"Eat like a caveman for a healthy heart", is the headline in The Daily Telegraph today. It and several other newspapers report on a new study which claims that a "paleolithic" or caveman diet of berries, nuts, lean meat and fish "could help to reduce the risk of developing heart disease".

The story is based on a small, short study of 20 young healthy volunteers which had a 30% drop out rate, with complete data for only six people available. However, these six people reduced their calorie intake by about 900 calories to about 1500 calories a day and the whole group of 14 who managed to stick with the study, lost on average 5lbs (2.3kg) in three weeks. There was no control group, so it is not possible to say whether there is anything about a caveman diet compared with any other low calorie diet that produces the weight loss or the other changes noted.

Where did the story come from?

Dr Magnus O-sterdahl and colleagues from the Karolinska Institute in Stockholm, Sweden carried out this research. The study was funded by grants from the Stockholm County Council and was published in the peer-reviewed: European Journal of Clinical Nutrition.

What kind of scientific study was this?

This was an uncontrolled observational study. The researchers say that they did not intend to copy stone-age eating habits but wanted to eliminate the harmful aspects of modern diets. They recruited 10 men and 10 women aged between 20 and 40 via a medical students' association. They only included healthy people with a body mass index (BMI) of less than 30, who did not need hospital care, were not on prescription drugs, did not have an eating disorder or were already eating a special diet. Five of the 10 men and one of the 10 women did not complet

e the study either because of illness, an inability to complete the diet or they broke the study protocol for other reasons. The average weight of the 14 volunteers who completed the study was 10stone 3lb (65.2kg) with a BMI of 22.2, so they were not overweight at the start of the study. The researchers measured a range of other factors, such as blood pressure and heart rate and took blood tests for haemoglobin, glucose, cholesterol and other markers of inflammation or clotting in the body such as plasminogen activator inhibitor-1 (PAI-1). PAI-1 protein is involved in the pathways that cause clotting within blood vessels. These tests were run on three occasions over 21 days.

There were strict instructions about what volunteers were allowed to eat freely, eat in restricted quantities and foods that were prohibited. They were allowed to eat fresh or frozen fruit, berries or vegetables, lean meat, unsalted fish, canned tomatoes, lemon or lime juice, spices and coffee or tea without milk or sugar, for three weeks. All dairy products were banned as well as beans, salt, peanuts, pasta or rice, sausages, alcohol, sugar and fruit juice. However, participants were allowed up to two potatoes a day and were also given some dried fruit, cured meats and a portion of fatty meat as a weekly treat.

All volunteers were asked to record and, if possible, weigh everything they ate, but were given approximate weights of things they could not weigh. The researchers report that, unfortunately, there was a computer error when this food registration data was entered and only the data for one man and five women was available for analysis.

What were the results of the study?

There were significant reductions in five of the 19 parameters measured. Average weight decreased by 5lb (2.3kg), body mass index by 0.8, waist circumference by 0.2in (0.5cm), systolic blood pressure by 3mmHg and PAI-1 by 72%.

The researchers also note that energy intake decreased by 36% and that they observed other favourable effects such as a reduced fat composition of the diet and improved antioxidant content. However, they also point out the unfavourable effect on calcium intake - calcium levels in the blood fell by more than 50% (from 851mg to 395mg).

What interpretations did the researchers draw from these results?

The researchers say "this short-term intervention showed some favourable effects on diet, but that further studies, including control group, are needed".

What does the NHS Knowledge Service make of this study?

"Fad diets" are often promoted as a method of improving health, however they should be subject to appropriate scientific investigation by well designed and conducted, preferably randomised and, at least, controlled studies. There are several limitations to this study which mean that readers should not draw too many conclusions from it.

- The researchers suggest that the high drop out rate of six people in 20 (30%) has caused the study to be underpowered, that is, they were unable to detect a significant effect for some measures. However, it is also possible that they did not detect a significant effect in some measures because they were not there or because the effect was harmful. More importantly, a high dropout rate suggests that there is something about the diet that makes six out of 20 people disinclined to complete a three-week study.

- At least one of the dietary components of a "healthy diet" changed unfavourably during the study. Calcium content fell by more than 50% (from 851mg to 395mg) and this, over a long period of time, could have had harmful effects on bone strength.

- A control group is important in this sort of study for a number of reasons. One important statistical error that can show up in uncontrolled trials is known as "regression toward the mean". This refers to the fact that those with extreme scores on any measure at one point in time will, for purely statistical reasons, probably have less extreme scores the next time they are tested. This research is unable to exclude this effect.

- It is not possible to say which part of this diet contributed to the reduction PAI-1, though reduction in weight on its own is thought to effect blood levels of this protein.

- It is not clear if maintaining this diet for longer than three weeks is possible, or if it results in long term benefits or harms.

Low calorie, low salt diets are expected to have an effect on weight and blood pressure in people who are overweight or have high blood pressure. This somewhat extreme 1500 calorie diet in healthy young volunteers appears hard to tolerate. It is not clear if a "caveman diet" has any specific advantage beyond the modest weight loss. By excluding calcium it may also be harmful for some people too.

Links to the headlines

Eat like a caveman for a healthy heart. The Daily Telegraph, May 9 2008
'Caveman diet' lowers the risk of heart disease, new research shows. Daily Mail, May 9 2008

This news comes from NHS Choices