Cholesterol Associated With Dementia Following Stroke

College of Physicians and Surgeons at Columbia-Presbyterian Medical Center

Office of Public Relations

630 West 168th Street

P&S Box #37

New York, NY 10032

Media Contact:

College of Physicians and Surgeons

Carolyn Conway  212-305-3900

Characteristics of study population according to incidence of dementia


New York, NY July 21, 1999 - The "bad" form of cholesterol just got worse. Not only are high levels of the low-density lipoprotein (LDL) form of cholesterol a significant risk factor for heart disease, they also increase a person's risk of developing dementia after a stroke, according to a study conducted by researchers at the Taub Alzheimer's Disease Research Center of Columbia University College of Physicians & Surgeons. The findings, reported in the July 21 issue of the Journal of the American Medical Association, provide insight on how high cholesterol levels may contribute to vascular dementia.

"In our study, people with the highest levels of LDL cholesterol had as much as a three-fold increase in the risk of dementia with stroke, called vascular dementia," says Dr. Richard Mayeux, M.D. M.Sc., principal investigator of the study and Gertrude H. Sergievsky Professor of Neurology, Psychiatry and Public Health at Columbia. "Low-density lipoprotein cholesterol is an independent risk factor for vascular dementia, and that finding has important therapeutic implications."

Vascular dementia is the second most common form of dementia in the elderly, following Alzheimer's disease. Dr. Mayeux and his colleagues had previously reported that APOE-e4, one variant of a gene involved in lipid metabolism, is associated with an increased risk of vascular dementia. APOE-e4 also increases the risk of developing Alzheimer's disease. Those findings inspired Dr. Mayeux and Dr. Joan T. Moroney, assistant professor of neurology, to directly investigate the effects of cholesterol levels on the risk of Alzheimer's disease and vascular dementia.

"These people were robust and healthy at the start of the study," says Dr. Mayeux. "We collected data on several suspected risk factors in advance, then waited to see who would develop dementia."

Of the 1,111 people followed for an average of two years, 286 became demented. Of those, 225 developed Alzheimer's disease and 61 developed vascular dementia. Another 61 people had strokes but did not develop dementia. Risk of vascular dementia was associated with both total cholesterol and LDL cholesterol, but not any form of the APOE gene. The risk of Alzheimer's disease was not associated with lipids but was associated with the APOE-e4 gene.

The people who developed dementia following stroke had an average LDL cholesterol level of 130 milligrams per deciliter (mg/dl), while stroke patients who did not develop dementia had an average LDL level of 118 mg/dl. The National Cholesterol Education Program, which is endorsed by the American Heart Association, considers an LDL level of lower than 130 mg/dl to be healthy.

The patients who developed dementia following stroke were divided into four equal groups, according to their LDL levels. When adjusted for vascular disease risk factors, the group with the highest LDL levels, above 143 mg/dl, had three times the risk of developing dementia following stroke as compared with those with the lowest LDL levels, below 96 mg/dl. Complete data table

Dr. Mayeux and the study's lead author, Dr. Moroney, say that it will be up to future studies to explain how high cholesterol levels may contribute to vascular dementia.

"These findings raise lots of questions, more than they answer," says Dr. Moroney. She and Dr. Mayeux have begun a new study that includes more extensive monitoring of cholesterol levels and more extensive brain imaging of patients with stroke.

Both Drs. Mayeux and Moroney agree that lowering cholesterol may reduce the risk of vascular dementia. They are considering a test of that hypothesis with clinical trials of cholesterol-lowering drugs for stroke patients with high cholesterol levels.

The National Institute on Aging funded the study. The National Institute of Neurological Disorders and Stroke also provided support for Dr. Moroney.



Dementia with stroke (n=61)

Stroke without dementia (n=61)

Alzheimer's disease (n=225)

Control subjects (n=764)

Lipid levels (mg/dl)


217.1 ± 45.8

195.6 ± 40.8

191.9 ± 37.4

204.1 ± 38.7


189.2 ± 116.1

177.7 ± 83.7

152.0 ± 62.9

179.2 ± 100.9

HDL cholesterol

46.8 ± 16.8

45.9 ± 18.8

49.1 ± 14.5

47.8 ± 15.5

LDL cholesterol*

130.3 ± 38.9

118.2 ± 34.8

112.3 ± 33.7

120.6 ± 34.3

Corrected LDL-C§*

119.1 ± 37.1

105.4 ± 34.3

100.5 ± 32.9

110.1 ± 33.3


45.9 ± 44.2

46.6 ± 41.6

38.0 ± 32.2

36.1 ± 36.4

Body mass index


27.3 ± 6.5

27.6 ± 5.5

26.9 ± 5.2

27.8 ± 5.6

APOE genotype‡ (percentage)


0 (0.0)

0 (0.0)

12 (5.4)

12 (1.6)


14 (24.6)

18 (30.0)

66 (29.5)

183 (24.9)


43 (75.4)

42 (70.0)

146 (65.2)

541 (73.5)

Vascular risk factors (percentage)

Cardiac disease*

31 (52.5)

25 (41.0)

68 (30.5)

173 (22.5)


40 (67.8)

44 (72.1)

116 (51.8)

367 (48.1)


19 (32.2)

17 (27.9)

34 (15.2)

106 (13.9)

Smoking, current

7 (13.2)

12 (20.0)

29 (15.1)

99 (14.2)


Values in parentheses represent percentages. Some percentages based on an incomplete sample. Significance levels are based on ANOVA for continuous data and chi square tests for categorial data and denote the overall significance of differences among the four groups.

* p

‡ p

† Distributions of triglycerides and Lp(a) were skewed, thus values were log transformed prior to performing statistical testing

§ LDL levels were corrected for the cholesterol carried by Lp(a) by using the following formula: LDL Corrected = LDL - (0.3 x Lp(a))