International collaborations of enthusiasts can progress the application of nutrition research in the practical world. Good examples are IVACG, IDECG, ICCIDD and IOTF. In medical practice, the selection of nutritional knowledge and methodology that is used differs for the various specialities. For intensive medical care there are ASPEN, BAPEN, ESPEN, etc, the parenteral and enteral nutrition societies. For the last 10 y, the Heelsum collaboration has been researching and discussing ways of helping family doctors use nutrition in a more appropriate way in their work with patients. ‘Evidence-based nutrition advice and counselling’ is in the end its objective. An essential first step is to understand better the role and the challenges of general practice and primary care and the role of nutrition in it. There are three features of nutrition that stand out.

First is the disease perspective. Family physicians deal with a large range of illness and disease, acute and chronic, from ‘trivial’ to life threatening, with nutrition often related to its prevention or intervention. In other words: there is no dominating single group of diseases or nutritional intervention, it transverses the scope of diseases, each with its own sense and nonsense: from the supposed role of chocolate-eating in migraine to folic acid suppletion during pregnancy and from the facts of a sugar-free diet in diabetes type II to the fictions of unlimited use of healthy unsaturated fats. In supporting the quality of nutritional advice, the evidence of all these interventions must be rated in the context of outcome of (primary) care.

But nutrition in family practice is more than its evidence in the treatment of diseases. The second feature is that nutrition advice is directed at individual patients' need to change their lifestyle, with ‘eating’ often only one of the aspects involved. For example, in losing weight bickering over eating might be less effective than promoting exercise. Lifestyle advices—including nutrition—are often strategies of individual change, rather than disease interventions.

Lifestyle is a community characteristic as much as an individual one, and the third feature of nutrition in family practice is the interface of the consulting room and the community. Counselling that might achieve, with limited efforts and effect at population level, may not yield the same for an individual patient—or at the expense of much higher costs.

These three features—the disease-based evidence, the individual strategy of change and the population interface of primary care—took central stage in the 4th workshop of the Heelsum Collaboration on Nutrition in General Practice held during 13–15 December 2004, and characterise this supplement (of the EJCN) that publishes the papers (and summarised discussions) from this workshop.

The first of these international workshops, ‘Nutritional Attitudes and Practices of Primary Care Physicians’, was held in Heelsum, near Wageningen, The Netherlands, in December 1995 and published in Truswell (1997). The second workshop was held in the same place, Heelsum, The Netherlands, 3 y later, with the title ‘Family Doctors and Patients: Is Effective Nutrition Interaction Possible’. It was published in Truswell (1999). A shorter version of this workshop, condensed by Truswell (2000) as a special Article in the American Journal of Clinical Nutrition.

The third workshop was held in December 2001, again in Heelsum with the title ‘Nutrition Guidance of Family Doctors’. It was published in Truswell (2003).

Between these 3-day invited meetings, participants from the Heelsum collaboration have spoken at specific 2-h workshops: in 1997 at the 16th International Nutrition Congress, Montreal; in 1999 at the Eighth European Nutrition Conference, Lillehammer; in 2001 at the 17th International Nutrition Congress, Vienna; in 2003 at the Ninth European Nutrition Conference, Rome.

Throughout these meetings, participants have used the words ‘family physician’, ‘primary care physician’ and ‘general practitioner (GP)’ as meaning essentially the same basic speciality of medicine.

The participants of the Heelsum collaboration are scientists in the fields of nutrition, health promotion, (nutrition) communication, general practice, epidemiology and methodology, as well as researchers interested in the interface between nutrition education and medicine in general practice. The headquarters of the international scientific committee is at Wageningen University, which has one of Europe's leading Nutrition departments.

The earlier workshops had concluded that GPs are in the best position to give nutrition advice. They are readily available in the community, generally trusted and often consulted for problems with a (potential) nutrition component. It also reflects their role in secondary and tertiary prevention. Giving nutrition advice is facilitated largely for GPs when the health-care system promotes continuity of care, and a role for practice nurses. But, even when conditions are optimal, GPs do not always capitalise on their excellent position—for which lack of time and training in nutrition and patients' poor compliance with dietary prescriptions are held responsible. These barriers have come forward in particular in the treatment of obesity, which is one of the most common health problems in primary care. This emphasised the need of more ‘nutrition’ input in the undergraduate and speciality training. In addition, the workshops had concluded a need of developing methods of nutrition advice that can be used in a variety of primary health-care settings and for a variety of patient groups and health problems in general practice. Innovations in communication methods and the use of information technology were identified as particularly promising, as was the packaging of research evidence for application in daily practice.

This was the starting point of the Fourth Heelsum Workshop, at which speakers came from the Netherlands, USA, UK, Australia, Italy, Spain, Croatia and Canada.

A session on The individual and his/her lifestyle dealt with empowering patients to discuss matters they are concerned about with their doctor. The concept that doctors and patients work as partners, each with their own contribution, has been one of the themes at the Heelsum workshops. The doctor has his/her medical knowledge, but the patient is his own expert on his habits and beliefs about foods. The GP also needs to be empowered. Some medical school lecturers have (hopefully in the past) taught that GPs should only treat minor diseases themselves and refer all interesting cases to specialists! In this section, other papers dealt with doctors' different communication styles, with the use of simplified tools for nutrition communication (eg, the different food group pyramids) and with the surprising rarity of reported nutritional deficiency in general practice.

The conference revisited the involvement of the computer and software in the primary physician's nutrition world. The Dutch College of GPs for example is developing software that will give the doctor prompts when to ask a nutrition question or record body weight (or BMI). Software is also being developed with dietary prescriptions that can be printed out, discussed and handed to the patient. On the other hand, patients have access to the internet. They can and do surf to find information about their symptoms, or disease and dietary recommendations. Medical information on the internet varies enormously in scientific reliability. Family physicians as well as their patients are going to have to be able to recommend the reliable web sites and discourage visiting the maverick, biased and potentially dangerous sites.

The individual–population interface: In earlier Heelsum workshops, management of obesity has had much discussion as a rather gloomy problem that doctors tended to avoid because it is much easier to write a drug prescription than try and persuade someone to eat less food. Our consensus was that the GP alone cannot eliminate obesity any more than (s)he can eliminate smoking. Society has to support the doctor's efforts. This is clearly happening now with smoking and it is starting to happen with overweight/obesity. Doctors should recognise and draw attention to increasing adiposity, early overweight, but if they have done this and the patient does not take the appropriate action the doctor should not feel a failure. In this 2004 workshop, the prospects for more action (and success) on overweight/obesity in general practice look much brighter. The social and political environment is changing. Obesity has become widely discussed as a general concern; hence, if a GP raises the subject with an individual and suggests some lifestyle changes, he now has society on his side. An exciting paper here is a report on Counterweight (Laws, 2005) a well-designed and large trial of obesity management in general practices in England and Scotland. The methods were evidence-based, 80 general practices participated and the results are both encouraging and plausible. Obesity can now be treated in general practice but it requires a team effort, with one of the partners interested in the problem, with an in-house practice nurse and with a dietitian as consultant associated part-time with the practice.

Disease-directed evidence and the packaging of available research: The final section of the workshop was a special session on Cochrane reviews and nutrition. GPs, like everyone else, would like evidence-based nutrition advice, but most of the concepts on nutrition and aetiology are not based on randomised controlled trials, the only evidence Cochrane deals with. Most of our theory of nutrition has to be built on animal experiments and observational epidemiology, etc. In our Cochrane session, a nutritionist presented two Cochrane reviews that were thought to conflict with the bulk of evidence and to be unhelpful for nutrition education. The other five speakers in this session describe what the Cochrane network is and what Cochrane reviewers do. The main goal should be to ensure that nutritionists are not left behind in the move towards the evidence-based approach. To reach that target, joining the well-established Cochrane Library seems to be of great importance. Areas of tension are the necessity of more weight to the pathophysiological background, the value of intermediary endpoints, a critical interpretation of observational studies rather than randomised controlled trials and, last but not least, the developing of general accepted templates, gold standards.

The participants discussed whether it would be useful to apply for and set up a so-called Diet and Nutrition Sub-Field within the existing Cochrane Primary Health Care Field.