Introduction: Risk factors for cardiovascular disease (CVD) also increase the risk of dementia. However, whether commonly used CVD risk scores are associated with dementia risk in older adults who do not have a history of CVD, and potential gender differences in this association, remains unclear. The aim of this study was to determine whether CVD risk scores are prospectively associated with cognitive decline and dementia in initially healthy older men and women. Methods: A total of19,114 participants from a prospective cohort of individuals aged 65+ years without known CVD or dementia were recruited. The atherosclerotic cardiovascular disease risk score (ASCVDRS), Systematic Coronary Risk Evaluation 2-Older Persons (SCORE2-OP), and the Framingham risk score (FRS) were calculated at baseline. Risk of dementia (according to DSM-IV criteria) and cognitive decline (defined as a >1.5 standard deviation decline in global cognition, episodic memory, psychomotor speed, or verbal fluency from the previous year) were assessed using hazard ratio. Results: Over a median follow-up of 6.4 years, 850 individuals developed dementia and 4,352 cognitive decline. Men and women in the highest ASCVDRS tertile had a 41% (95% CI 1.08, 1.85) and 45% (1.11, 1.89) increased risk of dementia compared to the lowest tertile, respectively. Likewise, men and women in the highest SCORE2-OP tertile had a 64% (1.24, 2.16) and 60% (1.22, 2.11) increased risk of dementia compared to the lowest tertile, respectively. Findings were similar, but the risk was slightly lesser when examining risk of cognitive decline for both ASCVDRS and SCORE2-OP. However, FRS was only associated with the risk of cognitive decline among women (highest vs. lowest tertiles: 1.13 [1.01–1.26]). Conclusion: These findings suggest the utility of the ASCVDRS and SCORE2-OP in clinical practice, to not only assess future risk of CVD, but also as potential early indicators of cognitive impairment, even in relatively healthy older men and women.

Dementia is the seventh leading cause of death among all diseases and one of the major causes of disability and dependency among older people globally [1]. Currently more than 55 million people live with dementia worldwide, and there are nearly 10 million new cases every year [1]. This increase in dementia prevalence is attributed to population growth and aging. Dementia will continue to be a challenge until there are research breakthroughs in prevention or cures [2]. Identifying at-risk individuals and addressing modifiable risk factors are crucial to prevent dementia and delay the rate of cognitive decline.

A number of modifiable risk factors for dementia are also associated with cardiovascular disease (CVD) risk [3]. For example, CVD risk factors of hypertension, obesity, and diabetes are individually associated with an increased risk of dementia [4‒6]. These risk factors, however, seldom occur in isolation. A systematic review of 18 studies involving >40,000 participants reported a clear relationship between a greater number of risk factors, which included CVD risk factors such as blood pressure, cholesterol, hypertension, and an increased risk of dementia [7]. A meta-analysis of six of these studies indicated that the presence of three modifiable risk factors doubled the risk of dementia compared to having just one risk factor [7].

CVD risk calculators which incorporate multiple risk factors produce a single risk score and are a standard tool to identify an individual’s cardiovascular risk over a specified period. There is some evidence that these CVD risk scores may also be associated with cognitive decline [8‒16], especially in middle-age [8, 9, 12, 15, 16]. Evidence of a relationship between CVD risk scores and cognition in later life (aged >70 years) is mixed [10, 11, 13], and very few studies have looked at the association between CVD risk scores and the incidence of dementia [17‒19]. Furthermore, potential gender differences in the association between the risk scores and dementia have not been investigated, despite well-established gender differences in CVD and dementia risk [20, 21]. The aim of this study was, therefore, to examine the association between commonly used CVD risk scores and the risk of dementia and cognitive decline in older men and women without a prior history of CVD.

Study Population

Participants were from the ASPirin in Reducing Events in the Elderly (ASPREE) study, details of which have been reported previously [22]. Briefly, it was a double-blind, randomized, placebo-controlled clinical trial investigating the effect of low-dose aspirin (100 mg daily) on a composite endpoint of disability-free survival in initially healthy older adults free from dementia, significant cognitive impairment, independence-limiting physical disability, or prior CVD events at enrolment (baseline). Participants were from Australia (n = 16,703) and USA (n = 2,411) aged 70+ years (65+ for African American and Latino individuals in the USA) and recruited between March 2010 and December 2014. Australian participants were recruited through general practices, and in the USA, participants were recruited through primary care practices, mailing lists, and electronic records. Eligibility criteria included having no prior CVD events or diagnosis of dementia, and a global cognition score of >77/100 on the Modified Mini-Mental State Exam (3MS) [23]. The ASPREE study medication ceased in June 2017, and the participants were followed up till December 2018 [24]; however, the ASPREE eXTension (ASPREE-XT) follow-up study was established in 2018 and continues collecting observational data from these participants post-trial. The ASPREE study has multiple Institutional Review Board approvals in the USA and Australia. All participants provided written informed consent before enrolment.

CVD Risk Scores

Three 10-year CVD risk scores were calculated using data at baseline: the atherosclerotic cardiovascular disease risk score (ASCVDRS) [25], Systematic Coronary Risk Evaluation 2-Older Persons (SCORE2-OP) [26], and the modified Framingham risk score (FRS) [27]. The FRS is the first and most popular CVD risk score. It was developed in the 1970s using data from 8,491 participants aged 30–74 years from the Framingham Heart Study [27]. This score was developed using a population that was predominantly white; however, it has been validated for different ethnic groups and it performed across ethnicities [28]. This risk score, however, tends to underestimate the risk of CVD in an older adult (mean age: 73.50 ± 2.85; range 70–79 years) population [29]. The ASCVDRS is an updated risk score developed in 2013 by the American College of Cardiology and the American Heart Association, using data from individuals aged 40–79 years, from four large community-based cohorts (total n = 24,626) that were racially and geographically diverse [25]. This score too was less effective in predicting CVD events in an older adult (75+) population [30, 31].

The SCORE2-OP is the most recently developed CVD risk score, used to estimate risk of CVD in individuals aged over 70 years in four geographical risk regions [26]. This score has also been validated in different external cohorts of older adults [32]. The SCORE2-OP equation incorporates participant age (years), sex, diabetes, smoking, systolic blood pressure, total cholesterol, high-density lipoprotein. The ASCVDRS and FRS include the same measures as SCORE2-OP except that they also include antihypertensive medication use and the ASCVDRS includes ethnicity. While the SCORE2-OP was developed for individuals aged over 70 years, the ASCVDRS was developed for participants aged up to 79 years and FRS for those aged 74 or under. For participants with an age higher than the maximum age for the corresponding risk prediction model, we substituted the maximum age for that model.

Dementia

Full details describing the assessment and adjudication of dementia have been reported previously [33]. Briefly, participants were triggered for a dementia endpoint assessment if they had a 3MS score below 78, a drop of more than 10.15 points from the predicted 5-year score (based on their baseline 3MS score, age, and level of education), a referral to a specialist for memory concerns or other cognitive problems as noted on the participant’s medical records, a clinician diagnosis of dementia, or prescription of cholinesterase inhibitors (in Australia). Upon triggering, additional cognitive and functional assessments consisting of the Alzheimer’s Disease Assessment Scale – cognitive subscale [34], color trails [35], Lurian overlapping figures [36], Confusion Assessment Method (to eliminate delirium) [37], and the Alzheimer Disease Cooperative Study Activities of Daily Living Scale [38] were then performed. This information was then reviewed, together with their detailed medical history, the results of clinical exams and blood measures, where available, by a specialist committee composed of geriatricians and neurologists, who adjudicated dementia according to DSM-IV criteria [39].

Cognitive Decline

Cognitive assessments were administered regularly by trained staff over a maximum of 9 years. The cognitive battery included 3MS, a measure of global cognition, the Hopkins verbal learning test-revised delayed recall task for episodic memory, the single letter (F)-controlled oral word association test for executive function and verbal fluency, and the symbol digit modalities test to measure psychomotor speed [22]. Cognitive decline was defined as a >1.5 standard deviation (SD) decline in cognitive score from their own baseline value on any of the four cognitive tests. This definition did not include participants with evidence of only a transient decline (e.g., those with a >1.5 SD drop at 1 follow-up, but scoring above this threshold at a subsequent follow-up) [33].

Covariates

Potential confounding factors were those known to be related to cognition and dementia that were not included in the calculation of the CVD risk scores. Socio-demographic factors included years of education (<12 years or ≥12 years) and living situation (living alone or with family/others), health-related factors included baseline alcohol consumption status (current vs. former or never) and obesity (BMI ≥30 kg/m2), and clinical factors included depressive symptoms (measured by the Center for Epidemiologic Studies Depression Scale 10 item [CES-D 10] with a score of 8+) and chronic kidney disease (CKD; eGFR <60 mL/min per 1.73 m2 or urinary albumin:creatinine ratio ≥3 mg/mmol with eGFR ≥60 mL/min per 1.73 m2). Age was part of the CVD risk score equation, and this was not additionally included as covariates.

Statistical Analyses

All of the analyses were stratified by gender due to well-established differences in CVD risk scores and dementia rates. Descriptive characteristics were presented as percentages for categorical variables and mean ± SD or median with interquartile range for continuous variables. Further, χ2 tests and ANOVA were also done for both categorical and continuous variables, respectively. Histograms were used to graphically show the distribution of CVD risk scores for men and women. Participants who did not have any in-person cognitive assessments (n = 1,028, 5.4%) were excluded from the cognitive decline analysis. Separate Cox proportional hazards regression models, for men and women, with time-to-event analysis were used to calculate hazard ratios (HRs) and corresponding 95% confidence interval (CI) for incident dementia and cognitive decline, using CVD risk scores as a continuous variable, and then according to the tertile of CVD risk scores. Tertiles were defined separately for men and women for the three CVD risk scores and were categorized as lowest, medium, and highest. Cumulative incidences were used to show the risk of dementia with allowance for the competing risk of death. The proportional hazards assumption was checked using Schoenfeld residuals. The duration of study follow-up was used as the time scale with the date of randomization as the entry date. We constructed a first model which adjusted for education, living status, baseline alcohol consumption, obesity, depressive symptoms, and chronic kidney disease. A second model included baseline cognitive measures in addition to all the variables from the first model, to account for the variability in baseline cognitive scores among the study participants.

Sensitivity analyses were performed by re-running the models to only include participants within the age range for calculating the respective CVD risk scores. This meant including only individuals aged 79 or below for ASCVDRS, and only participants aged 74 or below for FRS. We additionally excluded individuals with dementia from the cognitive decline analysis, to ensure these individuals were not driving the results. Additionally, given the strong association between stroke with cognitive decline and dementia [40], a sensitivity analysis was undertaken for the cohort after excluding those who had incident stroke during follow-up. All the analyses were performed in STATA version 17.0 (StataCorp, College Station, TX, USA).

We calculated CVD risk scores for all except 488 of the 19,114 ASPREE participants (97.7%; Fig. 1). Exclusions (n = 488) were mainly due to the unavailability of baseline high-density lipoprotein (n = 447) or total cholesterol (n = 1). Men had on average higher CVD risk scores (n = 8,125; ASCVD – mean [SD]: 0.31 [0.11]; SCORE2-OP: 0.23 [0.07]; FRS: 0.30 [0.09]) than women (n = 10,541; ASCVD: 0.23 [0.11]; SCORE2-OP: 0.17 [0.08]; FRS: 0.17 [0.07]) (online suppl. Fig. S1; for all online suppl. material, see https://doi.org/10.1159/000535284). Participant baseline characteristics according to tertiles of ASCVDRS are presented in Table 1. The higher CVD risk score groups were more likely to have individuals with fewer years of education, with obesity and with CKD (Table 1). Baseline characteristics of participants according to SCORE2-OP (online suppl. Table ST1) and FRS (online suppl. Table ST2) were similar to ASCVDRS and are included in the online supplementary materials. Over a median follow-up of 6.4 years (interquartile range, 5.3–7.8 years), there were 850 cases of incident dementia (4.9% of men and 4.3% of women). Among the 18,666 participants, a total of 1,028 (5.5%) did not have in-person cognitive assessments after the baseline assessments (during the follow-up period) and thus could not be included in the cognitive decline analysis. Among the remaining 17,638 participants, 4,352 (23.3%) had incident cognitive decline (Fig. 1).

Fig. 1.

Flow diagram of participants included in the analysis.

Notes: ASPREE, ASPirin in Reducing Events in the Elderly; CVD, cardiovascular disease; HDL, high-density lipoprotein cholesterol; ASCVDRS, atherosclerotic cardiovascular disease risk score; FRS, Framingham risk score; n, sample size.

Fig. 1.

Flow diagram of participants included in the analysis.

Notes: ASPREE, ASPirin in Reducing Events in the Elderly; CVD, cardiovascular disease; HDL, high-density lipoprotein cholesterol; ASCVDRS, atherosclerotic cardiovascular disease risk score; FRS, Framingham risk score; n, sample size.

Close modal
Table 1.

Descriptive characteristics of study participants by gender and ASCVDRS tertiles (n = 18,666)

Men (n = 8,125)Women (n = 10,541)
ASCVDRS tertilesASCVDRS tertiles
low, n = 2,709medium, n = 2,708high, n = 2,708p valuelow, n = 3,514medium, n = 3,513high, n = 3,514p value
10-yr ASCVDRS 0.21 (0.03) 0.30 (0.02) 0.43 (0.08) na 0.12 (0.03) 0.21 (0.03) 0.36 (0.08) na 
Age, years 72.2 (2.6) 74.8 (3.9) 77.9 (4.7) <0.001 71.7 (2.0) 74.7 (3.5) 79.2 (4.3) <0.001 
3MS 93.5 (4.5) 92.9 (4.6) 91.8 (4.9) <0.001 94.7 (4.1) 94.1 (4.5) 93.1 (4.7) <0.001 
HVLT-R 7.6 (2.8) 7.2 (2.8) 6.7 (2.8) 0.93 8.6 (2.6) 8.3 (2.7) 7.6 (2.8) <0.001 
COWAT 11.8 (4.5) 11.6 (4.5) 11.1 (4.3) 0.44 12.9 (4.6) 12.5 (4.4) 12.2 (4.5) 0.10 
SDMT 37.5 (9.7) 35.6 (10.0) 33.0 (9.6) 0.08 40.9 (9.6) 38.1 (9.9) 34.6 (10.1) 0.02 
Education <12 yr 1,045 (38.6%) 1,224 (45.2%) 1,295 (47.8%) <0.001 1,415 (40.3%) 1,628 (46.3%) 1,813 (51.6%) <0.001 
Living alone 472 (17.4%) 555 (20.5%) 668 (24.7%) <0.001 1,224 (34.8%) 1,470 (41.8%) 1,720 (49.0%) <0.001 
Alcohol 2,280 (84.2%) 2,283 (84.3%) 2,193 (81.0%) <0.001 2,632 (74.6%) 2,540 (72.3%) 2,371 (67.5%) <0.001 
Obese 621 (22.9%) 719 (26.6%) 802 (29.6%) <0.001 1,058 (30.1%) 1,220 (34.7%) 1,137 (32.4%) <0.001 
Depression* 185 (6.8%) 216 (8.0%) 219 (8.1%) 0.16 378 (10.8%) 429 (12.2%) 417 (11.9%) 0.14 
CKD 401 (15.8%) 590 (23.5%) 970 (37.8%) <0.001 602 (18.3%) 806 (24.5%) 1,278 (38.7%) <0.001 
Treatment arma 1,312 (48.4%) 1,348 (49.8%) 1,396 (51.6%) 0.07 1,733 (49.3%) 1,786 (50.8%) 1,732 (49.3%) 0.33 
Men (n = 8,125)Women (n = 10,541)
ASCVDRS tertilesASCVDRS tertiles
low, n = 2,709medium, n = 2,708high, n = 2,708p valuelow, n = 3,514medium, n = 3,513high, n = 3,514p value
10-yr ASCVDRS 0.21 (0.03) 0.30 (0.02) 0.43 (0.08) na 0.12 (0.03) 0.21 (0.03) 0.36 (0.08) na 
Age, years 72.2 (2.6) 74.8 (3.9) 77.9 (4.7) <0.001 71.7 (2.0) 74.7 (3.5) 79.2 (4.3) <0.001 
3MS 93.5 (4.5) 92.9 (4.6) 91.8 (4.9) <0.001 94.7 (4.1) 94.1 (4.5) 93.1 (4.7) <0.001 
HVLT-R 7.6 (2.8) 7.2 (2.8) 6.7 (2.8) 0.93 8.6 (2.6) 8.3 (2.7) 7.6 (2.8) <0.001 
COWAT 11.8 (4.5) 11.6 (4.5) 11.1 (4.3) 0.44 12.9 (4.6) 12.5 (4.4) 12.2 (4.5) 0.10 
SDMT 37.5 (9.7) 35.6 (10.0) 33.0 (9.6) 0.08 40.9 (9.6) 38.1 (9.9) 34.6 (10.1) 0.02 
Education <12 yr 1,045 (38.6%) 1,224 (45.2%) 1,295 (47.8%) <0.001 1,415 (40.3%) 1,628 (46.3%) 1,813 (51.6%) <0.001 
Living alone 472 (17.4%) 555 (20.5%) 668 (24.7%) <0.001 1,224 (34.8%) 1,470 (41.8%) 1,720 (49.0%) <0.001 
Alcohol 2,280 (84.2%) 2,283 (84.3%) 2,193 (81.0%) <0.001 2,632 (74.6%) 2,540 (72.3%) 2,371 (67.5%) <0.001 
Obese 621 (22.9%) 719 (26.6%) 802 (29.6%) <0.001 1,058 (30.1%) 1,220 (34.7%) 1,137 (32.4%) <0.001 
Depression* 185 (6.8%) 216 (8.0%) 219 (8.1%) 0.16 378 (10.8%) 429 (12.2%) 417 (11.9%) 0.14 
CKD 401 (15.8%) 590 (23.5%) 970 (37.8%) <0.001 602 (18.3%) 806 (24.5%) 1,278 (38.7%) <0.001 
Treatment arma 1,312 (48.4%) 1,348 (49.8%) 1,396 (51.6%) 0.07 1,733 (49.3%) 1,786 (50.8%) 1,732 (49.3%) 0.33 

n, sample size; na, not applicable; p, significance value; ASPREE, Aspirin in Reducing Events in the Elderly; ASCVDRS, atherosclerotic cardiovascular risk score; yr, years; SD, standard deviation; 3MS, Modified Mini-Mental State Exam; HVLT-R, Hopkins verbal learning test – revised delayed recall; COWAT, controlled oral word association test; SDMT, symbol digit modalities test; alcohol, alcohol consumption [current]; CKD, chronic kidney disease.

*Depression defined as CESD-10 score of 8+/30.

aTreatment arm is randomized aspirin versus placebo.

Cardiovascular Risk Scores and Risk of Dementia

The cumulative incidence of dementia over the follow-up period by CVD risk score tertiles is shown in Figure 2. For men, those in the highest ASCVDRS tertile had an increased risk of dementia compared to the lowest (fully adjusted HR: 1.41 [95% CI: 1.08, 1.85]; online suppl. Table S3). When ASCVDRS was treated as a continuous variable, a 10% or 0.1 (approximately 1 SD) increase in ASCVDRS was associated with an increased risk of dementia for men (fully adjusted – HR: 1.08 [95% CI: 1.00, 1.21]; online suppl. Table S3). Women in the highest ASCVDRS tertile also had an increased risk of dementia (fully adjusted – HR: 1.45 [95% CI: 1.11, 1.89]; online suppl. Table S3), and when ASCVDRS was treated as a continuous variable, a 10% or 0.1 (approximately 1 SD) increase was associated with a 12% increase in the risk of dementia (fully adjusted – HR: 1.13 [95% CI: 1.04, 1.23]; online suppl. Table S3).

Fig. 2.

Cumulative incidence of dementia according to CVD risk score tertile in men and women.

Notes: Cumulative incidence rates after taking competing risk of death into account. HR, hazard ratio; CI, confidence interval; ASCVDRS, atherosclerotic cardiovascular disease risk score; SCORE2-OP, SCORE2 – Older Persons; FRS, Framingham risk score.

Fig. 2.

Cumulative incidence of dementia according to CVD risk score tertile in men and women.

Notes: Cumulative incidence rates after taking competing risk of death into account. HR, hazard ratio; CI, confidence interval; ASCVDRS, atherosclerotic cardiovascular disease risk score; SCORE2-OP, SCORE2 – Older Persons; FRS, Framingham risk score.

Close modal

The results for SCORE2-OP were similar to those for ASCVDRS. Men in the highest SCORE2-OP tertile had an increased risk of dementia (fully adjusted – HR: 1.64 [95% CI: 1.24, 2.16]; online suppl. Table S3), and when SCORE2-OP was treated as a continuous variable, a 10% or 0.1 (approximately 1 SD) increase was associated with a 31% increase in the risk of dementia (fully adjusted – HR: 1.31 [95% CI: 1.15, 1.49]; online suppl. Table S3). For women, those in the highest SCORE2-OP tertile had an increased risk of dementia compared to the lowest (fully adjusted – HR: 1.60 [95% CI: 1.22, 2.11]; online suppl. Table S3). When SCORE2-OP was treated as a continuous variable, a 10% or 0.1 (approximately 1 SD) increase in SCORE2-OP was associated with an increased risk of dementia for women (fully adjusted – HR: 1.22 [95% CI: 1.10, 1.35]; online suppl. Table S3). There was no association between FRS (categorical or continuous variable) and incident dementia for men or women (online suppl. Table ST3).

Cardiovascular Risk Scores and Risk of Cognitive Decline

The cumulative incidence of cognitive decline over the follow-up period by CVD risk scores is shown in Figure 3. For men, the highest ASCVDRS tertile had an increased risk of cognitive decline compared to the lowest tertile (fully adjusted – HR: 1.28 [95% CI: 1.13, 1.43]; online suppl. Table S4), and when ASCVDRS was treated as a continuous variable, there was a 12% increased risk of cognitive decline for every 10% or 0.1 (approximately 1 SD) increase in ASCVDRS (fully adjusted – HR: 1.12 [95% CI: 1.07, 1.17]; online suppl. Table S4). Women in both the medium (fully adjusted – HR: 1.16 [95% CI: 1.04–1.29]; online suppl. Table S4) and the highest ASCVDRS tertile (fully adjusted – HR: 1.47 [95% CI: 1.32–1.64]; online suppl. Table S4) had an increased risk of cognitive decline compared to those in the lowest tertile. When ASCVDRS was treated as a continuous variable, a 10% or 0.1 (approximately 1 SD) increase in ASCVDRS was associated with a 17% increased risk of cognitive decline for women (fully adjusted – HR: 1.18 [95% CI: 1.13–1.22]; online suppl. Table S4).

Fig. 3.

Cumulative incidence of cognitive decline according to CVD risk score tertile in men and women.

Notes: Cumulative incidence rates after taking competing risk of death into account. HR, hazard ratio; CI, confidence interval; ASCVDRS, atherosclerotic cardiovascular disease risk score; SCORE2-OP, SCORE2 – Older Persons; FRS, Framingham risk score.

Fig. 3.

Cumulative incidence of cognitive decline according to CVD risk score tertile in men and women.

Notes: Cumulative incidence rates after taking competing risk of death into account. HR, hazard ratio; CI, confidence interval; ASCVDRS, atherosclerotic cardiovascular disease risk score; SCORE2-OP, SCORE2 – Older Persons; FRS, Framingham risk score.

Close modal

For men, the highest SCORE2-OP tertile had an increased risk of cognitive decline compared to the lowest tertile (fully adjusted – HR: 1.43 [95% CI: 1.27, 1.61]; online suppl. Table S4), and when it was treated as a continuous variable, there was a 32% increased risk of cognitive decline for every 10% or 0.1 (approximately 1 SD) increase in SCORE2-OP (fully adjusted – HR: 1.32 [95% CI: 1.23, 1.41]; online suppl. Table S4). Similarly, women in the highest tertile had an increased risk (fully adjusted – HR: 1.51 [95% CI: 1.40–1.68]; online suppl. Table S4) of cognitive decline compared to those in the lowest tertile. When SCORE2-OP was treated as a continuous variable, a 10% or 0.1 (approximately 1 SD) increase in ASCVDRS was associated with a 31% increased risk of cognitive decline for women (fully adjusted – HR: 1.31 [95% CI: 1.25–1.38]; online suppl. Table S4).

FRS (categorical and continuous measures) was not associated with cognitive decline for men. For women, the highest FRS tertile had approximately a 12% increased risk of cognitive decline (fully adjusted – HR: 1.12 [95% CI: 1.01–1.25]; online suppl. Table S4), and when FRS was treated as a continuous variable, a 10% or 0.1 (approximately 1 SD) increase in FRS was associated with an increased risk of cognitive decline for women (fully adjusted – HR: 1.08 [1.02–1.15]; online suppl. Table S4). A summary of all of the results when the CVD risk scores are treated as tertiles is provided in Figure 4.

Fig. 4.

Summary results showing the risk of dementia and cognitive decline with medium and high CVD risk.

Notes: Results are shown as HR [95% CI]. HR, hazard ratio; CI, confidence interval; ASCVDRS, atherosclerotic cardiovascular disease risk score; SCORE2-OP, SCORE2 – Older Persons; FRS, Framingham risk score.

Fig. 4.

Summary results showing the risk of dementia and cognitive decline with medium and high CVD risk.

Notes: Results are shown as HR [95% CI]. HR, hazard ratio; CI, confidence interval; ASCVDRS, atherosclerotic cardiovascular disease risk score; SCORE2-OP, SCORE2 – Older Persons; FRS, Framingham risk score.

Close modal

Sensitivity Analyses

In the sensitivity analysis excluding individuals above the age for each of the CVD risk scores, the results were largely similar, but with wider CIs due to fewer events and smaller sample size (online suppl. Table S5–S8). In the sensitivity analysis for cognitive decline after excluding 820 dementia cases, the association of all three CVD risk scores with cognitive decline remained similar to the main findings (online suppl. Table S9). Sensitivity analysis excluding those with incident stroke (online suppl. Tables S10, S11) demonstrated similar results comparable to our main findings.

In this longitudinal cohort study of participants who were predominantly aged 70 years and above, without a prior CVD event or major cognitive impairments, ASCVDRS and SCORE2-OP were associated with an increased risk of dementia and cognitive decline among both men and women. The FRS, however, was not associated with the risk of dementia, but was associated with an increased risk of cognitive decline for women only. Our study demonstrated clear differences in the associations depending on the risk score examined and, in particular, a lack of evidence that FRS was associated with risk of dementia. The factors used to calculate these CVD risk scores are mostly similar although differing weights were given to the factors included in the risk equations. ASCVDRS and FRS, but not SCORE2-OP, include antihypertensive medication use, while only ASCVDRS includes ethnicity. Additionally, the three CVD risk scores were developed for different cohorts. A prior study that compared ASCVDRS and FRS found the performance of ASCVDRS was slightly better than FRS for predicting CVD in Australian men and women aged between 40 and 74 (mean [SD]: 53.9 [9.3]) years) [41] and therefore supports the findings of our study. SCORE2-OP is a recently developed risk score specifically for people aged >70 years and thus the population included in our study. Indeed, the lack of association with the FRS in our study may be, in part, due to the fact that it was developed using a younger cohort (up to 74 years only). Since age is an integral part of the risk score calculation, applying a maximum threshold might underestimate the risk. However, the results from sensitivity analyses for FRS after excluding 41.5% of the participants older than 74 years were similar to the main results for both risk of dementia and cognitive decline.

Ours is the first study that has investigated the association between these CVD scores, specifically ASCVDRS and SCORE2-OP, and the risk of dementia in an older adult population. Previous studies have investigated the association between CVD risk scores and cognition in the older adult population (>65 years) [10, 11, 13, 42‒44]. The FRS is the most popularly investigated risk score, and while studies [10, 11] have reported an association between the FRS and decline in cognitive function, they did not explore the association separately for men and women. Our findings indicate the association with FRS, maybe specific to women, and the two studies reporting significant overall results included a higher proportion of women (75.8% [10] and 60.1% [11]) which could be driving the overall results. Another study [43] reported higher CVD risk burden (measured using ASCVDRS) increased the risk of cognition impairment and accelerated its progression over time which is similar to the findings of our study. Additionally, this is the first time that the association between SCORE2-OP and cognitive decline has been investigated.

This study has two main clinical implications. The first one is the potential clinical utility of CVD risk scores as an easy cost-effective way to identify older adults at risk for dementia or accelerated cognitive decline. Both the ASCVDRS and SCORE2-OP use variables that can be collected easily in clinical and research settings and present a convenient way to identify those individuals who are at a higher risk of incident dementia and cognitive decline. The second implication is that, once the at-risk population has been identified, targeted interventions, such as maintaining a healthy lifestyle (healthy diet, regular physical exercise, quit smoking, etc.), that are associated with slower rate of decline in older adults can be suggested [45]. A strength of this study is the prospective follow-up of a large cohort of individuals without any CVD events or major cognitive impairment at baseline. The study followed rigorous protocols with all measurements, baseline, and follow-up, collected in a systematic and standardized manner by trained and accredited staff. Due to the regular cognitive assessment follow-up, cognitive decline was captured almost at the time of the event. Additionally, the adjudication of dementia was conducted in a robust manner by expert committees of clinicians. Lastly, this study collected data on a wide range of risk factors, allowing us to adjust for multiple covariates that were not included in the CVD risk score calculation.

One important limitation of this study is that the ASCVDRS and FRS were developed using a younger cohort, and maximum age limits needed to be applied for both these risk score calculations. However, when we performed sensitivity analyses with participants aged <79 years (for ASCVD) and <74 years (for FRS), the results were consistent with the main analysis. Furthermore, the SCORE2-OP was developed specifically for individuals aged 70 years and over. Another limitation is that the study findings are largely restricted to a Caucasian white population; hence, we were unable to analyze the results by ethnicity and/or race. These results need to be replicated in more ethnically diverse populations. Given that all participants were relatively healthy at baseline, there might be a healthy volunteer bias for participant recruitment to this clinical trial, as evidenced by a lower incidence of dementia in this cohort compared to other population studies. However, with more people reaching older age in generally good health, our findings might become increasingly generalizable to older adults. Though our results are associative and before recommendations can be made for use within a dementia/cognitive decline risk prediction framework, further validation and comparative work are needed especially among ethnically diverse subgroups.

In conclusion, in this large prospective cohort of relatively healthy older adults, higher ASCVDRS and SCORE2-OP were associated with an increased risk of dementia and cognitive decline in both men and women. However, FRS was only associated with an increased risk of cognitive decline among women. These results provide some first evidence that CVD risk scores (ASCVDRS and SCORE2-OP) can be used in the context of assessing dementia and cognitive decline risk in older adults. Further research is needed to validate these results in different, ethnically diverse, populations and to explore which CVD risk scores will best identify dementia risk in relatively healthy older adults.

We are most grateful to the ASPREE participants who volunteered for this clinical trial, the general practitioners, and the medical clinics that supported the participants in the study. We also thank the dedicated and skilled ASPREE staff in both Australia and the USA.

The ASPREE study complies with the Declaration of Helsinki and was approved by multiple Institutional Review Boards (www.aspree.org). All participants signed written informed consent on participation. Ethics approval was obtained from Monash University HREC (2006/745MC and 12771), Alfred Hospital HREC (593/17), Royal Australian College of General Practitioners (NREEC 02/022b), HREC (Tasmania) Network (H0008933 and H0017149), ACT Health HREC (ETH.11.07.997 and ETH.3.18.037E), University of Adelaide HREC (H-250-2011 and 32802), and the University of Iowa IRB (201904807).

R.C.S. reports being the site principal investigator or sub-investigator for Alzheimer’s disease clinical trials for which his institution (Rush University Medical Center) is compensated (Amylyx Pharmaceuticals, Inc.; Athira Pharma, Inc.; Edgewater NEXT, Eli Lilly & Co., Inc.; and Genentech, Inc.). T.C. has received honoraria for lectures from Roche. All other authors report no disclosures relevant to the manuscript.

This was mainly supported by grants from the National Institute on Ageing and the National Cancer Institute at the US National Institutes of Health (Grant No. U01AG029824 and U19AG062682); the National Health and Medical Research Council of Australia (Grant No. 334047 and 1127060); Monash University (Australia); and the Victorian Cancer Agency (Australia). Other funding resources and collaborating organizations of the ASPREE study are listed at https://aspree.org/. S.V. is supported by the Monash Faculty (School of Public Health and Preventive Medicine) International Tuition Scholarship and the stipend offered as part of a PhD exchange program by Monash University and King’s College London. J.R. is supported by a National Health and Medical Research Council Research Leader Fellowship (1135727).

S.V., I.H., and J.R. were involved in the study conceptualization; S.V. led the analysis, interpreted the results, and drafted the manuscript. E.C. provided the CVD risk score estimates for the analysis. I.H., E.C., R.W., R.F.-P., C.R. A.T., A.M., R.C.S., T.T.-J.C., R.L.W., J.Mc.N., S.O., M.N., C.S., and J.R. reviewed, edited, and contributed to the writing. All authors made significant contributions to the paper and have read and approved the final version of the manuscript.

All the data analyzed during the current study are publicly available but due to ethical reasons will be shared with approved researchers through a secure web-based data portal Safe Haven (www.aspree.org). Further inquiries can be directed to the corresponding author.

1.
World Health Organisation
Dementia
. Available from: https://www.who.int/news-room/fact-sheets/detail/dementia.
2.
GBD 2016 Dementia Collaborators
Szoeke
CE
,
Vollset
SE
,
Abbasi
N
,
Abd-Allah
F
,
Abdela
J
.
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
(
1
):
88
106
.
3.
Leritz
EC
,
McGlinchey
RE
,
Kellison
I
,
Rudolph
JL
,
Milberg
WP
.
Cardiovascular disease risk factors and cognition in the elderly
.
Curr Cardiovasc Risk Rep
.
2011
;
5
:
407
12
.
4.
Ou
Y-N
,
Tan
C-C
,
Shen
X-N
,
Xu
W
,
Hou
X-H
,
Dong
Q
.
Blood pressure and risks of cognitive impairment and dementia: a systematic review and meta-analysis of 209 prospective studies
.
Hypertension
.
2020
;
76
(
1
):
217
25
.
5.
Albanese
E
,
Launer
LJ
,
Egger
M
,
Prince
MJ
,
Giannakopoulos
P
,
Wolters
FJ
.
Body mass index in midlife and dementia: systematic review and meta-regression analysis of 589,649 men and women followed in longitudinal studies
.
Alzheimers Dement
.
2017
;
8
:
165
78
.
6.
Chatterjee
S
,
Peters
SA
,
Woodward
M
,
Mejia Arango
S
,
Batty
GD
,
Beckett
N
.
Type 2 diabetes as a risk factor for dementia in women compared with men: a pooled analysis of 2.3 million people comprising more than 100,000 cases of dementia
.
Diabetes Care
.
2016
;
39
(
2
):
300
7
.
7.
Peters
R
,
Booth
A
,
Rockwood
K
,
Peters
J
,
D’Este
C
,
Anstey
KJ
.
Combining modifiable risk factors and risk of dementia: a systematic review and meta-analysis
.
BMJ Open
.
2019
;
9
(
1
):
e022846
.
8.
Sedaghat
S
,
Lutsey
PL
,
Ji
Y
,
Empana
JP
,
Sorond
F
,
Hughes
TM
.
Association of change in cardiovascular risk factors with incident dementia
.
Alzheimers Dement
.
2022
;
19
(
5
):
1821
31
.
9.
Torres
S
,
Alexander
A
,
O’Bryant
S
,
Medina
LD
.
Cognition and the predictive utility of three risk scores in an ethnically diverse sample
.
J Alzheimers Dis
.
2020
;
75
(
3
):
1049
59
.
10.
Song
R
,
Xu
H
,
Dintica
CS
,
Pan
K-Y
,
Qi
X
,
Buchman
AS
.
Associations between cardiovascular risk, structural brain changes, and cognitive decline
.
J Am Coll Cardiol
.
2020
;
75
(
20
):
2525
34
.
11.
Harrison
SL
,
de Craen
AJ
,
Kerse
N
,
Teh
R
,
Granic
A
,
Davies
K
.
Predicting risk of cognitive decline in very old adults using three models: the Framingham stroke risk profile; the cardiovascular risk factors, aging, and dementia model; and oxi-inflammatory biomarkers
.
J Am Geriatr Soc
.
2017
;
65
(
2
):
381
9
.
12.
Llewellyn
DJ
,
Lang
IA
,
Xie
J
,
Huppert
FA
,
Melzer
D
,
Langa
KM
.
Framingham Stroke Risk Profile and poor cognitive function: a population-based study
.
BMC Neurol
.
2008
;
8
(
1
):
12
8
.
13.
Wang
R
,
Fratiglioni
L
,
Kalpouzos
G
,
Lövdén
M
,
Laukka
EJ
,
Bronge
L
.
Mixed brain lesions mediate the association between cardiovascular risk burden and cognitive decline in old age: a population-based study
.
Alzheimers Dement
.
2017
;
13
(
3
):
247
56
.
14.
DeRight
J
,
Jorgensen
RS
,
Cabral
MJ
.
Composite cardiovascular risk scores and neuropsychological functioning: a meta-analytic review
.
Ann Behav Med
.
2015
;
49
(
3
):
344
57
.
15.
Kaffashian
S
,
Dugravot
A
,
Elbaz
A
,
Shipley
MJ
,
Sabia
S
,
Kivimäki
M
.
Predicting cognitive decline: a dementia risk score vs the Framingham vascular risk scores
.
Neurology
.
2013
;
80
(
14
):
1300
6
.
16.
Zheng
F
,
Xie
W
,
Li
C
,
Gao
D
,
Liang
J
.
Prediction abilities of SCORE2 risk algorithms for incident dementia and all-cause mortality: results from the UK biobank cohort study
.
J Gerontol
.
2022
;
78
(
4
):
704
10
.
17.
von Cederwald
BF
,
Josefsson
M
,
Wåhlin
A
,
Nyberg
L
,
Karalija
N
.
Association of cardiovascular risk trajectory with cognitive decline and incident dementia
.
Neurology
.
2022
;
98
(
20
):
e2013
22
.
18.
Fayosse
A
,
Nguyen
D-P
,
Dugravot
A
,
Dumurgier
J
,
Tabak
AG
,
Kivimäki
M
.
Risk prediction models for dementia: role of age and cardiometabolic risk factors
.
BMC Med
.
2020
;
18
:
107
10
.
19.
Song
R
,
Pan
KY
,
Xu
H
,
Qi
X
,
Buchman
AS
,
Bennett
DA
.
Association of cardiovascular risk burden with risk of dementia and brain pathologies: a population-based cohort study
.
Alzheimers Dement
.
2021
;
17
(
12
):
1914
22
.
20.
Anstey
KJ
,
Peters
R
,
Mortby
ME
,
Kiely
KM
,
Eramudugolla
R
,
Cherbuin
N
.
Association of sex differences in dementia risk factors with sex differences in memory decline in a population-based cohort spanning 20–76 years
.
Sci Rep
.
2021
;
11
(
1
):
7710
0
.
21.
Appelman
Y
,
van Rijn
BB
,
Ten Haaf
ME
,
Boersma
E
,
Peters
SA
.
Sex differences in cardiovascular risk factors and disease prevention
.
Atherosclerosis
.
2015
;
241
(
1
):
211
8
.
22.
ASPREE Investigator Group
.
Study design of ASPirin in Reducing Events in the Elderly (ASPREE): a randomized, controlled trial
.
Contemp Clin Trials
.
2013
;
36
(
2
):
555
64
.
23.
McNeil
JJ
,
Woods
RL
,
Nelson
MR
,
Murray
AM
,
Reid
CM
,
Kirpach
B
.
Baseline characteristics of participants in the ASPREE (ASPirin in reducing events in the elderly) study
.
J Gerontol A Biol Sci Med Sci
.
2017
;
72
(
11
):
1586
93
.
24.
McNeil
JJ
,
Wolfe
R
,
Woods
RL
,
Tonkin
AM
,
Donnan
GA
,
Nelson
MR
.
Effect of aspirin on cardiovascular events and bleeding in the healthy elderly
.
N Engl J Med Overseas Ed
.
2018
;
379
(
16
):
1509
18
.
25.
Goff
DC
,
Lloyd-Jones
DM
,
Bennett
G
,
Coady
S
,
D’Agostino
RB
,
Gibbons
R
.
2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American heart association task force on practice guidelines
.
Circulation
.
2014
129
25 Suppl 2
S49
73
.
26.
SCORE2-OP working group and ESC Cardiovascular risk collaboration
.
SCORE2-OP risk prediction algorithms: estimating incident cardiovascular event risk in older persons in four geographical risk regions
.
Eur Heart J
.
2021
;
42
(
25
):
2455
67
.
27.
D’Agostino
RB
,
Vasan
RS
,
Pencina
MJ
,
Wolf
PA
,
Cobain
M
,
Massaro
JM
.
General cardiovascular risk profile for use in primary care: the Framingham Heart Study
.
Circulation
.
2008
;
117
(
6
):
743
53
.
28.
D’Agostino
RB
,
Grundy
S
,
Sullivan
LM
,
Wilson
P
CHD Risk Prediction Group
.
Validation of the Framingham coronary heart disease prediction scores: results of a multiple ethnic groups investigation
.
JAMA
.
2001
;
286
(
2
):
180
7
.
29.
Rodondi
N
,
Locatelli
I
,
Aujesky
D
,
Butler
J
,
Vittinghoff
E
,
Simonsick
E
.
Framingham risk score and alternatives for prediction of coronary heart disease in older adults
.
PLoS One
.
2012
;
7
(
3
):
e34287
.
30.
Dalton
JE
,
Rothberg
MB
,
Dawson
NV
,
Krieger
NI
,
Zidar
DA
,
Perzynski
AT
.
Failure of traditional risk factors to adequately predict cardiovascular events in older populations
.
J Am Geriatr Soc
.
2020
;
68
(
4
):
754
61
.
31.
Nanna
MG
,
Peterson
ED
,
Wojdyla
D
,
Navar
AM
.
The accuracy of cardiovascular pooled cohort risk estimates in US older adults
.
J Gen Intern Med
.
2020
;
35
(
6
):
1701
8
.
32.
Van Trier
T
,
Snaterse
M
,
Boekholdt
S
,
Scholte Op Reimer
W
,
Visseren
F
,
Hageman
S
.
External validation of systematic COronary risk evaluation 2-older persons (SCORE2-OP) in the EPIC-norfolk prospective population study
.
Eur J Prev Cardiol
.
2023
30
Supplement_1
324
.
33.
Ryan
J
,
Storey
E
,
Murray
AM
,
Woods
RL
,
Wolfe
R
,
Reid
CM
.
Randomized placebo-controlled trial of the effects of aspirin on dementia and cognitive decline
.
Neurology
.
2020
;
95
(
3
):
e320
31
.
34.
Rosen
WG
,
Mohs
RC
,
Davis
KL
.
A new rating scale for Alzheimer’s disease
.
Am J Psychiatry
.
1984
;
141
(
11
):
1356
64
.
35.
Pontón
MO
,
Satz
P
,
Herrera
L
,
Ortiz
F
,
Urrutia
CP
,
Young
R
.
Normative data stratified by age and education for the neuropsychological screening battery for hispanics (NeSBHIS): initial report
.
J Int Neuropsychol Soc
.
1996
;
2
(
2
):
96
104
.
36.
Alegret
M
,
Boada-Rovira
M
,
Vinyes-Junqué
G
,
Valero
S
,
Espinosa
A
,
Hernández
I
.
Detection of visuoperceptual deficits in preclinical and mild Alzheimer’s disease
.
J Clin Exp Neuropsychol
.
2009
;
31
(
7
):
860
7
.
37.
Wei
LA
,
Fearing
MA
,
Sternberg
EJ
,
Inouye
SK
.
The Confusion Assessment Method: a systematic review of current usage
.
J Am Geriatr Soc
.
2008
;
56
(
5
):
823
30
.
38.
Fish
J
.
Alzheimer’s disease cooperative study ADL scale
. In:
Kreutzer
JS
,
DeLuca
J
,
Caplan
B
, editors.
Encyclopedia of clinical neuropsychology
New York, NY
Springer New York
2011
. p.
111
2
.
39.
APA Meeting
Continuing medical education syllabus and scientific proceedings in summary form: annual meeting of the American psychiatric association
American Psychiatric Association
1994
.
40.
Kuźma
E
,
Lourida
I
,
Moore
SF
,
Levine
DA
,
Ukoumunne
OC
,
Llewellyn
DJ
.
Stroke and dementia risk: a systematic review and meta-analysis
.
Alzheimers Dement
.
2018
;
14
(
11
):
1416
26
.
41.
Albarqouni
L
,
Doust
JA
,
Magliano
D
,
Barr
EL
,
Shaw
JE
,
Glasziou
PP
.
External validation and comparison of four cardiovascular risk prediction models with data from the Australian Diabetes, Obesity and Lifestyle study
.
Med J Aust
.
2019
;
210
(
4
):
161
7
.
42.
Hazzouri
AZA
,
Haan
MN
,
Neuhaus
JM
,
Pletcher
M
,
Peralta
CA
,
López
L
.
Cardiovascular risk score, cognitive decline, and dementia in older M exican A mericans: the role of sex and education
.
J Am Heart Assoc
.
2013
;
2
(
2
):
e004978
.
43.
Wu
X
,
Wang
H
,
Chen
C
,
Xiong
Y
,
Zhu
L
,
Jia
J
.
The association between cardiovascular risk burden and cognitive function amongst the old: a 9-year longitudinal cohort study
.
Eur J Neurol
.
2021
;
28
(
9
):
2907
12
.
44.
Makino
K
,
Lee
S
,
Bae
S
,
Chiba
I
,
Harada
K
,
Katayama
O
.
Absolute cardiovascular disease risk is associated with the incidence of non-amnestic cognitive impairment in Japanese older adults
.
Front Aging Neurosci
.
2021
;
13
:
685683
.
45.
Jia
J
,
Zhao
T
,
Liu
Z
,
Liang
Y
,
Li
F
,
Li
Y
.
Association between healthy lifestyle and memory decline in older adults: 10 year, population based, prospective cohort study
.
BMJ
.
2023
;
380
:
e072691
.