5th international conference on dietary assessment methods. (Conference Report).

Expanding the horizon: dietary assessment in a multi-cultural world, Chang Rai, Thailand, 26-29 January 2003

This latest conference hosted by the Institute of Nutrition, Mahidol University, Thailand, followed those previously hosted by Arizona (2000), Wageninen Agricultural (1998), Harvard (1995) and Minnesota (1992) universities. The keynote speakers reflected the strength in nutritional epidemiology for which these meetings are renowned and included Walter Willett, Wija van Staveren, Susan Krebs-Smith and Sheila Bingham. Located in Thailand, the conference focused on dietary assessment in developing countries and provided a varied and highly stimulating program.

Pre-conference workshops were provided on measurement errors, disease associations, measuring infant feeding practices and self-administered dietary assessment tools for consumers. This last workshop generated an interesting discussion on the potential for self-administered methods in consumer nutrition education and on the role of the food industry in their development.

The conference was formerly opened by Her Royal Highness the Crown Princess of Thailand in a beautiful and inspiring ceremony. Introductory keynote speakers, Kraisid Tontisirin and Barry Popkin reported on FAO methodology for determining adequate nutrition and on in sights into current nutrition transitions respectively. In keeping with his publications in this area (1,2) Dr Popkin provided a wealth of data on shifts in global food consumption patterns (see also www.nutrans.org). This session lead to powerhouse presentations on population-based diet--disease analysis by Walter Willett from Harvard and Victor Kipnis from the US National Cancer Institute. These plenary sessions provided the backdrop to the debate on the Pros and Cons of FFQs' held at the end of the second day. In his lecture, Dr Willett stated that changes in diet and lifestyle could result in enormous reductions in the rate of coronary heart disease and type 2 diabetes mellitus. To this end, nutrition theory still relied mainly on cross-secti onal, cohort and, to a lesser extent, case control studies using population based survey methods. Noting that 'skilful analytic processes cannot replace strong data', Dr Willett also argued for the contentious issue of energy adjustment on the basis of its equivalence to isocaloric conditions in experimental studies.

On the other hand, Dr Kipnis contended that given reporting bias, adjusting for energy is an additive 'contaminating' move, which he demonstrated through sophisticated statistical reasoning. A vexing theoretical position on measurement error and the associated difficulty in detecting true disease risk emerged. This called for more precise validation studies of dietary assessment methods using the only currently recoverable biomarkers of intake, energy (via the doubly labelled water method) and protein (from urinary nitrogen). A number of presentations on the OPEN study using these approaches were presented throughout the conference, and the debate on FFQs remained unresolved.

Geoff Marks, from Australia, opened the second day with a review of nutrition survey systems throughout the world, and a reiteration of the goals of the conference to seek best methods and achieve harmonisation in data collection. Dr Marks presented a careful analysis of Australian dietary survey data from 1983 and 1995 to demonstrate that sample design and method differences are important considerations in making comparisons between surveys. He listed a number of key challenges for dietary survey methodology, such as closer integration between survey programs, the development of database structures, and accommodating changes in the food supply and shifts in the nutritional sciences. Advances were apparent albeit in the context of evolving research and development agendas. Dr Gail Harrison reiterated some of these views, noting that harmonisation had been achieved with the development of INFOODS, the European EFCOSUM project and many of the WHO initiatives. Underreporting remains a universal problem, with re search required on the impact of cultural aspects, pattern recall, dietary variation and portion size perceptions. Equally, studies of diet and physical activity are required, with both involving self-reporting of repetitive, variable behaviours. This will be a new focus for the future.

In between plenary sessions, participants could attend an array of smaller sessions with researchers from a range of areas, too many to cover here. The qualitative /ethnographic session reminded audiences of the need to understand study populations before developing methods. Dr Vongsvat Kosulwat presented on the unique identity of Thai food based on in-depth interviews with 20 Thai food experts. This identity could be classified through its remarkable appearance, taste, and variety, fresh spices and herbs and the 'family style' set meal. In converting this knowledge to dietary advice, identifying several dishes rather than food groups may be a more appropriate approach. Another way to categorise foods was presented in a Swedish study on the effects of migration where staple carbohydrate foods (bread, rice, pasta, bulgur) as the core culturally identifying food are the last to change, compared to the next level of complementary foods (fish, meat, eggs, cheese, fruit and vegetables) and finally accessories (dr inks, snacks), the most susceptible to outside influence.

In other sessions, Mary Anthony demonstrated a US web/CD-based computer assisted dietary assessment program (for a preview see http://stg.centrax.com.Diethistory_tutor), and Linda Tapsell and Gina Martin reported on dietary assessment in clinical practice and experimental research. The 2006 conference in Copenhagen will take up this new theme.

Overall, the 2003 conference was a real credit to the team from Mahidol University. It will certainly have its own special impact over time.

References

(1.) Popkin BM, Siega-Riz, AM, Haines PS, Jahns L. Where's the fat? Trends in US diets 1965-1996. Prev Med 200 1;32:245-54.

(2.) Popkin BM, Nielsen SJ Patterns and trends in food portion sizes 1977-1988. JAMA 2003;289:450-53.