Abstract
Dysmorphophobia, an over-concern with an imagined or slight defect in physical appearance, was first described by Morselli at the end of the 19th century [1]. Although this condition has a rich tradition in European psychiatry, the disorder received little attention elsewhere until recently [2]. Dysmorphophobia re-entered the official psychiatric classification system as an atypical somatoform disorder in DSM-III [3]. This was subsequently modified in the revised classification (DSM-III-R) with the introduction of body dysmorphic disorder (BDD) as a separate diagnostic status [4], which has been continued with DSM-IV [5].
Distressing and often debilitating, BDD is a preoccupation with an imagined or slight defect in appearance. It is relatively common in patients with depression, obsessive compulsive disorder and social phobia [6], as well as dermatological [7] and cosmetic surgery settings [8]. BDD is usually a chronic disorder consisting of timeconsuming and distressing preoccupations with a perceived appearance, as well as associated compulsive behaviours such as mirror checking, excessive grooming and skin picking.
Andreasen and Bardach suggested that BDD might account for an estimated 2% of all patients who request plastic surgery [8]. In a later study using a standardized instrument, 7% met the diagnostic criteria for BDD as defined by the body dysmorphic disorder examination [9, 10]. A number of researchers have noted that postoperatively the result is often regarded as poor or unsatisfactory from the patient's perspective [11, 12]. Phillips et al. found eight of 30 patients with BDD had undergone surgical procedures but only 4% reported that it had been of any help [12]. Of 43 patients with BDD who underwent an array of dermatological treatments including antibiotics, steroids and dermabrasion, only five (11%) considered the treatment useful [13].
BDD is often associated with other psychiatric disorders [6, 12]. In two studies of patients with dysmorphophobia attending dermatologists, depression was the most commonly associated disorder [7, 14], with 40% being moderately or severely depressed compared to none of the comparison subjects [14].
There has been less research on the prevalence and psychiatric comorbidity of BDD in patients attending plastic surgery clinics. Hay reported that 16 of 45 (35%) of a group of patients requesting cosmetic rhinoplasty had ‘severe neurosis’, and one had a psychotic disorder [15]. Beale et al. did a psychological study of patients seeking augmentation mammoplasty in 1980 [16]. The study consisted of interviews with 64 women who were to undergo, or had already undergone, an augmentation mammoplasty, and a control group of 28 women. They noted the principal difference was that women seeking augmentation had lower self esteem and considered themselves less ‘feminine’.
This paper describes the characteristics of body dysmorphic symptoms, and any associated psychiatric comorbidity, in cosmetic plastic surgery outpatients using the dysmorphic concern questionnaire (DCQ) [2]. The aim was to compare patients attending plastic surgery clinics for cosmetic or non-medically explained reasons with those attending for medically explained reasons. The hypotheses to be tested were that: (i) patients attending plastic surgery outpatient clinics for cosmetic reasons would have higher DCQ scores than the comparison group, and (ii) that inpatients with high DCQ scores would have significantly higher rates of psychiatric morbidity as measured by a standardized psychiatric instrument which would be independent of possible confounding factors on multivariate analysis.
Method
The study was undertaken at plastic surgery outpatient clinics of Fremantle Hospital, a large teaching hospital of approximately 500 beds in Western Australia. The ethics committee of the hospital approved the protocol.
As the aim of the study was to compare the characteristics of patients presenting for cosmetic (non-medically explained) and medically explained reasons, consecutive attenders who presented for cosmetic procedures were compared with an equal number of consecutive patients who presented at the same clinic for procedures as a result of underlying illness or organic pathology. These included patients with basal cell carcinomas, cysts, breast reconstruction following carcinoma excision, and trauma with associated reconstructive surgery.
This method was chosen because a survey of 48 case notes demonstrated that one-eighth of new referrals at the clinic were for cosmetic reasons (n = 6). Approaching all referrals to the clinic would therefore have led to the inclusion of a far greater number of subjects than was indicated by power calculations as being necessary to demonstrate a significant difference between the two groups (see below).
A two-stage screening strategy was used. Prior to the start of the clinic, all the patients’ files were reviewed by the treating consultant plastic surgeons who divided patients into those presenting for plastic procedures due to underlying physical illness or pathology, and those presenting for solely cosmetic reasons.
All patients presenting for cosmetic reasons, and an equal number of control subjects who presented for medically explained reasons, were administered the 12-item general health questionnaire (GHQ-12) [17], and the DCQ [2].
The DCQ consists of seven statements regarding the respondents’ over-concern with perceived physical defects and their attempts to deal with the perceived problem. An example of the questionnaire is given in an initial paper by Oozthuizen et al. [2]. The statements are derived from the literature on dysmorphic concern and target affective, cognitive and behavioural symptoms which have been shown to be valid and reliable [2]. The DCQ uses a four-point scale with responses ranging from ‘not at all’, ‘same as most people’, ‘more than most people’ to ‘much more than most people’. The total score for the questionnaire was the sum of the ratings for the seven questions.
The DCQ was chosen as it is a self-report questionnaire which was acceptable to patients in this setting [18], and takes only a few minutes to complete [2, 19]. It has been validated against the BDDE [19], which is a valid and reliable diagnostic instrument specifically for BDD but takes an hour to complete [19, 20].
Analysis was undertaken using the statistical package for social sciences (SPSS). Subjects with body dysmorphic symptoms were divided about the median into those with high and low DCQ scores. Psychiatric caseness was defined as a score of greater than four on the GHQ [21].
Occupational status was assessed using the Australian Standard Classification of Occupations (ASCO) [22]. Socio-demographic, clinical S. KISELY, D. MORKELL, B. ALLBROOK, P. BRIGGS, J. JOVANOVIC 123 and psychiatric features of subjects who presented for cosmetic surgery were compared with those presenting for medically explained reasons using the Odds Ratio and chi square test for categorical data, and the Students T or Mann–Whitney U-test for continuous data. The association between psychiatric morbidity as measured by the GHQ and dysmorphic concern on the DCQ was assessed using the Odds Ratio and chi square test for categorical data and Spearman's rho for continuous data. Spearman's rho is a non-parametric equivalent of Pearson's r and was used because the GHQ and DCQ are ordinal rather than interval scales. Significance was assessed using 95% confidence intervals and a p-value of less than 0.05. Factors associated with psychiatric caseness on univariate analysis were investigated further using logistic regression to determine whether dysmorphic concern was independently associated with psychiatric morbidity taking into account other possible confounding variables such as gender.
Although no data were available from a plastic surgery clinic, previous work done in dermatological and psychiatric settings indicated a 35–40% difference in the rate of psychiatric morbidity between patients with and without body dysmorphic symptoms [2, 14]. Power calculations indicated that to have an 80% chance of detecting such a difference at the 95% level of significance, a sample size of about 40 patients was required in each group, making a total of 80 surveyed subjects.
Results
Sociodemographic characteristics
Ninety subjects were eligible for inclusion in the study of whom 84 (93%) participated. The mean age of the sample was 45.5 and just under two-thirds were female (n = 52). Forty-three percent of the subjects were in paid employment (n = 36), of whom half were in classes 4 and 5 of the Australian Standard Classification of Occupations (ASCO). All but two of the cases (95%) were of Caucasian origin.
Referral procedure and diagnosis
Seven cases presented for rhinoplasty (8%), 37 cases for mammoplasty (44%), and the other 40 cases (48%) for a number of procedures including excisions of abscesses or cysts, tendon repairs, abdominoplasty, bilateral blepharoplasty and facelifts. Of the 42 patients who presented for medically explained symptoms, exactly half were following trauma, 40% (n = 17) for carcinoma and the remaining 10% for other conditions such as cysts or abscesses.
Rating instrument scores
The mean DCQ score was 10.2 with a median of 11.00 and a standard deviation of 4.63. The GHQ consisted of a mean score of 3.64 and a median of 3.00 and standard deviation of 3.40. Thirty-two percent of the sample were GHQ cases (n = 27).
Differences between cosmetic (non-medically explained) and medically explained presentations
There were no differences in socio-demographic characteristics between the two groups other than the fact females were significantly more likely to present for cosmetic surgery (Table 1). Mammoplasties accounted for two-thirds of cosmetic procedures as opposed to only a fifth of presentations for medically explained reasons (Table 1). Patients presenting for cosmetic procedures had significantly higher scores on the DCQ and GHQ (Table 1). They were six times as likely to be GHQ cases (Table 1).
Features of patients presenting for medically explained and cosmetic (non-medically explained) surgery
Dysmorphic concern and psychiatric comorbidity
There was a significant association between the dysmorphic concern as measured by the DCQ, and psychiatric symptoms on the GHQ (Tables 2,3). There was also a strong association between DCQ and GHQ scores (Tables 2,3). Patients who had high DCQ scores had a 32-fold increase in psychiatric comorbidity as measured by the GHQ. This result was confirmed by a correlation coefficient of 0.637 on Spearman's rho with a p-value of less than 0.0001. Other factors associated with dysmorphic concern included being female, and referral for mammoplasty (Table 2). These factors were also associated with GHQ caseness (Table 3). Patients with carcinoma or trauma were not significantly more likely to be GHQ cases (Table 3).
Features of patients with dysmorphic concern
Features of patients with psychiatric comorbidity
When factors associated with GHQ caseness on univariate analysis were entered into a logistic regression equation, only dysmorphic concern was independently associated with psychiatric morbidity (adjusted OR = 32.0, 95% CI = 6.5–156; log likelihood statistic = 34.5; d.f. = 1; p < 0.0001). Similarly, GHQ caseness and presentation for cosmetic surgery independently predicted DCQ score on logistic regression (adjusted OR = 23.7, 95% CI = 4.5–124; log likelihood statistic = 22.8; d.f. = 1, p < 0.0001; and adjusted OR = 6.3, 95% CI = 1.9–20.6; log likelihood statistic = 10.3, d.f. = 1, p = 0.001, respectively).
Discussion
The aim of the study was to compare the characteristics of patients presenting for cosmetic and non-cosmetic (medically explained) procedures, rather than measure the prevalence of each group. A review of the case notes of 48 consecutive referrals to the same clinic indicated that 60% of attenders at the clinic were male (n = 29). Twenty-nine per cent of referrals were for mammoplasty (n = 14), and four per cent for rhinoplasty, the remainder being for a wide range of other procedures. One-eighth were for cosmetic (non-medically explained) procedures (n = 6). The figures for this study differ to those of other studies because here equal numbers of patients with cosmetic (non-medically explained) and medically explained reasons for presentation were compared with each other.
Strengths and weaknesses of the study
This is the first study to compare the characteristics of patients presenting to plastic surgeons for cosmetic reasons with those presenting for medically explained reasons, in terms of dysmorphic concern and other psychiatric comorbidity. Dysmorphic and psychiatric symptoms were measured using standardized instruments rather than case reports or clinical interviews. Self report instruments were used to reduce observer bias that might have been introduced if raters who were not blind to the results of the initial screening by the plastic surgeons had made assessments of psychiatric and dysmorphic status.
To our knowledge, this is also the first study to investigate the association between dysmorphic concern and psychiatric morbidity using multivariate analysis to control for the effect of possible confounding variables.
Limitations of the study include the fact that psychiatric caseness was only measured using the GHQ and was not confirmed by a standardized psychiatric interview carried out blind to the plastic surgeon's assessments. In addition, there were insufficient subjects to compare the effect of different types of carcinoma (e.g. basal cell carcinoma as opposed to breast carcinoma) on dysmorphic concern and psychiatric status. The sample was recruited in a public hospital and so the results may not be generally applicable to private practice. The use of an equal number of consecutive rather than random controls may have introduced selection bias as cases and controls were not recruited over exactly the same time frame.
Dysmorphic concern and psychiatric morbidity
As expected, patients presenting for cosmetic reasons were more likely to have high DCQ and GHQ scores, and to present for mammoplasty. However, the finding that women were more likely to present for cosmetic reasons and to have higher rates of dysmorphic concern is in contrast to earlier studies where men and women were equally affected [2, 11], but consistent with a UK study where 75% of the patients were female [23].
There was a strong positive association between dysmorphic concern symptoms (the central component of BDD) and psychiatric morbidity as rated by the GHQ. Statistical analysis indicated that patients with high DCQ scores were 32 times more likely to be GHQ cases and had higher rates of psychiatric morbidity (93%) than those presenting for carcinoma (15%). This finding is consistent with previous work on BDD [13, 15, 24]. Veale et al. found that of 50 patients with BDD, 26 per cent had either dysthymia or a major depressive disorder [23], while Phillips et al. in a study of 130 cases of diagnosed BDD found 59% qualifying for a concurrent diagnosis of major depression [13].
The association between psychiatric morbidity and dysmorphic concern in this study persisted on multivariate analysis to control for the effect of any possible confounding variables such as gender or type of procedure. It is unlikely that the association between GHQ and DCQ scores was due to any overlap between the domains surveyed by the two instruments as the GHQ does not cover any aspect of somatic concern
These findings suggest that patients presenting for cosmetic surgery to plastic surgeons have high rates of dysmorphic concern and psychiatric morbidity. As surgical and dermatological interventions usually do not help patients with dysmorphic concern [13], cognitive behavioural therapy or selective serotonin re-uptake inhibitors might be alternative treatments [13, 25–28].
Footnotes
Acknowledgements
The authors would like to thank Dr David Castle for advice concerning the DCQ and Dr Andrew Wenzel for assistance with the patient interviews.