Tuesday August 30 2011
The study does not confirm that chocolate is protective
“Eating high levels of chocolate could reduce the risk of cardiovascular disease and stroke,” reported BBC news. According to the broadcaster, a study has found that the highest levels of chocolate consumption “were associated with a 37% reduction in cardiovascular disease”.
The news is based on an analysis that combined the results of seven previous studies. These studies had looked at how chocolate consumption related to the risk of heart disease, stroke and metabolic diseases. Although this analysis did show that the risk of cardiovascular disease was lowered by about a third in the high chocolate consumers compared with the low chocolate consumers, it does not confirm that chocolate is “good for you”. This is because the studies available for inclusion were limited by the designs and methods they employed. Also, each study categorised chocolate consumption differently, making their results hard to combine accurately.
Based on these studies it is not possible to say whether chocolate reduces the risk of cardiovascular disease and stroke. They also do not explain how chocolate might reduce risk, for example, whether chocolate contains chemicals that are protective, or whether eating chocolate causes people to be less stressed. Chocolate is high in calories, fat and sugar, and can lead to weight gain, which is a known risk factor for heart disease and diabetes. This study does not give enough evidence to suggest that chocolate is protective of the heart.
Where did the story come from?
The study was carried out by researchers from the University of Cambridge. It received no specific funding. The study was published in the peer-reviewed British Medical Journal.
The newspapers advised that it is not appropriate to eat large amounts of chocolate in an attempt to reduce the risk of heart disease. This is appropriate advice.
What kind of research was this?
This was a systematic review and meta-analysis that aimed to identify randomised controlled trials and observational studies that had looked at whether there was an association between chocolate consumption and the risk of developing heart and metabolic disorders (including diabetes).
The researchers said some previous laboratory and observational studies have suggested that a chemical found in chocolate, called flavonol, may have the potential to be good for the heart and prevent metabolic disorders. However, the researchers wanted to look at all of the available evidence from studies in humans to see whether there is any association between chocolate intake and the risk of developing ‘cardiometabolic disorders’. These include the following conditions:
- cardiovascular disease - stroke, heart failure and heart attack
- metabolic syndrome - a group of risk factors that occur together and increase the risk for coronary artery disease, stroke and type 2 diabetes
What did the research involve?
The researchers looked for all randomised controlled trials, cohort studies, case-control studies or cross-sectional studies that had looked at chocolate and cardiovascular disease or metabolic disorders in adults. To gather studies they searched various medical and scientific publication databases, which contained publications from 1950 to October 2010.
Two reviewers independently looked at the abstracts of the papers to decide whether they were suitable to be included in the study (based on the study type and the topic of the paper). Included papers were assessed for quality. For example, the researchers assessed whether the participants’ usual chocolate consumption was measured using a validated method, whether a diagnosis of cardiometabolic disease was made through objective examinations (rather than self-reporting by participants) and whether adjustments were made for factors such as age, gender, body mass index, smoking, physical activity and other dietary factors.
Where feasible they pooled all of the data together and looked at the relative differences between high and low chocolate consumption and outcomes such as diabetes, heart disease, cardiovascular disease, death following heart disease, and incidence of stroke and deaths from stroke.
They also performed statistical tests to see how variable the studies were (their heterogeneity) and they also assessed whether there had been ‘publication bias’. This is where studies with particular results (often positive ones) are more likely to be published than those without significant findings.
What were the basic results?
Out of 4,576 studies that were initially identified the researchers found that seven met their criteria and were included in the review. In total these seven studies provided data on 114,900 participants. One was a cross-sectional study carried out in the US, and the other six studies were cohort studies that had been carried out in Germany, the Netherlands, Sweden, Japan and North America. Most of the participants in the study were white, but one study also included Hispanic and African American people, and one study looked at an Asian population. The age of the participants across the studies varied between 25 and 93 years.
In three of the studies the participants were taking medication, including hormone replacement therapy drugs and drugs for cardiovascular disease.
All of the studies reported overall chocolate consumption, but did not report whether people had eaten white or dark chocolate. All of the studies reported chocolate consumption in a different manner, either by including ranges reflecting how often people ate chocolate or the grams of chocolate eaten a day. For example, one study grouped participants into three categories according to consumption levels, with the highest consumption category including people eating chocolate once a week or more. Another study categorised people into four groups, with people in the top quarter eating up to 7.5 grams a day. Given the differences in the ways each study reported and measured chocolate consumption, the researchers decided to use the highest and lowest categories in each study to measure the association of chocolate consumption and metabolic disorders.
The highest levels of chocolate consumption were associated with a 37% reduction in cardiovascular disease compared with the lowest levels (relative risk 0.63, 95% confidence interval 0.44 to 0. 90) and a 29% reduction in stroke compared with the lowest levels (relative risk 0.71, 95% confidence interval 0.52 to 0.98).
Only one of the studies evaluated the association between chocolate consumption and diabetes, and it reported a beneficial risk reduction associated with the highest level of consumption in Japanese men and women: compared with the lowest consumption they experienced risk reductions of 35% and 27%, respectively (hazard ratios 0.65, 95% CI 0.43 to 0.97, and 0.73, 95% CI 0.48 to 1.13, respectively).
How did the researchers interpret the results?
The researchers said that their findings confirm that “existing studies generally agree on a potential beneficial association of chocolate consumption and a lower risk of cardiometabolic disorders”. However, they did warn that eating too much chocolate can have harmful effects. They said that corroboration is now needed from further studies to assess whether chocolate caused the effects or was just associated with a lowered risk of cardiometabolic disorders.
This research reviewed the available evidence on whether there is an association between chocolate consumption and risk of cardiovascular disease, diabetes and metabolic syndrome. It found that people who consumed more chocolate had approximately a third lowered risk of cardiovascular disease.
However, the review is limited by the quality of the available studies. It only examined studies with cross-sectional and cohort designs rather than randomised controlled trials, which would provide the best method for assessing whether a defined level of chocolate consumption had an effect on later health outcomes. Observational studies are not able to establish a cause and effect relationship. The cross-sectional study in particular was not able to establish cause and effect as it simply questioned participants on chocolate consumption at the same time as assessing coronary heart disease.
Another key problem with combining the results of these seven studies was that they had each categorised chocolate consumption differently. For this reason it is not possible to say how much chocolate is “good” for you or assess the risk of eating “high levels” of chocolate relative to “low levels” in any context. It is not possible, for example, to judge whether people eating high levels of chocolate would be eating sufficient to gain weight over time, which may in turn lead to an increased risk of cardiometabolic disorders. Also, in some studies the amount of chocolate needed to be classed in the highest consumption groups could be considered to be relatively low, as in some cases it was the equivalent of just one standard bar per week. This would mean that whether a participant had one bar or ten bars per week they would be classified in the same group, potentially distorting the results.
The researchers themselves highlight that the available data on the topic were limited and each of the studies was very different. Therefore, it is not possible to establish a clear relationship between the amount of chocolate eaten and the risk of cardiometabolic disorders.
The researchers also said that their research would need to be followed up by other studies, not only to confirm whether there is an association but also to see whether chocolate was actually responsible for the decreased risk. For example, two theories that would require testing are whether chemicals such as flavonol cause a decreased risk, or whether not denying yourself chocolate is associated with decreased stress that leads to positive cardiometabolic effects. Neither of these theories was addressed directly by this research.
Other limitations to this study were that the population was predominantly white and did not contain British participants. It, therefore, may not be relevant for the British population as a whole.
Overall, the analyses presented by these researchers are worthy of follow-up but the limitations of the studies included in this pooled analysis make it too limited to draw firm conclusions on whether chocolate lowered the risk of cardiometabolic diseases.
Analysis by Bazian