Anxiety Disorders 20 (2006) 296–311
Is ‘‘shy bladder syndrome’’ (paruresis) correctly classified as social phobia? Philipp Hammelstein a,*, Steven Soifer b a
Department of Clinical Psychology, Heinrich-Heine-University, Universitaetsstrasse 1 (Geb.23.03), 40225 Duesseldorf, Germany b School of Social Work, University of Maryland, Baltimore, USA
Received 22 September 2004; received in revised form 3 December 2004; accepted 18 February 2005
Abstract Paruresis manifests in an inability to urinate in public restrooms followed by a considerable avoidance behavior. According to DSM-IV TR this disorder is classified as social phobia. A sample of N = 226 subjects completed different questionnaires concerning paruresis, social phobic symptoms, lower urinary tract symptoms and depressive symptoms. These individuals were divided into four groups: no symptoms, suffering primarily from paruresis, non-generalized social phobia and generalized social phobia. The paruretic group differs significantly in all symptom variables from both the non-generalized and the generalized social phobia groups. Regression analysis separated by groups shows that the interference with everyday life can be mainly explained by paruretic symptoms (in the paruretic group) or by social anxiety and depressive symptoms, respectively (in the social phobic groups). These results question the classification of paruresis as simply being a form of social phobia. # 2005 Elsevier Inc. All rights reserved. Keywords: Paruresis; Social phobia; Shy bladder; Diagnosis; Comorbidity
Paruresis manifests in a fear and inability to urinate in public restrooms when other persons are present or may enter the room (Soifer, Zgourides, Himle, & Pickering, 2001; Zgourides, 1987). This disorder was first labeled as ‘‘paruresis’’ and described as a ‘‘disorder of micturition’’ by Williams and Degenhardt (1954). According to DSM-IV-TR (American Psychiatric Association, 2003) this disorder * Corresponding author. Tel.: +49 211 8113482; fax: +49 211 8114261. E-mail address:
[email protected] (P. Hammelstein). 0887-6185/$ – see front matter # 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.janxdis.2005.02.008
P. Hammelstein, S. Soifer / Anxiety Disorders 20 (2006) 296–311
297
falls under the category of a social phobia. However, most descriptions, books and articles concerning social phobias do not refer to paruresis (Hofmann & DiBartolo, 2001). Fifty years after the first systematic description, there is still little knowledge about the origin and specific features of this disorder. Unfortunately, most literature is based on treatment case reports (Hatterer, Gorman, Fyer, & Campeas, 1990; Lamontagne & Marks, 1973; McCracken & Larkin, 1991; Nicolau, Toro, & Perez Prado, 1991; Sagar & Ahuja, 1988; Zgourides & Warren, 1990; Zgourides, Warren, & Englert, 1990), while other studies examine voiding dysfunction in sub-clinical samples such as college students (Gruber & Shupe, 1982; Rees & Leach, 1975). Based on these descriptions of paruresis, two primary classifications can be identified (even though the existing literature did not consider these classifications). On the one hand, the functional nature of paruresis was emphasized and the syndrome was consequently described as a ‘‘functional disorder of micturition’’ (Williams & Degenhardt, 1954; Zgourides, 1987) or as ‘‘psychogenic urinary retention’’ (Christmas, Noble, Watson, & Turner-Warwick, 1991; Lamontagne & Marks, 1973; Wheeler & Renshaw, 1995). On the other hand, presence of anxiety and avoidance were pointed out, and paruresis was described as a social anxiety disorder (Malouff & Lanyon, 1985; Soifer et al., 2001). Supporting evidence exists for both views. In the case of paruresis as a functional disorder, the patients’ subjective complaints are the substantiation: patients emphasize their inability to urinate and often negate the experience of fear (Hammelstein, Ja¨ntsch, & Barnett, 2003; Hammelstein, Pietrowsky, Merbach, & Bra¨hler, in press). Yet, analysis of social phobia subtypes raise concerns about classification of paruresis solely as a social anxiety disorder. In analyzing the items on the ‘‘Liebowitz Social Anxiety Scale’’ (LSAS; Liebowitz, 1987), Heimberg et al. found that the item ‘‘urinating in (use of) a public restroom’’ was shown to be the most different from other items on the LSAS (Heimberg, Holt, Schneier, Spitzer, & Liebowitz, 1993). Based on this result Heimberg et al. postulated that paruresis should be eliminated as an example of social phobia in the text of DSM-IV. Furthermore, medical treatments which are effective for some forms of social anxiety (like beta-blockers or MAO-inhibitors) do not improve paruretic symptomatology (Hatterer et al., 1990; Zgourides, 1988, 1991; Zgourides & Warren, 1990). Classification of paruresis as social anxiety is a good match to the features of the disorder itself: the pronounced avoidance behavior and the fear of negative evaluation (people recognizing and ridiculing the paruretic’s voiding dysfunction). Studies in sub-clinical samples have found higher values of self-reported interpersonal anxiety and performance anxiety in subjects suffering from paruretic symptoms in contrast to control subjects (Malouff & Lanyon, 1985). When examining male college students (n = 90), Gruber and Shupe (1982) found a strong relationship between paruretic symptoms and body shyness and found
298
P. Hammelstein, S. Soifer / Anxiety Disorders 20 (2006) 296–311
that more paruretic symptoms are associated with increased feelings of fear and self-centeredness. According to the affect model of depression and anxiety (Clark, Watson, & Mineka, 1994; Watson, Clark, & Carey, 1988), Hammelstein and Meyer (submitted for publication) found a higher negative affect but no differences in positive affect when comparing subjects suffering from paruresis with controls which further support the classification of paruresis as a form of anxiety. Vythilingum, Stein, and Soifer (2002) reported a high comorbidity between paruresis and other forms of social anxiety and took that as a symbol of the association between paruresis and social anxiety. However, according to DSM-IV-TR (American Psychiatric Association, 2003) this disorder is designated as a social phobia and—as far as we know—no study exists that compares subjects suffering from paruresis with subjects suffering from other forms of social phobia. The authors’ study tries to fill this gap by comparing paruresis with different forms of social phobia symptomatology (paruretic symptoms, social anxiety symptoms, depressive symptoms, lower urinary tract symptoms) and their interference with everyday life. Examining possible similarities and differences between paruresis and social phobias, not only mean differences of self-reported complaints, are of interest to see the level of interference with everyday life within these groups. The authors assume that paruresis is a form of social phobia. That means we hypothesize that paruretic subjects differ from pure social phobic subjects only in the severity of paruretic symptoms, not in the severity of social anxiety symptoms or lower urinary tract symptoms. We assume that in contrast to pure social phobic subjects, the level of interference with everyday life in paruretic subjects can be predicted mainly by the severity of paruretic symptoms.
1. Methods 1.1. Subjects To ensure a maximal variance of paruretic and social phobic symptoms, subjects were recruited through different Internet self-help forums. This onlinerecruiting seems appropriate because subjects suffering from paruresis rarely visit general practitioners or therapists. The questionnaire was placed on nine German self-help forums (three forums for social phobia, three forums for anxiety disorders, two forums for paruresis and one forum for male-specific problems). These self-help forums which are dedicated to specific clinical problems (social phobia, anxiety disorders, paruresis) have similar compositions. On all these forums the web user can gather information about the etiology of the disorder and about different treatment possibilities. On every forum there is a kind of virtual message board to leave notes or questions for other web users. The male-specific forum is an informational service that is non-profit and provides information about different male-specific themes (intimate relationships, sexuality, education,
P. Hammelstein, S. Soifer / Anxiety Disorders 20 (2006) 296–311
299
alimony, child custody, etc.). On this forum there is no virtual message board. Our sample consisted of 226 subjects (female subjects: n = 82, 36.3%) between the ages of 15 and 55 (M = 29.2, S.D. = 8.2). Level of education was quite high; with the average amount of education ranging from 8 to 18 years (M = 11.6 (S.D. = 1.7). Male subjects were older than female subjects (F(1, 221) = 7.23, P < .01, Cohen’s d = .38), but they did not differ in years of education (F(1, 179) = .36, P = .55, Cohen’s d = .09). 1.2. Measures In addition to standard demographic information, the participants were given the following questionnaires. 1.2.1. Anxiety Disorders Inventory Schedule—Revised (ADIS-R) This scale allows for classification of subjects in diagnostic categories concerning social phobia on the ADIS-R (DiNardo, Moras, Barlow, & Rapee, 1993; German version: Margraf, Schneider, & Ehlers, 1994). At the moment there is no German translation of the most current ADIS-IV (DiNardo, Brown, & Barlow, 2004), but to our knowledge there are no substantial differences concerning the items in the section about social anxiety. We removed ‘‘the use of public rest rooms’’ from the list of possible feared situations. That was done to ensure that no subject would get the ‘‘diagnose’’ of social phobia based on paruretic symptoms. If the subject met all DSM-criteria for social phobia, he or she was ‘‘diagnosed with’’ it. However, this diagnose was surely not validated by an interview. This diagnosis indicates that the subjects described themselves as persons who (a) have a marked and persistent fear of social or performance situations in which embarrassment may occur, (b) experience concerns about embarrassment and are afraid that others will see them as being anxious, weak, ‘‘crazy,’’ or stupid, (c) experience at least moderate fear in one of the listed situations, (d) experience anxiety whenever they are exposed to this social situation, (e) experience significant interference with everyday life caused by the distress or the corresponding avoidance behavior and (f) recognize that their fear is excessive or unreasonable. 1.2.2. Paruresis Checklist (PCL) The PCL was first described as a screening-instrument for self-administration by Soifer et al. (2001), however, the original version has no statistical validity. A slightly modified Germen version was used in another study (Hammelstein et al., in press). The PCL consists of 10 dichotomous items which are geared towards the DSM-IV criteria of Social Phobia, and thereby are specific to paruretic symptoms (e.g., ‘‘Do you have a marked and persistent fear of using public restrooms while others are present?’’ or ‘‘Do you avoid urinating in public restrooms and/or do you endure the public restroom situation with intense anxiety or distress?’’). To get an
300
P. Hammelstein, S. Soifer / Anxiety Disorders 20 (2006) 296–311
index of the severity of symptoms, the first eight items are summed up (the last two are diagnostic control items), leaving us with a score ranging from 0 to 8. The German version has shown a sufficient reliability in a representative male sample (Cronbach’s a = .75; Hammelstein et al., in press). A cut-off-value of 5 has been postulated by Hammelstein et al. (in press) because this sum value is associated with a marked increase in the relative risk of significant interference with everyday life. The coefficient alpha in the present study was .88. 1.2.3. Paruresis-Scale (PARS) The Paruresis-Scale is a German alternative instrument for measuring paruretic symptoms (Hammelstein & Pietrowsky, in press). The PARS consists of 13 items concerning paruretic symptoms and corresponding avoidance behavior which are rated on a 5-point scale (1 = not at all, 5 = extremely). It has proven to be reliable (Cronbach’s a = .94) and has a higher discriminative power when used exclusively in subjects with low or high paruretic symptoms compared to the PCL. The PCL and PARS have a correlation of r = .84. In the present study PARS shows a high internal consistency with a = .96. 1.2.4. International Prostate Symptom Score (I-PSS) This questionnaire consists of seven symptoms (incomplete emptying, frequency, intermittency, urgency, weak stream, hesitancy and nocturia) to be rated on a 5-point scale (0 = not at all, 4 = always) (Barry et al., 1992). The sum value allows the severity of lower urinary tract symptoms (LUTS) to be assessed. Though it was developed to assess and measure voiding symptoms in men with benign prostatic hyperplasia, several studies have shown that the I-PSS is neither prostate- nor gender-specific (Chai, Belville, McGuire, & Nyquist, 1993; Lepor & Machi, 1993). Therefore, the I-PSS is deemed to be a measure of LUTS. With a = .81 (Schneider, Ludwig, Weidner, & Bra¨hler, 2003), the reliability of the German version has been shown. The coefficient alpha in the present study was .72. 1.2.5. Social Phobia Scale (SPS) In this study, the German version of the Social Phobia Scale (SPS; Mattick & Clarke, 1998; German version: Stangier, Heidenreich, Berardi, Golbs, & Hoyer, 1999) was used. The SPS assesses fears of being scrutinized during routine activities. It consists of 20 items to be rated on a 5-point Likert scale. The German version has shown to be a reliable and sensitive instrument (Heinrichs et al., 2002; Stangier et al., 1999) with a high internal consistency (a .90). In the present study the coefficient alpha is .93. 1.2.6. Social Interaction Scale (SIAS) The parallel instrument of the SPS is the SIAS (Mattick & Clarke, 1998; German version: Stangier et al., 1999). It also consists of 20 items to be rated on a 5-point scale and assesses a person’s fear of more general social interactions.
P. Hammelstein, S. Soifer / Anxiety Disorders 20 (2006) 296–311
301
With Cronbach’s a .90 (Heinrichs et al., 2002; Stangier et al., 1999), the German version of the SIAS seems to be as reliable as it is in the present study (a = .90). For the SIAS as well as the SPS, the more conservative German cut-off values of Heinrichs et al. (2002) are used. 1.2.7. Center for Epidemiological Studies Depression Scale (CES-D) The German version of the slightly modified CES-D (German version: Hautzinger, 1993; original version: Radloff, 1977) was used. It consists of items designed for the self-evaluation of depressive symptoms to be rated on a 4-point scale (0 = rarely, 3 = mostly). Since the German factor solution does not correspond to the original version, only the sum value is considered. The German average sum values are significant higher than the American ones. Therefore, the German cut-off value is 23 not 16 (Hautzinger, 1993). In the present study alpha coefficient is high with a = .92. 1.2.8. Subjective interference with everyday life The subjective interference with everyday life was assessed using six items concerning different areas of life (family commitments, recreation, social activities, job, sexual satisfaction, and keeping the house). Concerning each area of functioning the subject had to rate the amount of impairment on an 11-point scale ranging from ‘‘no interference at all’’ (0) to ‘‘absolute interference (11). ‘‘In order to make the rating easier, the six areas were clarified (e.g., the area ‘‘recreation’’ was explained by: ‘‘This area comprises hobbies, sport and other leisure time activities’’). Internal consistency of these six items seems sufficient (Cronbach’s alpha = .76) for building a sum value.
2. Results 2.1. Assigning to groups Assigning the survey responders to groups was done in several steps. First, subjects were diagnosed with ‘‘social phobia’’ on the basis of ADIS-R items (see above). Next, those subjects who reported the greatest level of suffering from paruresis were diagnosed with ‘‘paruresis’’, independent of their status concerning the social phobia diagnosis. Subjects who did not meet the criteria for social phobia and did not state that they suffer from paruresis much were assigned to the control group. The diagnostic group of social phobia was further broken down into the groups ‘‘non-generalized social phobia’’ (NG-SP) and ‘‘generalized social phobia’’ (G-SP) according to the postulation of Heimberg et al. (1993). For this purpose the feared situation items were assigned to the four situational domains (formal speaking and interaction; informal speaking and interaction; assertive interaction; observation of behavior by others). If a subject reported at least moderate anxiety in at least one situation from all four categories,
302
P. Hammelstein, S. Soifer / Anxiety Disorders 20 (2006) 296–311
he or she got the classification of ‘‘generalized social phobia’’; otherwise, they were assigned to the ‘‘non-generalized social phobia’’ group. Of the total sample (n = 226, female subjects: n = 82, 36.3%) subjects were assigned to the group NG-SP (n = 103, portion of female subjects: n = 40), the group G-SP (n = 50, thereof 30 female subjects) and the group of ‘‘paruresis’’ (PAR; n = 37, thereof three female subjects). Thirty-six subjects got no diagnosis (‘‘control subjects’’; thereof nine female subjects; see Table 1). Within the group of subjects suffering from paruresis, n = 26 subjects received the diagnosis of ‘‘non-generalized social phobia’’ and n = 2 subjects were diagnosed with ‘‘generalized social phobia’’ in addition to their paruretic symptoms. 2.2. Comparison of groups concerning demographic aspects As expected, the gender ratio is different between groups (x2 = 27.14, P < .001), and age of the subjects is slightly different between the groups (F(3, 219) = 2.45, P = .06). Groups also differ in years of education (F(3, 177) = 5.12, P < .01). Pairwise comparisons after Bonferroni adjustment show that control subjects have significantly more years of education than subjects with NG-SP and subjects with G-SP, respectively (P < .05). Fifty percent of the subjects with any diagnose (n = 95) had visited a general practitioner regarding their symptoms. On this item, the groups do not differ significantly (x2 = 2.05, P = .36). However, they do differ concerning the question of consultation with a therapist (x2 = 8.11, P < .05): with about 54% of the social phobic groups (n = 44 of NG-SP subjects and n = 25 of G-SP subjects) having had psychotherapy, but only 26% (n = 9) of the paruretic group. However, former consultation has no effect on current interference with everyday life (rpbis = .02 concerning physician; rpbis = .12 concerning psychotherapist). 2.3. Comparison of groups concerning pathological symptoms Since n = 20 subjects had missing values in I-PSS, a covariance analysis was calculated separately for I-PSS as a dependent variable and the other remaining dependent variables. Gender and age were included as covariates. For I-PSS there is a tendentially significant effect of the factor group (F(3, 200) = 2.61, P = .05), with a significant influence of the covariate age (F(1, 200) = 16.35, P < .001) but not of the covariate gender (F(1, 200) = .05, P = .82). Pairwise comparisons show the strongest differences between the groups of control subjects and generalized social phobia and between the groups of generalized and non-generalized social phobia. The MANCOVA was run with group as factor, sex and age as covariates, and with the following dependent variables: paruretic symptoms (PCL, PARS), social phobia symptoms (SIAS, SPS), depressive symptoms (CES-D) and interference with everyday life. There is a significant main effect of group (F(18, 609) = 25.91,
Paruresis
Age PARS PCL I-PSS SPS SIAS CES-D Subj. Interf.
Non-generalized SP
Generalized SP
Control subjects
Female (n = 3)
Male (n = 34)
Female (n = 40)
Male (n = 63)
Female (n = 30)
Male (n = 20)
Female (n = 9)
Male (n = 27)
28.33 28.67 6.67 8.00 23.33 24.33 11.67 3.56
31.65 36.59 7.18 7.81 21.15 26.97 14.59 2.70
28.20 4.25 2.05 6.58 28.55 44.35 28.62 4.43
30.38 11.65 3.48 6.39 29.13 45.05 27.87 5.17
26.03 8.07 2.20 7.52 45.27 54.27 31.27 5.51
27.85 18.45 4.00 8.35 46.00 59.10 35.30 6.83
26.52 2.00 1.44 5.38 4.22 7.78 11.88 1.00
30.19 16.63 2.52 5.83 9.55 15.80 14.44 1.81
(2.88) (24.83) (1.53) (1.41) (8.15) (9.29) (5.03) (1.08)
(9.47) (8.63) (1.14) (5.10) (11.00) (12.05) (8.96) (1.21)
(7.86) (7.91) (1.94) (4.57) (13.37) (16.01) (13.36) (1.97)
(8.13) (11.95) (2.44) (4.369) (14.24) (13.82) (10.13) (1.70)
(7.93) (11.49) (1.88) (4.82) (13.43) (11.77) (10.73) (1.58)
(8.57) (15.29) (2.38) (6.06) (12.83) (11.01) (9.42) (1.45)
(2.43) (4.09) (1.24) (3.96) (3.49) (5.89) (10.34) (2.35)
(7.55) (16.36) (2.71) (4.36) (10.35) (12.07) (9.30) (2.34)
P. Hammelstein, S. Soifer / Anxiety Disorders 20 (2006) 296–311
Table 1 Mean values and standard deviations (in brackets) of age, paruretic symptoms (PARS, PCL), lower urinary tract symptoms (I-PSS), social phobic symptoms (SPS, SIAS), depressive symptoms (CES-D) and subjective interference with everyday life (Subj. Interf.)
303
304
P. Hammelstein, S. Soifer / Anxiety Disorders 20 (2006) 296–311
Table 2 Comparison of univariate effects (via MANCOVA) of group on the dependent variables and the results of the following pairwise comparisons Univariate comparisons
Pairwise comparisons
df
F
P
PARS
3, 220
37.65
