The causes of overeating

Page 154
Black blue
154 HEALTH PSYCHOLOGY
groups of dieters and non-dieters either a high-calorie preload or a low-calorie preload.
The results are illustrated in Figure 6.5 and indicated that whereas the non-dieters
showed compensatory regulatory behaviour, and ate less at the taste test after the
high-calorie preload, the dieters consumed more in the taste test if they had had the
high-calorie preload than the low-calorie preload.
This form of disinhibition or ‘the what the hell effect’ illustrates overeating in
response to a high-calorie preload. Disinhibition in general has been defined as ‘eating
more as a result of the loosening restraints in response to emotional distress, intoxication
or preloading’ (Herman and Polivy 1989: 342), and its definition paved the way for a
wealth of research examining the role of restraint in predicting overeating behaviour.
The causes of overeating
Research has explored possible mechanisms for the overeating shown by restrained
eaters. These are described below and include the causal model of overeating, the
boundary model of overeating, cognitive shifts, mood modification, denial, escape theory,
overeating as relapse and the central role for control.
i) The causal analysis of overeating
The causal analysis of eating behaviour was
first described by Herman and Polivy (Herman and Mack 1975; Herman and Polivy
1980, 1988; Polivy and Herman 1983, 1985). They suggested that dieting and bingeing
were causally linked and that ‘restraint not only precedes overeating but contributes
to it causally’ (Polivy and Herman 1983). This suggests that attempting not to eat,
paradoxically increases the probability of overeating; the specific behaviour dieters are
Fig. 6-5 Overeating in dieters in the laboratory (after Herman and Mack 1975)
Page 154
Black blue
Page 155
Black blue
EATING BEHAVIOUR 155
attempting to avoid. The causal analysis of restraint represented a new approach to
eating behaviour and the prediction that restraint actually caused overeating was an
interesting reappraisal of the situation. Wardle further developed this analysis (Wardle
1980) and Wardle and Beales (1988) experimentally tested the causal analysis of overeating. They randomly assigned 27 obese women to either a diet group, an exercise
group or a no treatment control group for seven weeks. At weeks four and six all subjects
took part in a laboratory session designed to assess their food intake. The results showed
that subjects in the diet condition ate more than both the exercise and the control group
supporting a causal link between dieting and overeating. From this analysis the overeating shown by dieters is actually caused by attempts at dieting.
ii) The boundary model of overeating In attempt to explain how dieting causes
overeating, Herman and Polivy (1984) developed the ‘boundary model’, which represented an integration of physiological and cognitive perspectives on food intake. The
boundary model is illustrated in Figure 6.6.
According to the model, food intake is motivated by a physiologically determined
hunger boundary and deterred by a physiologically determined satiety boundary.
In addition, the boundary model suggests that the food intake of restrained eaters is
regulated by a cognitively determined ‘diet boundary’. It indicates that dieters attempt to
replace physiological control with cognitive control which represents ‘the dieters selected
imposed quota for consumption on a given occasion’ (Herman and Polivy 1984: 149).
Herman and Polivy (1984) described how after a low calorie preload the dieter can
maintain her diet goal for the immediate future since food intake remains within the
limits set by the ‘diet boundary’. However, after the dieter has crossed the diet boundary
(i.e. eaten something ‘not allowed’), they will consume food ad lib until the pressures
Fig. 6-6 A boundary model explanation of overeating in dieters
Page 155
Black blue
Page 156
Black blue
156 HEALTH PSYCHOLOGY
of the satiety boundary are activated. The boundary model proposes a form of dual
regulation, with food intake limited either by the diet boundary or the satiety boundary.
The boundary model has also been used to examine differences between dieters, binge
eaters, anorexics and normal eaters. This comparison is shown in Figure 6.7.
iii) Cognitive shifts The overeating found in dieters has also been understood in
terms of shifts in the individual’s cognitive set. Primarily this has been described in
terms of a breakdown in the dieter’s self control reflecting a ‘motivational collapse’ and
a state of giving in to the overpowering drives to eat (Polivy and Herman
1983). Ogden and Wardle (1991) analysed the cognitive set of the disinhibited dieter
and suggested that such a collapse in self control reflected a passive model of overeating
and that the ‘what the hell effect’ as described by Herman and Polivy (1984) contained
elements of passivity in terms of factors such as ‘giving in’, ‘resignation’ and ‘passivity’.
In particular, interviews with restrained and unrestrained eaters revealed that many
restrained eaters reported passive cognitions after a high calorie preload including
thoughts such as ‘I’m going to give into any urges I’ve got’ and ‘I can’t be bothered,
it’s too much effort to stop eating’ (Ogden and Wardle 1991). In line with this
model of overeating, Glynn and Ruderman (1986) developed the eating self-efficacy
Fig. 6-7 A comparison of the boundaries for different types of eaters
Page 156
Black blue
Page 157
Black blue
EATING BEHAVIOUR 157
questionnaire as a measure of the tendency to overeat. This also emphasized motivational collapse and suggested that overeating was a consequence of the failure of
this self-control.
An alternative model of overeating contended that overeating reflected an active
decision to overeat and Ogden and Wardle (1991) argued that implicit within the ‘What
the hell effect’ was an active reaction against the diet. This hypothesis was tested using a
preload/taste test paradigm and cognitions were assessed using rating scales, interviews
and the Stroop task which is a cognitive test of selective attention. The results from two
studies indicated that dieters responded to high calorie foods with an increase in an
active state of mind characterized by cognitions such as ‘rebellious’, ‘challenging’
and ‘defiant’ and thoughts such as ‘I don’t care now in a rebellious way, I’m just going
to stuff my face’ (Ogden and Wardle 1991; Ogden and Greville 1993 see Focus on
research 6.1, page 160). It was argued that rather than simply passively giving in to
an overwhelming desire to eat as suggested by other models, the overeater may
actively decide to overeat as a form of rebellion against self-imposed food restrictions.
This rebellious state of mind has also been described by obese binge eaters who report
bingeing as ‘a way to unleash resentment’ (Loro and Orleans 1981). Eating as an
active decision may at times also indicate a rebellion against the deprivation of other
substances such as cigarettes (Ogden 1994) and against the deprivation of emotional
support (Bruch 1974).
iv) Mood modification
Dieters overeat in response to lowered mood and researchers
have argued that disinhibitory behaviour enables the individual to mask their negative
mood with the temporary heightened mood caused by eating. This has been called the
‘masking hypothesis’ and has been tested by empirical studies. For example, Polivy and
Herman (1999) told female subjects that they had either passed or failed a cognitive task
and then gave them food either ad libitum or in small controlled amounts. The results in
part supported the masking hypothesis as the dieters who ate ad libitum attributed more
of their distress to their eating behaviour than to the task failure. The authors argued
that dieters may overeat as a way of shifting responsibility for their negative mood from
uncontrollable aspects of their lives to their eating behaviour. This mood modification
theory of overeating has been further supported by research indicating that dieters eat
more than non-dieters when anxious regardless of the palatability of the food (Polivy
et al. 1994). Overeating is therefore functional for dieters as it masks dysphoria and this
function is not influenced by the sensory aspects of eating.
v) The role of denial Cognitive research illustrates that thought suppression and
thought control can have the paradoxical effect of making the thoughts that the
individual is trying to suppress more salient (Wenzlaff and Wegner 2000). This has been
called the ‘theory of ironic processes of mental control’ (Wegner 1994). For example,
in an early study participants were asked to try not to think of a white bear but to ring a
bell if they did (Wegner et al. 1987). The results showed that those who were trying not
to think about the bear thought about the bear more frequently than those who were
told to think about it. Similar results have been found for thinking about sex (Wegner
et al. 1999), thinking about mood (Wegner et al. 1993) and thinking about a stigma
Page 157
Black blue
Page 158
Black blue
158 HEALTH PSYCHOLOGY
(Smart and Wegner 1999). A decision not to eat specific foods or to eat less is central to
the dieter’s cognitive set. This results in a similar state of denial and attempted thought
suppression and dieters have been shown to see food in terms of ‘forbiddenness’ (e.g.
King et al. 1987) and to show a preoccupation with the food that they are trying to deny
themselves (Grilo et al. 1989; Ogden 1995a). Therefore, as soon as food is denied it
simultaneously becomes forbidden and which translates into eating which undermines
any attempts at weight loss. Boon et al. (2002) directly applied the theory of ironic
processes of thought control to dieting and overeating. They used a factorial design and a
standard preload taste test paradigm. Restrained and unrestrained eaters were given
a preload that they were told was either high or low in calories and then were either
distracted or not distracted. Their food intake was then measured in a taste test. The
results showed that the restrained eaters ate particularly more than the unrestrained
eaters in the high calorie condition if they were distracted. The authors argued that
this lends support to the theory of ironic processes as the restrained eaters have a
limited cognitive capacity, and when this capacity is ‘filled’ up by the distraction their
preoccupation with food can be translated into eating.
vi) Escape theory Researchers have also used escape theory to explain overeating
(Heatherton and Baumeister 1991; Heatherton et al. 1993, 1991). This perspective
has been applied to both the overeating characteristic of dieters and the more extreme
form of binge eating found in bulimics and describes overeating as a consequence of
‘a motivated shift to low levels of self awareness’ (Heatherton and Baumeister 1991). It
is argued that individuals prone to overeating show comparisons with ‘high standards
and demanding ideals’ (Heatherton and Baumeister 1991: 89) and that this results in
low self-esteem, self dislike and lowered mood. It is also argued that inhibitions exist
at high levels of awareness when the individual is aware of the meanings associated
with certain behaviours. In terms of the overeater, a state of high self awareness can
become unpleasant as it results in self criticism and low mood. However, such a state is
accompanied by the existence of inhibitions. The individual is therefore motivated to
escape from self awareness to avoid the accompanying unpleasantness but although
such a shift in self awareness may provide relief from self-criticism it results in a
reduction in inhibitions thereby causing overeating. Within this analysis disinhibitory
overeating is indicative of a shift from high to low self awareness and a subsequent
reduction in inhibitions.
vii) Overeating as a relapse Parallels exist between the undereating and overeating
of the restrained eater and the behaviour of the relapsing smoker or alcoholic. The
traditional biomedical perspective of addictive behaviours viewed addictions as being
irreversible and out of the individual’s control. It has been argued that this perspective
encourages the belief that the behaviour is either ‘all or nothing’, and that this belief is
responsible for the high relapse rate shown by both alcoholics and smokers (Marlatt and
Gordon 1985). Thus, the abstaining alcoholic believes in either total abstention or
relapse, which itself may promote the progression from lapse to full-blown relapse. In the
case of the restrained eater, it is possible that they too believe in the ‘all or nothing’
theory of excess which promotes the shift from a high calorie lapse to the ‘what the hell’
Page 158
Black blue
Page 159
Black blue
EATING BEHAVIOUR 159
relapse characterized by disinhibition. This transition from lapse to relapse and the
associated changes in mood and cognitions is illustrated in Figure 6.8.
These parallels have been supported by research suggesting that both excessive
eating and alcohol use can be triggered by high risk situations and low mood (Brownell
et al. 1986a; Grilo et al. 1989). In addition, the transition from lapse to relapse in both
alcohol and eating behaviour has been found to be related to the internal attributions
(e.g. ‘I am to blame’) for the original lapse (e.g. Ogden and Wardle 1990). In particular,
researchers exploring relapses in addictive behaviours describe the ‘abstinence violation
effect’ which describes the transition from a lapse (one drink) to a relapse (becoming
drunk) as involving cognitive dissonance (e.g. ‘I am trying not to drink but I have just
had a drink’), internal attributions (e.g. ‘It is my fault’) and guilt (e.g. ‘I am a useless
person’) (Marlatt and Gordon 1985). These factors find reflection in the overeating
shown by dieters (Ogden and Wardle 1990).
Fig. 6-8 The ‘what the hell’ effect as a form of relapse
viii) The role of control The interview data from a study of 25 women who were
attempting to lose weight provides further insights into the mechanisms behind overeating (Ogden 1992). The results from this study indicated that the women described
their dieting behaviour in terms of the impact on their family life, a preoccupation with
food and weight and changes in mood. However, the concept of self control transcended
these themes. For example, when describing how she had prepared a meal for her family
one woman said ‘I did not want to give in, but I felt that after preparing a three-course
meal for everyone else, the least I could do was enjoy my efforts’. The sense of not giving
in suggests an attempt to impose control over her eating. In terms of the preoccupation
with food, one woman said ‘Why should I deprive myself of nice food’ and another said
‘Now that I’ve eaten that I might as well give in to all the drives to eat’. Such statements
again illustrate a sense of self control and a feeling that eating reflects a breakdown in
this control. In terms of mood, one woman said that she was ‘depressed that something
as simple as eating cannot be controlled’. Likewise this role of self control was also
apparent in the women’s negative descriptions of themselves with one woman saying
Page 159
Black blue