The claims in today's papers that cocoa consumption could "stave off Alzheimer’s" (The Times) or "ward off dementia" (Daily Express) are hard to swallow.
They are based on a small study that found an association between cocoa consumption, improved blood flow to the brain and an "upturn" in memory. But the study did not involve any patients with Alzheimer’s disease or other types of dementia and is unable to support claims that cocoa may prevent either of these diseases.
In the study elderly people were told to drink two cups of cocoa each day for 30 days. About half drank cocoa rich in a compound called flavanol, while half drank cocoa without much flavanol.
The researchers looked at blood flood to the brain in both groups of elderly people. They were surprised to find no significant differences in the results for the two groups. Individuals in both groups who had impaired blood flow in response to brain activity at the start of the study were found to have improved blood flow after the study period. But flavanol-rich cocoa consumption in itself had no effect.
The researchers conclude that cocoa improves blood flow to the brain, but the lack of a control group in which participants didn't drink any cocoa means that we can’t assume that drinking cocoa caused the observed improvements
This study raises the possibility that there is something in cocoa – not necessarily flavanol – that may improve blood flow inside the brain. Based on this study it is not possible to predict whether this possibility will lead to an effective preventative treatment for dementia or cognitive decline.
The study was carried out by researchers from Harvard Medical School and was funded by the US National Institute on Aging and the National Heart, Lung and Blood Institute.
The study was published in the peer-reviewed journal Neurology.
The quality of the UK’s media reporting on the study was mixed. Some organisations, such as the BBC, Daily Mirror and The Daily Telegraph stuck to the facts – cocoa may improve some brain functions. But the Express, Times and the Daily Mail – at least in their headlines – exaggerated the findings to imply that a way to prevent dementia has been discovered, which is clearly not the case.
However, the Daily Mail made up for this by:
The cocoa tested in this study was rich in an antioxidant called flavanol, which has previously been shown to have benefits in terms of cognitive function as well as some vascular measures.
The brain requires a continuous supply of oxygen and sugar, which is delivered by blood, in order to function properly. Blood flow to the brain has been shown to change with variations in brain activity, increasing as the energy demands of the brain increase. This close relationship between brain activity and blood supply has been termed ‘neurovascular coupling’ (NVC).
The authors report that impaired NVC has been associated with several diseases, such as vascular dementia.
The researchers recruited 60 people over the age of 65 years who had a range of vascular conditions (high blood pressure, treatment with high blood pressure medications or well controlled type 2 diabetes).
The participants were randomised to one of two groups. The first group drank two cups of flavanol-rich cocoa each day for 30 days. The second group drank two cups of flavanol-poor cocoa each day. All study participants were asked not to eat chocolate during the study, and to avoid caffeine on days when study measurements of mental and vascular functioning were taken.
Several tests were completed at the beginning of the study, one day into cocoa consumption and after 30 days of cocoa consumption. These tests assessed vascular function, blood supply to the brain and cognitive function.
Originally, the researchers compared outcomes between participants drinking flavanol-rich cocoa and those drinking flavanol-poor cocoa. This was to analyse whether flavanol-rich cocoa consumption had any effect on neurovascular coupling.
A secondary data analysis ignored the randomised groups and examined changes in vascular and cognitive outcomes across the entire group. The researchers also conducted a subgroup analysis looking at changes in these outcomes in participants with intact NVC and those with impaired NVC at the beginning of the study.
These analyses were observational and can only tell us about associations between cocoa and vascular or cognitive functioning. They cannot prove that cocoa consumption causes any of the observed differences.
Of the 60 study participants, approximately 90% had well controlled blood pressure, half had type 2 diabetes and three-quarters were overweight or obese.
Nearly a third of the participants were found to have impaired neurovascular coupling at the start of the study. These participants were found to have significantly worse scores on some of the cognitive function tests at the beginning of the study (baseline), compared with participants with intact NVC.
When examining the effect of cocoa on NVC, researchers found that changes in blood flow and blood pressure were not significantly different between people drinking flavanol-rich versus flavanol-poor cocoa. When pooling results across all of the participants, the researchers found no significant change in NVC over time.
When assessing the association between cocoa consumption and NVC status overall, the researchers found that a significantly greater proportion of individuals with impaired NVC at baseline had increased NVC at follow-up compared with people with intact NVC at baseline (89% versus 36%).
Among those with impaired NVC, cocoa consumption was associated with a 10.6% increase over 24 hours, and an 8.3% increase after 30 days. No significant changes were seen among participants with intact NVC at the start of the study.
A similar pattern of changes was seen when researchers assessed performance on one of the cognitive function tests. Changes in test scores were dependent upon NVC status at baseline, with participants showing intact NVC exhibiting no changes in test performance over the course of the study, while those with impaired NVC at baseline showed significantly improved performance after 30 days. Performance between the two groups was significantly different as well.
The researchers conclude that neurovascular coupling could be modified, and that consumption of cocoa was associated with improvements in this relationship among individuals with impaired NVC.
This study suggests that cocoa consumption may be associated with the manner in which blood flow and brain function interact among elderly people with vascular conditions.
A major limitation of this study is the method of analysis. While randomised controlled trials are considered to be the best method to determine the effect of a treatment or intervention (in this case, consumption of flavanol-rich cocoa) on a health outcome (neurovascular coupling), this strength depends on the ability to analyse outcomes in the intervention compared with the control group. In this study, the analysis of the intervention group compared with the control group revealed no significant difference in NVC.
After this finding, the researchers conducted further analyses by considering all the participants together. This analysis did not have the benefit of randomisation, which means that we cannot determine if cocoa consumption was, in fact, responsible for changes in NVC.
The researchers suggest multiple reasons for the lack of a significant difference between the effect of flavanol-rich and flavanol-poor cocoa on NVC. First, they suggest that it may not be flavanol, but another component of cocoa that is responsible for the observed changes in NVC in both randomised groups. Alternatively, they suggest that NVC is extremely sensitive to flavanols, and that the low concentrations seen in the flavanol-poor cocoa group were sufficient to cause an improvement in NVC. The researchers did not take into account the dietary changes that were made during this month to accommodate the calories from cocoa, and did not record how much chocolate or cocoa the participants usually consumed.
A proper control group, consuming neither cocoa nor flavanol-containing drinks, would be needed to test these various hypotheses.