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

Stephanie L. Harrison, PhD; Anton J. M. de Craen, PhD; Ngaire Kerse, PhD; Ruth Teh, PhD; Antoneta Granic, PhD; Karen Davies, PhD; Keith A. Wesnes, PhD; Wendy P. J. den Elzen, PhD; Jacobijn Gussekloo, PhD; Thomas B. L. Kirkwood, PhD; Louise Robinson, MD; Carol Jagger, PhD; Mario Siervo, PhD; Blossom C. M. Stephan, PhD


J Am Geriatr Soc. 2017;65(2):381-389. 

In This Article

Abstract and Introduction


Objectives: To examine the Framingham Stroke Risk Profile (FSRP); the Cardiovascular Risk Factors, Aging, and Incidence of Dementia (CAIDE) risk score, and oxi-inflammatory load (cumulative risk score of three blood biomarkers—homocysteine, interleukin-6, C-reactive protein) for associations with cognitive decline using three cohort studies of very old adults and to examine whether incorporating these biomarkers with the risk scores can affect the association with cognitive decline.

Design: Three longitudinal, population-based cohort studies.

Setting: Newcastle-upon-Tyne, United Kingdom; Leiden, the Netherlands; and Lakes and Bay of Plenty District Health Board areas, New Zealand.

Participants: Newcastle 85+ Study participants (n = 616), Leiden 85-plus Study participants (n = 444), and Life and Living in Advanced Age, a Cohort Study in New Zealand (LiLACS NZ Study) participants (n = 396).

Measurements: FSRP, CAIDE risk score, oxi-inflammatory load, FSRP incorporating oxi-inflammatory load, and CAIDE risk score incorporating oxi-inflammatory load. Oxi-inflammatory load could be calculated only in the Newcastle 85+ and the Leiden 85-plus studies. Measures of global cognitive function were available for all three data sets. Domain-specific measures were available for the Newcastle 85+ and the Leiden 85-plus studies.

Results: Meta-analysis of pooled results showed greater risk of incident global cognitive impairment with higher FSRP (hazard ratio (HR) = 1.46, 95% confidence interval (CI) = 1.08–1.98), CAIDE (HR = 1.53, 95% CI = 1.09–2.14), and oxi-inflammatory load (HR = 1.73, 95% CI = 1.04–2.88) scores. Adding oxi-inflammatory load to the risk scores increased the risk of cognitive impairment for the FSRP (HR = 1.65, 95% CI = 1.17–2.33) and the CAIDE model (HR = 1.93, 95% CI = 1.39–2.67).

Conclusion: Adding oxi-inflammatory load to cardiovascular risk scores may be useful for determining risk of cognitive impairment in very old adults.


Risk factors for cardiovascular disease (CVD) and stroke, such as hypertension and high cholesterol, have been associated with cognitive decline and dementia.[1] Many cardiovascular risk factors co-occur, and risk prediction models such as the Framingham Stroke Risk Profile (FSRP)[2] have been developed to predict an individual's risk of future CVD or stroke. The Cardiovascular Risk Factors, Aging, and Incidence of Dementia (CAIDE) risk score was specifically developed for the midlife population to predict future risk of dementia based on cardiovascular and lifestyle factors.[3] Blood biomarkers such as C-reactive protein (CRP), homocysteine, and interleukin (IL)-6 are independent risk factors for CVD.[4]

Previous longitudinal studies have suggested a positive association between higher cardiovascular risk and faster rate of cognitive decline, but these studies have been conducted in midlife and younger-old populations;[5–12] this association has not been explored in very old adults. Cardiovascular biomarkers, such as homocysteine, CRP, and IL-6, have also been associated with cognitive decline,[13–15] but previous research has not focused on very old adults.

Identifying individuals at the highest risk of dementia is important for developing targeted intervention strategies for the primary prevention of dementia. Determining risk of cognitive decline and dementia in cognitively healthy individuals is difficult because of the numerous risk factors and individual variability in different populations. For example, certain risk factors may affect age groups differently. Nevertheless, several dementia risk prediction models have been proposed and investigated in population-based longitudinal studies, for example the Australian National University—Alzheimer's Disease Risk Index, the CAIDE model, the Brief Dementia Screening Indicator, and the Late-Life Dementia Risk Index, which was developed in an older population (mean age 76), but a recent systematic review concluded that the current models available were not adequate for discriminating individuals who later developed dementia from those who did not.[16]

Studying very old adults presents an opportunity to identify factors associated with cognitive impairment at the extreme end of aging that are potential targets for intervention to maintain cognitive performance. Predicting what affects cognitive health may also have implications for prolonging healthy life expectancy.[17]

The objectives of this study were to determine whether there was a prospective association between the FSRP,[18] the CAIDE risk model,[3] or oxi-inflammatory load (a sum score of three biomarkers: homocysteine, IL-6, CRP) and cognitive function in very old individuals. Whether combining oxi-inflammatory load with the FSRP and CAIDE risk model could strengthen the association with cognitive function was also examined. Population-based data from three of the largest cohort studies in this age group from different regions of the world were used: Newcastle 85 + Study (United Kingdom), Leiden 85-plus Study (the Netherlands) and Life and Living in Advanced Age: A Cohort Study in New Zealand (LiLACS NZ) Study (non-Māori cohort only).