Introduction

Presence of the apolipoprotein allele APOEe4 is a recognized risk factor for Alzheimer disease, and a possible role of the low-density lipoprotein (LDL) receptor-related protein in Alzheimers has been demonstrated. Scientists from the Boston Collaborative Drug Surveillance Program have recently reported on a study they conducted of the possible effect of 'statin' drugs (hydroxy-methylglutaryl-coenzyme A, or HMGCoA reductase inhibitors), and other lipid-lowering drugs, on the occurrence of dementia.

Method

The General Practice Research Database, which consists of data from patients enrolled by family practitioners in the UK, provided the study material. Out of more than 3 million subjects enrolled since 1987, three separate groups were selected: Group I were all those aged 50 to 89 years with at least one prescription for a statin drug or another type of lipid-lowering agent at any time.¹ Group II was those patients diagnosed as having hyperlipidemia who did not receive any lipid-lowering treatment. Group III was a random sample of 25,000 people from the database who didnt fall into one of the other two groups. All selected persons were followed for 5 years, from January 1992 to January 1998. However, all those with a diagnosis of possible interfering conditions² before the diagnosis of dementia was made were excluded.

Within these subjects, all those who developed dementia of any kind were identified. Up to four matched controls were selected for each case of dementia. Matching was done based on a number of factors, including age, sex, and calendar time of dementia diagnosis. Conditional logistic regression analysis was done to establish the risk of dementia according to the type of drug exposure and untreated hyperlipidemia.

Results

The three groups were constituted as follows: Group I: 24,480 individuals; Group II: 11,421 individuals, and Group III 25,000 individuals selected at random. Overall, there were 284 people with a first-time diagnosis of dementia during the 5-year period, and they were matched with 1,080 controls.

After adjusting for age, sex, calendar time, smoking, body mass index (BMI), previous cardiovascular disease, and diabetes, relative risk estimates (odds ratio, or OR) were calculated as follows:

• Dementia, with untreated hyperlipidemia: OR 0.72 (95% CI, 0.45-1.14)
• Dementia, with current statin use: OR 0.29 (95% CI, 0.13-0.63)
• Dementia, with other lipid-lowering drugs OR 0.96 (95% CI, 0.47-1.97)

The roughly 3-fold lowered risk for dementia in those taking statins was statistically significant (p=0.002), while there was no significant association for the use of other lipid-lowering drugs or untreated hyperlipidemia (p=0.94 and p=0.16, respectively).

There were no significant effects of age and sex on these findings, nor was the type of statin taken relevant. Independently associated risk factors for dementia were a history of bypass surgery, smoking, and the lowest body mass category compared to the highest (i.e. a BMI below 28 vs. > 27.9).

Comment

These findings show that, at least for people living in the UK, the risk of dementia is lowered by approximately 70% in those prescribed statins, compared with those treated with non-statin lipid-lowering drugs or those without hyperlipidemia. Moreover, people with hyperlipidemia but who did not receive treatment had no reduced risk of dementia.

Observational epidemiological studies, such as this one, are sometimes criticized as not proving causality. However, many causal links are first suggested by the results of observational studies, which then lead to prospective clinical studies. Examples are the negative association between aspirin and myocardial infarction, and occurrence of increased venous thromboembolism in women taking estrogens.

What is the mechanism by which statins could have a preventive effect on the development of dementia? This is not obviously apparent from this study. The desired effect of statins is to inhibit LDL cholesterol levels; however, this was probably not relevant for the effect on dementia, as the other lipid-lowering agents were without a protective effect. Moreover, other studies have failed to show an association between lipid levels and Alzheimer disease.

Statins have been shown to have benefits beyond their lipid-lowering effect. One of these is improved microvascular blood flow, related to increased endothelial nitric oxide synthase and reduced endothelin-1, leading to capillary dilatation. This possible mechanism would imply likely benefits in preventing vascular dementia, without any obvious effect on Alzheimer disease; however, the majority of the study population examined in this study had Alzheimer disease.

These findings, without substantiation by results from prospective studies, are not sufficiently strong to recommend prescribing statins to all older persons, or indeed to all APOE-e4 allele subjects. However, they are sufficiently interesting to justify a large, prospective, well-designed study. As it is, they provide further data on the wide potential benefits of this class of drugs, and should stimulate new approaches to preventing dementias.