Tuesday July 1 2008
Using olive oil instead of other fats may help control cholesterol levels
“Poor diet may increase risk of dementia,” The Daily Telegraph reports today. Low levels of high-density lipoproteins (HDL) in the blood – ‘good cholesterol’ – were “associated with a greater risk of memory loss”, the newspaper says. It suggests that foods such as oily fish and oats can increase levels of good cholesterol.
The story is based on a UK study. However, the way the data was analysed means it cannot prove that lower HDL cholesterol levels cause memory loss. In addition, ‘memory loss’ was assessed only by a single 20-word recall test. This is not a diagnosis of Alzheimer’s disease or any other subtype of dementia.
The causes behind general memory loss and diseases such as Alzheimer’s are largely unclear, and this study indicates the complex effects of other medical and lifestyle factors. Although these findings do suggest some association between HDL cholesterol levels and memory loss, further research is needed to clarify this relationship.
Where did the story come from?
This research was carried out by Archana Singh-Manoux and colleagues from the Department of Epidemiology and Public Health at UCL in London, and the Hôpital Ste Périne in France. Individual researchers obtained funding from the European Science Foundation, the Academy of Finland and the Medical Research Council. The Whitehall II study received support from various national organisations. It was published in the peer-reviewed journal from the American Heart Association: Arteriosclerosis, Thrombosis and Vascular Biology.
What kind of scientific study was this?
This study analysed data from the Whitehall II study both as cross-sectional and as longitudinal data. It aimed to examine the relationship between fasting blood lipid levels and short-term verbal memory.
The Whitehall II study began in 1985 and has been following 10,308 civil servants who were members of 20 London-based government departments, aged between 35 and 55 at the start of the study. For this publication, the current researchers used data obtained between 1995-97 and 2002-04. During the first time-period, fasting blood samples were collected to analyse blood lipid (fat) levels and measurements of total cholesterol, HDL cholesterol and levels of triglycerides were taken. The participants’ short-term memory was also assessed using a 20-word recall test where, after being shown a list of 20 words, they had to write down in order as many of them as they could remember. The researchers took into account possible confounding factors such as age, sex, occupation and educational level, and numerous medical factors such as blood pressure, coronary heart disease (past heart attacks or angina), stroke, diabetes, smoking, alcohol consumption and medications (including lipid-lowering medications and hormone treatments which can affect lipid levels). The same tests were repeated in 2002-04.
The researchers carried out statistical analyses to determine whether there was a relationship between lipid levels and memory from the first time period. They then looked at data from the second time period and examined the link between change in lipids with change in memory between the first and second time-period five to seven years later. They categorised HDL cholesterol as either decreasing in level, increasing in level, or remaining broadly the same over both time periods.
What were the results of the study?
The researchers had information available for 3,673 participants at both time periods (73.2% male). During 1995-97, 34.8% of the sample were in the high-risk category for total cholesterol (≥240mg/dL), 11.9% in the high-risk category for HDL cholesterol (<40mg/dL) and 9.9% in the high-risk category for triglycerides (≥200mg/dL).
When analysing the data in a cross-sectional way, the researchers found no significant links between triglycerides or total cholesterol level and poor memory during the first time-period. However, they found that, compared with having the highest HDL cholesterol levels, having low HDL cholesterol levels increased the odds of poor memory by 60% after adjusting for age and sex. When they adjusted further for age, sex, education, employment and medical factors, smoking and alcohol, there was no increase in the odds of poor memory with low HDL cholesterol compared with high levels.
A cross-sectional analysis of memory and lipid levels from 2002-04 gave a similar pattern. There were no significant links between poor memory test results and triglycerides or total cholesterol. But there was a significantly increased risk of having a poor memory test with lowest HDL cholesterol levels compared with the highest levels when the researchers adjusted for age, sex, employment and education. However, there was no significant link when they also adjusted for medical factors, smoking and alcohol.
When they looked at whether the change in lipid levels between the two time periods was associated with decline in memory function, they found no relationship between total cholesterol levels or triglycerides and memory function. However, they found that decreasing HDL cholesterol levels between the two time periods was significantly linked with a two-thirds increase in the odds of a decline in memory, compared with people whose HDL cholesterol levels had remained high (when adjusted for all confounders).
What interpretations did the researchers draw from these results?
The researchers say their findings suggest a “robust association between low HDL cholesterol and poor memory” and that decreasing levels of HDL cholesterol were associated with a decline in memory over five years of follow-up.
What does the NHS Knowledge Service make of this study?
This study demonstrates the consistent relationship between HDL cholesterol and poor recall memory over two time periods, distinct from the non-significant relationships with both triglycerides and total cholesterol. However, there were some limitations.
- The relationships between memory and HDL cholesterol in both time periods were non-significant when a full adjustment was made for medical factors such as heart disease, blood pressure and diabetes, and lifestyle factors such as smoking and alcohol. This shows that these other factors could possibly be having a much larger effect on both HDL cholesterol and memory.
- The cross-sectional analysis used at two separate time points cannot prove conclusively that lower HDL cholesterol causes poor memory recall. For example, as the authors acknowledge, dementia itself may modify cholesterol levels through changes in diet or metabolism.
- The evidence from the longitudinal data is more reliable. However, this may have been confounded by other factors that were not taken into account in the adjustments. Other lifestyle factors such as exercising regularly may have played a part in this relationship.
- ‘Memory loss’ was determined by a single 20-word recall test, therefore the significance of scoring poorly on this test of immediate recall alone is questionable. It is not a diagnosis in itself, and should not be confused with a diagnosis of Alzheimer’s disease or any other subtype of dementia.
- Although still a large sample size, data was analysed for only 36% of the entire Whitehall cohort and results may have been different if all had been included.
- This study included mainly male London civil servants, and as such may not be representative of the general population.
Although these findings do suggest some association between low HDL cholesterol levels and memory loss, further research would be needed to clarify any relationship. For now, this story again demonstrates that a healthy diet and active lifestyle may be the best routes to health.
Sir Muir Gray adds...
I won't change my diet until more research becomes available.