Claims that 'butter is safe' and 'margarine deadly' are simplistic

"Butter unlikely to harm health, but margarine could be deadly," The Daily Telegraph reports. A major analysis of data found no link between saturated fats and heart disease, stroke or diabetes, but there was a link with trans fats.

Saturated fats are found in dairy products such as butter and cheese, as well as in meat and some fish, such as salmon. Some trans fats may come from natural animal sources, but most are from changes made to plant oils during the industrial manufacturing process.

This latest research, which pools the results of around 70 previous studies, found no evidence that eating higher amounts of saturated fat (compared to low amounts) raised the risk of death, heart disease, stroke or diabetes. Meanwhile, eating more trans fats was linked to an increased risk of death or heart disease.

However, the researchers warned that the results are not clear-cut and future research could change the picture. These were all observational studies, which cannot prove cause and effect. Consuming high levels of saturated fat can increase the risk of obesity – a condition that can adversely impact on quality of life.

The researchers make the important point that, rather than focusing on a single food source, a person's whole diet is important. They say that any future guidelines about healthy diets that recommend reducing fat need to be clear about what people should eat as an alternative.

Where did the story come from?

The study was carried out by researchers at McMaster University, the University of Toronto, St Michael's Hospital Toronto, Hamilton Health Sciences, and the Hospital for Sick Children Toronto, all in Canada. It was funded by the World Health Organization. It was published in the peer-reviewed medical journal The BMJ, as an open-access article, which means anyone can read it for free online.

The study was reported accurately in The Independent, which included the researchers’ warnings about the study’s limitations. The Daily Mirror concentrated on the risks from trans fats, while the Telegraph accompanied its article with encouragement to eat more butter – a recommendation not backed by the current available evidence.

The Daily Express’s warning that "low fat versions [of spread] could kill you" is just plain daft.

Much of the tone of the reporting seems out of date, as many sources seemed to imply that margarine contained high levels of trans fats. In fact, due to the negative publicity surrounding trans fats in recent years, food manufactures have virtually removed trans fats from the UK food chain.

Most brands of margarine now contain no, or only trace elements, of trans fats.

A recent factsheet released by the British Nutrition Foundation (PDF, 23kb) estimated that, on average, trans fats account for 1% of total energy intake per person in the UK, which is thought to be well within safe limits.

What kind of research was this?

The researchers carried out a systematic review of observational studies about saturated fat and trans fats, and health outcomes. They performed a meta-analysis by pooling the best-quality studies to see what the results showed overall.

A systematic review and meta-analysis is an excellent way of summing up the state of research on a topic at any given time. However, the results are only as good as the existing studies on the topic. Observational studies can show links between things (in this case, saturated fat and health outcomes such as chances of death) but cannot prove that one thing causes another.

What did the research involve?

The researchers searched databases to find all the research done so far on this topic. They pooled the results of prospective cohort studies in a meta-analysis. One big prospective cohort study on saturated fat could not be pooled because of the way the data was presented, but the researchers compared it to the results of the meta-analysis, to see if the results agreed. They also looked at the results of other studies with different designs. They aimed to see whether there was a link between eating more or less saturated fat or trans fats and outcomes, including death from any cause, heart disease, stroke and type 2 diabetes, and deaths from these specific causes.

In most of the studies included, the researchers compared what happened to the people who ate the most of the type of fat being studied, compared to the least. This means that the amount of fat eaten could vary a lot between studies. Most of the studies measured how much fat people ate by asking them to fill in a questionnaire about how much of different types of food they had eaten in the last day, week or month.

What were the basic results?

The researchers found no link between eating more saturated fat and the chances of dying from any cause, dying from cardiovascular disease (heart disease or stroke) or heart disease specifically, or getting heart disease, stroke or type 2 diabetes.

They found evidence that eating more trans fats increased the chances of dying from any cause (relative risk (RR) 1.34, 95% confidence interval (CI) 1.16 to 1.56), dying from heart disease (RR 1.28, 95% CI 1.09 to 1.5) and having heart disease (1.21, 95% CI 1.1 to 1.33).

They then checked the studies and their findings against a quality system called GRADE (Grading of Recommendations Assessment, Development and Evaluation), to see how reliable the results are likely to be. The system found that the certainty of the results was likely to be very low for saturated fat. That is, that while no link was found between saturated fat intake and these findings, we can’t have much confidence in this finding. However, it was moderate for the link between trans fats and heart disease or death from heart disease, suggesting there is stronger evidence for this link.

How did the researchers interpret the results?

The researchers were cautious about their results (unlike the media). They said that "further research is likely to have an important effect on our confidence in the estimation of association and could change the estimate." This means that, because of the limitations of the research published so far, better research in the future might give a different answer to the question about whether saturated fat is bad for our health.

Conclusion

This careful systematic review and meta-analysis of research into the effects of saturated and trans fat on health found no evidence that eating more saturated fat raises the risk of death from any cause, death from cardiovascular disease, or risk of getting heart disease, stroke, or type 2 diabetes. However, the authors say they can only have "very low" confidence in their findings, because of the methodological limitations of the individual studies that contributed data.

The study did find a link between eating more trans fats and death from any cause, from heart disease or risk of getting heart disease. In the study, trans fats from industrial sources (rather than natural animal sources) were more strongly linked to heart disease or the chances of dying from heart disease. However, this may just be because people in the studies had eaten more industrial trans fats than natural trans fats.

The researchers say they only have low confidence in their findings for several reasons. A meta-analysis is only as good as the studies that you can include in it. The included studies had considerable differences in their methods and results. The researchers could not include one big study because the way that the data had been collected and analysed made it impossible to pool with the other studies in the meta-analysis. The meta-analysis results showed no link between saturated fat and risk of death, while the big study they could not include showed that eating more saturated fat was linked to an increased risk of death. We don't know whether including the results of this study would have changed the overall results of the meta-analysis, if that had been possible.

There could be various reasons for the differences in findings between the different studies. They might have included different study populations, or varied in how they recorded their diet or assessed and followed up the health outcomes. The studies are also likely to vary in how well they took account of different confounding factors that might have affected the results (for example, people's age, occupation, total food consumption, exercise, smoking, income). The problem with studies looking at one part of people’s diets (in this case, saturated fat or trans fat) is that the rest of their lifestyle may also have a big impact on their health.

Finally, the researchers make an important point about what people are eating instead of saturated fat. Refined carbohydrates such as white bread and sugar may be no healthier than fat. And we don't know whether some other types of fats, such as monounsaturated fat, are healthier than saturated fat. This is especially important for people setting national guidelines about diet. If guidelines tell people to eat less saturated fat, or less trans fats, they should also say what people should eat instead, if they are to have a positive effect on health.

While this study does not show strong evidence that saturated fat is harmful, it does not rule out the possibility that it may be harmful.

It cannot be concluded that people can eat as much saturated fat as they like, without any effect on their health.