This issue of the International Journal of Obesity contains an article by Elbel et al. and an associated commentary by Dr David Allison addressing the effectiveness, or otherwise, of public health interventions concerned with the labelling of nutrient/energy content of food provided in catering outlets. Such interventions, and other public health policies such as the regulation of advertising of foods or beverages to children, appear to be motivated by an understandable desire among local/national politicians and health activists to do something about the burgeoning problem of obesity, and especially childhood obesity. However, such public health initiatives should be based on evidence of likely efficacy, and this has not always been a major factor influencing such policy developments. The article by Elbel et al. illustrates how such an approach can lead to difficulties when an objective study fails to confirm the effectiveness of such a policy.

Dr Allison eloquently covers the issues surrounding this topic, and we will not re-iterate the points he makes. However, we are concerned that public health initiatives in relation to preventing and reversing obesity should be firmly based on scientific evidence. We are reminded that the motivation behind the landmark Swedish Obesity Surgery (SOS) Study was the request by the Swedish government's health department for evidence to be produced that weight loss in obesity was of benefit in terms of disease prevention and reduced mortality. The fact that the SOS study lasted 15 years before such data were forthcoming shows how long it sometimes takes to gather the evidence needed to be able to develop evidence-based healthcare or public health policy.

The use of diet and lifestyle to improve health and prevent disease is the cornerstone of public health in many countries. Such approaches are almost always based on sound nutritional and behavioural science, as they should be. However, before interventions such as menu labelling and advertising control are widely introduced in an attempt to supplement such policies, we feel that a clear evidence base of the effectiveness of these approaches is needed. This is likely to be a substantial challenge, as recent long-term diet and lifestyle interventions have failed to produce convincing evidence of their effectiveness as obesity treatments. Obtaining evidence concerning the benefits of interventions such as food labelling will not be easy, or cheap, but it is essential if robust, effective public health interventions are to be produced that will really make a difference to the problem of obesity.