Villejuif Cedex, France - A low level of HDL cholesterol in middle age is associated with poor memory and could lead to dementia later in life, a new study suggests [1].

The study, published in the July 1, 2008 issue of Arteriosclerosis, Thrombosis, and Vascular Biology, was conducted by a team led by Dr Archana Singh-Manoux (INSERM, Villejuif Cedex, France).

They explain that there is some evidence suggesting that dementia itself modifies lipid levels, leading to low total-cholesterol or LDL-C levels among those with dementia, so examination of the effect of lipids on cognition in the elderly is likely to yield spurious results. They add that associations between midlife lipid levels and late-life dementia appear to be robust, although the precise lipid level that might be important is unclear; some studies have implicated high levels of LDL-C or total cholesterol, whereas others have implicated low levels of HDL-C. They note that HDL-C is critical for the maturation of synapses and the maintenance of synaptic plasticity; it can influence the formation of amyloid, the major component of the protein plaques found in the brains of Alzheimer's patients. Low HDL-C has also been shown to be associated with lower hippocampal volume.

In the current study, the researchers investigated the association between lipids and short-term verbal memory in 3673 participants in the Whitehall II study, a long-term health examination of more than 10 000 British civil servants working in London, UK. They measured total cholesterol, HDL-C, and triglycerides and assessed short-term memory at two time points: first between 1995 and 1997, when participants had a mean age of 55 years; and again between 2002 and 2004, when participants had a mean age of 61 years. Memory was assessed using 20 one- or two-syllable words read aloud at two-second intervals. Study participants were then given two minutes to write down as many of the words as they could remember. Memory deficit was defined as the recall of four or fewer words.

Analyses were adjusted for education, occupational position, coronary heart disease, stroke, hypertension, use of medication, diabetes, smoking, and alcohol consumption. Results showed that, compared with people with high HDL-C (>60 mg/dL), those with low HDL-C (<40 mg/dL) were more likely to have memory deficit at both time points; this association was independent of the presence of the apoE4 allele, a potent risk factor for Alzheimer's disease.

When lipid levels at the two time points at which memory was measured were analyzed, only changed levels of HDL-C were found to be associated with a decline in memory. Compared with people with high levels of HDL-C, those with lower HDL-C had a greater risk of memory decline, defined as a decrease of two or more words recalled (OR 1.61, 95% CI 1.19-2.16).

Other key findings include the fact that men and women did not differ significantly in the link between lipids and memory loss, the fact total cholesterol and triglycerides did not show a link with memory decline, and the fact that the use of statins to raise HDL-C and lower LDL-C showed no association with memory loss.

"We found that a low level of HDL-C may be a risk factor for memory loss in late midlife. This suggests that low HDL-C might also be a risk factor for dementia," Singh-Manoux said. The authors note that the precise mechanism linking HDL-C to dementia remains unclear, but it is possible that HDL-C prevents the formation of beta-amyloid or that HDL-C affects memory through its influence on atherosclerotic disease or vascular injury or through its anti-inflammatory and antioxidant effects.

"Many previous investigations into the association between lipids and memory in the elderly have focused on total or LDL cholesterol because of their status as proven risk factors for cardiovascular disease. Our findings emphasize the need to expand the focus to HDL cholesterol. Thus, physicians and patients should be encouraged to monitor levels of HDL cholesterol," they conclude.

In an accompanying editorial [2], Drs Anatol Kontush and John Chapman (INSERM and University Pierre and Marie Curie, Paris, France) note that the association between low plasma concentrations of HDL-C and dementia have been repeatedly reported. They also note that high HDL-C levels, possibly mediated by low cholesteryl-ester-transfer-protein (CETP) activity, have been linked to longevity, improved cognition, and dementia-free survival and that CETP polymorphisms resulting in low HDL-C are prevalent in people with Alzheimer's disease.

But the editorialists point out that none of these studies suggests causality, because plasma lipid levels can change considerably during the development of dementia, making the time point of the measurement critically important.

They report that there are many complex and variable biochemical mechanisms potentially linking HDL-C to Alzheimer's disease, adding that it is "tempting to speculate that increasing levels of HDL-C might protect our good memories." However, they note that "we should remain extremely cautious when proposing therapeutic intervention on the basis of observational studies that do not imply causation. This is particularly true for a study with a number of important limitations, such as that of Singh-Manoux et al.

"Thus, HDL-C remains a potentially promising but still remote target in the prevention of dementia and memory loss. Nonetheless, these studies demand that we focus more effort on research at the interface between HDL-C and brain function," they conclude.