Feature Nutrition

Sticky labels

BMJ 2010; 341 doi: http://dx.doi.org/10.1136/bmj.c5861 (Published 26 October 2010) Cite this as: BMJ 2010;341:c5861
  1. David Payne, editor, bmj.com
  1. 1BMJ, London WC1H 9JR, UK
  1. dpayne{at}bmj.com

Europe is forcing huge numbers of food companies to drop health claims for their products because they fall short of new standards of scientific proof. But David Payne asks if this crackdown is helpful

Last week the makers of fermented milk drink Yakult were dealt a blow when Europe’s food safety regulator ruled that the company could not claim that the product maintained immune defences against the common cold.

The European Food Safety Authority (EFSA) panel charged with evaluating health claims considered a dossier of 12 studies submitted by Yakult for its strain of probiotic bacteria Lactobacillus casei shirota.

The Yakult ruling was one of a tranche of probiotic claims considered by EFSA’s independent panel of scientists. According to manufacturers, probiotic products can strengthen the body’s defences, improve immune function, and reduce gastrointestinal problems. But EFSA found the claims were either so general as to be inadmissible or could not be shown to be true based on the evidence submitted by the food industry.

European nutrition and health claims legislation came into effect in July 2007 and applies to all claims, including those displayed on product labels, marketing, and advertising material.

Food companies wanting to make a new health claim or amend an existing one in an EU member state must submit an application to the country’s relevant body together with scientific evidence to support their case. The information is then forwarded to EFSA based in Parma, Italy, for assessment.

Member states were also required to submit a list of established claims, compiled in consultation with the food industry, with supporting references. To qualify for the EU list of permitted claims, EFSA must be assured that the claims are based on generally accepted scientific evidence and be well understood by the consumer. Both authorised and rejected health claims are displayed on the European Commission website.

Industry unease

But it is the volume of claims submitted, and the number rejected by EFSA, that has taken both the food industry and campaigners by surprise. Member states submitted more than 44 000 general function health claims after the regulation was adopted in 2006. These were whittled down to just over 4000 for consideration by EFSA.

So far the authority has examined around 1300 and rejected 80% of them, mostly because the applications contained insufficient scientific evidence and claims were not supported by high quality human studies.

Rejected claims include that glucosamine can reduce the risk of osteoarthritis, cranberry juice can reduce the risk of urinary tract infection in women, and prunes can help maintain normal bowel function.

Helen Crawley, a public health nutritionist and reader in food policy at City University, London, says the food industry is “outraged” by EFSA’s robust approach. “The fear we had was that EFSA would fall for the claims, but they’ve been doing the opposite.”

Food industry representatives say that in principle they support the drive to harmonise claims across Europe but are frustrated at EFSA’s reluctance to enter into a dialogue with companies while their claims are being considered.

Anne Heughan, who chairs the UK Food and Drink Federation’s nutrition committee and is head of external affairs at Unilever, says: “The industry was very committed to having EU health claims legislation on the basis that there were an increasing number of claims on the market at different levels of science.

“Enforcement authorities said they couldn’t work out if something was scientifically accurate or not. We trust and respect EFSA in terms of what it is doing, and it should be done on a European basis. If it was done at member state level it would require duplication of effort.

“The difficulty for the industry is we would want to have a bit more dialogue with them. The process isn’t easy. We are trying to understand what level of science they expect. This is a particular problem with new areas of science, such as immune function or weight management.

“We’d also like a pre-market dialogue, when we’re putting a research and development programme together, so we can be told by EFSA ‘we believe this will sustain your claim.’ ”

EFSA says it is addressing these concerns by holding stakeholder meetings to discuss the process. The first took place in Parma in June and was attended by more than 400 people from industry and governments of member states.

Albert Flynn, professor of nutrition at the University of Cork and chair of EFSA’s evaluation panel, said the June meeting discussed the evaluation process. Before the meeting, the authority issued guidance and consultation documents about which principles the authority should adhere to when it considers opinions. A second meeting is planned for December, which will focus on claims relating to gut and immune functions, including probiotics.

Ioannis Misopoulos, director general of the International Probiotics Association, told the BBC in July: “The regulation is killing this industry and the job losses are already being felt.”1

Professor Flynn says the independent evaluations provided by the panel are on an unprecedented scale. The authority is acting as a “referee” between the industry and the consumer after years of claims being unregulated.

He adds: “The legislation is complicated, and it’s a challenge for everybody to get their heads around it. But it is also very explicit. The legislation says all claims should be based on generally accepted science. Claims are an important part of marketing for the industry and they are highly competitive, so by their very nature will be controversial.

“It’s important that we are consistent, careful, and transparent. Every word we write will be scrutinised by the applicant and their competitors. If we were slipshod we would be picked up very quickly.

“Every claim will have a target population, and with health claims this is a general healthy population. We are not talking about the treatment of disease.”

As a manufacturer of probiotic products, Danone will be attending the December meeting. Connie Hersch, the company’s director of healthcare professional relations, said it is a problem that EFSA is not allowed to consult experts that have worked with the company.

“We have been very supportive of the process and have been investing in the science. But if you are a major player, as we have been for 15-20 years, you have probably worked with most of the international experts. There are not endless amounts of experts in the field.

“That’s especially important for an emerging science like probiotics. The panel members don’t have much expertise. We need people with research or clinical experience.”

Professor Flynn argues that the EFSA process subjected probiotic studies to an independent evaluation for the first time and that the panel is made up of 20 expert members, with access to more than 30 others.

Campaign group Baby Milk Action, however, questions the independence of EFSA’s panel, given that all are part time and many will have been involved in industry funded research at some point in their careers. Patti Rundall, its policy director, says the panel’s chair should not have any conflicts of interest. Professor Flynn responded: “We are part timers, yes. EFSA could of course have been set up with in-house employed experts with no other interests, similar to the US Food and Drug Administration, but that would have been expensive.

“I have two vice chairs, one of whom takes over the meeting if I have to leave the room because of a conflict. We have an elaborate conflicts of interest policy, and our interests are displayed on the EFSA website.”

Evidence thresholds

Baby Milk Action is also concerned that the panel does not consider scientific evidence unless it is submitted as part of the claim. Professor Flynn argues that the onus for providing scientific evidence has to rest with the industry and it isn’t the authority’s role to go searching for other studies.

But is EFSA’s science bar too high, and the kinds of studies it allows too restrictive? Ms Hersch says Danone would also be seeking clarity about the standards of studies at the December meeting. “Can you allow diseased populations? When is it appropriate that the study be done in a healthy population?

“If you look at EFSA’s opinion on prunes, for instance, does it mean they don’t work? Is the highest possible standard of proof the best possible standard? It may mean that you reject lots of things that actually do work. Remember, we are talking about foods here, not drugs.”

“The same goes for cranberry juice. Most women in this country have had a urinary tract infection, and you used to be told by your doctor to drink some cranberry juice. What does a doctor say now?”

The Alliance for Natural Health is a UK based not for profit alliance of doctors, scientists, lawyers, and consumers that “promotes natural health as the primary approach to the management of human health.”

It accuses EFSA of neutering freedom of speech and gagging industry at a time when most consumers need more, not less information, to make the right food choices to protect them against developing largely preventable chronic diseases.

According to an article about the process on its website: “When you dig into the conclusion for each claim evaluation, the phrase ‘a cause and effect relationship has not been established between [x] and [y]’ is far more frequent than many of us would have hoped.

“Herein lies the problem: causal relationships are extremely difficult to establish, as testified by the decades of research required to make the cast iron link between smoking tobacco and developing lung cancer. The list of failed claims applications is seemingly endless . . . and it makes very disturbing reading when you see the level of proof that EFSA is relying on. If such a level of proof were used for drugs, most would not make it to market.”2

A rejected claim can be resubmitted, along with new scientific evidence to support it. But Dr Crawley predicts the industry may bypass the process and decide to include “softer” claims that bypass the regulations. In May 2010 a conference was held in Brussels that discussed how to deliver positive product claims without EFSA regulations.3 Delegates discussed alternative solutions, what the current regulations allow, and what the implications of withdrawing from the whole EFSA claims process would be.

Dr Crawley says: “If you have a health claim rejected it’s a harder position to deal with. The whole world knows about it. We will see the industry regrouping.”

One option, she says, might be to mimic some of the subtler approaches used by the supplements industry when fish oils were first advertised on TV, communicating claims about IQ by showing shoals of fish swimming in the sea.

If companies find alternative ways of communicating health benefits that do not rely on the EFSA process, do they need to make health claims?

“I have a problem with people spending their money on a food because it makes a claim, and to have a special ingredient for that claim to be made. EFSA passed Benecol [a plant stenol ester], and there is good evidence about its benefit for people with raised blood cholesterol. But not for people who don’t have it.”

Food companies should not add anything to a food unless there is a clear health benefit (such as for fibre), Dr Crawley says.

“I certainly think we should not have health claims for children. You can go into a shop now and buy yoghurts that say ‘paediatric dietician approved.’ I get emails from parents asking if it’s OK to give their children mashed potato. Years ago food for children didn’t exist. The industry wants you to believe you need special food for children.”

Health claims defined

EU regulation 1924/2006 covers both nutrition and health claims. Nutrition claims are claims such as “low fat” and “no added sugars.” Health claims are claims such as “reduces blood cholesterol for a healthy heart.”

A health claim, according to regulation 1924/2006, is “any claim that states, suggests or implies that a relationship exists between a food category, a food or one of its constituents and health, eg this drink will help you feel more energetic/this food will help you concentrate.”4

Claims should not:

  • Be false or misleading

  • Give rise to doubt about the safety or nutritional adequacy of other foods

  • Encourage or condone excess consumption of a food

  • State, suggest, or imply that a properly balanced and varied diet cannot provide adequate quantities of nutrients

  • Refer to changes in bodily function that could give rise to, or exploit fear in, the consumer either textually or through pictorial, graphic, or symbolic representation

EFSA and glucosamine

One of the claims submitted to EFSA linked glucosamine to a reduced risk of developing osteoarthritis in healthy adults. EFSA’s scientific opinion, published in October 2009, was based on 13 randomised controlled trials, two meta-analyses, one systematic review, two animal studies, and one in vitro study.

It concluded that “a cause and effect relationship has not been established between the consumption of glucosamine hydrochloride and a reduced rate of cartilage degeneration in individuals without osteoarthritis.”5

A meta-analysis in the BMJ in September 2010 supported this conclusion, finding that joint supplements glucosamine and chondroitin did not have any beneficial effect on osteoarthritis of the hip or knee, compared with placebo.6

The authors said health authorities and insurers should be discouraged from funding treatment, but there was no reason for patients not to keep taking them if “they perceive a benefit and cover the cost of treatment themselves.”

Two months earlier, a randomised placebo controlled trial in JAMA concluded that six months’ treatment with oral glucosamine did not reduce pain related disability in patients with chronic lower back pain and degenerative lumbar osteoarthritis.7

Reporting on this study in his BMJ blog, general practitioner Richard Lehman, referred to glucosamine as the “doctor’s best friend.” He said: “It doesn’t do anything, but people believe it might, and so if you have trouble treating their joint symptoms you can always say ‘have you tried glucosamine?’ Note that you are not obliged to lie: the most you have to say is that it seems to work for some people. [In the JAMA study] it was as good as placebo with slightly fewer adverse effects.”8

Professor Flynn agrees. “Some people accept glucosamine has been mildly beneficial for people with osteoarthritis, but that’s an issue for the European Medicines Agency, and we never come into conflict with the EMA. What we have said is that for the general population, the evidence doesn’t substantiate the claim.”

EFSA and prunes

In reaching its opinion on prunes, EFSA’s panel considered 16 references, three of them to human intervention studies.

The first study did not find any significant difference between the two treatment groups or between different time points within each group. The second showed some effect on faecal bulk but not on stool frequency or consistency. The third study used prune juice not whole prunes. The remaining references provided only background information, the panel ruled.

It concluded that “the evidence provided is insufficient to establish a cause and effect relationship between the consumption of dried plums or ‘prune’ cultivars . . . and maintenance of normal bowel function.”9

“No bowel can resist prunes and figs—eat some daily,”10 medical agony aunt Miriam Stoppard said in an advice column 11 years ago. Based on EFSA’s evaluation, what could she say today?


Cite this as: BMJ 2010;341:c5861


  • Competing interests: The author has completed the unified competing interest form at www.icmje.org/coi_disclosure.pdf (available on request from him) and declares no support from any organisation for the submitted work; DS worked as web editor at the UK Food Standards Agency from 2001-8; and no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned; not externally peer reviewed.


View Abstract