Framingham Risk Scores Best Predictors of Cognitive Decline?

Pauline Anderson

April 05, 2013

Two Framingham vascular risk scores are more strongly associated with cognitive decline than is a dementia risk score, and they could be a better fit for use in primary prevention to target modifiable risk factors, a new study suggests.

The study compared the Framingham general cardiovascular disease (CVD) risk profile and the Framingham stroke risk profile with the Cardiovascular Risk Factors, Aging and Dementia (CAIDE) risk score and showed that the 2 Framingham scores, particularly the stroke score, were more strongly associated with10-year cognitive decline.

The findings provide physicians with more information about the link between midlife cardiovascular disease risk factors and later cognitive decline to pass on to their patients, said lead author, Sara Kaffashian, PhD, a doctoral fellow at INSERM Center for Research in Epidemiology and Population Health, Paris, France.

"Physicians can tell patients who have hypertension or high cholesterol levels that they are at high risk of having heart disease or stroke — but they can also tell these patients that their brain may be at risk, too," she told Medscape Medical News. "Cardiovascular risk factors should be managed and treated early on, in middle age for example, to have a more favorable effect on cognition."

However, Miia Kivipelto, MD, PhD, professor, deputy head, Senior Geriatrician Aging Research Center Karolinska Institutet, Stockholm, Sweden, who helped develop the CAIDE dementia risk score, said that although she is pleased to see increasing interest in risk scores that predict cognitive decline and dementia, she doesn't believe the scores are conducive to comparison because they have different outcomes.

The new research is published in the April 2 issue of Neurology.

Cardiovascular Risk

The 2 Framingham scores include age, systolic blood pressure, hypertension treatment, smoking, and diabetes. The Framingham CVD risk score also includes sex, high-density lipoprotein cholesterol, and total cholesterol. The Framingham stroke risk score incorporates prior CVD, atrial fibrillation, and left ventricular hypertrophy. This score includes 5 categories of systolic blood pressure.

The CAIDE risk score was developed to predict late-life dementia on the basis of midlife risk factors, including age, sex, systolic blood pressure, body mass index, total cholesterol, physical activity, APOE genotype (in 1 version), and education, a marker of cognitive decline.

Study participants were from Whitehall II, an ongoing prospective cohort study that was established in 1985 and enrolled male and female office-based employees aged 35 to 55 years. The study design includes self-administered questionnaires about every 2.5 years and a clinical examination every 5 years.

Cognitive tests were introduced at phase 5 (1997–1999, the baseline for this study). Cognitive function was assessed 3 times during the following 10-year period. Tests included standard cognitive assessment of inductive reasoning, short-term verbal memory, phonetic verbal fluency, semantic verbal fluency, and vocabulary. A global cognitive score was created by using all 5 tests.

The comparison of the Framingham CVD score and CAIDE dementia risk score was based on 4374 participants, and the comparison of the Framingham stroke and CAIDE score included 5157 participants.

The analysis showed that all 3 risk scores were associated with a 10-year decline in multiple cognitive tests, but the CVD and stroke risk scores had stronger associations with cognitive decline than did the dementia risk score.

Global Cognition

For example, for the 10-year change in global cognition, the CVD risk score had a standardized risk of –0.06 (95% confidence interval [CI], –0.08 to –0.05) compared with –0.03 (95% CI, –0.04 to –0.01); P < .01) for the dementia risk score.

And the standardized risk for the stroke score was –0.04 (95% CI, –0.05 to –0.03) for the 10-year change in global cognition, while for the dementia risk score it was –0.02 (95% CI, –0.03 to –0.01; P < .001).

For individual cognitive categories, higher CVD risk was associated with faster decline compared with lower risk in all tests except memory. Dementia risk was associated with faster decline in reasoning and vocabulary. Comparing dementia risk and stroke risk provided similar results for individual cognitive categories.

Of the 2 Framingham scores, the stroke risk score showed slightly stronger associations with 10-year cognitive decline.

The study results reinforce the importance of prevention and early treatment of cardiovascular disease risk factors, said Dr. Kaffashian. "A reduction in these risk factors will decrease the incidence of heart disease and stroke and may also prevent or delay cognitive decline. This should provide an added incentive for both physicians and the public to act and improve their cardiovascular health."

Diabetes — a component of both Framingham risk scores — showed the strongest independent association with 10-year cognitive decline. In the dementia risk score, education had the strongest association with cognitive performance at baseline even though it was not associated with 10-year cognitive decline.

Although the CAIDE risk score does not incorporate diabetes, an as-yet unpublished multiethnic study from Kaiser Permanente in San Francisco, California, found that including diabetes as a component of CAIDE did not increase its predictive value, said Dr. Kivipelto.

Integrating the dementia risk score into the primary care setting may not be realistic or practical, said Dr. Kaffashian. For one thing, although it's not intended to tell a patient whether or not he or she will become demented down the road, some patients may use it for that purpose and this may create undue anxiety. Also, in an already overtaxed general practice setting, "it would be unrealistic to expect clinicians to add yet another screening tool to their practice and patient care," she said.

Comparison Not Relevant?

But according to Dr. Kivipelto, the CAIDE score, which was developed in 2006, is used in primary care settings in Finland, Sweden, and other European countries. An online version is available for use by patients in midlife to predict their dementia risk later on.

It's important to refrain from concluding that the Framingham scores are superior to the dementia risk score, stressed Dr. Kaffashian. These scores were developed to predict different events, are composed of different risk factors and populations and are calibrated differently, so it's not surprising that their predictive values are different, she said.

"The Framingham scores showed stronger associations with cognitive decline because they're composed of cardiovascular risk factors that have in many studies been shown to be linked with several markers of brain pathology and cognitive function. So this may make them more sensitive in distinguishing individuals with subclinical brain pathology," she said. "The dementia risk score includes putative risk factors for dementia. Education level, for example, predicts dementia, but in this middle-aged population it was not associated with rate of cognitive decline. This does not mean that it's not important in prediction of dementia. "

For Dr. Kivipelto, comparing the risk scores isn't relevant; "It's more the fact that we are starting to have tools to predict dementia and cognitive decline. At this moment, you can't say which risk score is better because it depends on what your population is, how long the follow-up is, what your age range is, and what the outcome is, whether it's dementia or cognitive decline."

The outcome for CAIDE was clinically diagnosed dementia or Alzheimer's disease, while this current study used cognitive decline as an outcome, she pointed out. "We know that cognitive decline or cognitive impairment is a different thing; some people with cognitive decline may develop dementia, while others will not."

As for follow-up time, the current study looks at a 10-year period, while the CAIDE dementia risk score predicts dementia risk during a 20-year period, said Dr. Kivipelto.

The study was supported by Région Ile-de-France, the Medical Research Council, the British Heart Foundation, the Health and Safety Executive, the French Department of Health, the National Heart, Lung, and Blood Institute, the National Institutes of Health, National Institute on Aging, Agency for Healthcare Research and Quality, and the John D. and Catherine T. MacArthur Foundation. Dr. Kaffashian is supported by the Region Ile de France.

Neurology. 2013;80:1300-1306. Abstract