AQA A-level Psychology Eating Behaviour

The A-level psychology specification states you need to know the following for eating behaviour and the explanations for food preferences:

• The evolutionary explanation for food preferences including reference to neophobia and taste aversion
• The role of learning in food preference, including social and cultural influences.

Evolutionary explanations propose an adaptive function for a particular behaviour that has been shaped by natural selection over millions of years. The early diets of early humans that were hunter-gatherers would have included animals and plants that were part of their natural environment. The preferences for fatty food would have been adaptive because conditions in the EEA (environment of evolutionary adaptation) would have meant that energy resources were crucial in order to stay alive and find the next meal.

Fossil evidence suggests a preference for meat by early hunter-gatherers derived primarily from animal-based foods. Some theories suggest a meat diet which is densely packed with nutrients provided the catalyst for the growth of the human brain. Milton (2008) claims that without animals, it is unlikely that early humans would have gained enough nutrition from a vegetarian diet to evolve into the intelligent creatures we have become.

The human preference for sweet foods has been associated with a high concentration of quickly available sugar in readily available calories. In the EEA, a source of sugar in calories would therefore be ripe fruit which is subsequently characterised by its sweet taste. Fruits also provide vitamins and minerals which are necessary for bodily functions and growth and it would therefore have been adaptive for early humans to have evolved an innate preference for sweet foods. Mennella (2014) found children who preferred sweet solutions over salty ones tended to be taller for the age. This would make sense from an evolutionary perspective as within the EEA, children that sought more calories were more likely to grow and thus survive.

Taste aversion is a learned response to eating food that is toxic, spoiled or poisonous. Taste aversion results in the animal avoiding the food that made it ill or they associate with the illness. It's discovery was made by farmers that were trying to rid themselves of rats. The farmers found that they were difficult to kill using poisonous bait as the rats would only eat a small amount, become ill and then rapidly learn to avoid such.

Garcia et al. (1955) were the first to research taste aversion finding rats that were made ill through radiation shortly after eating saccharin subsequently developed an aversion to it and quickly associated the illness with eating the saccharin.

In addition to taste aversion, researchers also found that the odour a food can be linked to illness and consequently to the development of a food aversion. From an evolutionary perspective, the development of taster versions would have helped our early ancestors to survive because if they did eat poisonous food and survive, they would not repeat the same mistake. Once taste aversions have been learnt, they are difficult to shift demonstrating an adaptive quality for survival.

Neophobia, or food neophobia, is the reluctance to consume new or unusual foods and is a naturally occurring reaction which protects animals from the risk of being poisoned by consuming something that may be potentially harmful. This response is believed to be an important survival strategy when animals are presented with potential new foods whose safety is uncertain (Prescott, 2013).

Animals that have specialised diets which are restricted to just a few specific food sources tend not to exhibit food near phobia, where are species have a wide range of diet do displayed food neophobia (Ratcliffe et al. 2003). Rozin (1976) found rats were extremely neophobic and if they became ill after eating both are familiar food and an unfamiliar food, they will then avoid the unfamiliar food in future.

Neophobia in humans results in an individuals reluctance to eat new or unusual food, based on their culture and current diet (Rozin, 1997). Individuals will have expectations of what septa bull food should look and smell like, and therefore unfamiliar food that does not fall within this perception will be rejected (Dovey et al. 2008). Food neophobia has been found to be particularly strong in response to animal meat or animal products compared to non-animal products (Martin's et al, 1997). This is thought to have evolve due to the greater risk of illness from rotting meat and other animal products compared to non-animal products (Fessler, 2002).

• While many food preferences can be argued to have evolved due to the adaptive pressures of the environment of evolutionary adaptation (EEA), some aspects of food preference appear maladaptive for survival. A trait that is beneficial today such as the consumption of low cholesterol foods would not have evolved because of its beneficial effects for our ancestors. Many aspects of food such as saturated animal fats are nowadays harmful for our health and we are more likely to avoid them to survive have a healthier lifestyle. Krebs (2009) highlights a 'mismatch' between our evolved preferences (e.g. for fatty and sweet foods) and modern environments. This mismatch is proposed to contribute to many of the major global health issues which have emerged such as obesity and type 2 diabetes.

  ---

• There is research support for evolved preferences for sweet foods. Research by Bell et al (1973) has shown that early exposure to a sweet taste is not necessary for children to develop a preference for such foods. The Inupiat people of Northern Alaska that had no experience of sweet foods have been shown to develop a preference for such despite no previous exposure. Across the research, no culture without previous exposure to sweet food and drink has rejected such when exposed to it suggesting an innate response.

  ---

• Newborn infants have been shown to display an acceptance response when tasting something sweet for the first time suggesting an innate response. This response involves smiling, licking of the upper lip and sucking on which suggests that any substance which elicits this response will tend to be eaten (Grill and Norgren, 1978).

  ---

• There is real-world applications in understanding how taste aversion occurs. Understanding the origins of taste aversion has been helpful in understanding how food avoidance can occur during the treatment of cancer. Some cancer treatments such as chemotherapy and radiation can cause gastrointestinal illness. When this is paired with food consumption, taste aversions can result. Bernstein and Webster (1980) gave patients receiving chemotherapy a novel-tasting ice cream prior to their treatment and found they subsequently acquired an aversion to the ice cream. Such findings have resulted in the development of the 'scapegoat technique' which involves giving cancer patients a novel food along with some familiar food prior to the chemotherapy treatment. Patients have been shown to form an aversion to the novel food and not to the familiar food and this is consistent with an adaptive avoidance of unfamiliar foods known as neophobia.

The specification states it wants you to know about the role of learning in social and cultural influences in food preferences.

While the previous section proposed our food preferences and aversions where innate, this section proposes that our preferences may be acquired through experiences, and there are several ways in which they can be learned.

Social learning theory (Bandura 1961) explains social influence in terms of modelling and imitation and the same concept is thought to be relevant with food preferences. Children are thought to acquire their food preferences through observing role models such as their parents and their choices. The parents may also manipulate the availability of certain foods, whether this is a reward such as sweets or a take-away, or because certain foods are seen as having health benefits such as reduced fat or sugar. Children may also be influenced by role models if they appear to be rewarded from consuming the food (by showing enjoyment or being praised by others). This modelling may be an adaptive function to ensure children eat food that is safe as without this modelling , toddlers may attempt to eat potentially dangerous foods (Shutts et al. 2013).

Family influences are obvious within the family setting and in childhood as parents food directly affect what children eat as they are responsible for food preparation and choice. The preferences of parents to prepare certain meals will inevitably expose their children to the same foods. Peer influences are also thought to influence eating habits within children as Birch (1980) demonstrated. When placing a child at school lunchtimes next to 3 or 4 other children who had different vegetable preferences from them, after four days the participant child had changed their food preferences in response to observing other childrens food choices.

Media influences can also shape food preferences. As children get older and more independent of their parents food choices, other models outside of family and peers may become more important. Television advertising is one such example in with various adverts for foods that are generally considered 'unhealthy'. Such adverts often have fun related themes and products promoted by models that viewers can identify with.

Rozin (1984) suggests cultural influences are the single most reliable predictor of food preference, particularly family eating patterns. Around the table we learn when, what and how much to eat (Hansen et al. 2014). A persons culture will determine to a large extent the foods that is put on the table and what foods the children are exposed to. Cultural norms or ideals also exert influence on food preferences. An example of this is what is considered 'a proper meal'. For example, a common ideal in British households is that the main Sunday meal has to be a roast dinner. Many cultures dictate a view on meat and its consumption with some advocating vegetarianism or avoiding pork. In countries like Britain and France, eating offal (kidneys, liver, heart etc) is commonly accepted however in other countries such as the USA, there is a strong aversion to this.

The context of meals has changed significantly in some countries. In societies such as the UK, grazing rather than eating meals is prevalent as well as the desire for convenience foods such as takeaway meals. Maguire et al. (2015) found the number of takeaway restaurants in the UK has risen by 45% in the last 18 years, with areas of the highest deprivation seeing the highest increases. Gilman (2000) highlighted the decline of the family meals in western cultures, with more young people opting to eat while watching TV and this being associated with more salty, unhealthy foods. 

There are limitations to the role parental influence has on food preferences and, mostly because research into this area is limited. Studies have generally been small-scale and carried out on highly selective samples of white Americans which makes it difficult to generalise the findings to the populations. Contradicting research comes from Robinson et al. (2001) where almost 800 eight to nine year olds from different backgrounds were studied. This study found a complex association between the behaviour of parents and the food preferences of their children, with girls being more influenced by parental modelling and control compared to boys.

Other research studies have shown that the foods parents consume and make available to children are good predictors for the food children consume. For example, Wardle et al. (2005) found that fruit and vegetable consumption by parents is a strong predictor of the children's fruit and vegetable consumption. Jansen and Tenney, 2001 found seeing significant others modelling and eating of sugar-free yoghurts and drinks led to a preference for them in primary-age children. An increase in at the same behaviour in the presence of others (Known as social facilitation) likely serves as a means to ensure food is consumed have been demonstrated to be safe by others.

Research suggests social influences into food preferences vary in short-term and long-term effects. Family influences on food may potentially last a lifetime, however the effects of television programmes and advertisements are much less persistent. Helle Hare-Bruun et al. 2011) studied a group of 8-10 year old Danish boys and girls over a period of six years. Researchers found that the children who watched the most TV also tended to have the most unhealthy diet. The link was much weaker in the researchers six-year follow-up and had disappeared completely for girls. The conclusions drawn were that the effects on food Preferences from television are mostly short-term and as children get older, close friends and peers exert a more powerful social influence on children's long-term food choices.

There is research support for cultural influences on food preferences too and this is evident in western societies. One of the biggest cultural changes has seen the increasing availability of food outside the home, accompanied by a decline in cooking and family meal times together. Approximately 46% of food spend in the USA goes towards food that is eaten in places other than the home. American adolescents eat up to 30% of the meals outside of their homes, half of which are from fast food restaurants.

Cultural eating practices often reflect local environmental conditions such as the availability of different types of food. Research by Chen and Yang (2014) provides evidence for the role of cultural influences on food preferences and a local food environment. They studied Twitter tweets made over five weekends in Columbus, Ohio. Tweets were analysed for evidence of food activities such as shopping in high quality grocery stores or eating at fast-food outlets. The quality of food choices were also examined and results found a significant association between healthy food choices and a number of available grocery stores around them. Contrary to expectations, no association was found between the number of fast-food outlets and healthy unhealthy food choices. They concluded that cultural influences do you have an affect on learned food preferences, but people are able to resist development of unhealthy foods if a healthy alternative choice is available.

• For eating behaviour, you will need to know about the Neural and hormonal mechanisms involved in the control of eating behaviour, including the role of the hypothalamus, ghrelin and leptin.

The hypothalamus is a pea-sized structure within the brain that has a crucial role in integrating the nervous and endocrine systems. It is also involved in homeostasis, the mechanism by which an organism maintains a steady internal environment via balancing bodily processes within certain limits. This neural mechanism regulates the level of glucose in the blood, which is the most basic of sugars and the bodies main source of energy and helps detect whether the body has enough nutrients and correcting this when required.

The body has evolved two systems for achieving this, one of which turns eating 'on' and one for turning it 'off'. For humans, glucose levels likely play a significant role in producing feelings of hunger as hunger increases as glucose levels decrease. As glucose levels in the blood decline, this activates a part of the brain called the lateral hypothalamus which results in the feelings of hunger. This leads the individual  to search for and consume food which increases glucose levels once again. The increase in glucose levels activates the ventromedial hypothalamus, which causes feelings of satiation (feeling full) and inhibits further eating.

The lateral hypothalamus, when stimulated has been found to elicit feeding behaviour. Researchers have discovered that damage to the lateral hypothalamus in rats cause a condition called aphagia (Greek for 'absence of eating'). A neurotransmitter found in the hypothalamus, called neuropeptide Y (NPY), has been found to be important in 'turning on' eating behaviour. When neuropeptide Y is injected into the lateral hypothalamus of rats, it causes them to immediately begin feeding, even when full (Reynolds and Wickens, 2000). Repeated injections of neuropeptide Y into the hypothalamus produced obesity in just a few days (Stanley et al. 1986)

Research into the ventromedial hypothalamus has found that when damaged, this caused rats to overeat. Stimulation of this area inhibited feeding which led researchers to conclude that the ventromedial hypothalamus is responsible for the signal to stop eating based on the many glucose receptors in this area.

This mechanism became known as dual control theory.

Ghrelin is a hormone secreted by the stomach and acts as a hormonal marker of how long since we have last eaten as the amount produced is closely linked to how empty our stomach is. The longer we go without eating, the more Ghrelin is released which is then detected by receptors in a part of the hypothalamus called the arcuate nucleus. The arcuate nucleus then sends signals to the lateral hypothalamus to secrete neuropeptide Y.

In humans, Ghrelin is known to be an appetite stimulant and research has found that when given intravenously, it caused a short-term increase in the amount of food eaten. The amount of Ghrelin circulating in the bloodstream before a meal doubles and decreases very quickly after meals. It is also closely correlated with subjective feelings of hunger.

Leptin is a hormone that plays a crucial role in appetite and weight control that is normally produced by adipose fat cells and secreted into the bloodstream. Within the bloodstream, it travels to the brain (venture-medial hypothalamus) and decreases appetite with circulating leptin levels acting as a long-term signal on the amount of fat stored in adipose tissue. Short-term fluctuations in leptin levels provide information regarding changes in calorie intake.

Leptin is believed to have two major functions: by binding to receptors in the hypothalamus, it counteract the effects of neuropeptide Y which is responsible for turning feed 'on' and secreted in the gut and hypothalamus. Secondly, leptin increases sympathetic nervous system activity, which stimulates fat tissue to begin burning energy.

• There are limitations to dual control theory, and the homoeostatic explanation. For a hunger mechanism to be adaptive, it must be able to anticipate and prevent energy deficits rather than simply react to them. Therefore, the theory that hunger and eating are only triggered when energy resources fall below their desired level is not compatible with realit where such systems would have evolved. For a mechanism to be truly adaptive, it must be able to promote levels of food consumption that maintains Bali resources above the optimal level to act as a buffer against the lack of food availability in the future. Therefore explanations of food intake based solely on the homeostatic mechanism offers a limited explanation on eating behaviour.

  ---

• The idea that the lateral hypothalamus served as the on switch for eating behaviour has encountered numerous problems. For example, when the lateral hypothalamus is damaged, this caused deficits in other aspects of behaviour such as thirst and sex rather than just hunger. More recent research has also shown that eating behaviour is controlled by neural circuits which run throughout the brain, not just by the hypothalamus. Although it is accepted that the lateral hypothalamus undoubtedly plays a crucial role in controlling eating behaviour, it does not play the central role (Sakurai et al. 1998).

  ---

• Much of the research studies into neural and hormonal mechanisms for eating behaviour comes from non-human animal research studies, particularly rats. The biology between humans and rats is significantly different and therefore it is difficult to generalise the findings between species as the same effects may not apply.

  ---

• There are real world practical applications available from better understanding the neural and hormonal mechanisms controlling eating behaviour. Example, this offers therapeutic possibilities for both obesity and anorexia as we better understand the complex interactions between the nervous system and the hormones regulating eating. Research by Licinio et al (2004) is a good example of this as his research into an extremely rare genetic condition examined people who were unable to produce leptin naturally. This abnormality is associated with severe obesity and treatment that involved leptin–replacement therapy over an 18 month period, found that this led to an average weight loss of more than 40% and reduction in food tech intake of almost half.

Biological explanations for anorexia nervosa focuses on genetic explanations and neural explanations.

Genetic explanations focus primarily on genes believed to be involved while neural explanations focus on the neurotransmitters serotonin and dopamine.

Genetic explanations for anorexia nervosa (AN) have been explored using family studies, twin studies and adoption studies in an attempt to identify a gene or specific genes involved in the disorder.

Evidence supporting the genetic explanation for anorexia nervosa comes from twin studies where genetically identical twins have been found to have higher concordance rates. Non-identical twins share only 50% of their DNA with each other and comparing them to genetically identical twins that share 100% of their DNA provides a measure of the relative contributions between genetic and environmental factors that may contribute in the development of anorexia nervosa.

Twin studies have suggested a moderate to high heritability of AN with estimate varying between 28% and 74% (Thornton et al. 2010). Research by Wade et al. (2000) interviewed over 2000 female identical and non-identical twins and evaluating them using the DSM criteria for AN, found a heritability rate of 58%. This suggests that some people are more genetically predisposed to develop anorexia nervosa than others.

Family studies have shown that eating disorders also run in families. The first-degree relatives (parent, sibling or child) of individuals with anorexia nervosa are approximately 10 times more likely of developing the order themselves compared to relatives of non-affected individuals (Strober et al. 2000). Relatives of those diagnosed with AN have also been found to be at an increased risk of developing other eating disorders such as bulimia nervosa. This suggests that people may be inheriting a more general vulnerability to eating disorders rather than specifically anorexia nervosa (Tozzi et al. 2005)

Adoption Studies are used because twin studies suffer from the problem that not only do the identical twins share the same genetics, but also the same environment which can make it difficult to disentangle the influence between genes or the environment. Adoption studies avoid this issue because they allow researchers to study biological relatives whom do not share the same environment. Klump et al. (2009) studied 123 adopted sibling pairs and 56 biological sibling pairs for disordered eating symptoms instead of AN due to it's low prevalence in the sample. Results found heritability estimates ranged from 59% to 82% for different aspects of disordered eating. Non-shared environmental factors i.e. experiences that were different for each sibling pair) accounted for the remaining variance which demonstrated a significant environmental element involved too.

• There are problems with the genetic explanation for anorexia nervosa, for example Fairburn et al (1999) highlighted that despite a number of studies supporting a genetic component for anorexia nervosa, the actual heritability of the disorder is still unknown and no conclusive genes I've been identified. In addition, studies have given a widely contrasting range of hereditary estimates as well as ignoring the fact that identical and non-identical twins may not be raised in the same environments. Some research has suggested that identical twins tend to be treated more similarly than non-identical twins and this may invalidate the claim that the greater concordance rate for a AN in identical twins compared to non-identical is due to greater genetic similarity.

  ---

• Another criticism of genetic explanations of AN, is the assumption that environmental factors such as the media are not important. The media have for a long time projected idealistic body types and body images, for example an ultra-thin female body type which has been viewed as an important risk factor for eating disorders. It may be that genetically vulnerable individuals may seek out search media representations of thin role models as a means to reinforce their body image (Bulik 2004). This argument is supported bye a longitudinal study found adolescent girls who is anorexia nervosa symptoms severely increased over a 16 month period when they displayed significantly greater fashion magazine reading over the same time period (Vaughn and Fouts, 2003).

  ---

• If genetics were solely responsible for anorexia nervosa, concordance rates between identical twins would be 100% however this is not the case. This would confirm that other non-genetic factors are involved and it is possible this occurs through an interaction between predisposed genetics and the right environmental triggers activating the disorder. In addition, genetic explanations cannot explain why the disorder is prevalent in more females or in large numbers of homosexual males. No explanation is offered for the heightened incidence of the disorder in modern society in comparison to previous generations where inheritance of genes has not greatly changed.

Neural explanations for anorexia nervosa have examined the role of the neurotransmitters serotonin and dopamine.

Bailer et al (2007) examined serotonin disturbances between individuals with anorexia nervosa and binge-eating/purge type behaviours and compared them with a healthy control group. The highest serotonin activity was apparent within anorexia nervosa sufferers and this is believed to suppress appetite and increase anxiety which are characteristics of the disorder. Therefore higher than normal levels of serotonin activity may trigger anxiety and explain anorexia nervosa.

Another biological explanation into neurotransmitters looks at dopamine overactivity. Kaye et al. (2005) compared 10 recovering anorexia nervosa sufferers to 12 healthy women using pet scans to measure dopamine activity within the brain. The AN women were found to have overactivity within their dopamine receptors within the basal ganglia which is associated with interpreting pleasure and harm. This part of the brain may affect the way people interpret rewards or pleasurable activities such as food and eating. Therefore higher than normal dopamine levels may explain AN as sufferers find it difficult to associate positive feelings with pleasurable acts such as the consumption of food.

Other neural explanations have considered the idea that anorexia nervosa is linked to defective brain structures. Early research examined the possibility of damage to the hypothalamus however recent interest has considered the insula dysfunction hypothesis as the root cause. The insula brain area is part of the cerebral cortex and this theory proposes that it develops different in anorexics. Various symptoms of AN are associated with dysfunction in several brain areas, with the common link being the insula, which is responsible for the most neural connections than any other area of the brain, including areas associated with anorexia nervosa. Lipman et al. (2015) suggests as the normal function of the insular cortex is to regulate emotion, this may lead to deficits in emotional processes which may lead to the pathological thoughts and behaviours that are typical of individuals with anorexia nervosa.

• A problem for explanations that suggest anorexia nervosa is caused by serotonin imbalance is that SSRIs, which are drugs used to alter brain serotonin levels, have been found to be ineffective when used with patients. If serotonin was involved, we would expect to see changes to AN however this is not the case. Ferguson et al. (1999) found no difference in symptom outcomes between patients that took SSRIs and matched control groups with AN patients not taking them. Kaye et al. (2001) found that SSRIs were helpful  with recovering AN patients in preventing relapse. It may be that malnutrition-related changes in serotonin levels negate the effects of SSRIs and they are only effective when weight is at normal levels.

  ---

• There is research support for dopamine affecting anorexia nervosa. For example, a relation between dopamine and food aversion, weight loss, menstrual dysfunction and distorted body image cognitions have been linked to increased activity in dopamine pathways (Kaye, 2008). Barbato et al. (2006) found increased eye-blink activity in AN patients when compared to a control group. Higher levels of blinking is associated with higher levels of dopamine activity in the brain and this study also found a significant correlation between blink rate and the duration of anorexia nervosa. This suggests a relationship between dopamine activity and AN symptoms may develop over time.

  ---

• A weakness for neurotransmitter related explanations for AN are they are based on correlational findings where we cannot infer cause and effect for certain. Serotonin or Dopamine increase may in fact be a symptom of AN rather than a cause and such explanations and studies lack internal validity as we cannot conclusively say either neurotransmitter is the cause for certain but merely present in individuals with AN. Also with Bailer et als study, the women were already recovering from AN and its impossible to establish whether any serotonin imbalance observed was the cause of AN or a symptom.

  ---

• Kaye et als study into dopamine suffers from methodological problems such as gender bias and the fact that it was a small sample. The study focused only on women and we cannot say for certain imbalances in dopamine would be the same for men with AN. Also, the sample size was incredibly small meaning we may struggle to generalise the findings to the wider population. For these reasons the studies lack external validity and it may be that they only offer an explanation for AN for women which fall within a small sample rather than everyone with the disorder.

  ---

• Biological explanations offer the possibility for treatments through drug therapies to bring neurotransmitter levels into balance or even gene replacement therapy opening up possible routes for treating the disorder. Research into biological causes also has real world applications especially in the world of insurance cover for psychiatric conditions. Within the US, AN treatment was restricted as it was not recognised as having a biological cause however research into biological explanations may lead insurance companies to question this assumption in much the same way schizophrenia is accepted as being more biologically determined. This would help open up treatment and offer cover for possibly millions highlighting the research value for people.

For Psychological explanations for anorexia nervosa you will need to know the following:

• Family systems theory, including enmeshment, autonomy and control
• Social learning theory, including modelling, reinforcement and media
• Cognitive theory, including distortions and irrational beliefs.

Family systems theory Minuchin (1979) suggests individuals cannot be understood in isolation but rather as part of the family unit. Family systems theory argues that to understand the family system we must look at the family as a whole, not just at its individual members. When people have studied individually, the way they interact and communicate within the family unit is unclear.

Family systems theory, also referred to as the psychosomatic family model, suggests the prerequisite for the development of anorexia nervosa was a dysfunctional family unit occurring alongside physiological vulnerability in the child. Minuchin et al. argued that the psychosomatic family process was a necessary context for the development of anorexia nervosa and therefore treatment must be aimed to change the way the family functions.

Minuchin et al. (1978)  identified four main features for what they coined a typical anorexic family. As anorexia nervosa tends to affect more females than males, family systems theory has focused on the relationship between the mother and daughter when applied to anorexia nervosa:

• Enmeshment

Enmeshment refers to an extreme form of proximity and intensity in family interactions. Members of anorexic families are overly involved with one another with a lack of defined boundaries, poorly defined family roles and a lack of leadership. Family members may spend a lot of time together, to the exclusion of others, which may constantly impinge upon one another's privacy. They may also speak for each other on the assumption that they know what the others thinking or what they views would be. Barber and Buehler (1996) suggested that enmeshment stifled the development of children's skills in adequately dealing with common social stressors and therefore makes the development of eating disorders such as anorexia nervosa more likely.

• Autonomy & Control

As part of family systems theory, Minuchin et al. argued that the families that exhibited these features where they were overprotective, inflexible, overly involved and conflict avoidant were actively preventing members of the family from exercising autonomy and control. Enmeshed families place great constraints on its members that prevent them from becoming independent and exercising autonomy. In non-enmeshed families, as an individual reaches adolescence the family generally decreases their control and allows greater age-appropriate autonomy. Within enmeshed families, this does not occur and the family retains accustomed patterns of control preventing any change in the family dynamic. 

This view has been supported by psychoanalyst Hilde Bruch (1978) who suggested that anorexia was caused by the adolescent daughters struggle to achieve the autonomy and control she craved. The mother is seen as domineering, intrusive, and discouraging of separation while not accepting the daughters need for independence. Plus the outcome of this dynamic is confusion in the daughter which is expressed through three major symptoms of anorexia nervosa:

1. Distorted body image
2. Inability to identify internal body states such as hunger
3. An overwhelming feeling of a loss of control

Self starvation that is central to anorexia is, according to Bruch, a desperate attempt by the daughter to control her self-identity and gain some independence from the family, particularly the mother. This weight loss that follows in anorexia is therefore a means by the daughter to rebel and control her own body and seen as a visible measure of her success. The thinner the daughter becomes, the greater her degree of control is reinforced. Autonomy is also gained by the daughter by disrupting her dependent relationship with her mother.

• The concept of enmeshment is supported by some research findings. Manzi et al. (2006) demonstrated a distinction between family factors that promote positive emotional development and those that hinder it. Family cohesion, for example, was indicative of supportive interactions while enmeshment was rooted in control and manipulation. Manzi et al. found cohesion among family members was linked to positive outcomes and psychological well-being among adolescents. In contrast, enmeshment was seen to have the opposite effect and there was an inverse relationship between enmeshment and the same positive outcomes. These findings were in line with Minuchin et al's predictions and applied across different cultural groups.

  ---

• Research has found heightened tensions within families of anorexics however rather than being a precipitating trigger for the development of anorexia nervosa, these heightened tensions could easily be the result of having an anorexic within the family unit. The research by Manzi et al. above may simply be an example of this as it is difficult to establish cause and effect between two variables. For example, an overbearing or controlling parent may be the result of someone suffering from anorexia and their attempt to look after their child that is suffering from this disorder.

  ---

• There are problems with the psychosomatic family model that undermine this explanation. Research has tried without success to establish the characteristics within families of anorexia nervosa patients and subsequently make predictions. This has produced generally disappointing and inconsistent findings and there is growing evidence that families in which someone has an eating disorder are a diverse group in terms of the nature of family relationship, the emotional climate and patterns of family interaction (Kog and Vandereycken, 1989).

  ---

• Family-based therapies have also produced mixed results. Family-focused therapies have shown that families can play a key part in the recovery process and Carr (2009) concluded there is compelling evidence for the effectiveness of family interventions for adolescent sufferers. However, critics point out that well there is some evidence of family therapy accompanied by changes in family functioning are beneficial, these changes are not those predicted by the psychosomatic family model and may not occur in all families (Eisler et al, 2009).

  ---

• Another issue is the obvious gender bias in family systems theory. Gremillion (2003) highlighted the approach focused almost exclusively on the mother-daughter relationship with very little acknowledgement or explanation for how males may develop eating disorders such as anorexia nervosa. Intervention, in response, tends to focus on reforming 'dysfunctional' mothers rather than acknowledging the role played by fathers. Gremillion highlighted that fathers also contribute to the enmeshment process and their tendency to be overly controlling or demanding is often overlooked in the development of symptoms.

Social learning theory has also been applied to offer an explanation for anorexia nervosa.

Bandura's Social Learning Theory (1977) proposes that people learn by observing the behaviours of others, as well as observing the outcomes of those behaviours. This theory highlights that children will pay particular attention to significant role models in their life which may be parents, siblings, peers or those close within their social circle. If the attitudes and behaviours they have observed opposite leaved to produce a positive response from others, social learning theory proposes that they may imitate the same behaviours in the belief that they will also receive the same positive feedback from those around them.

• Modelling and Reinforcement

Key to social learning theory is modelled behaviour. In order for social learning to take place, a model, which may be a parent, sibling, peer or someone that can be related to in someway, must carry out a particular attitude or behaviour. These models therefore provide examples of attitudes to food or dieting that may be observed by the individual and then imitated by them.

Reinforcement occurs when an individual imitates models behaviour and others respond to the imitated behaviour. If their response is perceived as positive (e.g. positive statements on their weight loss), positive reinforcement makes the individual feel better about themselves and encourages them to continue losing weight. Such individuals may also witness others being reinforced for the thickness or weight loss and through a process of vicarious reinforcement, they may form the expectation that they would also receive similar positive acknowledgement from others if they did the same.

• Maternal Role Models

Social learning theory may offer another explanation for why research into family systems theory is mixed. Rather than playing an overbearing controlling role, it may be that mothers act as a maternal role model instead as research has shown that problematic eating behaviour in common in families with eating disorders. Researchers have suggested mothers 'model' weight concerns for their daughters and some studies finding similarities between mothers and daughters restraint and dieting behaviours among children as young as 10 (Hill et al. 1990). Other researchers demonstrated that mothers who complain about their own weight are more likely to have children who also have their own weight concerns (Smolak et al., 1999) and this may be transmitted through a process of social learning.

• Peer Influences

Peers also act as role models for social learning and are particularly influential important during adolescence. Research has found that dieting among friends was significantly related to unhealthy weight control behaviours, such as the use of diet pills or purging (Eisenberg et al,. 2005). One specific mechanism of peer influence that may affect eating behaviour is teasing as overweight girls and underweight boys have been found to be most likely to be teased by their peers which may enforce gender-based ideals (Jones and Crawford, 2006).

• Media Influences

The media is another major source of influence for body image attitudes by adolescent Westerners. For example, the portrayal of thin models on television, movies and magazines as well as social media is a significant contributing factor in body image concerns that pushes adolescent girls into a drive for thinness themselves. Media influences are not believed to affect everyone in the same way; for example, those with low self-esteem are more likely to compare themselves to idealised images that are portrayed in the media (Jones and Buckingham 2005). It is believed that low self-esteem plays a contributory factor in the development of eating disorders such as anorexia nervosa through the portrayal of abnormally thin women in the media. Research by Button et al. (1996) found that those with low self-esteem at ages 11-12yrs were at greater risk of developing an eating disorder by ages 15–17yrs.

The British medical Association (BMA, 2000) displayed concern about the use of very thin models to advertise products through the media. The report found the images of slim models in the media were in contrast to the actual body size and shape of most children and young women. The report highlighted that the degree of thinness displayed by these models is both unachievable and biologically inappropriate and provided 'unhelpful role models for young women'. A study by Health Magazine in 2002 concluded that in the USA 32% of female characters on TV were underweight, compared to just 5% of the female audience.

• Maternal influence is more complex than social learning theory suggests. Research into the role of mothers as models and their influence for eating disorders with their daughters has not always produced consistent results. Pike and Rodin (1991) found that there was no evidence for daughters imitating the concerns of their parents. Ogden and Steward (1991) found no associations for restrained eating or body dissatisfaction between mothers and daughters that were similar in their weight and BMI. These findings do not support the modelling hypotheses of transmission of eating disorders between a mother and her daughter. Ogden and Steward argued the mothers contribution towards the development of disordered eating may be more than simply as a role model. They suggested it may be the very nature of the mother-daughter relationship that is important and the degree to which they are enmeshed.

  ---

• The role of peer influences is supported by research. Social learning theory explanations for anorexia nervosa propose that adolescent girls will be influenced by the perceived weight of their peers and this is supported by research by Costa-Font and Jofre-Bonet (2013) who investigated the effect of pure weight on the likelihood of an individual developing anorexia nervosa. Their study found individuals who had peers with a larger BMI than their own had a lower likelihood of developing an eating disorder and this association what is even more marked in younger women. The findings of this research suggest that having peers with an average of higher than average BMI 'protects' an individual from developing eating disorders, however having peers with a lower than average BMI made the development of anorexia nervosa more likely.

  ---

• A diathesis-stress explanation may be better suited to explain the disorder. It may be that stressors could trigger anorexia, one of which may be media representations of thinness. If the media played a significant role in the cause of anorexia, we would expect to see many more cases as most young women are exposed to thin role models regularly however this is not the case. This would suggest another factor is involved which may be a diathesis, an underlying genetic vulnerability to develop anorexia or an environmental trigger such as a significant childhood trauma. Therefore an explanation that can account for both biological and environmental factors would be more valid than simply social learning alone.

  ---

• Social learning theory cannot explain why dieting continues after the point at which compliments for losing weight may stop, or indeed when negative comments commence. Social learning theory as an explanation for anorexia nervosa cannot also explain why only some women develop eating disorders despite all women being subjected to the same type of media images on ideal body types. A strength however, is social learning explanations for anorexia nervosa are better able to explain why the prevalence of the disorder has increased and why occurs more in females and within certain cultures such as western cultures

The cognitive explanation of anorexia nervosa sees the disorder occur due to a breakdown in rational thought processes, such as an individual wanting to attain an unrealistic level of perfection in order to be accepted. According to the cognitive explanation, the main cause of anorexia nervosa is cognitive distortions about body shape and weight with the most significant of these, the sufferer has a disturbed perception of their own body image.

Cognitive distortions are errors in thinking that may cause an individual to develop a negative body image about themselves. This may occur through comparisons or exposure to other models with more ideal body types in terms of how they look or the amount they eat. This may lead the individual to form a misperception about themselves being overweight which leads to feelings of self-disgust and attempts to lose weight.

Irrational beliefs are maladaptive ideas that lead to the development and maintenance of anorexia nervosa. Anorexics often misperceive their body image and see themselves as fatter than they actually are with flawed reasoning behind their eating habits. Irrational beliefs may become second nature and give rise to automatic negative thoughts, for example, all or nothing thinking: 'if I'm not thin, I'm fat'. Another example of negative thinking due to irrational beliefs is catastrophising and putting the worst possible perspective on even trivial events: 'I ate a bite of chocolate today, that means I have no willpower whatsoever'. 

The need for perfectionism is another key irrational belief in anorexia nervosa. This is the view that the individual must meet their most demanding standards all of the time, and failure to do so is harshly judged upon themselves. This may apply to different areas of the sufferers life such as academic success, relationships, career goals but most especially to eating, their body shape and striving for weight loss. Intensive record keeping usually accompanies such behaviour as the individual tries to make sure they are achieving their harsh goals and this subsequently makes them resistant to treatment. As part of the irrational belief system, perfectionist will not be satisfied even when they reach their goals instead opting to raise the standards even higher. Therefore, they are forever pursuing an unrealistic goal they can never achieve that trapped them in a vicious cycle of irrational perfectionism and starvation.

• There is research support for cognitive distortions being a factor in anorexia nervosa. Sachdev et al. (2008) used fMRI brain scans on healthy participants as well as those suffering with anorexia nervosa. Participants were shown images of their own and other peoples bodies and results found that the same brain areas were activated in both groups when they were shown non-self images. When participants were shown images of themselves, compared with the control group the anorexia nervosa patients showed little activation in parts of the brain thought to be involved in attention. Researchers concluded that this suggested that cognitive distortions do exist in sufferers, but they are limited to the individuals own body image and may not extend to other bodies in general.

  ---

• Lang et al. (2015) provided further support for the role of cognitive factors in anorexia nervosa. In this study, the performance of 41 children and adolescents diagnosed with anorexia nervosa were compared with 43 healthy control participants on a range of neuropsychological measures. No differences in IQ between the groups were identified however compared to the control group, the anorexia nervosa participants displayed a more inflexible and inefficient cognitive processing style. For example, they were less able to overcome previously held beliefs or habits in the face of new information and this inefficient cognitive processing was independent of any clinical (e.g. length of illness) or demographical factors (education). This suggests that it presented an underlying characteristic of the disorder.

  ---

• There are issues of causation with research studies that identify a link between cognition factors and anorexia nervosa. For example, the studies above (Lang et al. 2015 & Sachdev 2008) identify a link but do not conclusively show that one causes the other. There is very little research to demonstrate whether these cognitive factors existed before the onset of anorexia nervosa and whether they are a symptom of the disorder rather than a cause. Research by Megan Shott et al. (2012) found younger anorexic sufferers were no worse at set-shifting than non-anorexic controls, however older patients were. This suggests that cognitive flexibility may not make an individual vulnerable to developing the disorder, but instead be a symptom of it.

  ---

• If the cause of anorexia nervosa was cognitive in nature, it would stand to reason that treatments based on cognitive behavioural therapy would be effective as they attempt to change a patient's distorted cognitions and irrational beliefs about food, eating, weight loss and body shape and size. Research by Riccardo Dalle Grave et al (2014) showed that enhanced CBT lead to substantial increase in weight and decrease in concerns about body shape amongst 26 hospitalised patients suffering from anorexia nervosa. This was found to have been maintained one year after discharge in a follow-up also suggesting AN may be cognitive in nature.

Biological explanations for obesity also focus on genetic and neural explanations. 

Genetic explanations focus on genes often through family studies (identical and non-identical twins) as well as adoption studies. Specific genes may also be examined that may affect physiology and psychological features such as intelligence and mental orders.

Neural explanations can include dysfunctions of the brain and nervous system which may include structures such as the hypothalamus and neurotransmitters such as serotonin and dopamine.

Research seems to suggest a genetic basis for obesity as some individuals appear more genetically predisposed to become obese than others. There does not appear to be a single gene responsible but genetic explanations tend to suggest multiple genes that make an individual more vulnerable to developing the disorder.

Twin studies into obesity have reported heriability ranges of 40% to 75%. A meta-analysis by Maes et al. (1997) involving 75,000 individuals found heritability estimates for BMI in 74% of identical twins and only 32% in non-identical twins. Even when identical twins were reared apart and experienced similar environmental influences, they were more like in terms of their BMI Dan non-identical twins that were reared together in similar environments (Stunkard et al., 1990).

Adoption studies have looked at individuals who have been adopted as infants and raised by biologically unrelated families. This allows researchers to attempt to separate the influence of biological parents (genetics) and adoptive parents (environment). Stunkard et al. (1986) examined 540 adult adoptees, their biological parents and also their adoptive parents. The research found a strong relationship between the weight category (i.e. underweight, overweight or obese) of adopted individuals and their biological parents weight category however no significant relationship with their adoptive parents weight category.

• Polygenic Determination

One explanation suggests the action of genes on obesity is polygenic and involves multiple genes interacting with one another in complex ways to determine different aspects of obesity such as waist-to-hip ratio, which focuses on abdominal fat to BMI. Adam Locke et al. (2015) examined the genomes of more than 300,000 individuals and identified 97 genes associated with variations in BMI which supports a polygenic argument for obesity. Another study found that these 97 genes accounted for only 2.7% of BMI variation which is a small fraction of the heritability of obesity. It is believed that as many as 400 missing heritability genes may in fact account for obesity that are yet to be identified (Watson 2009).

• Twin studies into how genetics affect obesity have their limitations such as the fact that even identical twins share the same environment. Identical twins are brought up in the same families which makes it difficult to separate environmental and genetic influences contributing to obesity. Twin studies attempt to overcome this issue by comparing concordance rates between identical and non-identical twins, because in both these scenarios the children are raised together. This approach and its validity is dependent on both types of twins being treated with equal degrees of similarity by friends, parents, teachers and those around them. There is an argument by researchers that this equal environment assumption may not be valid and identical twins maybe treated more similarly than non-identical twins who may be seen as individuals. Therefore twin studies may overestimate the extent of genetic influence and this may explain the higher concordance rates in identical twins.

  ---

• The genetic influences appear to vary with age and this may be explained due to environmental factors. Research has found that the genetic contribution to BMI is not stable across a persons lifetime. A meta analysis by Elks et al. (2012) found heritability estimates varied dependent on age group studied. Researchers found heritability was highest during childhood and then decreased during adulthood. This may be due to greater gene expression during childhood, compared to adulthood, where individuals may have adopted individual dietary and exercise habits due to greater independence and control over their lifestyles. This may contribute to the decrease in genetic contribution to their BMI and increase the environmental contribution.

  ---

• Genetic explanations cannot adequately explain the upsurge in obesity over the last 20 years. Genes for humans have not changed that much unlike environmental factors such as the variety, availability and affordability of food. This undermines genetic influences and suggests the environment may play a larger role.

  ---

• Another explanation that may account for both genetic and environmental factors is a diathesis-stress model. Genes may predispose an individual to develop obesity provided sufficient environmental stressors trigger their expression. This may be explained by the abundance of mass-produced food that is high in sugar and fat that is in turn triggering such genetic predispositions. Therefore this may explain the differences highlighted in younger age-groups where less control may be exercised and the increase in obesity in modern times due to increased food production.

  ---

• A meta-analysis by Paracchini et al. (2005) into genes thought to be responsible for regulating leptin have found mixed results. Leptin is believed to be responsible for weight regulation however researchers found no evidence to suggest a link between these genes and obesity. This would suggest that obesity does not have a sole genetic basis and there may be other environmental factors, potentially as the diathesis-stress model explanation suggests.

Neural explanations for obesity have focused on neurotransmitters such as serotonin and dopamine as well as the hypothalamus.

Research studies into human and non-humans show that obesity is associated with low levels of serotonin. Normal levels of serotonin regulate feeding behaviour by inhibiting activity in various areas of the hypothalamus, including the venture-medial hypothalamus. Serotonin for example, signals to the hypothalamus a person has eaten to satiety.

Dysfunctions of the serotonin system may be genetically inherited, due to stress or other disorders such as depression. It is thought that in such cases, levels of serotonin are abnormally low which creates inaccurate satiety signals to the hypothalamus, disinhibiting eating behaviour. Low levels of serotonin may create cravings for carbohydrates, energy-rich foods that include sugars, and therefore cause weight gain through consuming too many calories.

Dopamine plays a crucial role in the brains reward and motivation systems. At normal levels, dopamine stimulates brain areas such as the hypothalamus, hippocampus and amygdala, providing rewarding feelings of well-being and pleasure. This activity is associated to the pleasure we derive from eating and cues associated with eating, such as the smell of food. Dysfunction in the dopamine system has also been linked to obesity. Gene-Jack Wang et al. (2001) found obese individuals had significantly less dopamine D2 receptors in a part of the brain known as the striatum compared to 'normal' weight individuals.

When levels of dopamine are low, it is thought that the neurotransmitter cannot perform its usual pleasurable reward function in response to eating and the individual derives no positive feelings from eating. Overeating is therefore seen as an attempt to activate reward centres in the brain that would normally provide these feelings of pleasure, by increasing dopamine levels. This explanation proposes that obesity is caused by food addiction that operates neurochemically in the same ways as other addictions.

The hypothalamus plays a key role in regulating metabolism and energy usage and research has identified one part of it, the arcuate nucleus, as playing a fundamental role in appetite and obesity. The arcuate nucleus is a collection of neurons in the hypothalamus, which monitor circulating levels of sugar in the blood, and responds when energy levels are low. When activated, the arcuate nucleus sends messages to other parts of the body to produce the desire to eat and coordinate this with energy utilisation. It is therefore responsible for maintaining body weight, and adjusting food intake to physical activity and any malfunction with this area of the hypothalamus can lead to overeating and obesity. 

• There is research support for serotonin affecting obesity levels from animal studies. Sunny Ohia et al. (2013) highlighted the importance of one serotonin receptor, the 2C receptor, which when modified in mice to have no functioning 2C receptor, they developed late-onset obesity. This supports the link between obesity and dysfunctional serotonin levels being a potential factor. An obvious criticism of this research is it was based on animal studies and the findings may not generalise to humans in the same way. This could mean the process in humans is significantly different and this explanation may not apply.

  ---

• There is supporting evidence for dopamine affecting obesity levels and a genetic basis to dysfunctions of the dopamine reward system. Research into the DRD2 gene, which codes for the D2 receptor, has been implicated in obesity. Ritchie and Noble (2003) conducted a PET scan study and found people that inherited one version of the DRD2 gene, had 30-40% less D2 receptors compared to those with other versions. This supports the explanation that people with low dopamine levels (due to fewer D2 receptors) experience less dopamine-activated pleasurable reward from eating, which in turn may contribute to overeating.

  ---

• An issue with neural explanations for obesity that examine serotonin and dopamine levels is they tend to be purely correlational and it is not clear whether neural mechanisms are a cause or an effect of being obese. It may be that obesity causes fluctuations in neurotransmitters rather than the other way around.

  ---

• Biological explanations, such as neural explanations provide the opportunity for treatment to be developed. For example, identifying the role neurotransmitters play in obesity can help lead to the development of drug therapies that correct dopamine and serotonin levels to address deficiency.

Restraint theory is a cognitive theory of obesity developed by Peter Herman and Deborah Mack (1975). Restraint theory explains obesity occurring through people attempting to lose weight by placing a self-imposed food intake target and then indulging in reduced eating to meet this target. However, rather than the behaviour leading to weight loss, in approximately 80% of cases it generally leads to overeating, weight increase and thus obesity.

There is debate as to why this happens; one point of view is the self-imposed target is often unrealistic and attempting to diet in this way leads to negative moods and increased feelings of hunger that results in a motivation to eat more. Foods may be categorised into groups of 'good' or 'bad' with rules created by the individual on which they can eat and which are forbidden, as well as the amounts they believe are consistent with weight loss. A restrained diet is a highly organised way of imposing control which restrained eaters believe is the right way to lose weight. The result however is that the restraint dieter becomes more preoccupied with food rather than less. Through placing limits, the individual no longer eats when they are hungry or stops when they are full. They're eating behaviour subsequently is not under physiological control as they actively attempt to ignore physiological indicators that signal hunger and satiety, which leads to disinhibition of eating behaviour.

Another view is that the type of people who are likely to attempt restrained eating are those with less self-control and who are thus more likely to display disinhibited eating, in other words individuals who are more vulnerable to obesity.

The boundary Model was also created by Herman and Polivy (1984) in an attempt to explain why dieting may lead to overeating. According to the boundary model, food intake is regulated along a continuum with hunger at one end and fullness at the other. At either end of this continuum, food intake is driven by biological processes and when the bodies energy levels are low, we subsequently experience hunger which gives rise to eating. At the other extreme, when we have taken sufficient energy (food), we experience unpleasant feelings of fullness and we therefore stop eating. In between these two points is what Hermon and Pahlevi refer to as the zone of biological indifference. This is where the individual is neither hungry no full and food intake is more influenced by cultural and social factors.

Applying this to dieters (restrained eaters), they have a larger zone of biological indifference and a lower threshold for hunger and a higher threshold for satiation (feelings of fullness). This basically means they get hungry more easily and take longer to feel full. In addition, Herman and Polivy claimed that dieters have a cognitively determined self imposed diet boundary that is set lower than their satiation threshold and represents what they believe they should eat. If the dieter goes over this limit, they experience something called the 'what the hell' effect because as the diet boundary has already been surpassed, the person will continue to eat until they reach satiation or even beyond.

Disinhibition is another psychological explanation for obesity and refers to overeating, eating too quickly and repeated lack of success when dieting. Disinhibition can occur when an individual is presented with lots of palatable foods or is under emotional distress (Hays and Roberts, 2008). In such situations, the normal inhibitions that prevent an individual from eating too much are removed, and affected individuals respond to less to the feelings of satiation and are more vulnerable to overeating. Research has shown that disinhibition strongly associated with adult weight gain and BMI (Bellisle et al. 2004).

Bond et al. (2001) identified three types of disinhibition each with their own impact on overeating:

• Habitual disinhibition: the tendency to overeat in response to daily life circumstances
• Emotional disinhibition: the tendency to overeat in response to emotional states such as anxiety or depression
• Situational disinhibition: the tendency to overeat in response to specific environmental cues (e.g. social events such as weddings, parties or social gatherings).

From these three different types, habitual disinhibition has been shown to be the most important correlate of weight gain and obesity due to the high number of daily overeating opportunities people are faced with in the typical western food environment (Hays and Robert, 2008). Bryant et al. (2008) suggested that this type of eating is better described as 'opportunistic eating' and situational and emotional disinhibition, on the other hand, occurred less frequently, therefore contributed less to obesity.

Restraint Theory

There is research support for the claim that dietary restraint can lead to overeating (Wardle and Beals 1988). In this research, 27 obese women were assigned to either a diet group which focused and restrained eating patterns, an exercise group or a non-treatment group for seven weeks. Participants were assessed at weeks four and six under laboratory conditions. At week four, food intake and appetite was assessed before and after a 'preload' (a small snack, such as a milkshake or chocolate bar). At week six, food intake was assessed under stressful conditions. Results found that at both assessment sessions, women in the diet condition eight more than women in the exercise and non-treatment groups, which is in line with predictions by restraint theory.

There has also been criticism to restraint theory, particularly by Tomiyama et al. (2009) highlighting most research relating to it has been restricted to laboratory settings. Such laboratory research has typically shown that restrained eaters overeat after they violate their diet however, Tomiyama et al. challenged this of view of restraint, arguing outside of such artificial settings, restrained eaters would be able to control their eating. In two studies, they showed that dieters who tracked their food consumption over a period of days did not overeat after violating their diet. This would potentially dispel the notion that diet violations lead to overeating in everyday life.

Research by Savage et al. (2009) has also found contradictory evidence of restraint theory. In her longitudinal study into restrained eating, dietary restraint and disinhibition in 163 women at the start of the study and every two years afterwards over a six year period was measured. Changes in the women's weight was also measured. They found that increases in restrained eating was linked to a decrease in weight (a significant negative correlation). Researchers concluded that restrained eating while dieting leads to weight loss rather than weight gain, at least in the short term, which is an outcome that is the opposite of what restraint theory predicts.

The Boundary Model

There is criticism to the boundary model and the 'what the hell' effect. Herman and Polivy argued that when dieters go past the self-imposed dietary boundary, the what the hell effect is described as a breakdown in the dieters self control as the person simply gives in to the overpowering urge to eat. Ogden and Wardle (1991) argued that rather than passively giving in to the desire to eat, the individual may instead actively decide to overeat as a form of rebellion against the self-imposed food restrictions. The idea of overeating as an expression of rebelliousness is supported by Loro and Orleans (1981). This research study found that obese binge eaters frequently reported binging as a way of unleashing resentment against their diet.

The boundary model presents us with practical applications and ways to reduce weight gain. For example, the model predicts that food intake is consciously limited and the paradoxical outcome is disinhibition, overeating, weight gain and obesity. Despite this outcome, the advice for obesity is to encourage dietary restraint. Modjtaba Zandian et al. (2009) suggests that weight-loss advice should avoid recommending restraint but instead focus on training dieters to eat at a slower rate which can help contribute to less consumption.

Disinhibition

The majority of research into disinhibition has been based on women, particularly white women, which means conclusions about men or other racial groups cannot be made. Atlas et al. (2002) reported that restraint and disinhibition scores were significantly lower in African-American students compared to white college students suggesting that disinhibited eating behaviours may be less evident among African American women. Research by Bellisle et al. (2004) restraint and disinhibition scores were lower in men compared to women, suggesting disinhibition might be more a more important influence in adult weight gain in women compared to men.

There are a number of proposed models to explain why dieting succeeds or fails however we will focus on the following:

• The Spiral Model
• Ironic Process Theory
• Restraint Theory, Disinhibition and the Boundary Model (covered previously and can be applied as an explanation for why dieting fails)

The Spiral Model

Heatherton and Polivy (1992) proposed the Spiral Model to explain dieting behaviour in terms of a chain of linked events. Dieting often begins in adolescence overwhelmingly with young women who experience body dissatisfaction. Potential low self-esteem may also be a risk factor for dieting attempts as the individual may be more likely to make negative comparisons about themselves. First-time dieters may experience some initial success with temporary weight loss however lasting weight loss is rare, so ultimately weight is regained and the outcome is failure. At this point, some may simply give up trying to lose weight while others may attribute their failure to some personal deficiency, such as themselves not trying hard enough or lacking the willpower.

Instead of rethinking their approach to weight loss, dieters will attempt to 'stick to the plan' with a greater effort by simply eating even less in their next attempt. This even greater food restriction has physical and psychological effects with the dieter experiencing more frustration and emotional distress, which makes them more vulnerable to disinhibited eating. There are also changes in the metabolic processes within the body which results in weight loss becoming even more difficult to achieve. For example, ghrelin levels increase (making an individual hungrier) and leptin levels decrease (making them eat more until they feel full) after significant weight loss. This results in a further failure in dieting which may result in repeated attempts to diet harder. The result is a lowering of self-esteem and increase in depression over time with the individual now trapped in a destructive 'downward spiral' in which weight loss is less and less likely.

Ironic Processes Theory

Research has shown that when people deliberately restrict their food intake and begin to diet, they become more preoccupied with thoughts of food rather than less. One explanation for this was provided by Daniel Wegner (1994) as part of his theory of ironic processes. In one of his experiments (Wegner et al. 1987), he noticed that when people were asked not to think about a white bear, they were almost always guaranteed to think about one even more so than people specifically asked to think of one. Applying this to people dieting, they will label certain foods as forbidden which makes them become more salient (i.e. they stand out). This means they are more likely to think about these foods precisely for the same reasons, because they are trying not to think about them. For example, an individual who tells themselves not to think about chocolate will find that images of chocolate are easier to retrieve from memory. This ultimately needs to disinhibition of eating, loss of self-control, and excessive food intake and ultimately the diet failing.

In addition, when individuals attempt to distract themselves from thinking about forbidden foods (e.g. reading a book) this will require mental activity that will leave the individual lacking the cognitive processing required to suppress the thoughts of temptation. This would explain why dieting can be an all-consuming and yet self-defeating process as the dieter has to spend all their time, energy and undivided attention trying not to think about food which leads them into disinhibition.

• The spiral model proposes a number of ways in which dieting might be successful including ways in which individuals can 'break out' of the spiral at various points. Heatherton and Polivy point out that the key issue is to prevent lowering of self-esteem to avoid the worst consequences of diet failure. For example, individuals that diet to avoid putting on weight as opposed to losing it are less likely to experience disinhibited eating, possibly because their self-esteem may be higher (Lowe and Kleifield, 1988). One option to escape the consequences of failure is therefore to give up dieting altogether however this does nothing to address the issue of self-esteem that created the motivation to diet initially. In addition, evidence from Joel Yager et al. (1988) suggests disinhibited eaters who give up dieting are more likely to turn to substance abuse as a way of reducing their unhappiness. It may be that the most appropriate way to successfully diet may be to promote self-esteem or learning to accept oneself as they are.

  ---

• Research support for ironic processes theory comes from Adriaanse et al. (2011) who investigated snacking behaviour. Participants were female students who were attempting to reduce the intake of unhealthy snacks such as chocolate and crisps. They were presented with diet intentions expressed in a negative form (e.g. 'when I am sad, I will not eat chocolate'). Researchers found an ironic rebound effect just by being exposed to these statements as it reinforced an association between 'being sad' and 'eating chocolate'. This was seen to make the link between the two more accessible in memory and more likely to be recalled. Another key finding was that the 'ironic effect' was not just cognitive, but also behavioural. After being exposed to these statements, participants kept a snack diary during the following week and data showed that they ate unhealthy snacks more often and consumed more calories than a control group. This study demonstrated how difficult it was to suppress thoughts of eating once they had become accessible in memory. As ironic cognitive processes generalised to actual snacking behaviour, this demonstrated how thinking of oneself as dieting can lead to failure, which supports the validity of ironic processes theory as an explanation for dieting failure.

  ---

• The spiral model and ironic processes theory cannot explain why some dieters go on to succeed or account for individual differences that may lead to success. Ogden (2010) highlighted that theories that suggest dieting is counterproductive are unable to explain why many go on to lose weight successfully despite being preoccupied with food. This includes anorexics who lose weight through extreme dieting. One explanation that attempts to explain individual differences looks at 'locus of control'. Those with an internal locus of control believe weight loss is contingent on their own efforts and they have control over their own success or failure. Those that fail may have a greater external locus of control and attribute it to external forces that will determine their success or failure. This may also link in with studies into how low self-esteem may be linked in with those that fail as they may be more likely to have an external locus of control.