PSYCHOANALYTIC PSYCHOLOGY, 1987, 4(2), 145-159

Copyright© 1987, Lawrence Erlbaum Associates, Inc.

The Psychodynamics of Obesity:

A Review
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Joyce Slochower, Ph.D.

Hunter College
City University of New York

The psychoanalytic model of obesity continues to be controversial among

research-oriented theorists. Although empirical tests of the psychoanalytic
model initially yielded many failures to support the role of affect in overeating,
recent experimental data confirm the centrality of this dynamic. This article
presents evidence supporting the psychodynamic model of obesity and dis-
cusses some developmental implications for these data.

Despite numerous attempts to understand and treat eating disorders that re-
sult in obesity, there is little agreement among researchers and clinicians con-
cerning central etiology and treatment techniques. Current theories of over-
eating include psychodynamic (Bruch, 1973; Slochower, 1983), behavioral
(Leon, 1976; Stunkard, 1975) and physiological (Stunkard, 1980) models.
Additionally, some now argue that obesity is not a unique clinical entity
(Herman & Polivy, 1975) or that obesity is not related to overeating at all
(Wardle, 1985).
The psychodynamic theory of obesity is among the oldest and most contro-
versial of these models. Perhaps the most central tenet of the psychodynamic
model is that overeating is a primary response to anxiety caused by uncon-
scious conflict that may temporarily relieve affective distress (cf. Caldwell,
1965; Fenichel, 1945; Garetz, 1973; Kaplan & Kaplan, 1957; Kornhaber,
1970; Rakoff, 1967). Although virtually all other models of obesity have
vehemently rejected the notion that obesity is associated with anxiety, accu-
mulated clinical data and new experimental evidence now appear to support
this notion. This article reviev/s the empirical support for the role of affect in

Requests for reprints should be sent to Joyce Slochower, Ph.D., Department of Psychology,
Hunter College, 695 Park Avenue, New York, NY 10021.
 146 SLOCHOWER

overeating and then discusses some implications for a developmental, psy-

choanalytic model of obesity.

THE DATA
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

The causal relationship between painful affect and overeating is both the
This document is copyrighted by the American Psychological Association or one of its allied publishers.

most central and the most easily testable aspect of the psychoanalytic model
of obesity. Attempts to validate this model have therefore focused on
demonstrating that overeating does occur in response to anxiety or depres-
sion. This notion has received attention primarily from two sources: clinical
interview and questionnaire studies of obese patients' eating patterns, and
laboratory experiments using mildly obese samples and experimental manip-
ulations of anxiety. These data are briefly reviewed here.

Clinical Studies

Many early clinical investigations of the dynamics of overeating used the

technique of interviewing moderately obese patients recruited from treat-
ment settings and asking them under what circumstances they overate. In a
large-scale study, Freed (1947) asked 500 obese women whether they ate more
or less when they were nervous or worried. Seventy-five percent reported an
increase in eating when nervous; 20% said they ate more when bored; only
5% reported no relationship between emotions and eating. Hamburger
(1951), Hecht (1955), Atkinson and Ringuette (1967), Kalucy and Crisp
(1974), and Castelnuovo-Tedesco and Schiebel (1975) all reported similar
findings using much smaller samples of female obese patients recruited from
a variety of treatment settings. The emotions most frequently associated with
overeating were
overeating were depression
depression and
and anxiety.
anxiety.
In an interview study of 125 obese patients, Stunkard (1959) described
three eating patterns within the obese group: night eating (especially during
periods of life stress), binge eating (again associated with life stress), and eat-
ing without satiation (apparently not related to stress).
A few studies incorporated a nonobese comparison group into the design.
Holland, Masling, and Copley (1970) interviewed 48 obese, superobese, and
normal weight women. Notably, a mental status test indicated no differences
between groups in the presence or severity of psychiatric disturbance. Both
overweight groups, however, reported significantly more eating when anx-
ious or depressed.
Leon and Chamberlain (1973) administered a questionnaire about eatng
to 51 obese and 39 nonobese subjects. Approximately 26% of the obese
groups (including both successful and unsuccessful dieters) reported eating in
 PSYCHODYNAMICS OF OBESITY 147

response to emotional distress, whereas only 8% of the nonobese group did

so.
Silverman (1976) developed the technique of subliminal presentation of
unconscious material to test various aspects of psychoanalytic theory. In an
attempt to relate obesity to an ungratified symbiotic merger fantasy,
Silverman, Martin, Ungaro, and Mendelsohn (1978) treated 30 obese women
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

in a behavior modification program for overeating and exposed them to one

This document is copyrighted by the American Psychological Association or one of its allied publishers.

of two subliminal messages. In the symbiotic condition, subjects received the

message, "Mommy and I are one," and in the control condition a neutral
message, "People are walking," was presented. There was no between group
difference in weight loss after 8 weeks, but the symbiotic group maintained
the weight loss significantly longer than did the control group. These results
suggest that shifts in emotional state may alter overeating patterns. However,
it is not clear precisely how subjects assimilated these subliminal stimuli, and
it is difficult to specify why the symbiotic message had an effect only on
maintenance of weight loss.
Plutchik (1976) administered a series of scales about eating to 60 male and
female, obese and nonobese subjects. Significant correlations were found be-
tween the degree of obesity and the frequency of eating in response to both
anxiety and depression. In another questionnaire study of 62 obese and nor-
mal weight men and women, Hudson and Williams (1981) reported that sig-
nificantly more obese than nonobese subjects ate when bored or depressed.
Using a structured interview paradigm, Edelman (1981) evaluated the fre-
quency of binge eating in 100 obese and nonobese people. Results showed
that significantly more women and obese subjects reported eating in response
to emotional distress and to environmental stimuli.
In an ambitious attempt to evaluate the psychoanalytic model of obesity
and to assess the effectiveness of psychoanalysis, Rand and Stunkard (1977,
1978, 1983) and Glucksman, Rand, and Stunkard (1978) solicited informa-
tion from 72 analysts about 84 obese and 63 nonobese patients. Analysts
were asked to describe the psychodynamics associated with their patients'
weight gains. Seventy-one percent of the obese group, but only 6% of the
normal weight patients were described as showing evidence of internal dy-
namics relating to eating.
Analysts also described patients' weight patterns. Weight gains when
stressed were reported for 79% of the obese patients, but for only 9°7o of the
nonobese group. Moreover, significantly more of the obese patients than
nonobese patients were described as eating in response to emotional upset.
These clinical studies strongly support the relationship between affect dis-
turbance and overeating. However, methodological as well as theoretical dif-
ficulties limit their explanatory power.
A most obvious weakness inherent in the early studies is the absence of a
 148 SLOCHOWER

nonobese comparison group that would have permitted a clearer association

of obesity with stress-related eating patterns. Because samples were drawn
largely from psychiatric sources, it is not clear to whar degree the patholog-
ical eating patterns described reflected some other personality difficulties
rather than characteristics uniquely associated with obesity. Additionally,
one must consider the possibility that patients' perceptions of their over-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

eating symptoms were distorted (or that their therapists' perceptions were in-
This document is copyrighted by the American Psychological Association or one of its allied publishers.

accurate). Also, the absence of data (e.g., projective measures) concerning

the internal conflicts related to overeating seriously limits the interpret ive
power of many studies (and prevents any developmental inferences from be-
ing made). The absence of overt, behavioral indexes of both obesity and of
overeating patterns further weaken the studies' power to test causal hypo-
theses. In fact, much of the reported data could be interpreted f rofli a
behavioral learning model of obesity without ever resorting to psychoana-
lytic theory at all!

Experimental Evidence
Researchers whose focus was to obtain clear-cut causal data concerning the
relationship between anxiety and obesity have mainly used laboratory
methodologies to do so. Those designs permitted the independent manipula-
tion of various emotional (and hunger) states, behavioral measures of over-
eating and body weight, and the selection of matched nonobese control
groups.
A classic test of the effects of emotional state was conducted by Schachter,
Goldman, and Gordon (1968). They varied the level of fear by threatening to
administer severe or mild electric shock and then measured the amount eai en
by obese and normal weight subjects in a bogus 'taste test." Obese subjects
were unresponsive to fear manipulations (and also to their hunger level).
Obese subjects also did not report any fear reduction following eating. Nor-
mal weight subjects ate less when frightened than when calm.
Several other studies were modeled on this basic paradigm bui varied ihe
fear manipulation. None provided support for the psychodynamic model.
Instead, obese subjects" eating patterns were found to be unaffected by exper-
imental manipulations of (a) neurotic anxiety and objective fear (Abramson
&Wunderlich, 1972), (b) boredom and interest (Abramson &Stinson, 197 7),
(c) stress (Meyer & Pudel, 1972, 1977), and (d) moderate or high anxiety
(Reznick & Balch, 1977; Ruderman, 1983). Only one study (McKenna, 1972)
revealed even a marginally significant relationship between anxiety and eat-
ing among obese subjects; however, no anxiety reduction occurred following
eating in that study.
In a different approach to the problem of eating disorders, Herman aid
Polivy (1975) proposed that apparent differences between obese and non-
 PSYCH0DYNAM1CS OF OBESITY 149

obese people are actually reflective of an underlying dimension called dietary

restraint. Dietary restraint refers to the individual's ability to restrain eating
in the presence of appetizing food. It was argued that obese individuals, like
many normal weight individuals, are characterized by high-restraint patterns
and that it is restraint, not obesity, that accounts for their eating difficulty.
Herman and Polivy (1975) tested their hypotheses using samples of pri-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

marily normal weight subjects classified as high or low restraint on the basis
This document is copyrighted by the American Psychological Association or one of its allied publishers.

of self-reported dietary history. In their study they evaluated the impact of

level of fear on eating. It was found that restrained subjects (presumably sim-
ilar to the obese), did not change their eating habits when fearful. Restrained
subjects also showed no fear reduction following eating. Unrestrained sub-
jects ate less when fearful than when calm. Restraint has also been found to
predict subject differences in external responsiveness that parallel those typi-
cally found among obese and normal subjects (Herman, Polivy, Pliner, &
Threlkeld, 1978).
Threlkeld, 1978).
Although these results support restraint theory —particularly with regard
to the behavior of the unrestrained eater —the relevance of these data to the
clinically obese subject group is not clear, because overweight subjects were
not included in those experiments. Also, because restrained subjects' re-
sponse to anxiety in those studies was rather weak, the results are not strongly
supportive of the psychodynamic model.
Baucom and Aiken (1981) attempted to extend the restraint model by
exposing both obese and nonobese, restrained and unrestrained subjects to
experimentally induced depression. Depression was induced by having some
subjects fail on a concept formation task. The results showed that restrained
eaters ate more than nonrestrained eaters when depressed, and the same held
true for obese and nonobese dieters. The nondieting subject group responded
inversely to depression; that is, they ate less. Baucom and Aiken suggested
that it is dieting, rather than obesity per se, that is predictive of emotional eat-
ing. The data offer indirect support for the role of emotionality in over-
eating, but, notably, fail to identify obesity as the crucial subject variable.
Some recent data, however (Ruderman & Christensen, 1983; Temarken &
Kirschenbaum, 1984), suggest that restraint alone does not account for
overeating patterns, and again point to the importance of obese-normal dif-
ferences.
The bulk of early experimental tests of the psychodynamic model, then,
provide little direct support for the presence of emotion-induced eating and
fail to find clear data linking this pattern with obesity. I (Slochower, 1975,
1976) attempted to reconcile the marked discrepancy between the findings of
the interview and experimental studies by pointing to several theoretical diffi-
culties inherent in the latter. I argued, in line with Leon and Roth (1977), that
experimental tests of the psychodyanmic model, and especially the use of
"taste tests" to measure eating, lacked external validity. It also seemed likely
 150 SLOCHOWER

that experimental studies of the psychodynamic hypothesis had not properly

created the kinds of anxiety states that result from internal unconscious con-
flict. Both fear and anxiety are usually consciously experienced, painful, af-
fective states. However, the causes of anxiety are often more difficult for the
individual to identify, because they are at least partially unconscious. The
presence of unconscious conflict in anxiety also often makes anxiety relief
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

less within the individual's control than fear relief.

This document is copyrighted by the American Psychological Association or one of its allied publishers.

It seemed likely, however, that experimental tests of the psychoanalytic

model actually manipulated objective fear rather than anxiety because in
these experiments, the fear manipulation (e.g., threats of electric shock) in-
volved an external stimulus whose source could be easily defined and
controlled. I reasoned that overeating patterns probably develop in ar at-
tempt to alleviate painful affect that is not easily deaf: with by other means.
Overeating patterns should therefore only be evident in situations where the
individual experiences painful anxiety that is not easily controlled and/or
whose source is unclear.
The hypothesis that it is only unlabeled (free floating) anxiety thai is re-
lated to overeating in obese people was tested using standard experimental
procedures and college students as subjects. The presence or absence of a la-
bel for anxiety was varied directly by using a false heart rate feedback para-
digm to induce arousal and then either providing or not providing a label for
that arousal. Eating was measured unobtrusively rather than by the taste :est
paradigm, to increase the external validity of the measure.
In that study (Slochower, 1975), obese participants ate significantly more
when experiencing high but unlabeled anxiety than when their anxiety was
high but clearly labeled. In subsequent studies (Slochower, 1983; Slochower
& Kaplan, 1980; Slochower, Kaplan, &Mann, 1981), it became clear that the
obese person's ability to control anxiety was central to this overeating pat-
tern. In each study, obese participants ate significantly more when they expe-
rienced high but uncontrollable anxiety than when their high anxiety was felt
to be controllable. Low anxiety did not increase their consumption, in all
studies, overeating also resulted in significant anxiety reduction for cbese,
but not for normal weight groups.
We consistently found these results in studies using two different labora-
tory manipulations of anxiety (false heart rate feedback and bogus test in-
structions that made it difficult for participants to control important test re-
sults; Slochower, 1983; Slochower & Kaplan, 1980). Similar results were
obtained in a study that involved the stress situation of final examinations
(Slochower etal., 1981).
We (Slochower & Kaplan, 1980; Slochower et al., 1981) also addressed tne
relationship between the psychodynamic model and externality theory. The
latter theory posits that obese people show a heightened degree of reactivity
to potent external cues, and a reduced level of responsiveness to internal cuss
 PSYCHODYNAMICS OF OBESITY 151

(cf. Schachter & Rodin, 1974). We reasoned, however, that this pattern of
heightened external reactivity may actually reflect the obese person's depend-
ence on external cues in the face of internal anxiety. We proposed that, rather
than representing alternative models of overeating, the psycho dynamic and
externality models act to potentiate each other so that overeating is greatest
in the presence of both high, uncontrollable anxiety and potent food cues. In
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

two studies that covaried anxiety level and food potency, it was found that
This document is copyrighted by the American Psychological Association or one of its allied publishers.

obese people's external responsiveness was significantly enhanced by high

levels of internal anxiety, whereas normal weight people's eating was not.
Finally, I (1983) considered an implication of Bruch's (1973) work in con-
junction with the finding that obese people respond "inappropriately" to anx-
iety. Bruch has described repetitive early feeding experiences in which food is
used to pacify, subdue, or comfort the child. She argued that this feeding pat-
tern interferes with the perceived association of hunger and eating so that the
child eventually becomes unable to differentiate between hunger and emo-
tional stress. Bruch believed that these obese individuals suffer from a deficit
in awareness of internal hunger cues.
I reasoned that the repeated association of painful affect with food and
eating may similarly stop the obese person from fully experiencing differenti-
ated affective states. This may happen because many emotional states are
quickly responded to in a similar way, by eating, thus prematurely cutting off
the affective experience. It seemed plausible that, over time, the reliance on
an external behavior (eating) to explain and cope with internal affect would
undermine the child's ability to accurately perceive and to differentially label
affect in the absence of a clear-cut external anchor.
Obese and nonobese people described their emotional reaction to a series
of affectively charged photographs. It was found that the obese group re-
sponded to the emotionally arousing cues with fewer, and more gross emo-
tional labels than did normals. The proposed hypothesis was thus confirmed,
indicating that obese people have difficulty describing, and perhaps also
perceiving, internal emotional states.
Lowe and Fisher (1983) extended my findings (Slochower, 1983) by testing
subjects in a naturalistic setting to see whether the experimental data were in-
fluenced by perceived demand characteristics on the part of obese subjects
(Krantz, 1978). Obese and nonobese women monitored their food intake and
their mood for 13 days. Obese subjects reported significantly more negative
moods than did normal weight subjects. Furthermore, obese subjects' eating
of snacks (but not of regular meals) was significantly correlated with the in-
tensity of their negative affect. These results, then, tended to support my lab-
oratory findings and to point to an important distinction between snacking
and mealtime eating patterns among obese people.
As part of a multidisciplinary research project on obesity carried out in the
Netherlands, Strien (1986) evaluated the relative validity of psychosomatic,
 152 SLOCHOWER

externality, and restraint theories. The Dutch Eating Behaviour Question-

naire (DEBQ) was constructed to separately assess emotional, external, and
restrained eating patterns. These scales showed, high internal consistency,
factorial validity, and high dimensional stability. In a series of studies in
which the DEBQ was administered to obese and nonobese Dutch women, it
was found that (a) obese subjects showed distinct responses to diffuse and la-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

beled emotions; (b) a positive relationship was found between emotional eat-
This document is copyrighted by the American Psychological Association or one of its allied publishers.

ing and external eating, suggesting that emotionality and external fooc cues
can act together to elicit eating; and (c) no significant relationship was found
between restrained eating and either emotional or external eating. Instead,
emotional eating appeared to precede restrained eating, thereby weakening
the predictive value of the restraint factor in overeating patterns. Taken to-
gether, these results strongly support my findings (Slochower, 1983) enhance
the generalizability of these data, and provide additional information con-
cerning the role of restraint theory in obesity.

SUMMARY OF THE FINDINGS

The data reviewed in this article largely validate the psychoanalytic model of
obesity and appear to resolve the discrepancy between results from early ex-
perimental and nonexperimental studies in favor of the latter. My results also
suggest (in line with the psychoanalytic model) that overeating is used in an
attempt to control overwhelming internal anxiety states, It may be argued
that my laboratory findings (Slochower, 1983) lack clear-cut external valid-
ity, and also that the findings of clinical investigators (e.g., Edelman, 1981;
Glucksmanet al., 1978; Lowe & Fisher, 1983; Rand &Stunkard, 1978; Strien
et al., 1985) lack strong internal validity. However, the convergence of data
from these varied empirical tests reduces the cogency of this critique and ulti-
mately strengthens the power of the psychodynamic hypothesis. Similarly,
the fact that parallel findings were obtained from psychiatric and
nonpsychiatric, male and female, and mildly and moderately
moderately obese
obese groups
groups
all
all point
point to
to the
the generalizability
generalizability of
of the
the anxiety-eating
anxiety-eating relationship.
relationship.
Other issues do remain puzzling. Few experimental or interview studies
made any attempt to delineate the type of obesity characteristic of its subjects
(e.g., early vs. late onset or stable vs. fluctuating obesity). It is thus unclear
whether there are obese groups to whom these findings do not apply. Also,
the absence of carefully controlled studies comparing the responses of ob ;se
and normal, restrained and unrestrained eaters, has left open the possibility
that dietary restraint was inadvertently confounded with weight differences
in many studies. Future studies that clearly delineate the type of obesity char-
acteristic of its subjects and that attempt a more careful definition of 'nor-
mal"
mal" subjects,
subjects, will
will clarify
clarify these
these issues.
issues.
 PSYCHO DYNAMICS OF OBESITY 153

THEORETICAL IMPLICATIONS

A question that was not directly addressed by either the interview or experi-
mental tests of the psychodynamic model concerns just how this overeating
pattern might develop. Classical psychoanalytic theories of obesity assumed
that overeating symptomotology was associated with conflicts related to ex-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

cessive or inadequate gratification of the oral drives. Masling, Rabie, and

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Blondheim (1967), for example, reported that obese patients gave more oral
dependent responses on Rorschach Tests and Thematic Apperception Tests
than did control subjects. In this apparently clear-cut connection between
oral stage fixations and overeating symptomotology, food and eating sym-
bolize emotional nurturance (Alexander,
(Alexander, 1934;
1934; Bychowski,
Bychowski, 1950;
1950; Conrad,
Conrad,
1952; Fenichel,
1952; Fenichel, 1945;
1945; Hamburger,
Hamburger, 1951;
1951; Schick,
Schick, 1947).
1947).
The causal connection between oral drive frustration and obesity, how-
ever, is not quite so straightforward, because oral phase fixations are associa-
ted with a wide variety of ego organizations and symptom choices, and many
of those are unrelated to eating disorders.
In 1957, Kaplan and Kaplan conducted a large-scale review of clinical case
studies of obesity reported in the psychoanalytic literature. They concluded
that many different interpretations of overeating were described in these
studies. Overeating was seen as an unconscious reflection of:

1. diminishing anxiety
2. achieving pleasure
3. achieving social success and acceptance
4. relieving frustration and deprivation
5. expressing hostility (conscious or unconscious)
6. self-indulgence
7. rewarding onself
8. expressing defiance
9. submitting (e.g., to parental authority)
10. self-punishment in response to guilt
11. exhibitionism
12. attaining attention and care
13. justifying failure in life
14. testing love
15. counteracting a feeling of being unloved
16. distorting reality
17. identifying with an overweight parent
18. sedating oneself
19. avoiding competition in life
20. avoiding changing the status quo
21. proving one's inferiority
 154 SLOCHOWER

22. avoiding maturity

23. diminishing fear of starvation
24. consciously fulfilling the wish to become fat
25. handling anxiety from infantile oral frustration
26. a diversion from monotony
27. diminishing feelings of insecurity
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Although many of these dynamics can be associated with classic oral con-
flicts, others appear to be reflective of later developmental issues, or of issues
related to the integration and differentiation of the self.
The apparent complexities of overeating have also raised questions about
interpersonal dynamics associated with this symptom choice. In her seminal
work on eating disorders, Bruch (1957, 1961, 1973) described the parents of
obese children as markedly unable to differentiate their own needs fiom
those of the child. These parents are therefore unable co offer the child ade-
quate emotional protection. Bruch also characterized the feeding experien. ces
of these children as markedly inappropriate because food is offered nor in re-
sponse to nutritional need but for a variety of other reasons. Bruch noted
that obese adults are actually unable to perceive hunger cues accurate!)'.
In line with Bruch's thinking, I would suggest that the potentially obese
child may also fail to develop the ability to perceive and appropriately moni-
tor emotional states. This weakened capacity for internal self-moniioring
would leave the obese child excessively dependent on external factors (ini-
tially the mother; later, upon a variety of external cues; to define internal af-
fect. That would account for the obese person's tendency to use food to cope
with psychic distress. I also suggest that there may be a dynamic explanation
for the pattern of external responsiveness described by Schachter, Rodin,
and others (Schachter & Rodin, 1974).
My data (Slochower, 1983) indicating that obese people are literally una o\e
to accurately label their emotional states (a primary ego capacity) suggests
that there may be very early precursors for the affect-eating connection de-
scribed by Bruch (1973). In this context, the work of object relations theorists
is relevant. In particular, Winnicott (1965) emphasized the crucial impor-
tance of an appropriately facilitating maternal environment in the early
months. Of central importance is the mother's capeicity to be a subjective ob-
ject to the infant and the infant's capacity to be satisfied by her.
In considering how early maternal-infant communication might set the
stage for the development of eating disorders, a crucial factor seems to be the
mother's ability to accurately perceive her infant's biologically based oral
needs, to differentiate those from needs that are not food-related, and to re-
spond appropriately to each. For example, because crying not due to hunger
in the early months may not always be easily subdued, excessive maternal
anxieties may result in the compulsive use of feedings in an attempt to qmet
 PSYCHODYNAM1CS OF OBESITY 155

the infant. Similarly, intense infantile hunger (or hunger accompanied by

gastric distress) sometimes results in an initial rejection of the breast or bot-
tle. In both cases, it would be necessary that the mother have sufficient confi-
dence in both the quality of her milk and in her ability to replenish a drained
supply to tolerate her infant's distress and rejection of the breast.
In addition to the importance of the mother having a stable, maternal
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

introject, the infant must also be able to experience satisfaction from feed-
This document is copyrighted by the American Psychological Association or one of its allied publishers.

ings and to communicate this to the mother. The infant's own internal devel-
opmental processes may interfere with this if, for example, severe gastric
symptoms (like colic) turn satisfactory feedings into painful internal states. It
is not only the infant who is deprived of a "good feed" here. The mother's
own ability to tolerate the infant's ruthlessness and to allow herself to offer
food on demand will be seriously strained or destroyed if her good feed pro-
duces distress rather than satisfaction. A destructive cycle may ensue in
which maternal empathy is largely replaced by frustration and rage, and
feedings are gratifying for neither mother nor infant. To the degree that the
mother is unable to differentiate between her own sense of deprivation and
her infant's distress, her own projections will intensify an objectively difficult
situation.
It is suggested that the painful, repeated mishandling of the feeding experi-
ence itself is likely to result in the development of an eating disorder rather
than in a different oral symptom. The infant's repetitively enacted oral frus-
tration combined with the heightened emotional charge given to the feeding
situation by the mother leave food and feedings as "spotlighted" aspects of
early development. Although actual obesity may not appear as a symptom
until later on in life, clinical signs of the powerful meaning of food should be
evident much earlier.
The oral drives are most developmentally central during the first 2 years of
life, but food-related interactions continue to be an important ground for
emotional communication throughout childhood and assume different sym-
bolic meanings. Early separation issues are often marked by the first appear-
ance of food refusal, clear food preferences, and intentional messing. Here,
the mother's ability to tolerate the child's apparently whimsical rejection,
messing, and demand for food is central. Her capacity to retain a sense of her
own feeding potential and her willingness to allow the child to assume control
over the feeding experience will be necessary. Otherwise, eating —or not
eating —will become a central arena for the enactment of separation issues.
As development continues, early symbolic communication around feeding
becomes increasingly laden with additional layers of meaning as the child's
psychological world expands from the two-person unit to triangular, and
then social relationships. Issues of competence, sexuality, and so forth, may
all be enacted around meals. Similarly, a variety of parental anxieties may be
expressed through the use of food in the home context. Both regressive and
 156 SLOCHOWER

progressive pulls may be reflected in eating at each phase, and the eating
arena may be selected as a central conflict focus at any point in the develop-
mental process. However, for this focus on eating to result in clinical obesity,
a serious and prolonged disruption of the connection between eating and
hunger would have to develop. It is suggested that this disruption has early
(preoedipal) origins.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

CONCLUSIONS

The empirical evidence regarding the psychoanalytic model of obesity has

shifted in recent years. It first appeared that only interview studies supported
the notion that certain types of anxiety states play a crucial role in overeating.
Now, however, both quasi-experimental and experimental designs report re-
sults consistent with the theory. This followed a more careful definition of
the emotional states associated with overeating. The power of these findings
should free future investigators to explore the developmental antecedents of
the affect-eating connection.
This review has attempted to clarify the power of the psychoanalytic model
of obesity, but it has not evaluated the relative weight of this model vis-a-vis
competing models, for instance, restraint and set point theories. The psycho-
analytic model also in no way contradicts the likelihood that genetic and/or
physiological factors play a role in obesity. It is hoped that future investiga-
tions will focus on assessing the developmental antecedents of the psychoan-
alytic model and on evaluating its relative weight within the larger picture of
factors causing pathological overeating.

ACKNOWLEDGMENT

I thank Michael Lowe for his helpful suggestions concerning the literature
reviewed in this article.

REFERENCES

Abramson, E.E., & Stinson, S. G. (1977). Boredom and eating in obese and non-obese
individuals. Addictive Behaviors, 2, 181-185.
Abramson, E. E., & Wunderlich, R. A. (1972). Anxiety, 'ear, and eating: A test of the
psychosomatic concept of obesity. Journal of Abnormal Psychology, 79, 317-321.
Alexander, F. (1934). The influence of psychologic factors upon gastrointestinal disturbances.
Psychoanalytic Quarterly, 3, 501-518.
Atkinson, R. ML, &Ringuette, E. L. (1967). A survey of biographical and psychological featuies
in extraordinary fatness. Psychosomatic Medicine, 29, 121-133.
 PSYCHODYNAMICS OF OBESITY 157

Baucom, D. H., & Aiken, P. A. (1981). Effect of depressed mood on eating among obese and
nonobese dieting and nondieting persons. Journal of Personality and Social Psychology, 41,
577-585.
Bruch, H. (1957). The importance of overweight. New York: Norton.
Bruch, H. (1961). Transformation of oral impulses in eating disorders. Psychiatric Quarterly,
35, 458-481.
Bruch, H. (1973). Eating disorders. New York: Basic Books.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Bychowski, G. (1950). On neurotic obesity. Psychoanalytic Review, 37, 301-319.

This document is copyrighted by the American Psychological Association or one of its allied publishers.

Caldwell, J. (1965). Lifelong obesity —A contribution to the understanding of recalcitrant

obesity. Psychosomatics, 6, 417-426.
Castelnuovo-Tedesco, P., & Schiebe'., D. (1975). Studies of superobesity: I. Psychological
characteristics of superobese patients. International Journal of Psychiatry in Medicine, 6,
465-480.
Conrad, S. (1952). The psychological causes and treatment of overeating and obesity. American
Practitioner, 3, 438.
Edelman, B. (1981). Binge eating in normal weight and overweight individuals. Psychological
Reports, 49, 739-746.
Fenichel, O. (1945). The psychoanalytic theory of neurosis. New York: Norton.
Freed, S. C. (1947). Psychic factors in development and treatment of obesity. Journal of the
American Medical Association, 133, 369-373.
Garetz, F. K. (1973). Socio-psychological factors in overeating and dieting with comments on
popular reducing techniques. The Practitioner, 210, 671-676.
Glucksman, M. L., Rand, C. S. W., & Stunkard, A. J. (1978). Psychodynamics of obesity.
Journal of the American Academy of Psychoanalysis, 6, 103-115.
Hamburger, W. W. (1951). Emotional aspects of obesity. Medical Clinics of North America, 35,
483-499.
Hecht, M. B. (1955). Obesity in women: A psychiatric study. Psychiatric Quarterly, 29, 203-231.
Herman, C. P., & Polivy, J. (1975). Anxiety, restraint, and eating behavior. Journal of
Abnormal Psychology, 43, 647-660.
Herman, C. P., Polivy, J., Pliner, P., & Threlkeld, J. (1978). Distractibility in dieters and
nondieters: An alternative view of externality. Journal of Personality and Social Psychology,
36, 536-548.
Holland, J., Masling, J., & Copley, D. (1970). Mental illness in lower class normal, obese and
hyperobese women. Psychosomatic Medicine, 32, 351-357.
Hudson, A., & Williams, S. G. (1981). Eating behavior, emotions, and overweight. Psycholog-
ical Reports, 48, 669-670.
Kalucy, R. S., & Crisp, A. H. (1974). Some psychological and social implications of massive
obesity: A study of some psychosocial accompaniments of major fat loss occurring without
dietary restrictions in massively obese patients. Journal of Psychosomatic Research, 18,
465-473.
Kaplan, H. I., & Kaplan, H. S. (1957). The psychosomatic concept of obesity. Journal of
Nervous and Mental Disease, 125, 181-201.
Kornhaber, A. (1970). The stuffing syndrome. Psychosomatics, 11, 580-584.
Krantz, D. S. (1978). The social context of obesity research: Another perspective on its place in
the field of social psychology. Personality and Social Psychological Bulletin, 4, 177-184.
Leon, G. (1976). Current directions in obesity research. Psychological Bulletin, 83, 557-578.
Leon, G., & Chamberlain, K. (1973). Emotional arousal, eating patterns, and body image as
differential factors associated with varying success in maintaining a weight loss. Journal of
Consulting and Clinical Psychology, 40, 474-480.
Leon, G., & Roth, L. (1977). Obesity: Psychological causes, correlations and speculations.
Psychological Bulletin, 84, 117-139.
 158 SLOCHOWER

Lowe, M. R., & Fischer, E. B., Jr. (1983). Emotional reactivity, emotional eating, and obesity:
A naturalistic study. Journal of Behavioral Medicine, 6, 135-149.
Masling, J., Rabie, L., & Blondheirn, S. H. (1967). Obesity, level of aspiration, and Rorschach
and TAT measures of oral dependence. Journal of Consulting Psychology, 31, 233-239.
McKenna, R. J. (1972). Some effects of anxiety level and food cues on the eating behavior of
obese and normal subjects: A comparison of the Schachterian and psychosomatic concepts of
obesity. Journal of Personality and Social Psychology, 22, 311-319.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Meyer, J. E., & Pudel, V. E. (1972). Experimental studies on food intake in obese and normal
This document is copyrighted by the American Psychological Association or one of its allied publishers.

weight subjects. Journal of Psychosomatic Research, 16, 305-308.

Meyer, J. E., & Pudel, V. E. (1977). Experimental feeding in man: A behavioral approach to
obesity. Psychosomatic Medicine, 39, 153-157.
Plutchik, R. (1976). Emotions and attitudes related to being overweight. Journal of Clinical
Psychology, 32, 21-23.
Rakoff, V. (1967). The psychiatric aspects of obesity. Modern Treatment, 4, III 1-1124.
Rand, C. S., & Stunkard, A. J. (1977). Psychoanalysis and obesity. Journal of the American
Academy of Psychoanalysis, 5, 459-497.
Rand, C. S., & Stunkard, A. J. (1978). Obesity and psychoanalysis. American journal of Psy-
chiatry, 135, 547-551.
Rand, C. S., & Stunkard, A. J. (1983). Obesity and psychoanalysis: Treatment and four-year
follow-up. American Journal of Psychiatry, 140, 1140-1144.
Reznick, H., & Balch, P. (1977). The effects of anxiety and response cost manipulations on the
eating behavior of obese and normal subjects. Addictive Behaviors, 2, 219-225.
Ruderman, A. (1983). Obesity, anxiety, and food consumption. Addictive behaviors, 8,
235-242.
Ruderman, A., & Christensen, H. (1983). Restraint theory and its applicability to overweight in-
dividuals. Journal of Abnormal Psychology, 92, 210-215.
Schachter, S., Goldman, R., & Gordon, A. (1968). The effect of fear, food deprivation and
obesity on eating. Journal of Personality arid Social Psychology, 10, 91-97.
Schachter, S., & Rodin, J. (Eds.). (1974). Obese humans and rats. Potomac, MD: Lawrence
Erlbaum Associates, Inc.
Schick, A. (1947). Psychosomatic aspects of obesity. Psychoanalytic Review, 34, 173-183.
Silverman, L. H. (1976). Psychoanalytic theory: "The reports of my death are greatly exaggt ra-
ted. "American Psychologist, 31, 621-637.
Silverman, L. H., Martin, A., Ungaro, R., & Mendelsohn, E. (1978). Effects of subliminal
stimulation of symbiotic fantasies on behavior modification treatment of obesity. Journal of
Consulting and Clinical Psychology, 46, 432-441.
Slochower, J. (1975). Emotionality and overeating in obese and normal weight individuals. Un-
published doctoral dissertation, Columbia University.
Slochower, J. (1976). Emotional labeling and overeating in obese and normal weight indivic u-
als. Psychosomatic Medicine, 38, 131-139.
Slochower, J. (1983). Excessive eating: The role of emotions and environment. New York: Hu-
man Sciences.
Slochower, J., & Kaplan, S. P. (1980). Anxiety, perceived control, and eating in obese and nor-
mal weight persons. Appetite, 1, 75-83.
Slochower, J., Kaplan, S. P., & Mann, L. (1981). The effects of life stress and weight on mood
and eating. Appetite, 2, 115-125.
Strien, T. V. (1986). Eating behaviour, personality traits, and body mass. Lisse (Netherlands):
Swets & Zeitlinger B. V.
Stunkard, A. J. (1959). Eating patterns and obesity. Psychiatric Quarterly, 33, 284-292.
Stunkard, A. J. (1975). Presidential Address, 1974: From explanation to action in psychoso-
matic medicine: The case of obesity. Psychosomatic Medicine, 37, 195-235.
 PSYCHODYNAMICS OF OBESITY 159

Stunkard, A. J. (1980). Obesity. Philadelphia: Saunders.

Teraarken, A. J., & Kirschenbaum, D. S. (1984). Effect of plans for future meals on counter-
regulatory eating by restrained and unrestrained eaters. Journal of Abnormal Psychology, 93,
458-472.
ent. Ap
Wardle, J. (1985). Review of Excessive eating: The role of emotions and environment. Appetite,
4, 348-349.
Winnicott, D. W. (1965). The maturational processes and the facilitating environment. New
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

York: International Universities Press.

This document is copyrighted by the American Psychological Association or one of its allied publishers.