European Psychiatry 22 (2007) 69e74

http://france.elsevier.com/direct/EURPSY/

Original article

Epidemiology of specific phobia subtypes: Findings from

the Dresden Mental Health Study
Eni S. Becker a,*, Mike Rinck a, Veneta Türke b, Petra Kause b, Renee Goodwin c,
Simon Neumer d, Jürgen Margraf e
a
Radboud University Nijmegen, Clinical Psychology, Behavioural Science Institute, PO Box 9104, 6500 HE, Nijmegen, Netherlands
b
University of Technology Dresden, Clinical Psychology and Psychotherapy, Dresden, Germany
c
Columbia University, Department of Epidemiology, New York, NY, USA
d
Regionscenter for barne og ungdomspsykiatrie, Oslo, Norway
e
University of Basel, Clinical Psychology and Psychotherapy, Basel, Switzerland
Received 12 September 2006; accepted 14 September 2006
Available online 8 December 2006

Abstract

This study determined the prevalence, age of onset, comorbidity, and impairment associated with specific phobia subtypes in the community.
Data were drawn from the Dresden Mental Health Study (N ¼ 2064), a representative community-based sample of young women in Dresden,
Germany. The lifetime prevalence of any specific phobia was 12.8%, with subtypes ranging in prevalence between 0.2% (vomiting, infections)
and 5.0% (animals). There were significant differences in the mean age of onset of specific phobias. Significant differences in comorbidity pat-
terns also emerged between subtypes. No significant differences were found in level of impairment associated with the subtypes. The findings
suggest that specific phobias are common among young women and that they differ in prevalence, associated comorbidity, and mean age of
onset. These data suggest significant differences in the phenomenology and clinical significance of specific phobia subtypes.
Ó 2006 Elsevier Masson SAS. All rights reserved.

Keywords: Specific phobias; Subtypes of phobias; Onset; Comorbidity

1. Introduction in life, and that they predict the subsequent onset of depressive
and substance use disorders in adolescence and adulthood [7].
For a long time, specific phobias (SPs) were considered Specific phobias all share a ‘‘marked and persistent fear of
a common, but inconsequential psychological problem. How- clearly discernible circumscribed object or situations’’ accord-
ever, increasing evidence has shown that SPs are clinically ing to the DSM-IV [3]. DSM-IV differentiates among four
significant and relatively understudied disorders. First, epidemi- types of SPs: animal type, natural environment type, blood-
ologic studies show that SPs are the most common mental disor- injection-injury type, and situational type. There is an addi-
ders in women [12,17], with lifetime prevalence rates doubling tional fifth category, ‘‘other’’, including phobias with fear of
those of men [12]. Second, studies have documented significant noise and costumes. These categories were arrived at by clus-
rates of impairment, distress, and comorbidity associated with tering according to content of fears, but there is little data
SPs [22]. Although comorbidity rates are assumed to be moder- about the coherence and validity of these clusters. There is
ate compared to other mental disorders, there are not enough some evidence that the various subtypes have different ages
data to draw firm conclusions. Third, longitudinal epidemio- of onset. However, data suggest that phobias might be hetero-
logic studies suggest that SPs have a mean age of onset early geneous even within DSM-IV categories. Unfortunately, rela-
tively few studies have specifically examined subtypes, and
* Corresponding author. Tel.: þ31 (0) 24 361 2668; fax: +31 (0) 24 361 5594. even fewer have differentiated between single SPs. Moreover,
E-mail address: [email protected] (E.S. Becker). interpretation of such comparisons is often aggravated by the

0924-9338/$ - see front matter Ó 2006 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.eurpsy.2006.09.006
70 E.S. Becker et al. / European Psychiatry 22 (2007) 69e74

inclusion of different SPs in subtypes, e.g. some studies of sit- axis I disorders according to DSM-IV, for lifetime and point-
uational phobias include only claustrophobia [16], others only prevalence. The F-DIPS is an earlier, almost identical version
driving phobia [2]. In sum, previous community-based studies of the DIPS [19], based on the ADIS-L [5]. The following
have mainly reported general categories of phobia prevalence disorders can be diagnosed: all anxiety disorders, all affective
and associated morbidity. Yet, relatively little is known about disorders, the research-diagnosis mixed anxiety-depression, hy-
SP subtypes in the community. pochondriasis, somatization disorder, conversion disorder and
Furthermore, several methodological features of previous pain disorder, substance abuse and dependence, bulimia, an-
studies have limited the ability to generalize the findings. First, orexia, and some children’s disorders (separation anxiety, atten-
findings from clinical studies can not be generalized to the tion-deficit and disruptive behavior disorders, elimination
population, especially when only a small percentage of those disorders). Furthermore, there is a socio-demographic section,
with SPs seek treatment. Second, very few epidemiological a screening for psychosis, a screening for the general medical
studies have looked at comorbidity of SPs [7], and there is condition and medication, a short section on family history of
even less information available on the comorbidity of SP sub- psychological disorders, and a section about treatment for psy-
types. Studies with clinical samples are difficult to interpret in chological disorders. The F-DIPS has good reliability (Kappa
terms of comorbidity, since it is over represented in clinical for anxiety disorders 0.64, affective disorders 0.71, somatoform
samples [15,20]. Yet, comorbidity is of great importance, be- disorders 0.66, Yule for substance abuse 0.85 and for eating dis-
cause it informs us of the relations between different disorders, orders 0.94; [9]). Interviewers were either psychologist, physi-
and it may also provide valuable information about differences cians, or psychology students in their last years of training.
between subtypes. More importantly, SPs are often the earliest All underwent an extensive one-week training. All interviewers
manifestation of psychological disorders, and prevention strat- attended supervision bi-weekly. Specially trained supervisors
egies will benefit from knowing more about comorbidities. proofread every interview.
Against this background, the goals of the current study In addition to the diagnoses, the severity of the disorders
were threefold. First, the study determined the prevalence of was rated on a scale ranging from 0 to 8 (0 ¼ no severity to
SP subtypes in the community among women. Second, the 8 ¼ maximum severity). Furthermore, the interviewer rated
study assessed the mean age of onset of each subtype. Third, the degree of distress and of impairment due to the diagnoses
the study investigated associations between SP subtypes and on scales ranging from 0 (no distress or impairment) to 8 (se-
psychiatric comorbidity and impairment. vere distress and impairment). Interviewers were extensely
trained in administering those scales. There were several ques-
2. Method tions to pinpoint the onset of the disorder. Due to the early on-
set of SP, the exact date was hard to determine. Participants
The Dresden Study of mental health is a prospective epide- not remembering the onset were not included in the onset anal-
miological study designed to collect data on the prevalence, yses. If participants claimed always to have suffered from
incidence, course, and risk factors of mental disorders. The re- a certain fear, the age of onset was set to 3 years.
sults presented here are from the baseline survey, which was
conducted from July 1996 to September 1997. 2.3. Analytic strategy

2.1. Participants Data were analyzed using the Statistical Package for Social
Sciences, SPSS, Windows, German Version 8.0. Statistical
The sample was drawn from the Dresden government regis- tests included analyses of variance and Chi-square tests. The
try of residents. 5204 women were located and eligible (age 18e bivariate comorbidities were obtained by estimating odds ra-
24, German) for the study. From this sample, 2064 filled out tios using the ‘‘risk’’ procedure in the SPSS software package.
questionnaires and took part in the structured interview of men- Cumulative incidence (in percent) curves are presented to il-
tal disorders, and 998 filled out questionnaires only, yielding lustrate age of onset.
a response rate of 58.8%. The demographic data of these women
did not differ from the data of those who did not reply. The re- 3. Results
sponse rate is somewhat lower than in other studies. One reason
might be that women of that age group usually have the lowest 3.1. Socio-demographic data
response rate (Wittchen, personal communication). Further-
more, the data were collected shortly after the German reunifi- The majority of the participants had a partner or spouse
cation, when people were still very mistrustful regarding (66.6%), but very few were married (4.5%), separated or di-
surveys. And third, participants could not be reimbursed. Data vorced (0.4%). About half of the women were living with their
of the 2064 interviewed participants are reported here. parents; about a quarter with a partner or spouse and about
14% lived alone. 7% had children of their own. A few women
2.2. Diagnostic assessment were still attending school (3.6%); about 40% were students at
the university. Very few dropped out of school without a degree
The diagnostic assessment was based on the F-DIPS [13]. (0.4%), consistent with mandatory German school law
The F-DIPS is a structured interview allowing the diagnosis of (school-leaving certificate). A minority of the participants
 E.S. Becker et al. / European Psychiatry 22 (2007) 69e74 71

went to a ‘‘Hauptschule’’ (3.7%), the lowest level of school (e.g., situation phobias). This is due to some participants re-
education, approximately one third went to the medium level porting situational phobias that did not fit into the given sub-
of schooling (Realschule und Polytechnische Oberschule), groups (e.g., fear of ships). Animal phobia (mean 6.2) and
and 58% ended schooling with a degree that allows them to environmental phobia (mean 6.4), start early in life. Interest-
attend university (Abitur). Almost half of the young women ingly, phobia of heights (mean 9.1) has a later onset than the
were working, 31.3% of the whole sample full-time, 16% other environmental phobias. Most SPs begin around the age
part time. Approximately 5% were unemployed. There were of 8, although situational phobias (mean 13.4) start much later.
no differences concerning sociodemographic data between Here phobia of lifts starts earlier around the age of 9 years. A
the different SP subtypes or fears. closer look at physical phobias shows that phobia of infections
starts later than phobia of doctors or of vomiting. Fig. 1 shows
3.2. Prevalence rates of SPs the age curves for SP subgroups. The age curve for the cate-
gory ‘‘other’’ shows a discontinuity. This is due to phobias
As shown in Table 1, 12.8% of the young women fulfilled of noise and costumes starting very early in life, whereas the
the criteria of a lifetime prevalence (LP) of at least one SP. The other phobias like exam phobia start much later. The curves
point prevalence (PP; disorder was present in the last 7 days up show that animal and environmental phobias cluster together;
to the interview) for all SPs amounted to 9.9%, and the 12- blood phobia being just slightly later. Phobia of heights lies in-
month prevalence to 10%. Animal phobias were reported between, and situational phobias start much later.
most frequently (PP 4.3%). They were followed by blood pho- Duration needs careful interpretation, since the participants
bias (1.8%), physical phobias (fear of seeing a doctor PP 1%, are still very young and therefore phobias starting later in life
vomiting PP 0.1%, and fears of contagion 0.1%), and heights will necessarily have shorter durations. Nevertheless, the mean
(1.6%). Heights are usually clustered with other environmental durations show that most phobias are persistent. They begin
phobias, as storms and water. However, phobias of storms (PP while the women are still children and persist into adulthood.
0.1%) and water (PP 0.3%) occurred rarely. Situational pho-
bias were evident in 1.9% of the women, with driving phobia 3.4. Comorbidity of SPs
being the most common (PP 0.9%), followed by phobias of
lifts and enclosures (PP 0.5%) and flights (PP 0.4%). The Comorbidity is illustrated in Table 3. Comorbidity was de-
‘‘other’’ category contains phobias of noise, costumes, and fined as the occurrence of disorders at a lifetime prevalence
exams and occurred hardly at all. The 12-month prevalence level. The lifetime prevalence of the disorders were: 28.3% for
and the lifetime prevalence showed a similar pattern of occur- anxiety disorders, 13.7% for affective disorders, 3.2% for soma-
rence. This suggests that SPs are rather stable since prevalence toform disorders, 2.2% for substance related disorders, 4% for
rates do not change much over the different time frames. eating disorders, and 9.8% for disorders usually occurring in
childhood. An odds ratio (OR) larger than 1 indicates a positive
3.3. Age of onset of SPs relationship between SP and other mental disorders. As the con-
fidence intervals (CIs) show, however, not all of these relation-
Table 2 shows the age of onset (median and mean) and the ships are reliable. Furthermore, the N of some cells were so
duration of SPs. The number of participants is sometimes small that the CIs are very wide. Thus, only ORs that are signif-
higher in the aggregated groups than in the specific groups icant with Ns bigger than 4 are interpreted and printed in bold in

Table 1
Prevalence rates in percent and N for specific fears and specific phobias
Life time prev. phobias 12 month prev. phobias Point prev. phobias
(N ¼ 326) (N ¼ 264) (N ¼ 243)
Animals 5.0 (110) 4.5 (93) 4.3 (89)
Heights 1.9 (39) 1.7 (35) 1.6 (34)
Environment (except heights) 0.7 (15) 0.6 (13) 0.5 (10)
Storms 0.3 (7) 0.3 (6) 0.1 (3)
Water 0.3 (7) 0.3 (7) 0.3 (7)
Blood, injuries and shots 2.4 (50) 1.9 (39) 1.8 (38)
Situational 2.6 (53) 2.2 (45) 1.9 (39)
Flights 0.4 (9) 0.4 (9) 0.4 (8)
Lifts 0.7 (17) 0.6 (13) 0.5 (11)
Driving 1.1 (23) 1.0 (20) 0.9 (18)
Physical 2.2 (46) 1.5 (31) 1.3 (27)
Doctors 1.7 (37) 1.1 (23) 1.0 (21)
Vomiting 0.2 (4) 0.2 (4) 0.1 (3)
Infections 0.2 (4) 0.2 (4) 0.1 (3)
Others 0.6 (13) 0.4 (8) 0.3 (6)
All 12.8 (265) 10.0 (206) 9.9 (204)
72 E.S. Becker et al. / European Psychiatry 22 (2007) 69e74

Table 2 with substance use disorders and childhood disorders, but the
Mean age of onset and mean duration of specific phobias wide CIs have to be taken into account, and therefore the signif-
N Age Duration icant associations have to be interpreted with caution. Height
Median Median phobia, the other environmental phobia, is only associated
Mean (SD) Mean (SD) with anxiety disorders, showing a different pattern from the
Animals 107 4 16
other environmental disorders. Phobia of blood and injury is re-
6.25 (4.62) 14.24 (5.28) lated to anxiety disorders, too, and also to disorders usually oc-
Heights 36 8 10 curring in childhood. Physical phobias are also related to the
9.17 (5.88) 11.14 (6.02) anxiety disorders. The situational phobias show significant asso-
Environment 15 3 14 ciations with affective disorders, disorders usually occurring in
6.47 (5.05) 13.27 (5.12)
Storms 7 3 14
childhood, and probably also with substance use disorders.
4.86 (2.67) 14.00 (4.04)
Water 7 8 14.5 3.5. SPs and their effect on quality of life
8.57 (6.50) 12.63 (6.12)
Blood 48 6 12.5 Table 4 shows indicators of the severity of SP and of the de-
7.39 (4.87) 12.42 (5.15)
Situational 50 15.5 4
gree of impairment in everyday life. All ratings were provided
13.44 (7.03) 6.64 (6.41) by interviewers. As the results show, there were almost no dif-
Flights 9 15 6 ferences concerning the severity ratings between the different
14.67 (6.10) 6.33 (4.77) SPs. On average, all disorders displayed moderate severity
Lift 15 4 9 (values slightly above 4 on a scale ranging from 0 to 8) with
9.53 (7.11) 9.60 (7.35)
Driving 22 19 2
comparable standard deviations, and an analysis of variance
16.27 (5.95) 4.55 (5.65) revealed no differences between them (F(6/307) ¼ 0.28,
Physical 46 7 11 n.s.). The same was true for the impairment caused by the
9.13 (5.18) 10.39 (5.97) SPs (F(6/324) ¼ 0.56, n.s.). However, differences emerged
Doctors 37 7 11 regarding the ratings of distress in everyday life (F(6/
8.73 (5.20) 10.51 (5.98)
Vomiting 4 8.5 13.5
324) ¼ 2.47, p ¼ 0.02), with phobias of heights being the least
7.50 (5.20) 12.75 (6.85) distressing and ‘‘other’’ SPs the most. To aid interpretation,
Infections 4 15.5 7 the current ratings may be compared to those of a group of
15.50 (1.29) 5.67 (2.31) women who had other mental disorders than SP at the time
Others 13 5 7.5 of the interview (N ¼ 155; all other diagnoses included).
8.77 (6.86) 9.40 (7.21)
These women showed similar ratings, with mean severity of
4.00 (SD 1.38), mean impairment 3.41 (SD 1.81), and mean
Table 3. Animal phobias are significantly associated with all distress 4.20 (SD 1.74).
other disorder groups except eating disorders. Interestingly,
the eating disorders show no relation to any specific phobia. En- 4. Discussion
vironmental phobias (of water and storm) might be associated
Specific phobias are a common problem for young women.
100%
The prevalence rates of SPs in our sample (12.4%) are consis-
90% tent with several previous studies (e.g. 10.2% [11]), 14.9%
80%
[12], and 10.1% [17]) with others report slightly higher rates
(e.g. 22.7% [6]), 25.9% [18], or 20.2% [10]). Comparison of
70%
prevalence rates reported here to those of other studies is dif-
60% ficult as studies differ with regard to the classification methods
50%
used. Age, gender, and a varying number of feared objects or
situations included in the survey also vary. Our findings show
40% animal phobias (LP 5.0%) to be the most prevalent, followed
30% by blood phobia (LP 2.4%) and phobia of heights (LP 1.9%).
This finding is consistent with those of Agras et al. [1] and
20%
Curtis et al. [6] who also found animal fears to be the most
10% prevalent. Taken together, situation phobias were prevalent
0% among 2.6% of the young women, with fear of driving being
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 the most common.
age
The mean age of onset for SPs was 7.9. Lindal and Stefans-
situational
animals environment (other
than heights)
son [11] found a slightly later onset of 9.6 years in their sam-
physical
heights blood etc. other
ple. In our study, animal phobias started early in life
(5.5 years). Similarly, Öst [16] found the mean age of onset
Fig. 1. Cumulative age for specific phobia subtypes. to be 6.9, while Starcevic and Bogojevic [21] found
 E.S. Becker et al. / European Psychiatry 22 (2007) 69e74 73

Table 3
Lifetime comorbidity of specific phobias with other mental disorders: N, Odds ratios, and 95%-CIs
Lifetime comorbidity Anxiety Affective Somatoform Substance Eating Childhood
N, OR, 95%CI disorders disorders disorders disorders disorders disorders
Animals 25 22 8 5 3 13
2.36 (1.44e3.88) 3.07 (1.83e5.15) 3.99 (1.83e8.74) 4.00 (1.51e10.58) 1.04 (0.32e3.40) 2.38 (1.28e4.43)
Heights 10 6 0 0 0 3
2.72 (1.23e5.99) 2.07 (0.83e5.19) 1.39 (0.42e4.69)
Environment 1 0 1 1 1 2
0.92 (0.11e7.94) 6.59 (0.76e57.38) 11.04 (1.25e97.24) 4.94 (0.57e42.85) 5.58 (1.01e30.71)
Blood, injuries 17 5 0 1 3 10
and shots 5.24 (2.59e10.60) 1.34 (0.51e3.52) 1.78 (0.24e13.45) 2.56 (0.76e8.59) 5.07 (2.36e10.91)
Situational 8 8 0 3 0 6
1.54 (0.69e3.46) 2.42 (1.07e5.45) 5.71 (1.66e19.72) 2.57 (1.04e6.35)
Physical 9 5 1 2 2 4
2.31 (1.03e5.19) 1.65 (0.62e4.40) 1.27 (0.17e9.51) 4.42 (1.00e19.45) 1.98 (0.46e8.51) 1.94 (0.66e5.69)
Others 3 5 1 1 0 3
2.31 (0.58e9.29) 9.07 (2.42e34.02) 4.12 (0.51e33.52) 6.90 (0.84e56.82) 5.58 (1.38e22.53)
The reference group for determining the OR was ‘‘no specific phobia’’.

a somewhat later onset at 9.7 years. In the Starcevic and Bogo- eating disorders, and animal phobias were related to all other
jevic study [21], blood phobias started quite late with mental disorders. Otherwise, all subgroups showed a different
21.1 years, whereas Öst [16] found a much earlier age of onset pattern of comorbidity. Overall, phobias start early in life, and
for blood phobia, 8.8, which is comparable to our finding of pose a risk for developing a second mental disorder. This, and
7.0 years. In contrast, situational phobias started later in life the rather long duration, make the need for treatment of SP in
around 15e17 years. Consistent with this, Starcevic and Bogo- childhood and adolescence obvious.
jevics [21] found a mean of 23.6 years. In addition, this study investigated how strongly women
In terms of age of onset, our data show that there is wide were impaired by their specific phobia. Specific phobias are
variation in specific situational phobias, with fear of elevators often considered less impairing than other disorders, since
starting much earlier than fear of flying, and fear of driving be- the feared object or situation is circumscribed and therefore
ing the latest phobia. The fear of infections appears to begin its avoidance is much easier than, for example, in social pho-
much later than the phobias of other physical illnesses and bia or agoraphobia [14]. Regarding ratings of severity, impair-
may be related to an early start of hypochondria rather than ment, and associated emotional stress, however, SPs did not
specific fears. Overall, phobias concerning situations or phys- differ from other mental disorders. Moreover, there were no
ical fears appeared to be acquired in young adulthood rather differences between the SP subtypes regarding impairment,
than during childhood. Phobias appear relatively stable and and only small differences regarding distress. Impairment in
have a long duration. Taken together, this study found some- everyday life seems to be as high as for other mental disorders
what earlier onset of SPs than other surveys. But since the in this young sample, thus there is no hint that SPs are
data are retrospective (as in most studies), the exact age of on- a smaller problem. Overall, the majority of the disorders in
set is hard to determine. this young sample is not extremely impairing.
This study also assessed specific relations between SP sub- Some limitations regarding the generalizability of this
types and other mental disorders. Almost all subtypes were study should be noted. First, only women were included in
associated with other anxiety disorders, except the environmen- this survey. Women usually show higher rates of SP [8]. Sec-
tal phobias. These phobias are probably related to substance use ond, our sample consisted of young women. Therefore, no
disorders and disorders occurring in childhood, but these rela- conclusions can be drawn for males or for an older sample. Al-
tions have to be interpreted with care because of the low number though phobias start early in life, this study showed that some
of cases. Interestingly, no specific phobia subtype was related to start much later than others (e.g., situational phobias), and
therefore some SPs might be underrepresented. Furthermore,
Table 4 all participants lived in Dresden, Germany. The ECA study
Impairment ratings: means and standard deviations showed that prevalence rates vary across cities [4]. Therefore,
Severity of Impairment Distress findings from one site might not be representative of findings
disorder
at another. As the socio-demographic data showed, more
Animals (N ¼ 110) 4.30 (1.28) 3.34 (1.84) 3.97 (1.50) highly educated women participated in our study, but data
Heights (N ¼ 39) 4.14 (1.22) 3.51 (1.76) 3.33 (1.87)
from the official census show that this is rather typical for
Environment (N ¼ 15) 4.21 (1.05) 3.40 (1.45) 3.67 (1.63)
Blood, injuries, injections 4.36 (1.36) 3.20 (1.91) 3.92 (1.78) Dresden, being a city with a large university.
(N ¼ 50) The data suggest that SPs are a heterogeneous group. Age of
Situational (N ¼ 52) 4.15 (1.42) 3.69 (1.91) 3.81 (1.68) onset varies widely for different subgroups, as do patterns of
Physical (N ¼ 46) 4.21 (1.17) 3.07 (2.30) 4.43 (1.72) comorbidity. Even the phobias within one subgroup (e.g., natural/
Others (N ¼ 13) 4.55 (1.21) 3.54 (1.85) 4.92 (1.19)
environment or situational phobias) defined by DSM-VI differ
74 E.S. Becker et al. / European Psychiatry 22 (2007) 69e74

regarding age of onset. Age of onset might in some cases be related [4] Boyd JH, Rae DS, Thompson JW, Burns BJ, Bourdon K, Locke BZ, et al.
to vulnerable times during normal development. It can be assumed Phobia: Prevalence and risk factors. Soc Psychiatry Psychiatr Epidemiol
1990;25:314e23.
that the likelihood of developing a phobia is related to the exposure [5] Brown TA, DiNardo PA, Barlow DH. Anxiety Disorders Interview
to an object or situation that children fear while growing up, thus Schedule for DSM-IV (ADIS-IV). Albany: Graywind Publications;
following a prepared pattern. Furthermore, some phobias are 1994.
clearly related to experience, like fear of driving. Since most par- [6] Curtis GC, Magee WJ, Eaton WW, Wittchen HU. Specific fears and pho-
ticipants fear driving as a driver and not as a passenger, this phobia bias. Br J Psychiatry 1998;173:212e7.
[7] Essau CA, Conradt J, Petermann F. Frequency, comorbidity, and psycho-
can not start earlier than at around 18 years (the legal age for driv- social impairment of specific phobia in adolescents. J Clin Child Psychol
ing in Germany). Interestingly, subgroups of phobias seem to have 2000;29(2):221e31.
their own patterns of comorbidity, also. Physical illness phobias [8] Fredrikson M, Annas P, Fischer H, Wik G. Gender and age differences in
and blood phobias, for instance, seem not to be related to affective the prevalence of specific fears and phobias. Behav Res Ther
disorders, in contrast to situational phobias are. Thus, age of onset 1996;34:33e9.
[9] Keller A. Die Klassifikation psychischer Störungen nach DSM-IV mit
and patterns of comorbidity show that SPs may not be a uniform Hilfe eines strukturierten Interviews (F-DIPS). Unpublished dissertation,
group, and where subgroups are not homogeneous, doubts about University of Heidelberg; 2000.
the validity of the groups are raised. [10] Kendler KS, Neale MC, Kessler RC, Heath AC, Eaves LJ. The genetic
epidemiology of phobias in women. Arch Gen Psychiatry 1992;
49:273e81.
5. Conclusions [11] Lindal E, Stefansson JG. The lifetime prevalence of anxiety disorders in
Iceland as estimated by the US National Institute of mental health diag-
These data show that SPs are common, impairing, and phenom- nostic interview schedule. Acta Psychiatr Scand 1993;88:29e34.
[12] Magee WJ, Eaton WW, Wittchen HU, McGonagle KA, Kessler RC. Ag-
enologically heterogeneous among young women in the commu-
oraphobia, simple phobia, and social phobia in the national comorbidity
nity. In addition, these data carry messages for prevention and survey. Arch Gen Psychiatry 1996;53:159e68.
classification of SPs. First, the data suggest that age of onset [13] Margraf J, Schneider S, Soeder U, Neumer S, Becker ES. F-DIPS: Diag-
may vary widely by phobia subtype and, as such, efforts aimed nostisches Interview bei Psychischen Störungen (Forschungsversion).
at primary prevention need to take into account these specific dif- Unpublished manuscript; 1996.
[14] Marks IM. Fears, phobias, and rituals. New York: Oxford University
ferences. Second, the results indicate that the use of subtypes em-
Press; 1987.
ployed in the current diagnostic classification system may obscure [15] McConaughy SH, Achenbach TM. Comorbidity of empirically based
phenomenological differences within and across subtypes as they syndromes in matched general population and clinical samples. Child
occur naturally in the community. Future research should look Psychol Psychiatry 1994;35(6):1141e57.
closely at the differences and their significance for treatment. [16] Öst LG. Age of onset in different phobias. J Abnorm Psychol
1987;96(3):223e9.
[17] Regier DA, Rae DS, Narrow WE, Kaelber CT, Schatzberg AF. Preva-
Acknowledgments lence of anxiety disorders and their comorbidity with mood and addictive
disorders. Br J Psychiatry 1998;173:24e8.
[18] Robins LN, Helzer JE, Weissman MM, Orvaschel H, Gruenberg E,
The research was supported by grant DLR 01EG9410, Min-
Burke JD, et al. Lifetime prevalence of specific psychiatric disorders in
istry of Science, Research and Education. We would like to three sites. Arch Gen Psychiatry 1984;41:949e58.
thank the many people who helped with this study. [19] Schneider S, Margraf J. DIPS: Diagnostisches Interview bei psychischen
Störungen. Berlin: Springer; 2006.
[20] Smoller JW, Lunetta KL, Robins J. Implications of comorbidity and
References ascertainment bias for identifying disease genes. Am J Med Genetics
(Neuropsychiatric Genetics) 2000;96:817e22.
[1] Agras S, Sylvester D, Oliveau D. The epidemiology of common fears and [21] Starcevic V, Bogojevic G. Comorbidity of panic disorders with agorapho-
phobia. Compr Psychiatry 1969;10(2):151e6. bia and specific phobia: Relationship with the subtypes of specific
[2] Antony MM, Brown TA, Barlow DH. Heterogenity among specific pho- phobia. Compr Psychiatry 1997;38(6):315e20.
bia types in DSM-IV. Behav Res Ther 1997;35(12):1089e100. [22] Wittchen HU, Nelson CB, Lachner G. Prevalence of mental disorders and
[3] American Psychiatric Association. Diagnostic and statistical manual of psychosocial impairments in adolescents and young adults. Psychol Med
mental disorders. 4th ed. Washington, DC: APA; 1994. 1998;28(1):109e26.