Periodicity of eating and human health: present perspective and future directions

Population dietary advice needs to be firmly rooted in cultural context. A very important part of that cultural context is the variation in food-consumption patterns evident within and between different populations. It follows that a knowledge of the factors influencing patterns of food consumption and of the impact of different patterns of food consumption on health is an essential element in devising culturally-appropriate dietary advice. The present workshop has, for the first time ever, brought together the many disciplines which study these factors: applied nutrition and dietetics, experimental nutrition, chronobiology, physiology, behavioural genetics, psychology, sociology, anthropology, clinical medicine and dentistry. Its achievement is in outlining our present state of knowledge and indicating how the diverse disciplines involved should interact in future research to fill the gaps in our present knowledge. It is evident from the papers presented and from the extensive debate that there is a deficiency in quantifying periodicity of eating. It should be noted that the title of the workshop excludes any value-laden terms such as ‘meals’ or ‘snacks’. Quantifying periodicity of eating in such terms will inevitably occur, but because such terms vary in their meaning between individuals, let alone between different cultures, their use should be limited in future research. Ingestive behaviour includes ingestion of foods, beverages and micronutrients. The definition of an eating occasion may vary in different studies, depending on the purpose of the investigation. However, it includes consideration of the minimal time span between intakes and the minimum level of intake. Thus, in future research on the periodicity of eating, an objective definition of what constitutes an eating occasion is needed. For example, for the study of energy intake an eating occasion may be defined as an event which provides at least 210kJ with a separation in time from a preceding or follow eating event of at least 15 min. Another problem in quantifying periodicity of eating is one shared by all studies which seek to measure food and nutrient intake, namely inaccurate reporting. Given the wellestablished trend of the overweight to be particularly likely to under-report, much of the older cross-sectional epidemiological data, elegant for its day, linking hypercholesterolaemia, CHD, diminished glucose tolerance and obesity to reduced periodicity of eating, must be questioned. This is especially relevant to work on energy balance, given the evidence presented at the present workshop that the overweight are biased in underreporting of what are commonly referred to as ‘snacks’. These methodological difficulties, therefore, necessitate cautious interpretation of the contribution of ‘snacks’ to nutrient intake, although on balance there are sufficient robust data to conclude that significant proportions of macronutrients and micronutrients are so derived. A final point to be made in this context is that notwithstanding the difficulty of measuring periodicity of eating, the value of the data obtained should not be underestimated. There is an abundance of data on average daily intakes of nutrients but comparatively little knowledge regarding individual temporal distribution or of the interrelationships of such distributions between nutrients. Such knowledge would greatly help our understanding of the metabolic and behavioural effects of different periodicities of