Left Ventricular Mass Index Is Associated With Cognitive Function in Middle-Age
Elevated cardiovascular disease risk factor burden is a recognized contributor to poorer cognitive function; however, the physiological mechanisms underlying this association are not well understood. We sought to assess the potential mediation effect of left ventricular (LV) remodeling on the association between lifetime systolic blood pressure and cognitive function in a community-based cohort of middle-aged adults.
Nine hundred sixty participants of the Bogalusa Heart Study (59.2% women, 33.8% black, aged 48.4±5.1 years) received 2-dimensional echocardiography to quantify relative wall thickness, LV mass, and diastolic and systolic LV function; and a standardized neurocognitive battery to assess memory, executive functioning, and language processing. Multivariable linear regression assessed the association of cardiac structure and function with a global composite cognitive function score, adjusting for traditional cardiovascular disease risk factors. Mediation analysis assessed the effect of LV mass index on the association between lifetime systolic blood pressure burden and cognitive function.
There were 233 (24.3%) and 136 (14.2%) individuals with concentric LV remodeling and concentric LV hypertrophy, respectively. Each g/m2.7 increment in LV mass index was associated with a 0.03 standardized unit decrement in global cognitive function (P=0.03). Individuals with concentric LV remodeling and isolated diastolic dysfunction had the poorest cognitive function, and a greater ratio between early mitral inflow velocity and early diastolic mitral annular velocity (E/e’) was associated with poorer cognitive function, even after adjustment for LV mass index (B=−0.12; P=0.03). A total of 18.8% of the association between lifetime systolic blood pressure burden and midlife cognitive function was accounted for by LV mass index.
Cardiac remodeling partially mediates the association between lifespan systolic blood pressure burden and adult cognition in individuals without dementia or clinical cardiovascular disease. Slowing or reversing the progression of cardiac remodeling in middle-age may be a novel therapeutic approach to prevent cognitive decline.
Cognitive dysfunction and dementia are multifactorial processes characterized by the loss of intellectual functions, including but not limited to impaired reasoning, memory loss, and altered language processing. Because of the nonreversible nature and poor prognosis associated with dementia, identification and minimization of upstream risk factors is important for prevention in the clinical setting. Although hypertension has been previously identified as a risk factor for dementia later in life, the potential mediating role of cardiac structure and function on the relationship between blood pressure and cognitive function has not been explored, especially in younger diverse cohorts. Here, we observed that elevated left ventricular mass index associates with poorer cognitive function in a dose-dependent fashion, independent of traditional cardiovascular disease risk factors, among middle-aged Black and White men and women. Likewise, left ventricular mass index partially mediated the association between lifetime systolic blood pressure burden and cognitive function. Individuals with diastolic dysfunction, concentric hypertrophy, and/or concentric remodeling had worse cognitive outcomes compared with individuals with normal left ventricular structure and function. Systolic dysfunction and eccentric hypertrophy, however, were not associated with cognitive function. Thus, higher left ventricular mass index and diastolic dysfunction uniquely confer an elevated risk of cognitive impairment, independent of blood pressure, among individuals with subclinical cardiovascular disease. Clinical recognition and treatment of cardiac remodeling in middle-age may be a novel approach to prevent dementia later in life.
See Editorial by Bella
Dementia affects >40 million individuals globally,1 and the absolute burden of the disease is projected to increase by 3-fold over the next several decades.2 A total of 4.9 million individuals above the age of 65 suffer from late-onset Alzheimer’s disease in the United States,3 and the dementia-attributable mortality rate has more than doubled since 2000.4 While dementia continues to adversely affect individual patients, families, as well as broader healthcare and economic systems,3 there are very few effective treatments for this disease. Thus, these trends and the nonreversible nature of dementia underline a need to improve disease prevention to help curb the dementia epidemic and facilitate the likelihood of healthy aging.
Subclinical cardiovascular disease (CVD) and its upstream risk factors strongly influence the risk of dementia, and a 10% reduction in modifiable CVD risk factor burden could prevent 2 million cases of dementia, worldwide, by 2050.5 Addressing the vascular contributions to dementia is 1 of 6 recently outlined scientific focus areas by the National Heart, Lung and Blood Institute.6 Vascular disease is a key determinant in up to one-half of all dementia cases,7,8 and intrabrain vascular dysregulation has been recognized as an initial pathological insult in the development of late-onset Alzheimer’s disease.9 These observations are at least partially driven by hypertension, as several observational studies10,11 and the notable SPRINT (systolic blood pressure intervention trial)12 have demonstrated that hypertension over the life course independently associates with a higher risk of cognitive dysfunction later in life. However, other studies have noted that persons with dementia have lower blood pressure13 suggesting an age-dependent relationship, and that intermediary CVD phenotypes, including cardiac remodeling, may perhaps also influence the relationship between CVD risk factors and cognitive function.
Although early evidence suggests that left ventricular remodeling may increase dementia risk,14,15 these studies have been limited to older populations and have not formally quantified the mediation effect of cardiac anatomy and physiology on the association between blood pressure and dementia. As such, the independent association between left ventricular remodeling and cognitive function has not been previously explored in middle-aged individuals before the onset of clinical dementia. This knowledge gap and the sustained presence of racial disparities in CVD risk factor burden16 underline the need to conduct such studies in younger, racially diverse cohorts to improve the timing and development of primary prevention interventions for both CVD and dementia.
We conducted the first known study to examine the mediation effect of echocardiography-derived measures of left ventricular structure on the association between lifetime systolic blood pressure burden and cognitive function in a biracial, community-based cohort of middle-aged adults. We hypothesized that individuals with adverse cardiac structure and function would have poorer cognition even at a young age, and that left ventricular mass index (LVMI) would partially mediate the relationship between lifetime blood pressure burden and cognitive function in midlife.
The authors declare that all supporting data are available within the article and in the Data Supplement.
The Bogalusa Heart Study (BHS) is a long-term epidemiological study that observes cardiovascular health across the lifespan.17 Between 1973 and 2016, 7 surveys of children aged 4 to 17 as well as 10 surveys of adults, who had been previously examined as children, were conducted. The current BHS cohort includes 1298 participants, born between January 1959 and June 1979 who were examined in both childhood and adulthood. For the present study, we selected the 960 individuals who underwent contemporaneous echocardiographic, neurocognitive battery, and covariable examinations, as well whom had at least 6 measurements of blood pressure from childhood to adulthood. The current study sample (n=960) had a similar demographic and cardiometabolic profile compared with the 338 BHS participants that were excluded due to missing echocardiographic and cognitive function data (Table I in the Data Supplement). All study participants provided written informed consent at each examination, and study protocols were approved the Institutional Review Board of the Tulane University Health Sciences Center.
Two-dimensional echocardiography was performed by trained cardiac sonographers at the BHS field office. Participants were placed in a partial left lateral decubitus position for echocardiographic assessment. Fifteen to 20 cycles of 2-dimensional image signals were recorded. Echocardiographic recordings were accomplished using an Aplio 300 ultrasound instrument (Toshiba America Medical Systems, Tustin, CA) with a linear array transducer of 7.5 mHz using a standard protocol.18,19 Cardiac parameter, RWT, was calculated as the ratio of twice the posterior wall thickness divided by the left ventricular internal diameter in diastole, and a value over 0.42 centimeters was defined as an elevated RWT.20 Parasternal long- and short- axis views were used for measuring left ventricular end-diastolic and end-systolic measurements in duplicate and then averaged. Left ventricular mass was calculated from a necropsy-validated formula based on a thick-wall prolate ellipsoidal geometry,21 and left ventricular mass was indexed to average height in meters2.7 to obtain the LVMI parameter.22 An elevated LVMI was defined as greater than 46.7 g/m2.7 and 49.2 g/m2.7 in women and men, respectively.22 Four categories of left ventricular geometry were defined as follows: (1) normal (normal RWT and LVMI); (2) eccentric hypertrophy (normal RWT and high LVMI); (3) concentric hypertrophy (high RWT and high LVMI); and (4) concentric remodeling (high RWT and normal LVMI).
Diastolic function parameters assessed included early left ventricular filling peak velocity (E), late left ventricular filling peak velocity (A), medial mitral annular velocity (e’), deceleration time of the E wave (DT), and isovolumic relaxation time (IVRT). The apical 4-chamber view was utilized to assess patterns of mitral inflow including, DT, E wave, A wave, and medial e’ wave velocities. The E and A wave velocities were assessed using pulsed-wave Doppler echocardiography of transmitral flow at the mitral valve leaflet tips, while medial e’ velocity was measured using pulsed-wave tissue Doppler echocardiography of the mitral annulus. The apical 5-chamber view was used to measure IVRT, specifically by placing the sample volume in the left ventricular outflow tract to concurrently evaluate aortic ejection and the onset of mitral inflow. Doppler sample volumes were placed between the mitral leaflet tips to measure DT. Using the European Association of Echocardiography/American Society of Echocardiography guidelines, individuals were then grouped as having normal left ventricular diastolic function (E/A≥0.8 and IVRT<100 milliseconds and DT<200 milliseconds), impaired left ventricular relaxation (E/A<0.8 and IVRT≥100 milliseconds and DT≥200 milliseconds), pseudonormal left ventricular filling (0.8<E/A<1.5 and medial e’<8 milliseconds or lateral e’<10 milliseconds or E/e’>10), and restrictive left ventricular filling (E/A≥2 and IVRT≤60 milliseconds and DT<160 milliseconds).23 Left ventricular diastolic dysfunction was defined having impaired relaxation, pseudonormal filling, or restrictive filling.23 Left ventricular ejection fraction was assessed by tracing the endocardial border in the 4-chamber view in end-systole and end-diastole. Individuals with an ejection fraction ≥55% were considered to have preserved systolic function, while those <55% were defined as having impaired systolic function.
Cognitive Function Assessment
Ten standardized tests were used to assess global cognitive function (Figure I in the Data Supplement) and were conducted by trained BHS staff. These tests included the following: (1) Logical Memory I (Weschler memory Scale IV), (2) Logical Memory II (Weschler Memory Scale IV), (3) Recognition (Weschler Memory Scale IV), (4) Digit Span Forward, (5) Digit Span Backward), (Weschler Adult Intelligence Scale IV), (6) Word and Letter Reading (Wide Range Achievement Test IV), evaluating decoding capability, (7) Vocabulary (Weschler Adult Intelligence Scale IV), (8) Digit Symbol Coding (Weschler Adult Intelligence Scale IV), (9) Trail Making Test A, (10) Trail Making Test B. Excluding the Trail Making Tests, a higher score represented superior cognitive function for all tests. Thus, after inverting the sign for the Trail Making Tests, raw scores of the components of the cognitive function assessment were transformed to z scores with a mean of zero and an SD of 1.0. These standardized scores were summed to formulate a global cognitive function assessment score. Assessment of cognitive performance across 4 domains was also assessed, including episodic memory, working memory, executive function, and language processing (Figure I in the Data Supplement).
General Clinical Examination
Stringent protocols were used to collect clinical data on Bogalusa Heart Study participants. Validated questionnaires were used to acquire demographic and lifestyle variables specifically, age, race, sex, cigarette smoking, alcohol consumption, educational attainment, and medication status. Fasting measures of low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), glucose, and triglycerides were collected using standardized methods.24 Blood pressure was measured in triplicate, while height and weight were measured in duplicate at the time of each physical exam. Weight in kilograms was divided by height in meters squared to calculate body mass index (BMI). Triplicate measures for blood pressure and duplicate measures for height and weight were averaged for each study participant.
Continuous variables were presented as mean+SD, while numbers and percentages were used to present categorical variables. Normality of continuous variables was assessed via the Kolmogorov-Smirnov test. The Student t-test and Wilcoxon signed-rank test were used to assess differences in normally and non-normally distributed continuous variables, respectively. Differences between categorical variables were evaluated using Pearson χ2 test. We conducted and reported race- and sex-stratified analyses when appropriate. All hypothesis tests were 2-sided and used an alpha threshold of 0.05.
Associations of left ventricular structure and function with global cognitive function, and its 4 domains, were assessed using multivariable-adjusted linear regression models. Relevant covariables were adjusted for in multiple regression models, including age, sex, race, education, cigarette smoking, alcohol drinking, systolic blood pressure, diastolic blood pressure, BMI, LDL-C, HDL-C, glucose, triglycerides, and blood pressure-, cholesterol-, and glucose-lowering medications. These models were employed cross-sectionally in the full sample of BHS participants with measures of LVMI, cognitive function, and respective covariables (n=960).
Mediation analysis was performed using linear regression models to assess the potential mediation effect of LVMI on the association of lifetime systolic blood pressure and cognitive function. Mediation analysis models were created as previously described by VanderWeele and Sobel25,26 and the assumptions of mediation analysis, including temporality were verified. We focused on lifetime systolic blood pressure rather than lifetime diastolic blood pressure for the mediation analysis, as previous randomized controlled trial evidence has demonstrated that systolic blood pressure is the primary vascular driver of both CVD and dementia.12,27 Long-term burden or lifetime systolic blood pressure was measured as the area under the curve using 6 to 16 measurements of systolic blood pressure from childhood to adulthood. Three linear regression models were used to assess mediation using sex- and race-specific standardized scores for all continuous variables: (1) lifetime systolic blood pressure predicting global cognitive function at follow-up, adjusting for age, education, cigarette smoking, alcohol drinking, lifetime diastolic blood pressure, BMI, LDL-C, HDL-C, glucose, triglycerides, and blood pressure-, cholesterol-, and glucose-lowering medications (ßdirect=direct effect); (2) lifetime systolic blood pressure predicting LVMI at follow-up; adjusting for age, education, cigarette smoking, alcohol drinking, lifetime diastolic blood pressure, BMI, LDL-C, HDL-C, glucose, triglycerides, and blood pressure-, cholesterol-, and glucose-lowering medications (ß1=indirect effect 1); and (3) LVMI at follow-up predicting cognitive function at follow-up, adjusting for age, education, cigarette smoking, alcohol drinking, lifetime systolic blood pressure, lifetime diastolic blood pressure, BMI, LDL-C, HDL-C, glucose, triglycerides, and blood pressure-, cholesterol-, and glucose-lowering medications (ß2=indirect effect 2). The total indirect effect (ßindirect) was calculated as the product of indirect effect 1 and indirect effect 2. The total effect (ßtotal) was the sum of the total indirect effect (ßindirect) and direct effect (ßdirect). The mediation effect was the quotient of the total indirect effect and the total effect (ßindirect/ ßtotal). We conducted one a priori sensitivity analysis, using area under the curve values for systolic blood pressure and diastolic blood pressure in the main linear regression analyses. We conducted 2 post hoc sensitivity analyses: (1) adjusting for LVMI in a linear regression model assessing the association between diastolic function parameters and cognitive function and (2) excluding individuals with an ejection fraction less than 55% when assessing the association between LVMI and cognitive function and when assessing the mediation effect of LVMI on the association of lifetime systolic blood pressure burden and cognitive function.
Table 1 summarizes sociodemographic, lifestyle, cognitive, and cardiovascular-related characteristics stratified by race and sex. Over one half of the study sample (60.7%) had hypertension and Blacks had a significantly higher proportion of hypertension compared with Whites, regardless of sex. We observed no sex or race differences according to diabetes status, and the prevalence of type 2 diabetes mellitus in the sample was 13.3%. There were several race and sex differences in global cognitive function, LVMI, left ventricular geometry, diastolic blood pressure, and systolic blood pressure. Women and Whites had significantly higher global cognitive function scores compared with men and Blacks. On the other hand, men and Blacks had significantly higher LVMI, systolic blood pressure, and diastolic blood pressure versus women and Whites. White women contributed the highest proportion of individuals with normal left ventricular geometry (72.3%), while we observed the greatest percentage of left ventricular concentric remodeling (34.1%) and concentric hypertrophy (29.3%) among Black men. Black women and men had a significantly higher proportion of individuals with diastolic dysfunction compared with White women and men (women: 26.2% versus 12.4; P<0.0001; men: 22.8% versus 10.7%; P<0.01). We did not observe significant sex nor race differences for LDL-C or fasting blood glucose.
|Variable||All (n=960)||Whites (635)||Blacks (n=325)||P-Value for Race Difference|
|White Men (n=270)||White Women (n=365)||Black Men (n=123)||Black Women (n=202)||Men||Women|
|Sociodemographic and lifestyle|
|Age, y*||48.4 (5.1)||49.1 (4.8)†||48.1 (5.1)||48.3 (5.6)||48.1 (5.1)||0.16||0.99|
|Post–high school education‡||487 (50.7)||142 (52.6)†||224 (61.4)||39 (31.7)||82 (40.6)||<0.0001||<0.0001|
|Never||488 (50.8)||134 (49.6)||194 (53.1)||45 (36.6)§||115 (56.9)|
|Former||283 (29.5)||87 (32.2)||109 (29.9)||35 (28.4)||52 (25.8)|
|Current||189 (19.7)||49 (18.2)||62 (17.0)||43 (35.0)||35 (17.3)|
|Never||101 (10.5)||10 (3.7)§||45 (12.3)||10 (8.1)§||36 (17.8)|
|Former||307 (32.0)||91 (33.7)||114 (31.3)||38 (30.9)||64 (31.7)|
|Current||552 (57.5)||169 (62.6)||206 (56.4)||75 (61.0)||102 (50.5)|
|Global Cognitive Function Score||0.4 (4.8)||1.1 (4.4)§||2.2 (4.6)||-2.9 (4.1)†||-1.9 (4.3)||<0.0001||<0.0001|
|Left ventricular mass, g*||157.8 (52.0)||171.5 (43.9)§||130.6 (32.9)||210.6 (71.9)§||156.8 (43.6)||<0.0001||<0.0001|
|Left ventricular mass index, g/m2.7*||38.5 (11.4)||36.6 (9.6)||35.5 (9.1)||45.6 (15.3)†||42.3 (11.6)||<0.0001||<0.0001|
|Relative wall thickness, cm*||0.4 (0.1)||0.4 (0.1)§||0.4 (0.1)||0.5 (0.1)||0.4 (0.1)||<0.0001||<0.0001|
|Left ventricular geometry‡||<0.0001||<0.0001|
|Normal||557 (58.0)||165 (61.1)§||264 (72.3)||35 (28.5)§||93 (46.0)|
|Concentric remodeling||223 (23.2)||80 (29.6)||61 (16.7)||42 (34.1)||50 (24.8)|
|Eccentric LV hypertrophy||44 (4.6)||7 (2.6)||17 (4.7)||10 (8.1)||10 (5.0)|
|Concentric LV hypertrophy||136 (14.2)||18 (6.7)||23 (6.3)||36 (29.3)||49 (24.2)|
|Ejection fraction ≥55%||932 (97.1)||263 (97.4)||359 (98.4)||112 (91.1)§||198 (98.0)|
|Ejection fraction <55%||28 (2.9)||7 (2.6)||6 (1.6)||11 (8.9)||4 (2.0)|
|Normal||805 (83.9)||241 (89.3)||320 (87.6)||95 (77.2)||149 (73.8)|
|Impaired relaxation||21 (2.2)||2 (0.7)||9 (2.5)||4 (3.3)||6 (3.0)|
|Pseudonormal filling||134 (13.9)||27 (10.0)||36 (9.9)||24 (19.5)||47 (23.2)|
|Cardiovascular disease risk factors|
|Hypertension‡||583 (60.7)||167 (61.9)§||173 (47.4)||94 (76.4)||149 (73.8)||<0.01||<0.0001|
|Proportion of individuals with hypertension treated‡||334 (57.3)||78 (46.7)§||97 (56.1)||56 (59.6)||103 (69.1)||0.04||0.02|
|Systolic blood pressure, mm Hg*||123.6 (16.8)||126.2 (13.5)†||118.0 (14.6)||131.0 (16.6)†||125.7 (21.1)||<0.01||<0.0001|
|Total AUC of systolic blood pressure, mm Hg*||113.1 (9.2)||115.1 (7.7)§||109.1 (8.1)||118.7 (9.5)§||114.1 (9.9)||<0.001||<0.0001|
|Diastolic blood pressure, mm Hg*||78.7 (11.5)||79.7 (10.2)§||75.6 (10.1)||83.0 (12.9)||80.4 (13.1)||0.01||<0.0001|
|Total AUC of diastolic blood pressure, mm Hg*||69.6 (7.6)||69.9 (7.5)†||68.5 (6.8)||70.7 (9.5)||70.4 (7.6)||0.38||<0.01|
|Blood-pressure lowering medication‡||334 (34.8)||78 (28.9)||97 (26.6)||56 (45.5)||103 (51.0)||<0.01||<0.0001|
|Cholesterol-lowering medication‡||131 (13.6)||50 (18.5)†||42 (11.5)||20 (16.3)||19 (9.4)||0.59||0.44|
|Glucose-lowering medication‡||101 (10.5)||21 (7.8)||37 (10.1)||17 (13.8)||26 (12.9)||0.06||0.32|
|BMI, kg/m2*||31.3 (7.7)||30.7 (6.0)||30.2 (7.5)||31.2 (9.1)§||34.2 (8.6)||0.57||<0.0001|
|LDL cholesterol, mg/dL*||115.0 (35.6)||117.3 (33.6)||116.2 (34.3)||109.3 (40.1)||113.2 (37.3)||0.06||0.35|
|HDL cholesterol, mg/dL*||51.7 (16.1)||43.2 (12.3)§||56.5 (16.5)||51.2 (15.6)||54.5 (15.6)||<0.0001||0.16|
|Serum triglycerides, mg/dL, median (Q1–Q3)||126.0 (78.0–153.0)||126.0 (91.0, 179.0)§||108.0 (81.0, 155.0)||106.0 (76.0, 145.0)§||88.0 (66.0, 124.0)||<0.01||<0.0001|
|Type 2 diabetes mellitus‡||128 (13.3)||34 (12.6)||39 (10.7)||23 (18.7)||32 (15.8)||0.11||0.08|
|Proportion of individuals with type 2 diabetes mellitus treated ‡||72 (56.3)||14 (41.2)||23 (59.0)||13 (56.5)||22 (68.8)||0.26||0.39|
|Blood glucose, mg/dL*||105.5 (33.0)||106.4 (25.5)||103.4 (31.7)||106.1 (29.3)||107.7 (44.4)||0.93||0.22|
Table 2 presents the multivariable-adjusted, cross-sectional relationship of left ventricular structure and function with cognition. LVMI (B=−0.03; P=0.03) and E/e’ ratio (B=−0.14; P=0.01) were inversely associated with global cognitive function. Upon disaggregating the global cognitive score into respective cognition domains, LVMI (B=−6.6×10−3; P=0.02) and E/e’ (B=−0.03; P=0.02) emerged as significant predictors of working memory. LVMI and E/e’ ratio remained significant negative predictors of global cognitive function and working memory, after replacing follow-up systolic and diastolic blood pressure measurements with their respective area under the curve, or long-term burden values (Table II in the Data Supplement). Likewise, E/e’ ratio maintained a significant inverse association with global cognitive function, even after adjusting for LVMI (Table III in the Data Supplement).
|Variable||Global Cognitive Function||Episodic Memory||Working Memory||Executive Function||Language|
|Beta (SE)||P Value||Beta (SE)||P Value||Beta (SE)||P Value||Beta (SE)||P||Beta (SE)||P Value|
|Left ventricular mass index, g/m2.7||−0.03 (0.02)||0.03*||−2.4x10−3 (0.00)||0.41||−6.6x10−3 (0.00)||0.02*||−8.8x10−4 (0.00)||0.60||−1.5x10−3 (0.00)||0.31|
|Relative wall thickness||−0.82 (0.68)||0.68||−0.70 (0.39)||0.07||−0.02 (0.38)||0.96||0.16 (0.22)||0.47||0.46 (0.19)||0.02*|
|Ejection fraction (%)||−0.03 (0.03)||0.25||−6.7x10−3 (0.01)||0.19||1.2x10−3 (0.01)||0.81||−2.6x10−3 (0.00)||0.37||−1.5x10−3 (0.00)||0.56|
|E/A ratio||0.13 (0.42)||0.76||0.07 (0.08)||0.42||−0.02 (0.08)||0.79||−6.5x10−3 (0.00)||0.89||4.5x10−3 (0.04)||0.91|
|E/e’ ratio||−0.14 (0.01)||0.01*||−0.02 (0.01)||0.06||−0.03 (0.01)||0.02*||5.6x10−3 (0.00)||0.38||−4.1x10−3 (0.01)||0.47|
|Isovolumic relaxation time, ms||−8.4x10−3 (0.01)||0.18||−5.0x10−4 (0.00)||0.69||−6.4x10−4 (0.00)||0.61||−9.8x10−4 (0.00)||0.17||−1.4x10−4 (0.00)||0.82|
|Deceleration time, ms||−2.3x10−3 (0.00)||0.41||1.8x10−4 (0.00)||0.74||−5.5x10−4 (0.00)||0.32||−3.2x10−4 (0.00)||0.32||−3.0x10−4 (0.00)||0.28|
|Isolated diastolic dysfunction||−0.74 (0.39)||0.05||−0.07 (0.08)||0.35||−0.08 (0.08)||0.32||−0.03 (0.04)||0.51||−0.06 (0.04)||0.11|
To further explore the relationship between the left ventricle and cognition, we assessed differences in global cognitive score across subclinical cardiac phenotypes (Figures 1 and 2). Compared with individuals with normal left ventricular geometry, individuals with concentric left ventricular hypertrophy (P<0.01) and concentric left ventricular remodeling (P<0.0001) had significantly lower global cognitive function (Figure 1). Contrastingly, eccentric left ventricular remodeling was not associated with global cognitive function (Figure 1). Individuals with both isolated diastolic dysfunction (P<0.001) and combined systolic and diastolic dysfunction (P<0.001) had poorer cognition versus those with preserved systolic and diastolic function (Figure 2).
Figure 3 shows the multivariable-adjusted mediation effect of adult LVMI on the lifetime systolic blood pressure burden-adult cognitive function association. The direct effect of lifetime systolic blood pressure on adult cognitive function was measured as the standardized regression coefficient (βtotal=−0.082; P<0.0001) estimated without adult LVMI in the model. The total indirect effect (βindirect=−0.019) through adulthood LVMI was defined as the product of indirect effect 1 (β1=0.353; P<0.0001) and indirect effect 2 (β2=−0.054; P<0.001). The mediation effect of adult LVMI on lifetime systolic blood pressure burden and adult cognitive function was 18.8%. Lifetime systolic blood pressure burden did not significantly predict adult cognitive function after accounting for adult LVMI. Lifetime diastolic blood pressure burden did not significantly predict adult cognitive function in an unadjusted model nor after considering adult LVMI in the multivariable model. Neither sex nor race modified the association of left ventricular structure and function with cognitive function.
Results from our sensitivity analyses excluding individuals with an ejection fraction <55% are presented in Table IV and Figure II in the Data Supplement. Beta estimates and P values in these sensitivity analyses were consistent with primary study results presented in Table 2 and Figure 2. LVMI (B=−0.04, P=0.03) maintained its effect size with global cognitive function after excluding individuals with an ejection fraction <55%. LVMI partially mediated a similar, albeit slightly higher proportion (19.6%) of the association between lifetime systolic blood pressure and cognitive function after excluding individuals with an ejection fraction <55%.
Although it is generally accepted that hypertension over the life course increases dementia risk through vascular injury in the brain, the degree of mediation effect of later life LVMI has never been reported, specifically by leveraging the cumulative burden of blood pressure from childhood to adulthood. We now provide mechanistic and novel evidence that a higher LVMI partially mediates the association between systolic blood pressure burden and cognitive function in middle-aged Black and White men and women without dementia, and that individuals with preclinical diastolic dysfunction and concentric left ventricular remodeling may have the worst cognitive outcomes. In particular, we discovered that nearly one-fifth of the negative association between systolic blood pressure burden and cognition in middle-age can be explained by a higher LVMI in adulthood and that LVMI holds a negative dose-response relationship with cognitive function, even among individuals without overt CVD. These findings suggest that the long-term influence of elevated systolic blood pressure28 from childhood on cognitive deficits in adulthood is partially mediated through a higher LVMI later in life, and that ventricular remodeling and filling abnormalities may hold a key role in the pathophysiology of dementia universally across sex and race groups. Thus, detection and treatment of left ventricular hypertrophy in middle-age may be a highly generalizable strategy to prevent late-life cognitive impairment and dementia.
Our study is the first to formally test inferences derived from previous epidemiological evidence that suggests an association between left ventricular physiology and dementia. The Multi-Ethnic Study of Atherosclerosis reported that a higher LVMI and LV-mass-to-volume ratio, as measured by cardiac magnetic resonance imaging, independently associated with dementia and impaired cognitive function in persons over 60 years of age.14 A proportion of Multi-Ethnic Study of Atherosclerosis participants, however, also had elevated systolic blood pressure and the association between diastolic dysfunction and dementia became nonsignificant after adjustment for LVMI. Our results further confirm the association between LVMI and cognitive function, but now in the novel context of long-term burden of systolic blood pressure adjustment and a mediation analysis in a younger study sample. Additionally, the association between left ventricular filling and cognitive function remained significant even after adjustment of LVMI in Bogalusa participants, suggesting that diastolic dysfunction itself may be an independent risk factor for dementia. Using Cornell voltage criteria to define left ventricular hypertrophy, an analysis in over 12 000 participants in the ARIC study found that left ventricular hypertrophy associated with lower cognitive function, but not cognitive decline.29 Such discordant results may be explained by distinctions between cardiac structure and function, or differences between echocardiography and electrocardiography in measuring subtle differences in left ventricular hypertrophy.
Earlier studies have demonstrated that midlife hypertension significantly associates with a higher risk of dementia in late adulthood. In the Atherosclerosis Risk in Communities (ARIC) cohort, baseline hypertension predicted an additional cognitive decline of nearly 5% compared with normal blood pressure over 20 years, and treatment of hypertension appeared to mitigate cognitive declines in nearly 14 000 Black and White study participants.30 Likewise, investigators in the Honolulu Heart Study observed a 7% and 9% higher risk for intermediate and poor cognitive function, respectively, for every 10-mm Hg increase in systolic blood pressure, but not diastolic blood pressure among 3735 individuals.31 We similarly found that lifetime systolic, but not diastolic, blood pressure, significantly associated with lower global cognitive function in the current analysis. Additionally, we observed that the lifetime systolic blood pressure burden-cognitive function association was partially mediated by adult LVMI after controlling for traditional CVD risk factors. However, lifetime systolic blood pressure was not significantly associated with cognitive function after accounting for LVMI and other traditional risk factors, suggesting that there may be several unique biological pathways that connect CVD and dementia. Thus, our findings underline a need to extend dementia research to younger populations to continue discovering subclinical CVD risk factors that may additively or synergistically influence the progression of dementia, heart failure, and broader CVD.
LVMI and the E/e’ ratio were the most robust predictors of cognitive function, suggesting that concentric hypertrophy and diastolic dysfunction may contribute to dementia progression and may also be therapeutic upstream targets, independent of blood pressure. Interestingly, these associations appeared to be driven by the working memory cognitive function domain, underlining that subclinical cardiac remodeling and impaired diastolic function may specifically and adversely affect the prefrontal cortex.32 In a similar fashion, others have also found that diastolic dysfunction predicts mild cognitive impairment, particularly in patients with congestive heart failure.33–35 We observed similar associations of LVMI and E/e’ ratio with cognitive function including or excluding individuals with an ejection fraction below 55%, suggesting that overt systolic heart failure may not be necessary for cognitive decline to occur. However, it is important to note that cerebral hypoperfusion may be driving the association between left ventricular function and cognition among individuals with overt heart failure, whereas microvascular dysfunction, insulin resistance, and inflammation could be modifying the relationship between preclinical diastolic dysfunction and poorer cognitive function and working memory among individuals without heart failure. We did not identify associations of left ventricular remodeling or function with episodic memory, executive function, or language, and further research is required to determine if specific regions of the brain are adversely affected by physiological sequelae of subclinical CVD.
Mechanistic connections between elevated LVMI and cognitive decline may be due to cardiac pump abnormalities resulting in altered pulse wave propagation36 through the cerebral arterial tree and subsequent cerebrovascular dysregulation, characterized by endothelial dysfunction, enlarged perivascular spaces, and enhanced blood vessel permeability in the brain.37 We thus hypothesize that elevated systolic blood pressure burden over the lifespan leads to increased LVMI and diastolic dysfunction, which in turn may disrupt cerebrovascular autoregulation and pulse wave propagation. Pathophysiological changes in cerebrovascular autoregulation and pulse wave propagation may then lead to chronic hypoxemia and the development of white matter lesions,38 ultimately resulting in cognitive decline. Furthermore, it is likely that target organ damage due to chronic hypertension manifests differently in the heart (concentric remodeling and hypertrophy) versus the brain (cerebrovascular dysregulation and enlarged perivascular spaces) and these observations necessitate further biological delineation in subsequent basic science and imaging research. Our results serve as an important foundation for future studies involving the heart-brain relationship and are highly relevant as one-in-nine adults over the age of 45 report problems in memory and daily mental tasks.39
The Bogalusa Heart Study provides a unique opportunity to study the natural course of CVD and dementia in the setting of a diverse, community-based cohort of free-living individuals. We were able to examine the formal mediation effect of adulthood LVMI on the relationship of lifetime systolic blood pressure and adulthood cognitive function in the third, fourth, and fifth decades of life, which is a strength of the study because of a dearth dementia research in younger populations. Unlike most prior studies that strictly assessed the relationship between left ventricular remodeling and cognition in multivariable regression, which is limited in biological pathway assessment, we conducted formal mediation analyses using sex- and race-specific standardized values for lifespan systolic blood pressure, LVMI, dementia, and CVD covariates to more comprehensively characterize the interconnected relationship between long-term systolic blood pressure burden, left ventricular remodeling, and cognitive function. Furthermore, we incorporated echocardiographic measures of both left ventricular structure and function in our study design and analyses, more accurately capturing the biological framework involving the heart and brain. Additionally, Black individuals and women, 2 groups traditionally underrepresented in scientific research, encompassed 33% and 60% of our study sample.
Limitations of our study include potential residual confounding common to all observational research. In particular, biological information related to pulse wave propagation and the structure of blood vessels in the brain will be important to consider in future similar studies that assess the association of left ventricular structure and function with cognitive function. We minimized temporal ambiguity by conducting a mediation analysis, leveraging area under the curve values from childhood to adulthood for systolic blood pressure to help isolate the LVMI-cognitive function relationship. Finally, due to missing echocardiographic and cognitive function data, certain BHS participants were not able to be included in the current analysis, increasing the potential for both random and systematic error. However, demographic and cardiometabolic characteristics between those excluded and the current study sample were highly comparable, suggesting that selection bias was not present.
In conclusion, we have demonstrated that adulthood LVMI partially mediates the association between lifetime systolic blood pressure burden and cognitive function. LVMI specifically mediated nearly one-fifth of this latter association, suggesting that lifestyle modifications and antihypertensive medications that slow or even reverse pathological cardiac remodeling may be important approaches for dementia prevention. Our study findings suggest the potential to intervene at subclinical stages of CVD, including left ventricular hypertrophy, to maintain cognitive health over the life course.
We thank all staff members and study personnel who help conduct, sustain, and continue the BHS (Bogalusa Heart Study). We are especially grateful to the BHS study participants.
Sources of Funding
This research was supported by the National Institute on Aging as well as the NHLBI of the NIH under grant numbers: 2R01AG041200 (Principal Investigator, Dr Bazzano), R01AG062309 (Principal Investigators, Drs Bazzano and Carmichael), R21AG057983, (Principal Investigator, Dr Bazzano), F30HL147486 (Principal Investigator, Dr Razavi), R21AG051914 (Principal Investigator, Dr Kelly), P20GM109036 (Principal Investigator, Dr He).
Ray Dorsey E, Elbaz A, Nichols E, Abd-Allah F, Abdelalim A, Adsuar JC, Ansha MG, Brayne C, Choi JYJ, Collado-Mateo D,. Global, regional, and national burden of Parkinson’s disease, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016.Lancet Neurol. 2018; 17:939–953. doi: 10.1016/S1474-4422(18)30295-3CrossrefMedlineGoogle Scholar
Livingston G, Sommerlad A, Orgeta V, Costafreda SG, Huntley J, Ames D, Ballard C, Banerjee S, Burns A, Cohen-Mansfield J,. Dementia prevention, intervention, and care.Lancet. 2017; 390:2673–2734. doi: 10.1016/S0140-6736(17)31363-6CrossrefMedlineGoogle Scholar
Nichols E, Szoeke CEI, Vollset SE, Abbasi N, Abd-Allah F, Abdela J, Aichour MTE, Akinyemi RO, Alahdab F, Asgedom SW,. Global, regional, and national burden of Alzheimer’s disease and other dementias, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016.Lancet Neurol. 2019; 18:88–106. doi: 10.1016/S1474-4422(18)30403-4CrossrefMedlineGoogle Scholar
Kramarow EA, Tejada-Vera B. Dementia mortality in the United States, 2000-2017.Natl Vital Stat Rep. 2019; 68:1–29.MedlineGoogle Scholar
Norton S, Matthews FE, Barnes DE, Yaffe K, Brayne C. Potential for primary prevention of Alzheimer’s disease: an analysis of population-based data.Lancet Neurol. 2014; 13:788–794. doi: 10.1016/S1474-4422(14)70136-XCrossrefMedlineGoogle Scholar
Goff DC, Buxton DB, Pearson GD, Wei GS, Gosselin TE, Addou EA, Stoney CM, Desvigne-Nickens P, Srinivas PR, Galis ZS,. Implementing the National Heart, Lung, and Blood Institute’s strategic vision in the division of cardiovascular sciences.Circ Res. 2019; 124:491–497. doi: 10.1161/CIRCRESAHA.118.314338LinkGoogle Scholar
Gaugler J, James B, Johnson T, Marin A, Weuve J. 2019 Alzheimer’s Disease facts and figures.Alzheimers Dement. 2019; 15:321–387.CrossrefGoogle Scholar
James BD, Bennett DA, Boyle PA, Leurgans S, Schneider JA. Dementia from Alzheimer disease and mixed pathologies in the oldest old.JAMA. 2012; 307:1798–1800. doi: 10.1001/jama.2012.3556CrossrefMedlineGoogle Scholar
Iturria-Medina Y, Sotero RC, Toussaint PJ, Mateos-Pérez JM, Evans AC, Weiner MW, Aisen P, Petersen R, Jack CR, Jagust W,. Early role of vascular dysregulation on late-onset Alzheimer’s disease based on multifactorial data-driven analysis.Nat Commun. 2016; 7:11934. doi: 10.1038/ncomms11934CrossrefMedlineGoogle Scholar
Elias MF, Wolf PA, D’Agostino RB, Cobb J, White LR. Untreated blood pressure level is inversely related to cognitive functioning: the Framingham Study.Am J Epidemiol. 1993; 138:353–364. doi: 10.1093/oxfordjournals.aje.a116868CrossrefMedlineGoogle Scholar
Skoog I, Lernfelt B, Landahl S, Palmertz B, Andreasson LA, Nilsson L, Persson G, Odén A, Svanborg A. 15-year longitudinal study of blood pressure and dementia.Lancet. 1996; 347:1141–1145. doi: 10.1016/s0140-6736(96)90608-xCrossrefMedlineGoogle Scholar
Williamson JD, Pajewski NM, Auchus AP, Bryan RN, Chelune G, Cheung AK, Cleveland ML, Coker LH, Crowe MG, Cushman WC,. Effect of intensive vs standard blood pressure control on probable dementia: a randomized clinical trial.JAMA. 2019; 321:553–561. doi: 10.1001/jama.2018.21442CrossrefMedlineGoogle Scholar
Tzourio C, Laurent S, Debette S. Is hypertension associated with an accelerated aging of the brain?Hypertension. 2014; 63:894–903. doi: 10.1161/HYPERTENSIONAHA.113.00147LinkGoogle Scholar
Moazzami K, Ostovaneh MR, Ambale Venkatesh B, Habibi M, Yoneyama K, Wu C, Liu K, Pimenta I, Fitzpatrick A, Shea S,. Left ventricular hypertrophy and remodeling and risk of cognitive impairment and dementia: MESA (Multi-Ethnic Study of Atherosclerosis).Hypertension. 2018; 71:429–436. doi: 10.1161/HYPERTENSIONAHA.117.10289LinkGoogle Scholar
Park CM, Williams ED, Chaturvedi N, Tillin T, Stewart RJ, Richards M, Shibata D, Mayet J, Hughes AD. Associations between left ventricular dysfunction and brain structure and function: findings from the SABRE (Southall and Brent Revisited) study.J Am Heart Assoc. 2017; 6:e004898. doi: 10.1161/JAHA.116.004898LinkGoogle Scholar
Roth GA, Johnson CO, Abate KH, Abd-Allah F, Ahmed M, Alam K, Alam T, Alvis-Guzman N, Ansari H, Ärnlöv J,; Global Burden of Cardiovascular Diseases Collaboration. The burden of cardiovascular diseases among US States, 1990-2016.JAMA Cardiol. 2018; 3:375–379. doi: 10.1001/jamacardio.2018.0385CrossrefMedlineGoogle Scholar
Berenson GS; Bogalusa Heart Study Investigators. Bogalusa heart study: a long-term community study of a rural biracial (Black/White) population.Am J Med Sci. 2001; 322:293–300.CrossrefMedlineGoogle Scholar
Bazzano LA, Belame SN, Patel DA, Chen W, Srinivasan S, McIlwain E, Berenson GS. Obesity and left ventricular dilatation in young adulthood: the Bogalusa heart study.Clin Cardiol. 2011; 34:153–159. doi: 10.1002/clc.20896CrossrefMedlineGoogle Scholar
Sahn DJ, DeMaria A, Kisslo J, Weyman A. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements.Circulation. 1978; 58:1072–1083. doi: 10.1161/01.cir.58.6.1072LinkGoogle Scholar
Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, Flachskampf FA, Foster E, Goldstein SA, Kuznetsova T,. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging.J Am Soc Echocardiogr. 2015; 28:1.e14–39.e14. doi: 10.1016/j.echo.2014.10.003CrossrefGoogle Scholar
Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, Sachs I, Reichek N. Echocardiographic assessment of left ventricular hypertrophy: comparison to necropsy findings.Am J Cardiol. 1986; 57:450–458. doi: 10.1016/0002-9149(86)90771-xCrossrefMedlineGoogle Scholar
de Simone G, Kizer JR, Chinali M, Roman MJ, Bella JN, Best LG, Lee ET, Devereux RB; Strong Heart Study Investigators. Normalization for body size and population-attributable risk of left ventricular hypertrophy: the Strong Heart Study.Am J Hypertens. 2005; 18:191–196. doi: 10.1016/j.amjhyper.2004.08.032CrossrefMedlineGoogle Scholar
Nagueh SF, Smiseth OA, Appleton CP, Byrd BF, Dokainish H, Edvardsen T, Flachskampf FA, Gillebert TC, Klein AL, Lancellotti P,; Houston, Texas; Oslo, Norway; Phoenix, Arizona; Nashville, Tennessee; Hamilton, Ontario, Canada; Uppsala, Sweden; Ghent and Liège, Belgium; Cleveland, Ohio; Novara, Italy; Rochester, Minnesota; Bucharest, Romania; and St. Louis, Missouri. Recommendations for the evaluation of left ventricular diastolic function by echocardiography: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging.Eur Heart J Cardiovasc Imaging. 2016; 17:1321–1360. doi: 10.1093/ehjci/jew082CrossrefMedlineGoogle Scholar
- 24. BioLINCC: Bogalusa Heart Study (BHS). National Heart, Lung, and Blood Institute: Bogalusa Heart Study.https://biolincc.nhlbi.nih.gov/studies/bhs/. Accessed December 13, 2019.Google Scholar
VanderWeele TJ. Mediation analysis: a Practitioner’s Guide.Annu Rev Public Health. 2016; 37:17–32. doi: 10.1146/annurev-publhealth-032315-021402CrossrefMedlineGoogle Scholar
Sobel ME. Asymptotic confidence intervals for indirect effects in structural equation models.Sociol Methods Res. 1982; 13:290–312.CrossrefGoogle Scholar
Wright JT, Williamson JD, Whelton PK, Snyder JK, Sink KM, Rocco MV, Reboussin DM, Rahman M, Oparil S, Lewis CE,. A randomized trial of intensive versus standard blood-pressure control.N Engl J Med. 2015; 373:2103–2116. doi: 10.1056/NEJMoa1511939CrossrefMedlineGoogle Scholar
Barnes DE, Yaffe K. The projected effect of risk factor reduction on Alzheimer’s disease prevalence.Lancet Neurol. 2011; 10:819–828. doi: 10.1016/S1474-4422(11)70072-2CrossrefMedlineGoogle Scholar
Norby FL, Chen LY, Soliman EZ, Gottesman RF, Mosley TH, Alonso A. Association of left ventricular hypertrophy with cognitive decline and dementia risk over 20 years: the atherosclerosis risk in communities-neurocognitive study (ARIC-NCS).Am Heart J. 2018; 204:58–67. doi: 10.1016/j.ahj.2018.07.007CrossrefMedlineGoogle Scholar
Gottesman RF, Schneider AL, Albert M, Alonso A, Bandeen-Roche K, Coker L, Coresh J, Knopman D, Power MC, Rawlings A,. Midlife hypertension and 20-year cognitive change: the atherosclerosis risk in communities neurocognitive study.JAMA Neurol. 2014; 71:1218–1227. doi: 10.1001/jamaneurol.2014.1646CrossrefMedlineGoogle Scholar
Launer LJ, Masaki K, Petrovitch H, Foley D, Havlik RJ. The association between midlife blood pressure levels and late-life cognitive function. The Honolulu-Asia Aging Study.JAMA. 1995; 274:1846–1851.CrossrefMedlineGoogle Scholar
Almeida OP, Garrido GJ, Beer C, Lautenschlager NT, Arnolda L, Flicker L. Cognitive and brain changes associated with ischaemic heart disease and heart failure.Eur Heart J. 2012; 33:1769–1776. doi: 10.1093/eurheartj/ehr467CrossrefMedlineGoogle Scholar
Sacre JW, Ball J, Wong C, Chan YK, Stewart S, Kingwell BA, Carrington MJ. Mild cognitive impairment is associated with subclinical diastolic dysfunction in patients with chronic heart disease.Eur Heart J Cardiovasc Imaging. 2018; 19:285–292. doi: 10.1093/ehjci/jex169CrossrefMedlineGoogle Scholar
Eggermont LH, de Boer K, Muller M, Jaschke AC, Kamp O, Scherder EJ. Cardiac disease and cognitive impairment: a systematic review.Heart. 2012; 98:1334–1340. doi: 10.1136/heartjnl-2012-301682CrossrefMedlineGoogle Scholar
Lee TC, Qian M, Liu Y, Graham S, Mann DL, Nakanishi K, Teerlink JR, Lip GYH, Freudenberger RS, Sacco RL,; WARCEF Investigators. Cognitive decline over time in patients with systolic heart failure: insights from WARCEF.JACC Heart Fail. 2019; 7:1042–1053. doi: 10.1016/j.jchf.2019.09.003CrossrefMedlineGoogle Scholar
Rabkin SW, Chan SH. Correlation of pulse wave velocity with left ventricular mass in patients with hypertension once blood pressure has been normalized.Heart Int. 2012; 7:e5. doi: 10.4081/hi.2012.e5CrossrefMedlineGoogle Scholar
Meissner A. Hypertension and the brain: a risk factor for more than heart disease.Cerebrovasc Dis. 2016; 42:255–262. doi: 10.1159/000446082CrossrefMedlineGoogle Scholar
Joutel A, Monet-Leprêtre M, Gosele C, Baron-Menguy C, Hammes A, Schmidt S, Lemaire-Carrette B, Domenga V, Schedl A, Lacombe P,. Cerebrovascular dysfunction and microcirculation rarefaction precede white matter lesions in a mouse genetic model of cerebral ischemic small vessel disease.J Clin Invest. 2010; 120:433–445. doi: 10.1172/JCI39733CrossrefMedlineGoogle Scholar
Taylor CA, Bouldin ED, Mcguire LC. Morbidity and mortality weekly report subjective cognitive decline among adults aged ≥45 years-United States, 2015-2016.MMWR Surveill Summ. 2018; 67:753–757. doi: 10.15585/mmwr.mm6727a1Google Scholar