Eating behaviour and obesity

Energy intake and energy expenditure are consequences of behaviours (e.g. choosing foods, eating foods, watching TV, playing sports) that are themselves influenced by a wide range of internal and external determinants. These include the characteristics of the food supply, the knowledge, attitudes, emotional state and experiences of the individual, and the social and cultural context in which the behaviour occurs. This brief review highlights psychological research into determinants of food intake. Studies in humans and animals show that energy intake is affected by characteristics of the food environment. In the typical experiment, participants are randomly allocated to groups that are offered different types of food or given food in different situations. Studies of this kind have shown that calorie consumption is higher when the meal consists of a variety of foods compared with a single food type (1), (see also contribution by Rolls, p. 67 (1)), when the food is more palatable (2), and when it is presented in more energy-dense formulations (3). Studies carried out over longer periods – in both animals and man – show that more palatable and energy-dense diets result in greater weight gain 4-6). As well as food composition per se, cues such as the sight or smell of food can stimulate appetite and promote higher consumption (7). Studies with animals demonstrate that they can learn to associate non-food cues with eating (for example, the conditioning seen in Pavlov’s dogs). Exposure to the conditioned stimuli then elicits salivation and changes in blood glucose levels, and can even initiate food intake in sated animals (8). Responsiveness to cues for eating has been hypothesized to prepare the body for food intake, allowing more rapid digestion when the food is consumed. Food preferences also change as a consequence of experience and learning. The taste of foods eaten just before being sick often becomes disliked, while tastes that are associated with good consequences, such as a ready supply of calories, tend to be more liked (9). This is probably the basis of the ‘familiarity’ effect, in which foods that have been eaten before – and whose consequences proved to be favourable – tend to be preferred, and this is strongly the case in children. There is now evidence that systematic exposure (including flavours transmitted through breast milk) can increase familiarity and promote acceptance of hitherto disliked vegetables (10, 11). These observations make sense in the historical context: survival in the face of food scarcity is likely to have been enhanced by behavioural repertoires that maximize food consumption when food is plentiful, and allow selection of safe and energy-dense foods (12). One environmental characteristic that has attracted a good deal of interest is stress. However, no simple relationship with eating behaviour has emerged. In animal studies, stress has been shown to cause both hyperphagia (over-eating) and hypophagia (under-eating), with some evidence that milder stresses enhance food intake, while more severe stresses depress it. Human research has produced equally varied findings, although most people show a tendency to increase intake of sweet and fatty foods relatively more than protein under stress, which may be part of a coping response (13). In wealthy societies, systems of food production, storage and distribution have virtually obliterated food scarcities and created an increasingly attractive, diverse and energy-dense food supply. Food is widely available, and promotion and advertising provide additional exposure to food cues. The cost of food, which might otherwise be a barrier to consumption, is at a record low level, with the lowest unit costs being for processed, energy-dense foods served in large portions. In parallel with the transformation of the food supply, social norms related to eating have changed. Grazing, snacking, eating on the go and eating outside of the home are common and contribute a substantial proportion of calories. Children are given more control over food choices. Portion sizes have increased for meals and snacks, with growing evidence that people eat more when presented with larger portions (3). The modern food environment therefore strikingly resembles the conditions that have been shown in the laboratory to increase calorie intake (14). Although obesity is often discussed as a distinct category, adiposity is a continuously distributed trait and it is likely that differences between individuals in how they respond to the food environment contribute to this. A series of experiments in the 1960s demonstrated obese humans (and rats) up-regulate their intake when offered more palatable foods but also fail to down-regulate after a preload (extra food given before the test meal) (15). These findings suggested that part of the cause of obesity could be higher responsiveness to ‘external’ food cues and lower responsiveness to ‘internal’ satiety cues. Findings with children have supported this hypothesis. Obese children’s intake at a test meal was less depressed by a preload and more stimulated by exposure to the smell of food than that of normal-weight children, and externally responsive children gained more weight over the duration of a school summer camp (16, 17). Recent work has extended the prospects for carrying out large-scale research into appetite and weight gain in children with the development of a reliable, valid psychometric measure of children’s eating styles, the Child Eating Behaviour Questionnaire (18). Importantly, there is also evidence that differences in appetitive responsiveness are heritable (19, 20). Therefore, weight gain is likely to be the result of an interaction between genetically determined individual responsiveness to food and eating and the environment to which individuals are exposed. There has been great interest in the extent to which restrained eating (deliberate control) is a successful strategy. Early studies showed that restrained eaters had stronger appetitive responses to palatability and lower responses to internal satiety signals (21). These surprising results were attributed to the counterproductiveness of attempts to override ‘natural’ signals for hunger, which were hypothesized to lead to insensitivity to internal hunger cues. However, recent studies have used research designs that are better able to tease out cause and effect by studying the effects of deliberately increasing dietary restraint as part of a programme of weight control. Rather than causing loss of control or higher binge-eating, restraint is associated with reductions in binge-eating (22). Restraint has also been associated with better eating habits, and, when it is sustained, with better weight control (23). Nevertheless, the fact that so many restrained eaters fail to control their weight indicates the need for more research into how ordinary people, without professional help, can exert effective control. The implication of behavioural research is that changes in the food environment have stimulated food intake beyond that required to match energy expenditure, and consequently people’s weights have gone up. In theory, this process could be reversed. Behavioural research into eating is currently restricted to a small number of psychology laboratories, although there is probably similar work in food industry laboratories to identify the compositions and presentations of foods that will maximize consumption. Researchers need to identify food compositions that result in better control (e.g. higher bulk, lower energy density and perhaps more limited availability) and model and test food environments that are hypothesized to result in better weight control. Closer integration of basic behavioural research with developments in the food industries should be able to achieve new approaches to managing successful food supplies with success in weight control. Attempts to halt population weight increases by imploring or badgering people to be more prudent in their food consumption, without modifying the environment, are likely to have limited impact (24). Nevertheless, there is an urgent need for research into the nature and content of public education that would support healthy eating. Given the success that has been achieved in controlling the smoking epidemic through a combination of health education, provision of clinical services, legislation and fiscal policies, there is no reason to think that we should not have equal success with food. But industry will need to engage with a health agenda that promotes less, rather than more, consumption. Because obesity confers substantial health risks, weight-gain prevention is a priority for health and prosperity, and commercial and academic scientists need to work together to come up with new ideas. Research into individual differences indicates that people who are already obese are not lacking in will power but burdened with responsive appetite systems, and this is probably largely genetic. Until the environment changes, not only will overall population weights continue to rise, but the greatest increase will be in the severely obese. This can only increase the urgency to identify effective treatments. Behavioural treatments that focus on helping people to control exposure to cues that trigger eating and to monitor and modify their behaviour have had consistent but modest success and are recognized to be most effective, apart from the drastic option of bariatric surgery (25). Drug treatments have not proved to be the promised panacea, but it is likely that treatments of the future are going to come from combining behavioural and pharmaceutical interventions, probably also taking into account genetic and behavioural differences. No conflict of interest was declared.