADS PubMed 3. Vloet LC , Pel-Little RE , Jansen PA , Jansen RW . High prevalence of postprandial and orthostatic hypotension among geriatric patients admitted to Dutch hospitals . J Gerontol A Biol Sci Med Sci . 2005 ; 60 : 1271 – 1277 . Google Scholar CrossRef Search ADS PubMed 4. Angelousi A , Girerd N , Benetos A , et al. Association between orthostatic hypotension and cardiovascular risk, cerebrovascular risk, cognitive decline and falls as well as overall mortality: a systematic review and meta-analysis . J Hypertens . 2014 ; 32 : 1562 – 1571 ; discussion 1571. doi: 10.1097/HJH.0000000000000235 Google Scholar CrossRef Search ADS PubMed 5. Frewen J , Finucane C , Savva GM , Boyle G , Kenny RA . Orthostatic hypotension is associated with lower cognitive performance in adults aged 50 plus with supine hypertension . J Gerontol A Biol Sci Med Sci . 2014 ; 69 : 878 – 885 . doi: 10.1093/gerona/glt171 Google Scholar CrossRef Search ADS PubMed 6. Frewen J , Savva GM , Boyle G , Finucane C , Kenny RA . Cognitive performance in orthostatic hypotension: findings from a nationally representative sample . J Am Geriatr Soc . 2014 ; 62 : 117 – 122 . doi: 10.1111/jgs.12592 Google Scholar CrossRef Search ADS PubMed 7. Rose KM , Couper D , Eigenbrodt ML , Mosley TH , Sharrett AR , Gottesman RF . Orthostatic hypotension and cognitive function: the Atherosclerosis risk in communities study . Neuroepidemiology . 2010 ; 34 : 1 – 7 . doi: 10.1159/000255459 Google Scholar CrossRef Search ADS PubMed 8. Feeney J , O’Leary N , Kenny RA . Impaired orthostatic blood pressure recovery and cognitive performance at two-year follow up in older adults: the Irish Longitudinal Study on Ageing . Clin Auton Res . 2016 ; 26 : 127 – 133 . doi: 10.1007/s10286-016-0340-3 Google Scholar CrossRef Search ADS PubMed 9. Yap PL , Niti M , Yap KB , Ng TP . Orthostatic hypotension, hypotension and cognitive status: early comorbid markers of primary dementia ? Dement Geriatr Cogn Disord . 2008 ; 26 : 239 – 246 . doi: 10.1159/000160955 Google Scholar CrossRef Search ADS PubMed 10. Holm H , Nagga K , Nilsson ED , et al. Longitudinal and postural changes of blood pressure predict dementia: the Malmo Preventive Project . Eur J Epidemiol . 2017 . doi: 10.1007/s10654-017-0228-0 11. Schmidt MI , Duncan BB , Mill JG , et al. Cohort profile: longitudinal study of adult health (ELSA-Brasil) . Int J Epidemiol . 2015 ; 44 : 68 – 75 . doi: 10.1093/ije/dyu027 Google Scholar CrossRef Search ADS PubMed 12. Mill JG , Pinto K , Griep RH , et al. Medical assessments and measurements in ELSA-Brasil . Rev Saude Publica . 2013 ; 47 ( suppl 2 ): 54 – 62 . doi: 10.1590/S0034-8910.2013047003851 Google Scholar CrossRef Search ADS PubMed 13. Morris JC , Heyman A , Mohs RC , et al. The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part I. Clinical and neuropsychological assessment of Alzheimer’s disease . Neurology . 1989 ; 39 : 1159 – 1165 . Google Scholar CrossRef Search ADS PubMed 14. Passos VM , Caramelli P , Benseñor I , Giatti L , Barreto SM . Methods of cognitive function investigation in the Longitudinal Study on Adult Health (ELSA-Brasil) . Sao Paulo Med J . 2014 ; 132 : 170 – 177 . doi: 10.1590/1516-3180.2014.1323646 Google Scholar CrossRef Search ADS PubMed 15. Greenlief CL , Margolis RB , Erker GJ . Application of the trail making test in differentiating neuropsychological impairment of elderly persons . Percept Mot Skills . 1985 ; 61 : 1283 – 1289 . doi: 10.2466/pms.1985.61.3f.1283 Google Scholar CrossRef Search ADS PubMed 16. Bertolucci PH , Okamoto IH , Brucki SM , Siviero MO , Toniolo Neto J , Ramos LR . Applicability of the CERAD neuropsychological battery to Brazilian elderly . Arq Neuropsiquiatr . 2001 ; 59 : 532 – 536 . Google Scholar CrossRef Search ADS PubMed 17. Shao Z , Janse E , Visser K , Meyer AS . What do verbal fluency tasks measure? Predictors of verbal fluency performance in older adults . Front Psychol . 2014 ; 5 : 772 . doi: 10.3389/fpsyg.2014.00772 Google Scholar CrossRef Search ADS PubMed 18. Rawlings AM , Sharrett AR , Schneider AL , et al. Diabetes in midlife and cognitive change over 20 years: a cohort study . Ann Intern Med . 2014 ; 161 : 785 – 793 . doi: 10.7326/M14-0737 Google Scholar CrossRef Search ADS PubMed 19. Bennett DA , Schneider JA , Buchman AS , Barnes LL , Boyle PA , Wilson RS . Overview and findings from the rush Memory and Aging Project . Curr Alzheimer Res . 2012 ; 9 : 646 – 663 . doi: 10.2174/156720512801322663 Google Scholar CrossRef Search ADS PubMed 20. Rose KM , Eigenbrodt ML , Biga RL , et al. Orthostatic hypotension predicts mortality in middle-aged adults: the Atherosclerosis Risk In Communities (ARIC) Study . Circulation . 2006 ; 114 : 630 – 636 . doi: 10.1161/CIRCULATIONAHA.105.598722 Google Scholar CrossRef Search ADS PubMed 21. Fessel J , Robertson D . Orthostatic hypertension: when pressor reflexes overcompensate . Nat Clin Pract Nephrol . 2006 ; 2 : 424 – 431 . doi: 10.1038/ncpneph0228 Google Scholar CrossRef Search ADS PubMed 22. World Health Organization . International Statistical Classification of Diseases and Related Health Problems . Vol. 1 . Geneva, Switzerland : World Health Organization ; 2004 . 23. Lewis G , Pelosi AJ , Araya R , Dunn G . Measuring psychiatric disorder in the community: a standardized assessment for use by lay interviewers . Psychol Med . 1992 ; 22 : 465 – 486 . Google Scholar CrossRef Search ADS PubMed 24. Perlmuter LC , Sarda G , Casavant V , Mosnaim AD . A review of the etiology, associated comorbidities, and treatment of orthostatic hypotension . Am J Ther . 2013 ; 20 : 279 – 291 . doi: 10.1097/MJT.0b013e31828bfb7f Google Scholar CrossRef Search ADS PubMed 25. Wolters FJ , Mattace-Raso FU , Koudstaal PJ , Hofman A , Ikram MA ; Heart Brain Connection Collaborative Research Group . Orthostatic hypotension and the long-term risk of dementia: a population-based study . PLoS Med . 2016 ; 13 : e1002143 . doi: 10.1371/journal.pmed.1002143 Google Scholar CrossRef Search ADS PubMed 26. Elmståhl S , Widerström E . Orthostatic intolerance predicts mild cognitive impairment: incidence of mild cognitive impairment and dementia from the Swedish general population cohort good aging in Skåne . Clin Interv Aging . 2014 ; 9 : 1993 – 2002 . doi: 10.2147/CIA.S72316 Google Scholar CrossRef Search ADS PubMed 27. Albert MS , DeKosky ST , Dickson D , et al. The diagnosis of mild cognitive impairment due to Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease . Alzheimers Dement . 2011 ; 7 : 270 – 279 . doi: 10.1016/j.jalz.2011.03.008 Google Scholar CrossRef Search ADS PubMed 28. Schoon Y , Lagro J , Verhoeven Y , Rikkert MO , Claassen J . Hypotensive syndromes are not associated with cognitive impairment in geriatric patients . Am J Alzheimers Dis Other Demen . 2013 ; 28 : 47 – 53 . doi: 10.1177/1533317512466692 Google Scholar CrossRef Search ADS PubMed 29. Mehrabian S , Duron E , Labouree F , et al. Relationship between orthostatic hypotension and cognitive impairment in the elderly . J Neurol Sci . 2010 ; 299 : 45 – 48 . doi: 10.1016/j.jns.2010.08.056 Google Scholar CrossRef Search ADS PubMed 30. Czajkowska J , Ozhog S , Smith E , Perlmuter LC . Cognition and hopelessness in association with subsyndromal orthostatic hypotension . J Gerontol A Biol Sci Med Sci . 2010 ; 65 : 873 – 879 . doi: 10.1093/gerona/glq068 Google Scholar CrossRef Search ADS PubMed 31. Alvarez JA , Emory E . Executive function and the frontal lobes: a meta-analytic review . Neuropsychol Rev . 2006 ; 16 : 17 – 42 . doi: 10.1007/s11065-006-9002-x Google Scholar CrossRef Search ADS PubMed 32. Sambati L , Calandra-Buonaura G , Poda R , Guaraldi P , Cortelli P . Orthostatic hypotension and cognitive impairment: a dangerous association ? Neurol Sci . 2014 ; 35 : 951 – 957 . doi: 10.1007/s10072-014-1686-8 Google Scholar CrossRef Search ADS PubMed 33. Kim JS , Oh YS , Lee KS , Kim YI , Yang DW , Goldstein DS . Association of cognitive dysfunction with neurocirculatory abnormalities in early Parkinson disease . Neurology . 2012 ; 79 : 1323 – 1331 . doi: 10.1212/WNL.0b013e31826c1acd Google Scholar CrossRef Search ADS PubMed 34. Guaraldi P , Poda R , Calandra-Buonaura G , et al. Cognitive function in peripheral autonomic disorders . PLoS One . 2014 ; 9 : e85020 . doi: 10.1371/journal.pone.0085020 Google Scholar CrossRef Search ADS PubMed 35. Poda R , Guaraldi P , Solieri L , et al. Standing worsens cognitive functions in patients with neurogenic orthostatic hypotension . Neurol Sci . 2012 ; 33 : 469 – 473 . doi: 10.1007/s10072-011-0746-6 Google Scholar CrossRef Search ADS PubMed 36. Idiaquez J , Roman GC . Autonomic dysfunction in neurodegenerative dementias . J Neurol Sci . 2011 ; 305 : 22 – 27 . doi: 10.1016/j.jns.2011.02.033 Google Scholar CrossRef Search ADS PubMed 37. Sperling R , Mormino E , Johnson K . The evolution of preclinical Alzheimer’s disease: implications for prevention trials . Neuron . 2014 ; 84 : 608 – 622 . doi: 10.1016/j.neuron.2014.10.038 Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Journals of Gerontology Series A: Biomedical Sciences and Medical Sciences Oxford University Press

Loading next page...
 
/lp/ou_press/orthostatic-hypotension-and-cognitive-function-cross-sectional-results-sHZEi0wWCE
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
1079-5006
eISSN
1758-535X
D.O.I.
10.1093/gerona/gly061
Publisher site
See Article on Publisher Site

Abstract

Abstract Background The association between orthostatic hypotension (OH) and cognitive impairment is controversial, and most studies have investigated older white adults from Western Europe and the United States. Therefore, we investigated the association between OH and cognitive performance in a large and racially diverse sample of adults using cross-sectional data from the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil). Methods OH was defined when systolic blood pressure decreased 20 mmHg and/or diastolic blood pressure decreased 10 mmHg from supine to standing position. We investigated the association between OH and composite global cognition, memory, verbal fluency (VF), and Trail Making Test z-scores, using multiple linear regression models. We also investigated the association of orthostatic hypertension and systolic blood pressure/diastolic blood pressure changes with cognitive performance, as well as the interaction between OH and compensatory heart rate after postural change on cognitive performance. Results We evaluated 12,826 participants (mean age = 51.5 ± 9.0 years, 46% male, 53% white). Participants with OH (4% of the sample) had poorer z-scores for VF (β = −0.108, 95% confidence interval = −0.189; −0.025, p = .01) than participants without OH. Orthostatic hypertension was also associated with worse performance on the VF test (β = −0.080, 95% confidence interval = −0.157; −0.003, p = .04). Systolic blood pressure orthostatic change had a nonlinear association with VF. The interaction terms between OH and compensatory increase in heart rate for the Trail Making Test z-score (p = .09) was borderline significant, suggesting that participants who lack compensatory heart rate after postural change might have poorer performance. Conclusion OH and orthostatic hypertension were associated with poorer performance on the VF test in participants from Brazil. Hypotension, Orthostatic, Hypertension, Cognition, Dementia Orthostatic hypotension (OH) occurs when blood pressure drops substantially as a result of postural change from supine to upright position. It is classically defined as a drop of 20 mmHg or more in systolic blood pressure (SBP) and/or a drop of 10 mmHg or more in diastolic pressure (DBP) after 3 minutes of standing (1). OH is more frequent among older adults and those with functional limitations (2,3). In a recent meta-analysis of the association of OH with different outcomes, OH was associated with increased risk of cardiovascular events and overall mortality, whereas the evidence on cognitive outcomes was less clear (4). In the baseline analysis of the Irish Longitudinal Study of Aging (TILDA), participants with OH had poorer global cognitive performance, particularly among women and when hypertension was present in the supine position (5,6). However, OH was not associated with cognitive decline in the follow-up of the TILDA and in the ARIC (Atherosclerosis Risk in Communities) studies (7,8). Moreover, most of the evidence showing the association between OH and cognitive impairment comes from the United States and Western Europe and from older participants (2,6,7), with limited data from low-middle income countries and middle-aged participants (9,10). Therefore, the goal in this study was to investigate the cross-sectional association between OH and cognitive function in a large sample of participants aged 35–74 years from the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil). Material and Methods Participants The ELSA-Brasil is a longitudinal study of 15,105 employees from public institutions in six Brazilian cities (São Paulo, Belo Horizonte, Porto Alegre, Salvador, Rio de Janeiro, and Vitória). Participants were aged 35–74 years old and did not have dementia at enrollment (2008–2010). The study was approved by the local institutional review boards, and all the participants signed an informed consent form at the time of enrollment. A detailed description of the study can be found elsewhere (11,12). For this analysis, we excluded participants with a self-reported medical diagnosis of stroke to minimize the effect of previous cerebrovascular disease on cognitive function. We also excluded participants who regularly used some medications (antipsychotic medications, antiparkinsonian agents, and anticonvulsants) that indicated the presence of active neurologic or psychiatric diseases that could directly affect cognitive performance. Finally, we excluded participants with incomplete data for the cognitive tests or OH. Cognitive Function Evaluation Cognitive function was assessed using the following tests: (a) the Consortium to Establish a Registry for Alzheimer’s Disease Word List Memory Test (CERAD-WLMT) (13); (b) the Phonemic Verbal Fluency Test (PVFT) (14); and (c) the Trail Making Test version B (TMT) (15). Trained examiners administered the tests in a fixed order during a single session, in a quiet room with good lighting and low levels of noise or other distractions. We used the Brazilian version of the CERAD-WLMT (16). This test comprises immediate word recall, delayed word recall, and word recognition and evaluates the memory domain (13). Participants were asked to read and learn 10 words after three exposures; the sum of the number of words recalled in each of these attempts was the score in the immediate word recall. After a 5-minute filled delay, the participants were given 60 seconds to record the words. The delayed word recall score was equal to the number of recalled words. Finally, 20 words were presented and the participants had to recognize the 10 words that were presented previously. In the PVFT, participants were asked to generate as many words as possible that start with the letter F to evaluate language and executive ability. The score on this test was the total number of generated words (17). Finally, the TMT is a test of executive function, processing speed, and visual-spatial organization. The participants were instructed to draw lines connecting letters and numbers in an order that alternated between increasing numeric values and alphabetic order. The test score was the total time taken to complete the task, in seconds. To facilitate comparisons across cognitive tests, we generated z-scores by subtracting the participant’s score in each test from the sample mean score and dividing the difference by the standard deviation (SD) of the sample (18,19). Therefore, a z-score of −1 describes a cognitive performance that is 1 SD below the mean sample score. We combined the three scores in the CERAD-WLMT (ie, immediate recall, delayed recall, and recognition) by averaging the z-scores for each test and then standardizing this averaged score. We used the same process to generate z-scores for the PVFT and TMT. We inverted the TMT z-score signal because higher scores indicate poorer performance in the original TMT, whereas higher scores on the other tests are related to better performance. Finally, we averaged and standardized the z-scores of all tests as a measure of composite global cognitive score. Orthostatic Hypotension and Blood Pressure Assessment We followed the procedures of the ARIC study to assess OH (7,20). Blood pressure and heart rate (HR) were measured in the right arm using an automatic oscillometric device (Omron HEM 705CPINT) after a rest period of 5 minutes in a seated position, in a quiet, temperature-controlled room (20°C–24°C). Three measurements were taken at 1-minute intervals, and the mean of the two latest measurements was considered as the seated blood pressure. Blood pressure after postural change was measured in another moment, after evaluation of the ankle–brachial index. Blood pressure and HR were measured in the right arm using the automatic oscillometric device after 20 minutes of rest. The participant was then asked to stand up, and both measurements were reassessed at 2, 3, and 5 minutes after standing. OH was defined by a decrease of at least 20 mmHg in SBP and/or a decrease of at least 10 mmHg in DBP for any of the three measurements after standing. On the other hand, orthostatic hypertension was defined by an increase of at least 20 mmHg in SBP and/or an increase of at least 10 mmHg in DBP for any of the three measurements after standing (21). Changes in orthostatic SBP and DBP were calculated by the mean of the differences between each of the three standing position blood pressure measurements and the supine position measurement. The mean change in SBP measurement between these positions was then standardized. For that, the SBP change z-score was calculated by subtracting the participant’s SBP difference in response to postural change from the sample mean SBP change and dividing this difference by the sample SD for SBP change. The same calculation was performed for the mean difference in DBP. Therefore, an SBP change z-score of −1 would describe an SBP change that is 1 SD below the sample mean for SBP change. Moreover, we calculated the maximum HR response after postural change (ie, maximum difference between any of the three standing measurements and the supine measurement). Cognitive evaluation and OH assessment were performed within a 3-hour interval on the same day. Other Measurements We collected information on sociodemographic variables: age, sex, race, and education. Race and level of education were self-reported. Race was classified in white, black, brown (admixed race of white and black), or other race (ie, Asian and indigenous), and education was classified in high school or less versus college or more. Diabetes was self-reported and categorized into three groups: no history of diabetes, history of diabetes ≤5 years (median time of diabetes diagnosis in our sample), and history of diabetes >5 years. The following clinical information was self-reported: cardiovascular disease (myocardial infarction, myocardial revascularization, or heart failure), use of lipid-lowering drugs, alcohol use, smoking, and physical leisure activity. Alcohol use and smoking were categorized in never, former, and current user. Leisure activity was categorized into three groups: no physical activity, ≤150 min/wk, and >150 min/wk). Weight and height were measured in participants with light clothing and without shoes, and body mass index were calculated in kilogram per square meter. Low-density lipoprotein cholesterol was determined by the enzymatic colorimetric model. The diagnosis of depression was based on the International Classification of Diseases (ICD-10) criteria (22) and used the participant’s response to the Clinical Interview Schedule Revised (23). The use of antidepressants was reported by the participant. Thyroid stimulating hormone levels (TSH) and free thyroxine (FT4) were determined using a third generation immunoenzymatic assay (Siemens, Deerfield, IL). Thyroid function was classified as follows: (a) normal: TSH from 0.4 to 4 mIU/L and FT4 levels from 0.8 to 1.9 ng/dL; (b) hypothyroidism: TSH > 4.0 mIU/L and FT4 < 0.8 ng/dL; and (c) hyperthyroidism: TSH < 0.4 mIU/dL and FT4 > 1.9 ng/dL. We also investigated the use of drugs that could cause OH among participant’s self-reported medication: antihypertensive drugs, vasodilators, muscle relaxants, anticholinergics, sedatives, and alpha-blockers (8,24). Statistical Analysis We compared participants with and without OH using the unpaired t test for continuous variables and chi-square test for categorical variables. The independent variable was the OH status. We investigated the association between the dependent and independent variables using linear regression models. In these models, the dependent variables were the z scores for composite global, memory, verbal fluency, and TMT. We presented the beta coefficients for the crude model, as well as for the hierarchical linear regression models adjusted for (a) age and sex; (b) age, sex, race, and education; (c) age, sex, race, education, seated SBP, use of drugs that could cause OH, diabetes, low-density lipoprotein cholesterol, use of lipid-lowering drugs, cardiovascular disease, body mass index, alcohol use, smoking, physical activity, depression, use of antidepressants, and thyroid function. We also tested for the association between orthostatic hypertension and cognitive performance. We determined the association of cognitive performance with changes in orthostatic SBP and DBP (ie, mean change in SBP and DBP from supine to standing position in 2, 3, and 5 minutes) continuously by SD increase, using multiple linear regression models adjusted for the same set of possible confounders cited above. We tested for nonlinear associations of postural changes in SBP and DBP z-scores with cognitive performance by adding a quadratic terms for SBP/DBP change z-scores. Model selection between the linear and the quadratic models was based on the Akaike information criterion. Alternatively, we also investigated the association of cognitive performance with the absolute values of SBP and DBP change (per 10 mmHg) at 2, 3, and 5 minutes after standing. In addition, we tested whether the association between OH and cognitive test performance was modified by the HR response after postural change, using an interaction term between the OH and HR response in multiple linear models adjusted for the same set of confounders. To facilitate interpretation, HR response after postural change was used as a binary variable, using the median in the sample as the cutoff (≤14 bpm vs. >14 bpm). Moreover, we tested whether age (<65 vs. ≥65 years), sex, race (white vs. nonwhite, excluding Asian and other races), hypertension, diabetes, cardiovascular disease, depression, and drugs that may cause OH (25,26) would change the association between OH and cognitive performance by including an interaction term between OH and these variables in multiple linear regression models adjusted for the set of variables described above. As sensitivity analysis, we investigated the association of cognitive performance with early OH (ie, evaluated at 2 minutes after postural change), classical OH (ie, evaluated at 3 minutes after postural change), and delayed OH (ie, evaluated at 5 minutes after postural change), using the hierarchical linear models described above. Finally, we investigated the clinical meaning of our findings by testing the association between OH and cognitive impairment, defined as scores on cognitive tests 1 SD below the sample mean (27). We used hierarchical logistic regression models adjusted for the same set of possible confounders. We used Stata 13.0 software (StataCorp, College Station, TX) to perform the statistical analyses. The alpha level was set at .05, and tests were two tailed. Results Of the 15,105 eligible participants, 12,826 fulfilled the study criteria and 2,279 were excluded from this study (Supplementary Figure 1). The mean age of the sample was 51.5 ± 9.0 years, 46% were male, 53% were white, and 56% had college education or more. Hypertension was present in 34% of the sample, OH in 4%, orthostatic hypertension in 5%. OH was diagnosed in 1.7% of participants at 2 minutes of standing, 2.8% at 3 minutes, and 2.4% at 5 minutes. The mean SBP was 120.6 ± 16.6 mmHg, and mean DBP was 56.6 ± 14.4 mmHg (Supplementary Table 1). Participants recorded an average of 7.0 ± 2.0 words in the delayed word recall, 12.8 ± 4.4 words on the verbal fluency test, and completed the TMT in 121.5 ± 83.7 seconds. Participants with OH were older, more likely to be female, and had less education than participants without OH. The diagnosis of hypertension was more frequent among participants with OH, as well as the fact that the SBP and DBP measured in the seated position were higher in the OH group compared with those without OH. In addition, use of lipid-lowering drugs, cardiovascular disease, and depression were also more frequent among participants with OH (Table 1). Table 1. Baseline Characteristics of the Study Population by Orthostatic Hypotension Status (n = 12,826) Variable Without OH With OH p n = 12,293 n = 533 Immediate word recallc, mean (SD)a 21.3 (3.8) 20.8 (4.1) .0007 Delayed word recallc, mean (SD)a 7.0 (2.0) 6.8 (2.0) .001 Word recognitionc, mean (SD)a 9.6 (0.8) 9.5 (0.9) .08 Verbal fluencyc, mean (SD)a 12.8 (4.4) 11.9 (4.2) <.0001 Trail making test (s), mean (SD)a 120.9 (83.0) 135.1 (97.8) .001 Age (y), mean (SD)a 51.4 (8.9) 54.5 (9.3) <.0001 Male, %b 45.7 41.3 .04 Race, %b .12  White 53.6 51.2  Black 15.0 18.6  Brown 27.9 27.4  Other 3.5 2.8 College education or more, %b 56.1 48.8 .0009 Hypertension, %b 33.4 44.5 <.0001 Seated SBP (mmHg), mean (SD)a, d 120.4 (16.7) 124.8 (20.6) <.0001 Seated DBP (mmHg), mean (SD)a, d 76.0 (10.6) 77.3 (12.3) .02 SBP change(mmHg), mean (SD)a, e −4.3 (7.6) 15.0 (9.8) <.0001 DBP change(mmHg), mean (SD)a, e −7.7 (5.2) 3.7 (7.8) <.0001 SBP change z-score, mean (SD)a, e −0.1 (0.9) 2.1 (1.1) <.0001 DBP change z-score, mean (SD)a, e −0.1 (0.9) 1.9 (1.3) <.0001 Use of drugs that can cause OH, %b 31.3 41.5 <.0001 Diabetes, %b .16  No 91.2 88.7  Yes, ≤5 y 5.0 6.4  Yes, >5 y 3.8 4.9 LDL cholesterol (mg/dL), mean (SD)a 130.8 (34.7) 131.3 (36.3) .72 Use of lipid-lowering drugs, %b 12.2 16.7 .002 Cardiovascular diseasef, n (%)b 2.3 6.2 <.0001 Body mass index (kg/m2), mean (SD)a 26.9 (4.7) 26.9 (4.8) .94 Smoking, %b .52  Never 57.7 58.2  Former 29.6 30.8  Current 12.7 11.1 Alcohol use, %b .13  Never 9.8 12.2  Former 18.6 19.7  Current 71.6 68.1 Physically active, %b .06  No 61.9 66.8  Yes, <150 min/wk 12.8 12.2  Yes, ≥150 min/wk 25.3 21.0 Depression, %b 3.7 6.0 .006 Use of antidepressants, %b 4.9 5.6 .45 Thyroid function status, %b .12  Normal 85.9 83.3  Clinical or subclinical hypothyroidism 12.2 13.9  Clinical or subclinical hyperthyroidism 1.8 2.8 Variable Without OH With OH p n = 12,293 n = 533 Immediate word recallc, mean (SD)a 21.3 (3.8) 20.8 (4.1) .0007 Delayed word recallc, mean (SD)a 7.0 (2.0) 6.8 (2.0) .001 Word recognitionc, mean (SD)a 9.6 (0.8) 9.5 (0.9) .08 Verbal fluencyc, mean (SD)a 12.8 (4.4) 11.9 (4.2) <.0001 Trail making test (s), mean (SD)a 120.9 (83.0) 135.1 (97.8) .001 Age (y), mean (SD)a 51.4 (8.9) 54.5 (9.3) <.0001 Male, %b 45.7 41.3 .04 Race, %b .12  White 53.6 51.2  Black 15.0 18.6  Brown 27.9 27.4  Other 3.5 2.8 College education or more, %b 56.1 48.8 .0009 Hypertension, %b 33.4 44.5 <.0001 Seated SBP (mmHg), mean (SD)a, d 120.4 (16.7) 124.8 (20.6) <.0001 Seated DBP (mmHg), mean (SD)a, d 76.0 (10.6) 77.3 (12.3) .02 SBP change(mmHg), mean (SD)a, e −4.3 (7.6) 15.0 (9.8) <.0001 DBP change(mmHg), mean (SD)a, e −7.7 (5.2) 3.7 (7.8) <.0001 SBP change z-score, mean (SD)a, e −0.1 (0.9) 2.1 (1.1) <.0001 DBP change z-score, mean (SD)a, e −0.1 (0.9) 1.9 (1.3) <.0001 Use of drugs that can cause OH, %b 31.3 41.5 <.0001 Diabetes, %b .16  No 91.2 88.7  Yes, ≤5 y 5.0 6.4  Yes, >5 y 3.8 4.9 LDL cholesterol (mg/dL), mean (SD)a 130.8 (34.7) 131.3 (36.3) .72 Use of lipid-lowering drugs, %b 12.2 16.7 .002 Cardiovascular diseasef, n (%)b 2.3 6.2 <.0001 Body mass index (kg/m2), mean (SD)a 26.9 (4.7) 26.9 (4.8) .94 Smoking, %b .52  Never 57.7 58.2  Former 29.6 30.8  Current 12.7 11.1 Alcohol use, %b .13  Never 9.8 12.2  Former 18.6 19.7  Current 71.6 68.1 Physically active, %b .06  No 61.9 66.8  Yes, <150 min/wk 12.8 12.2  Yes, ≥150 min/wk 25.3 21.0 Depression, %b 3.7 6.0 .006 Use of antidepressants, %b 4.9 5.6 .45 Thyroid function status, %b .12  Normal 85.9 83.3  Clinical or subclinical hypothyroidism 12.2 13.9  Clinical or subclinical hyperthyroidism 1.8 2.8 Note: DBP = diastolic blood pressure; LDL = low-density lipoprotein; OH = orthostatic hypotension; SBP = systolic blood pressure. aInpaired t test. bChi-square test. cNumber of recorded words. dMean of the two last measurements of blood pressure in the seated position. eMean of the differences between each of the three blood pressure measurements in the standing position and the one in the supine position. fMyocardial infarction, myocardial revascularization, or heart failure. View Large Table 1. Baseline Characteristics of the Study Population by Orthostatic Hypotension Status (n = 12,826) Variable Without OH With OH p n = 12,293 n = 533 Immediate word recallc, mean (SD)a 21.3 (3.8) 20.8 (4.1) .0007 Delayed word recallc, mean (SD)a 7.0 (2.0) 6.8 (2.0) .001 Word recognitionc, mean (SD)a 9.6 (0.8) 9.5 (0.9) .08 Verbal fluencyc, mean (SD)a 12.8 (4.4) 11.9 (4.2) <.0001 Trail making test (s), mean (SD)a 120.9 (83.0) 135.1 (97.8) .001 Age (y), mean (SD)a 51.4 (8.9) 54.5 (9.3) <.0001 Male, %b 45.7 41.3 .04 Race, %b .12  White 53.6 51.2  Black 15.0 18.6  Brown 27.9 27.4  Other 3.5 2.8 College education or more, %b 56.1 48.8 .0009 Hypertension, %b 33.4 44.5 <.0001 Seated SBP (mmHg), mean (SD)a, d 120.4 (16.7) 124.8 (20.6) <.0001 Seated DBP (mmHg), mean (SD)a, d 76.0 (10.6) 77.3 (12.3) .02 SBP change(mmHg), mean (SD)a, e −4.3 (7.6) 15.0 (9.8) <.0001 DBP change(mmHg), mean (SD)a, e −7.7 (5.2) 3.7 (7.8) <.0001 SBP change z-score, mean (SD)a, e −0.1 (0.9) 2.1 (1.1) <.0001 DBP change z-score, mean (SD)a, e −0.1 (0.9) 1.9 (1.3) <.0001 Use of drugs that can cause OH, %b 31.3 41.5 <.0001 Diabetes, %b .16  No 91.2 88.7  Yes, ≤5 y 5.0 6.4  Yes, >5 y 3.8 4.9 LDL cholesterol (mg/dL), mean (SD)a 130.8 (34.7) 131.3 (36.3) .72 Use of lipid-lowering drugs, %b 12.2 16.7 .002 Cardiovascular diseasef, n (%)b 2.3 6.2 <.0001 Body mass index (kg/m2), mean (SD)a 26.9 (4.7) 26.9 (4.8) .94 Smoking, %b .52  Never 57.7 58.2  Former 29.6 30.8  Current 12.7 11.1 Alcohol use, %b .13  Never 9.8 12.2  Former 18.6 19.7  Current 71.6 68.1 Physically active, %b .06  No 61.9 66.8  Yes, <150 min/wk 12.8 12.2  Yes, ≥150 min/wk 25.3 21.0 Depression, %b 3.7 6.0 .006 Use of antidepressants, %b 4.9 5.6 .45 Thyroid function status, %b .12  Normal 85.9 83.3  Clinical or subclinical hypothyroidism 12.2 13.9  Clinical or subclinical hyperthyroidism 1.8 2.8 Variable Without OH With OH p n = 12,293 n = 533 Immediate word recallc, mean (SD)a 21.3 (3.8) 20.8 (4.1) .0007 Delayed word recallc, mean (SD)a 7.0 (2.0) 6.8 (2.0) .001 Word recognitionc, mean (SD)a 9.6 (0.8) 9.5 (0.9) .08 Verbal fluencyc, mean (SD)a 12.8 (4.4) 11.9 (4.2) <.0001 Trail making test (s), mean (SD)a 120.9 (83.0) 135.1 (97.8) .001 Age (y), mean (SD)a 51.4 (8.9) 54.5 (9.3) <.0001 Male, %b 45.7 41.3 .04 Race, %b .12  White 53.6 51.2  Black 15.0 18.6  Brown 27.9 27.4  Other 3.5 2.8 College education or more, %b 56.1 48.8 .0009 Hypertension, %b 33.4 44.5 <.0001 Seated SBP (mmHg), mean (SD)a, d 120.4 (16.7) 124.8 (20.6) <.0001 Seated DBP (mmHg), mean (SD)a, d 76.0 (10.6) 77.3 (12.3) .02 SBP change(mmHg), mean (SD)a, e −4.3 (7.6) 15.0 (9.8) <.0001 DBP change(mmHg), mean (SD)a, e −7.7 (5.2) 3.7 (7.8) <.0001 SBP change z-score, mean (SD)a, e −0.1 (0.9) 2.1 (1.1) <.0001 DBP change z-score, mean (SD)a, e −0.1 (0.9) 1.9 (1.3) <.0001 Use of drugs that can cause OH, %b 31.3 41.5 <.0001 Diabetes, %b .16  No 91.2 88.7  Yes, ≤5 y 5.0 6.4  Yes, >5 y 3.8 4.9 LDL cholesterol (mg/dL), mean (SD)a 130.8 (34.7) 131.3 (36.3) .72 Use of lipid-lowering drugs, %b 12.2 16.7 .002 Cardiovascular diseasef, n (%)b 2.3 6.2 <.0001 Body mass index (kg/m2), mean (SD)a 26.9 (4.7) 26.9 (4.8) .94 Smoking, %b .52  Never 57.7 58.2  Former 29.6 30.8  Current 12.7 11.1 Alcohol use, %b .13  Never 9.8 12.2  Former 18.6 19.7  Current 71.6 68.1 Physically active, %b .06  No 61.9 66.8  Yes, <150 min/wk 12.8 12.2  Yes, ≥150 min/wk 25.3 21.0 Depression, %b 3.7 6.0 .006 Use of antidepressants, %b 4.9 5.6 .45 Thyroid function status, %b .12  Normal 85.9 83.3  Clinical or subclinical hypothyroidism 12.2 13.9  Clinical or subclinical hyperthyroidism 1.8 2.8 Note: DBP = diastolic blood pressure; LDL = low-density lipoprotein; OH = orthostatic hypotension; SBP = systolic blood pressure. aInpaired t test. bChi-square test. cNumber of recorded words. dMean of the two last measurements of blood pressure in the seated position. eMean of the differences between each of the three blood pressure measurements in the standing position and the one in the supine position. fMyocardial infarction, myocardial revascularization, or heart failure. View Large Although OH and performance on all the cognitive tests were associated in simple linear models, only the PVFT z-score (adjusted β = −0.107, 95% confidence interval = −0.189; −0.025, p = .01) was associated with OH in multivariable analyses (Table 2). The associations of OH with cognitive impairment in all tests were not significant (Supplementary Table 2). In sensitivity analysis, we investigated whether cognitive performance would be associated with early, classical, or delayed OH. We found borderline associations of PVFT with classical OH (adjusted β = −0.081, 95% confidence interval = −0.179; 0.018, p = .11) and delayed OH (adjusted β = −0.097, 95% confidence interval = −0.204; 0.010, p = .08) in fully adjusted models (Supplementary Table 3). Similarly, orthostatic hypertension was associated with worse performance in all tests in simple linear models, but only the association with the PVFT z-score remained significant in the fully adjusted model (adjusted β = −0.080, 95% confidence interval = −0.157; −0.003, p = .04; Table 3). Table 2. Association Between Orthostatic Hypotension and Cognitive Performance (n = 12,826) Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.239 (−0.326; −0.152) <.0001 −0.159 (−0.242; −0.076) <.0001 −0.083 (−0.156; −0.010) .02 −0.071 (−0.143; 0.001) .05 Memory −0.150 (−0.237; −0.063) .001 −0.093 (−0.176; −0.010) .03 −0.048 (−0.127; 0.032) .24 −0.050 (−0.129; 0.030) .22 Verbal fluency −0.205 (−0.292; −0.118) <.0001 −0.169 (−0.255; −0.0.82) <.0001 −0.119 (−0.201; −0.037) .004 −0.107 (−0.189; −0.025) .01 Trail making −0.169 (−0.256; −0.083) <.0001 −0.087 (−0.171; −0.002) .04 −0.015 (−0.091; 0.060) .69 0.0001 (−0.075; 0.075) 1.00 Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.239 (−0.326; −0.152) <.0001 −0.159 (−0.242; −0.076) <.0001 −0.083 (−0.156; −0.010) .02 −0.071 (−0.143; 0.001) .05 Memory −0.150 (−0.237; −0.063) .001 −0.093 (−0.176; −0.010) .03 −0.048 (−0.127; 0.032) .24 −0.050 (−0.129; 0.030) .22 Verbal fluency −0.205 (−0.292; −0.118) <.0001 −0.169 (−0.255; −0.0.82) <.0001 −0.119 (−0.201; −0.037) .004 −0.107 (−0.189; −0.025) .01 Trail making −0.169 (−0.256; −0.083) <.0001 −0.087 (−0.171; −0.002) .04 −0.015 (−0.091; 0.060) .69 0.0001 (−0.075; 0.075) 1.00 Note: CI = confidence interval; LDL = low-density lipoprotein; OH = orthostatic hypotension. Model 1: linear regression model, adjusted for age and sex. Model 2: Model 1 + race and education. Model 3: Model 2 + seated systolic blood pressure, use of drugs that can cause OH, diabetes, LDL cholesterol, use of lipid-lowering drugs, cardiovascular disease, body mass index, alcohol use, smoking, physical activity, depression, antidepressant use, and thyroid function. View Large Table 2. Association Between Orthostatic Hypotension and Cognitive Performance (n = 12,826) Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.239 (−0.326; −0.152) <.0001 −0.159 (−0.242; −0.076) <.0001 −0.083 (−0.156; −0.010) .02 −0.071 (−0.143; 0.001) .05 Memory −0.150 (−0.237; −0.063) .001 −0.093 (−0.176; −0.010) .03 −0.048 (−0.127; 0.032) .24 −0.050 (−0.129; 0.030) .22 Verbal fluency −0.205 (−0.292; −0.118) <.0001 −0.169 (−0.255; −0.0.82) <.0001 −0.119 (−0.201; −0.037) .004 −0.107 (−0.189; −0.025) .01 Trail making −0.169 (−0.256; −0.083) <.0001 −0.087 (−0.171; −0.002) .04 −0.015 (−0.091; 0.060) .69 0.0001 (−0.075; 0.075) 1.00 Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.239 (−0.326; −0.152) <.0001 −0.159 (−0.242; −0.076) <.0001 −0.083 (−0.156; −0.010) .02 −0.071 (−0.143; 0.001) .05 Memory −0.150 (−0.237; −0.063) .001 −0.093 (−0.176; −0.010) .03 −0.048 (−0.127; 0.032) .24 −0.050 (−0.129; 0.030) .22 Verbal fluency −0.205 (−0.292; −0.118) <.0001 −0.169 (−0.255; −0.0.82) <.0001 −0.119 (−0.201; −0.037) .004 −0.107 (−0.189; −0.025) .01 Trail making −0.169 (−0.256; −0.083) <.0001 −0.087 (−0.171; −0.002) .04 −0.015 (−0.091; 0.060) .69 0.0001 (−0.075; 0.075) 1.00 Note: CI = confidence interval; LDL = low-density lipoprotein; OH = orthostatic hypotension. Model 1: linear regression model, adjusted for age and sex. Model 2: Model 1 + race and education. Model 3: Model 2 + seated systolic blood pressure, use of drugs that can cause OH, diabetes, LDL cholesterol, use of lipid-lowering drugs, cardiovascular disease, body mass index, alcohol use, smoking, physical activity, depression, antidepressant use, and thyroid function. View Large Table 3. Association Between Orthostatic Hypertension and Cognitive Performance (n = 12,826) Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.197 (−0.279; −0.116) <.0001 −0.133 (−0.211; −0.055) .001 −0.066 (−0.134; 0.003) .06 −0.057 (−0.125; 0.011) .10 Memory −0.133 (−0.215; −0.051) .001 −0.091 (−0.169; −0.013) .02 −0.051 (−0.126; 0.023) .18 −0.055 (−0.130; 0.020) .15 Verbal fluency −0.162 (−0.244; −0.081) <.0001 −0.133 (−0.214; −0.051) .001 −0.089 (−0.166; −0.012) .02 −0.080 (−0.157; −0.003) .04 Trail making −0.137 (−0.219; −0.056) .001 −0.068 (−0.148; 0.011) .09 −0.004 (−0.075; 0.067) .91 0.009 (−0.061; 0.80) .80 Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.197 (−0.279; −0.116) <.0001 −0.133 (−0.211; −0.055) .001 −0.066 (−0.134; 0.003) .06 −0.057 (−0.125; 0.011) .10 Memory −0.133 (−0.215; −0.051) .001 −0.091 (−0.169; −0.013) .02 −0.051 (−0.126; 0.023) .18 −0.055 (−0.130; 0.020) .15 Verbal fluency −0.162 (−0.244; −0.081) <.0001 −0.133 (−0.214; −0.051) .001 −0.089 (−0.166; −0.012) .02 −0.080 (−0.157; −0.003) .04 Trail making −0.137 (−0.219; −0.056) .001 −0.068 (−0.148; 0.011) .09 −0.004 (−0.075; 0.067) .91 0.009 (−0.061; 0.80) .80 Note: CI = confidence interval; LDL = low-density lipoprotein; OH = orthostatic hypotension. Model 1: linear regression model, adjusted for age and sex. Model 2: Model 1 + race and education. Model 3: Model 2 + seated systolic blood pressure, use of drugs that can cause OH, diabetes, LDL cholesterol, use of lipid-lowering drugs, cardiovascular disease, body mass index, alcohol use, smoking, physical activity, depression, antidepressant use, and thyroid function. View Large Table 3. Association Between Orthostatic Hypertension and Cognitive Performance (n = 12,826) Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.197 (−0.279; −0.116) <.0001 −0.133 (−0.211; −0.055) .001 −0.066 (−0.134; 0.003) .06 −0.057 (−0.125; 0.011) .10 Memory −0.133 (−0.215; −0.051) .001 −0.091 (−0.169; −0.013) .02 −0.051 (−0.126; 0.023) .18 −0.055 (−0.130; 0.020) .15 Verbal fluency −0.162 (−0.244; −0.081) <.0001 −0.133 (−0.214; −0.051) .001 −0.089 (−0.166; −0.012) .02 −0.080 (−0.157; −0.003) .04 Trail making −0.137 (−0.219; −0.056) .001 −0.068 (−0.148; 0.011) .09 −0.004 (−0.075; 0.067) .91 0.009 (−0.061; 0.80) .80 Crude Model 1 Model 2 Model 3 z-Score β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p Composite global −0.197 (−0.279; −0.116) <.0001 −0.133 (−0.211; −0.055) .001 −0.066 (−0.134; 0.003) .06 −0.057 (−0.125; 0.011) .10 Memory −0.133 (−0.215; −0.051) .001 −0.091 (−0.169; −0.013) .02 −0.051 (−0.126; 0.023) .18 −0.055 (−0.130; 0.020) .15 Verbal fluency −0.162 (−0.244; −0.081) <.0001 −0.133 (−0.214; −0.051) .001 −0.089 (−0.166; −0.012) .02 −0.080 (−0.157; −0.003) .04 Trail making −0.137 (−0.219; −0.056) .001 −0.068 (−0.148; 0.011) .09 −0.004 (−0.075; 0.067) .91 0.009 (−0.061; 0.80) .80 Note: CI = confidence interval; LDL = low-density lipoprotein; OH = orthostatic hypotension. Model 1: linear regression model, adjusted for age and sex. Model 2: Model 1 + race and education. Model 3: Model 2 + seated systolic blood pressure, use of drugs that can cause OH, diabetes, LDL cholesterol, use of lipid-lowering drugs, cardiovascular disease, body mass index, alcohol use, smoking, physical activity, depression, antidepressant use, and thyroid function. View Large We tested for quadratic associations between SBP/DBP standardized orthostatic changes and cognitive performance. SBP change z-score was nonlinearly associated with the PVFT z-score. Both decrease and increase in SBP after standing were associated with worse performance on the PVFT (p = .03; Figure 1). SBP and DBP change z-scores were not associated with other cognitive tests (p > .05 for all testes). When we considered the absolute values of SBP and DBP change at 2, 3, and 5 minutes, we found and association of SBP change after 2 minutes of standing with verbal fluency z-score (Supplementary Table 4), and an association of SBP and DBP change after 3 minutes with memory z-scores (Supplementary Tables 4 and 5). Figure 1. View largeDownload slide Association between phonemic verbal fluency test z-score and systolic blood pressure (SBP) standardized orthostatic change. SBP change was calculated as the mean of the SBP difference of each SBP measurements after 2, 3, and 5 min of standing and the supine position measurement. Regression model included a quadratic term for SBP change (p = .03), and it was adjusted for age, sex, race, education, use of drugs that can cause orthostatic hypotension, diabetes, LDL-cholesterol, use of lipid-lowering drugs, cardiovascular disease, body mass index, alcohol use, smoking, physical activity, depression, antidepressant use, and thyroid function. Figure 1. View largeDownload slide Association between phonemic verbal fluency test z-score and systolic blood pressure (SBP) standardized orthostatic change. SBP change was calculated as the mean of the SBP difference of each SBP measurements after 2, 3, and 5 min of standing and the supine position measurement. Regression model included a quadratic term for SBP change (p = .03), and it was adjusted for age, sex, race, education, use of drugs that can cause orthostatic hypotension, diabetes, LDL-cholesterol, use of lipid-lowering drugs, cardiovascular disease, body mass index, alcohol use, smoking, physical activity, depression, antidepressant use, and thyroid function. We observed a borderline interaction term between OH and HR change and the TMT z-score (p value for interaction = .09), suggesting that participants who did not have a compensatory increase in HR had poorer performance on this test (Figure 2). We did not find evidence of effect modification of age, sex, race, hypertension, diabetes, cardiovascular disease, depression, and drugs that may cause OH on the association between OH and cognition (Supplementary Table 6). Figure 2. View largeDownload slide Predicted z-scores for (A) composite global cognition, (B) memory, (C) verbal fluency, and (D) trail making tests, considering an interaction term between the presence of orthostatic hypotension and heart rate (HR) response after postural change. The sample median of HR response was used as the cutoff to dichotomize the HR variable (HR ≤ 14 bpm and HR > 14 bpm). p Values are for the interaction terms. Figure 2. View largeDownload slide Predicted z-scores for (A) composite global cognition, (B) memory, (C) verbal fluency, and (D) trail making tests, considering an interaction term between the presence of orthostatic hypotension and heart rate (HR) response after postural change. The sample median of HR response was used as the cutoff to dichotomize the HR variable (HR ≤ 14 bpm and HR > 14 bpm). p Values are for the interaction terms. Discussion In the first large study with a racially diverse sample from a low-middle income country (mean age = 51 years, 53% white), participants with OH and orthostatic hypertension had lower PVFT scores compared with participants without orthostatic changes in blood pressure. In addition, we found a nonlinear association between orthostatic SBP standardized change and verbal fluency performance. The interaction terms between OH and compensatory increase in HR on the TMT was borderline significant, suggesting that participants who lack compensatory HR after postural changing might have poorer performance on an executive function test. The association between OH and cognitive impairment is controversial. The cross-sectional results from TILDA (mean participant age = 63 years) showed that OH was associated with poorer performance on global cognition and in the memory domain in women (6), and orthostatic hypertension was associated with poorer performance on global cognition and executive function (5). We found a similar association between orthostatic hypertension and poor performance on the PVFT, which is also a test related to executive function. However, OH was not associated with cognitive decline after two years of follow-up in TILDA (8). Similarly, OH did not predict cognitive impairment when adjusted for demographics and cardiovascular risk factors in the ARIC study after 3 and 10 years of follow-up (7). In a convenience sample of 184 frail older adults with history of falls (mean age = 80 years), OH was not related to cognitive impairment, as evaluated by the Mini–Mental State Examination (MMSE) and the Cambridge Examination for Mental Disorders of the Elderly (CAMCOG) (28). On the other hand, in some longitudinal studies, OH was associated with dementia and mild cognitive impairment (9,10,25,26). Among 6,204 participants in the Rotterdam study (mean age = 69), OH and SBP variability were associated with an increased risk of dementia after 15 years of follow-up. Similar to our findings for borderline interaction between compensatory HR and OH, participants who did not have a compensatory increase in HR after standing had an increased risk of dementia in the Rotterdam study (25). The lack of HR increase in the presence of OH may lead to more severe cerebral hypoperfusion due to a further decrease in the cardiac output. In the Malmo Preventive Project, 18,240 Swedish participants (mean age = 45) were followed for an average of 23 years. DBP decrease on standing at baseline predicted a 22% higher risk of dementia, whereas SBP and OH at baseline did not predict dementia (10). Similarly, OH predicted the incidence of mild cognitive impairment in a Swedish cohort of 1,480 participants (mean age = 68) (26). In another cross-sectional study with 495 older patients (mean age = 76), OH was related to poorer performance on global cognitive function and to cognitive status and was more prevalent among patients with mild cognitive impairment, Alzheimer’s disease, and vascular dementia (29). We found consistent results for the association of worse performance on the PVFT with OH and orthostatic hypertension. The U-shaped association with SBP standardized change reinforces these findings, and it has been previously described in a small sample of individuals with asymptomatic OH (30). The PVFT evaluates executive function, which is related to frontal lobe function (31). Executive dysfunction is commonly seen in vascular cognitive impairment, which may be the result of hypoperfusion related to orthostatic changes in blood pressure. Indeed, the association between OH and cognitive impairment can result from three possible mechanisms (32). First, OH may lead to episodes of cerebral hypoperfusion, causing permanent cognitive deterioration (33). Second, cognitive impairment may be a transient symptom of OH with no permanent effects. The presence of OH was associated with poorer performance on global and executive function tests in small samples of patients with neurogenic OH and autonomic disorders who underwent cognitive evaluation during postural changes, supporting this hypothesis (34,35). Finally, OH and cognitive impairment may be the result of a common neuropathological process, which affects brain areas related to cognition and cardiovascular autonomic control (eg, hypothalamus, brain stem, telencephalic structures) (36). Most evidence about the association between OH and cognitive impairment is based on studies with older participants with high prevalence of OH and related diseases (5,6). On the other hand, ELSA-Brasil participants were mostly middle aged and had low prevalence of OH and cardiovascular disease (5,6). Our study has several strengths; it was the first to show evidence about the association between OH and cognitive impairment in a large sample of most middle-aged Brazilian participants. Our sample size was comparable to large studies on this topic, such as the ARIC (n = 12,702) and the Malmo Preventive Project (n = 18,240) (7,10). This large sample size allowed us to investigate interactions between OH and certain variables on cognitive performance; however, we did not find evidence of effect modification by any of these variables in our sample. Furthermore, we included most middle-aged participants; studies with middle-aged participants are necessary to determine risk factors for cognitive impairment, and to develop preventive strategies against dementia, because neuropathological lesions may start decades before the clinical onset of dementia (37). Moreover, we presented evidence in a racially diverse sample (11), whereas most studies included mainly whites from Western European and American cohorts (7,10,25). We were able to adjust our analyses for race, and also test for the interaction between race and OH on cognitive performance, which has not been investigated previously, even in other racially diverse cohort (7). The association of OH with memory and trail making scores lost significance after adjustment for race and education, suggesting that these socioeconomic variables were important confounders. Finally, we used standardized neuropsychological tests, not only screening tools, like the MMSE (9). However, our study should be considered in light of its limitations. This is a cross-sectional study, which limits any considerations about the effect of OH on cognitive function. In addition, although we excluded participants with previous history of stroke, we did not have neuroimaging exams that could explore a possible mediating effect of cerebrovascular lesions in the association between OH and cognitive performance. Although OH and cognition were not evaluated simultaneously in the ELSA-Brasil study as in previous small studies, they were evaluated within a short interval on the same day (34,35). In addition, OH is classically evaluated within 3 minutes of standing (1), but our definition of OH also included delayed orthostatic blood pressure response at 5 minutes. When we included a sensitivity analysis on the association of cognitive performance with early, classical, and delayed OH, only borderline association was found with classical and delayed OH. In addition, ELSA-Brasil is an occupational cohort with higher educational attainment than the Brazilian population, which could be related to lower prevalence of OH and cognitive impairment, limiting the study external validity. However, it is important to note that enrollment in the ELSA-Brasil study was based on occupation (ie, one third of participants were unskilled, one third were technicians, and one third were professionals/faculties), which brings a diverse socioeconomic background to our sample. Finally, the association between OH and poorer performance on PVFT was based on small effects and may have limited clinical meaning, as the association of OH with cognitive impairment was not significant. These findings were expected because we examined a large sample of mostly middle-aged individuals who were free of dementia at baseline. Conclusion OH was weakly associated with poorer performance on verbal fluency scores in a large sample of mostly middle-aged Brazilians. Further longitudinal studies with this racially diverse population will be important to confirm whether OH is associated with cognitive decline and dementia. Supplementary Material Supplementary data is available at The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences online. Funding This work was supported by the Brazilian Ministry of Health and CNPq (grant numbers 01060010.00RS, 01060212.00BA, 01060300.00ES, 01060278.00MG, 01060115.00SP, and 01060071.00RJ). Conflict of Interest None reported. References 1. Freeman R , Wieling W , Axelrod FB , et al. Consensus statement on the definition of orthostatic hypotension, neurally mediated syncope and the postural tachycardia syndrome . Clin Auton Res . 2011 ; 21 : 69 – 72 . doi: 10.1007/s10286-011-0119-5 Google Scholar CrossRef Search ADS PubMed 2. Viramo P , Luukinen H , Koski K , Laippala P , Sulkava R , Kivelä SL . Orthostatic hypotension and cognitive decline in older people . J Am Geriatr Soc . 1999 ; 47 : 600 – 604 . Google Scholar CrossRef Search ADS PubMed 3. Vloet LC , Pel-Little RE , Jansen PA , Jansen RW . High prevalence of postprandial and orthostatic hypotension among geriatric patients admitted to Dutch hospitals . J Gerontol A Biol Sci Med Sci . 2005 ; 60 : 1271 – 1277 . Google Scholar CrossRef Search ADS PubMed 4. Angelousi A , Girerd N , Benetos A , et al. Association between orthostatic hypotension and cardiovascular risk, cerebrovascular risk, cognitive decline and falls as well as overall mortality: a systematic review and meta-analysis . J Hypertens . 2014 ; 32 : 1562 – 1571 ; discussion 1571. doi: 10.1097/HJH.0000000000000235 Google Scholar CrossRef Search ADS PubMed 5. Frewen J , Finucane C , Savva GM , Boyle G , Kenny RA . Orthostatic hypotension is associated with lower cognitive performance in adults aged 50 plus with supine hypertension . J Gerontol A Biol Sci Med Sci . 2014 ; 69 : 878 – 885 . doi: 10.1093/gerona/glt171 Google Scholar CrossRef Search ADS PubMed 6. Frewen J , Savva GM , Boyle G , Finucane C , Kenny RA . Cognitive performance in orthostatic hypotension: findings from a nationally representative sample . J Am Geriatr Soc . 2014 ; 62 : 117 – 122 . doi: 10.1111/jgs.12592 Google Scholar CrossRef Search ADS PubMed 7. Rose KM , Couper D , Eigenbrodt ML , Mosley TH , Sharrett AR , Gottesman RF . Orthostatic hypotension and cognitive function: the Atherosclerosis risk in communities study . Neuroepidemiology . 2010 ; 34 : 1 – 7 . doi: 10.1159/000255459 Google Scholar CrossRef Search ADS PubMed 8. Feeney J , O’Leary N , Kenny RA . Impaired orthostatic blood pressure recovery and cognitive performance at two-year follow up in older adults: the Irish Longitudinal Study on Ageing . Clin Auton Res . 2016 ; 26 : 127 – 133 . doi: 10.1007/s10286-016-0340-3 Google Scholar CrossRef Search ADS PubMed 9. Yap PL , Niti M , Yap KB , Ng TP . Orthostatic hypotension, hypotension and cognitive status: early comorbid markers of primary dementia ? Dement Geriatr Cogn Disord . 2008 ; 26 : 239 – 246 . doi: 10.1159/000160955 Google Scholar CrossRef Search ADS PubMed 10. Holm H , Nagga K , Nilsson ED , et al. Longitudinal and postural changes of blood pressure predict dementia: the Malmo Preventive Project . Eur J Epidemiol . 2017 . doi: 10.1007/s10654-017-0228-0 11. Schmidt MI , Duncan BB , Mill JG , et al. Cohort profile: longitudinal study of adult health (ELSA-Brasil) . Int J Epidemiol . 2015 ; 44 : 68 – 75 . doi: 10.1093/ije/dyu027 Google Scholar CrossRef Search ADS PubMed 12. Mill JG , Pinto K , Griep RH , et al. Medical assessments and measurements in ELSA-Brasil . Rev Saude Publica . 2013 ; 47 ( suppl 2 ): 54 – 62 . doi: 10.1590/S0034-8910.2013047003851 Google Scholar CrossRef Search ADS PubMed 13. Morris JC , Heyman A , Mohs RC , et al. The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part I. Clinical and neuropsychological assessment of Alzheimer’s disease . Neurology . 1989 ; 39 : 1159 – 1165 . Google Scholar CrossRef Search ADS PubMed 14. Passos VM , Caramelli P , Benseñor I , Giatti L , Barreto SM . Methods of cognitive function investigation in the Longitudinal Study on Adult Health (ELSA-Brasil) . Sao Paulo Med J . 2014 ; 132 : 170 – 177 . doi: 10.1590/1516-3180.2014.1323646 Google Scholar CrossRef Search ADS PubMed 15. Greenlief CL , Margolis RB , Erker GJ . Application of the trail making test in differentiating neuropsychological impairment of elderly persons . Percept Mot Skills . 1985 ; 61 : 1283 – 1289 . doi: 10.2466/pms.1985.61.3f.1283 Google Scholar CrossRef Search ADS PubMed 16. Bertolucci PH , Okamoto IH , Brucki SM , Siviero MO , Toniolo Neto J , Ramos LR . Applicability of the CERAD neuropsychological battery to Brazilian elderly . Arq Neuropsiquiatr . 2001 ; 59 : 532 – 536 . Google Scholar CrossRef Search ADS PubMed 17. Shao Z , Janse E , Visser K , Meyer AS . What do verbal fluency tasks measure? Predictors of verbal fluency performance in older adults . Front Psychol . 2014 ; 5 : 772 . doi: 10.3389/fpsyg.2014.00772 Google Scholar CrossRef Search ADS PubMed 18. Rawlings AM , Sharrett AR , Schneider AL , et al. Diabetes in midlife and cognitive change over 20 years: a cohort study . Ann Intern Med . 2014 ; 161 : 785 – 793 . doi: 10.7326/M14-0737 Google Scholar CrossRef Search ADS PubMed 19. Bennett DA , Schneider JA , Buchman AS , Barnes LL , Boyle PA , Wilson RS . Overview and findings from the rush Memory and Aging Project . Curr Alzheimer Res . 2012 ; 9 : 646 – 663 . doi: 10.2174/156720512801322663 Google Scholar CrossRef Search ADS PubMed 20. Rose KM , Eigenbrodt ML , Biga RL , et al. Orthostatic hypotension predicts mortality in middle-aged adults: the Atherosclerosis Risk In Communities (ARIC) Study . Circulation . 2006 ; 114 : 630 – 636 . doi: 10.1161/CIRCULATIONAHA.105.598722 Google Scholar CrossRef Search ADS PubMed 21. Fessel J , Robertson D . Orthostatic hypertension: when pressor reflexes overcompensate . Nat Clin Pract Nephrol . 2006 ; 2 : 424 – 431 . doi: 10.1038/ncpneph0228 Google Scholar CrossRef Search ADS PubMed 22. World Health Organization . International Statistical Classification of Diseases and Related Health Problems . Vol. 1 . Geneva, Switzerland : World Health Organization ; 2004 . 23. Lewis G , Pelosi AJ , Araya R , Dunn G . Measuring psychiatric disorder in the community: a standardized assessment for use by lay interviewers . Psychol Med . 1992 ; 22 : 465 – 486 . Google Scholar CrossRef Search ADS PubMed 24. Perlmuter LC , Sarda G , Casavant V , Mosnaim AD . A review of the etiology, associated comorbidities, and treatment of orthostatic hypotension . Am J Ther . 2013 ; 20 : 279 – 291 . doi: 10.1097/MJT.0b013e31828bfb7f Google Scholar CrossRef Search ADS PubMed 25. Wolters FJ , Mattace-Raso FU , Koudstaal PJ , Hofman A , Ikram MA ; Heart Brain Connection Collaborative Research Group . Orthostatic hypotension and the long-term risk of dementia: a population-based study . PLoS Med . 2016 ; 13 : e1002143 . doi: 10.1371/journal.pmed.1002143 Google Scholar CrossRef Search ADS PubMed 26. Elmståhl S , Widerström E . Orthostatic intolerance predicts mild cognitive impairment: incidence of mild cognitive impairment and dementia from the Swedish general population cohort good aging in Skåne . Clin Interv Aging . 2014 ; 9 : 1993 – 2002 . doi: 10.2147/CIA.S72316 Google Scholar CrossRef Search ADS PubMed 27. Albert MS , DeKosky ST , Dickson D , et al. The diagnosis of mild cognitive impairment due to Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease . Alzheimers Dement . 2011 ; 7 : 270 – 279 . doi: 10.1016/j.jalz.2011.03.008 Google Scholar CrossRef Search ADS PubMed 28. Schoon Y , Lagro J , Verhoeven Y , Rikkert MO , Claassen J . Hypotensive syndromes are not associated with cognitive impairment in geriatric patients . Am J Alzheimers Dis Other Demen . 2013 ; 28 : 47 – 53 . doi: 10.1177/1533317512466692 Google Scholar CrossRef Search ADS PubMed 29. Mehrabian S , Duron E , Labouree F , et al. Relationship between orthostatic hypotension and cognitive impairment in the elderly . J Neurol Sci . 2010 ; 299 : 45 – 48 . doi: 10.1016/j.jns.2010.08.056 Google Scholar CrossRef Search ADS PubMed 30. Czajkowska J , Ozhog S , Smith E , Perlmuter LC . Cognition and hopelessness in association with subsyndromal orthostatic hypotension . J Gerontol A Biol Sci Med Sci . 2010 ; 65 : 873 – 879 . doi: 10.1093/gerona/glq068 Google Scholar CrossRef Search ADS PubMed 31. Alvarez JA , Emory E . Executive function and the frontal lobes: a meta-analytic review . Neuropsychol Rev . 2006 ; 16 : 17 – 42 . doi: 10.1007/s11065-006-9002-x Google Scholar CrossRef Search ADS PubMed 32. Sambati L , Calandra-Buonaura G , Poda R , Guaraldi P , Cortelli P . Orthostatic hypotension and cognitive impairment: a dangerous association ? Neurol Sci . 2014 ; 35 : 951 – 957 . doi: 10.1007/s10072-014-1686-8 Google Scholar CrossRef Search ADS PubMed 33. Kim JS , Oh YS , Lee KS , Kim YI , Yang DW , Goldstein DS . Association of cognitive dysfunction with neurocirculatory abnormalities in early Parkinson disease . Neurology . 2012 ; 79 : 1323 – 1331 . doi: 10.1212/WNL.0b013e31826c1acd Google Scholar CrossRef Search ADS PubMed 34. Guaraldi P , Poda R , Calandra-Buonaura G , et al. Cognitive function in peripheral autonomic disorders . PLoS One . 2014 ; 9 : e85020 . doi: 10.1371/journal.pone.0085020 Google Scholar CrossRef Search ADS PubMed 35. Poda R , Guaraldi P , Solieri L , et al. Standing worsens cognitive functions in patients with neurogenic orthostatic hypotension . Neurol Sci . 2012 ; 33 : 469 – 473 . doi: 10.1007/s10072-011-0746-6 Google Scholar CrossRef Search ADS PubMed 36. Idiaquez J , Roman GC . Autonomic dysfunction in neurodegenerative dementias . J Neurol Sci . 2011 ; 305 : 22 – 27 . doi: 10.1016/j.jns.2011.02.033 Google Scholar CrossRef Search ADS PubMed 37. Sperling R , Mormino E , Johnson K . The evolution of preclinical Alzheimer’s disease: implications for prevention trials . Neuron . 2014 ; 84 : 608 – 622 . doi: 10.1016/j.neuron.2014.10.038 Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

The Journals of Gerontology Series A: Biomedical Sciences and Medical SciencesOxford University Press

Published: Mar 20, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off