Social Psychiatry and Psychiatric Epidemiology

, 44:1

Social phobia in Swedish adolescents

Prevalence and gender differences


    • Dept. of Clinical and Experimental Medicine, Division of Child and Adolescent PsychiatryLinköping University
  • Maria Tillfors
    • Dept. of Behavioural, Social and Legal SciencesÖrebro University
  • Tomas Furmark
    • Dept. of PsychologyUppsala University
  • Gunilla Bohlin
    • Dept. of PsychologyUppsala University
  • Gerhard Andersson
    • Dept. of Behavioural Sciences and Learning, The Swedish Institute for Disability ResearchLinköping University
    • Dept. of Clinical NeuroscienceKarolinska Institutet
  • Carl Göran Svedin
    • Dept. of Clinical and Experimental Medicine, Division of Child and Adolescent PsychiatryLinköping University

DOI: 10.1007/s00127-008-0400-7

Cite this article as:
Gren-Landell, M., Tillfors, M., Furmark, T. et al. Soc Psychiat Epidemiol (2009) 44: 1. doi:10.1007/s00127-008-0400-7



The aim of this study was to investigate the prevalence of self-reported social phobia in a community sample of Swedish adolescents in junior high school, at the risk-period for developing social phobia. Of particular interest was to investigate gender differences in prevalence across ages. Prevalence of sub-threshold social phobia was also studied.


Students in grades 6–8 (aged 12–14) from seventeen schools in five Swedish municipalities were screened by means of a self-report questionnaire, the social phobia screening questionnaire-for children (SPSQ-C).


Data from a sample of 2,128 students were analysed and showed a point-prevalence rate of 4.4% (95%CI 3.5–5.2) and a significant gender difference (6.6% girls vs. 1.8% boys, P < 0.001). No significant differences in prevalence of probable cases emerged across the ages. At sub-threshold level, marked social fear of at least one social situation was reported by 13.8% of the total group. “Speaking in front of class” and “calling someone unfamiliar on the phone” were the most feared social situations. In the social phobia group, 91.4% reported impairment in the school-domain due to their social fear.


Social phobia is a common psychiatric condition in Swedish adolescents, especially in girls. As impairment in the school-domain is reported to a high degree, professionals and teachers need to recognize social phobia in adolescents so that help in overcoming the difficulties can be offered.


prevalencesocial phobiaadolescentsgender


Social phobia, or social anxiety disorder, is described in the DSM-IV [3] as a marked and persistent fear of one or more social and performance situations in which embarrassment may occur. Exposure to unfamiliar people or possible scrutiny of others typically evokes anxiety and tends to be avoided or endured under great distress. Social phobia is argued to exist on a continuum, varying in severity and number of feared situations [19]. Lifetime prevalence rates of social phobia range from 3.9–13.7% in Western societies [17, 19, 23, 24, 32]. The notable variation in prevalence can largely be explained by methodological factors in addition to true cultural differences [19].

Lifetime prevalence rates of DSM-III-R and DSM-IV defined social phobia in community samples of children and adolescents also vary considerably, ranging from 0.9 to 13.1% [26]. The highest rates have been noted in studies of older adolescents and young adults [14, 24, 27, 36, 41, 44].

Onset of social phobia is reported to be in early- to mid-adolescence with a median age ranging from 12 to 16.6 years [8, 11, 17, 34], thus starting earlier in life compared to most other mental disorders [23]. Earlier onset is associated with a higher number of social fears [32] and with fear of more than one social- or performance situation, typical for the generalized subtype of social phobia [44].

Time of onset and the higher prevalence seen in adolescents compared to children can be associated with various aspects of cognitive maturation, e.g. an increased ability to take others viewpoints. While this ability is useful, it can also mean that the adolescent becomes more aware of others’ attention and the social evaluation and the opinions of peers might cause distress in some individuals [6]. Other explanations for the higher rates in adolescents compared to children could be increasing academic and social demands and encounters with a wider range of social situations [30]. During the teenage years it becomes more important to be part of a group and the risk of being excluded from the social group can be perceived as particularly distressing during this period [6].

In adult populations, a majority of studies report a female preponderance of social phobia as well as sub-threshold social phobia [911, 24, 34]. The gender ratio seems to vary somewhat across different levels of social anxiety [13]. Although not as extensively studied, older adolescents show similar patterns as adults [14, 31, 41, 44].

Social phobia is associated with serious negative consequences, such as impaired quality of life and socio-economic disadvantages [17, 36]. Considerable impairment and distress have also been noted in sub-threshold social phobia, i.e. individuals reporting excessive social anxiety without meeting the full diagnostic criteria for social phobia [13, 16]. It can therefore be argued that sub-clinical levels of the disorder deserve more attention from researchers and clinicians. Moreover, as there is no exact diagnostic threshold for social phobia, epidemiological research may benefit from studying prevalence rates across different levels of impairment and distress [19].

Most of the published prevalence studies on social phobia have been conducted on adult samples in Western but non-European countries [17]. There is to date no published study on the prevalence of social phobia in Swedish adolescents. The only Scandinavian study so far is a report on social phobia among Finnish children in the ages of 8- to 9 years with an estimated prevalence of 0.1% [1]. Reports on mental health in the Swedish general population and in other countries during the last decades, show an increase of internalizing symptoms such as anxiety and depression, in all age-groups and for both genders [33]. These symptoms are however more common among adolescent girls [33]. Increasing female excess in general psychological distress is reported from the age of eleven to fifteen [38].

Given the reports of increasing mental health problems, especially in girls and the shortage of studies, there is a need for European studies on the prevalence of social phobia and gender differences in samples under the age of sixteen [43]. Thus, the aim of the present study was to examine the prevalence of social phobia, and sub-threshold social phobia, among Swedish adolescents in early- to mid-adolescence, at the typical period for onset for this disorder. In light of the female preponderance in adult community samples, we expected girls to report social phobia to a higher degree than boys. We also expected an increase in prevalence rates to appear over the ages as higher rates are seen in studies of older adolescents.



Students from 17 comprehensive schools in five Swedish municipalities participated. The total sample consisted of 3,037 students in grades 6–8. In the Swedish school-system students enter school at the age of 6 years. Thus, students in grade 6 are normally at the ages of 11–12 years, grade 7 in the ages of 12–13 and students in grade 8 are normally in the ages of 13–14. A total of 1,104 boys, (47.1%) and 1,241 girls (52.9%) agreed to participate, resulting in an acceptable response rate of 77%. Reasons for not responding could be absence from school or disagreement to participate. Data from the non-responders were not collected. The response rate according to grade was 82% out of 1,005 students in grade six, 77% out of 888 students in grade seven and 79% out of 1,144 students in grade eight. The response rate ranged from 73–87% in the five municipalities.

Due to incompletely filled out questionnaires, final data for analysis consisted of 2,128 respondents (992 boys, 46.6% and 1,136 girls, 53.4%) aged 11–15 years with a mean age of 13.1 years. None of the responders with incomplete questionnaires reported much distress in the phobic situations covered and thus they would not have been diagnosed with social phobia if included.


The schools were recruited from five municipalities which were representative of the Swedish population by socio-economic composition such as gender distribution, proportion of inhabitants of foreign origin, proportion of inhabitants under the age of 17 and proportion of marriages [35]. Three municipalities were medium-sized (20–50.000 inhabitants, i.e. Nässjö, Värnamo and Västervik), and two were large-sized (50–200.000 inhabitants, i.e. Jönköping and Linköping) according to Swedish demographics [35]. Twelve of the seventeen schools were situated in medium-sized municipalities and five schools in large-sized. The schools were recruited to obtain a representative sample of schools with small, medium and large number of students and with small, medium and large number of students from families of foreign origin.

Data collection took place in 2005. The study was approved by the Central Ethics Committee concerning Human Research in Sweden.

Parents and students were given written information administered by the school-staff. Participation was voluntary and parents could contact the principal investigator in case they did not want their child to participate (n = 31). Students were also informed about the voluntary investigation directly in the classroom either by a person from the project or by the teacher who had been given written information to read for the students. The questionnaire was administered to the students in their classroom.


A DSM-IV [3] diagnosis of social phobia (probable case) was established by means of the Social Phobia Screening Questionnaire for Children (SPSQ-C), which is a modified version of the social phobia screening questionnaire (SPSQ) [20]. The SPSQ has been used in adult studies [7, 20, 39]. Psychometric evaluation of the SPSQ-C has shown a test–retest of r = 0.60 over 3 weeks, and a sensitivity of 71% and specificity of 86% (unpublished data) in relation to SCID-I [18]. In the present study, the eight phobic situations in the SPSQ-C had an alpha coefficient of 0.78 reflecting an acceptable internal consistency.

The diagnostic section of the SPSQ-C was based on 8 potentially phobic situations. These situations were: “speaking in front of the class”, “raising your hand during a lesson”, “being together with others during breaks”, “initiating a conversation with someone unfamiliar”, “looking someone in the eyes during a conversation”, “making a phone-call to someone unfamiliar”, “going to a party”, and lastly “eating together with others during the lunch-break”. The respondent rated their perceived social fear in these potentially phobic situations on a three-point scale ranging from no fear, or some fear to marked fear. Five diagnostic questions followed, assessing whether the individual met the DSM-IV social phobia criteria A, B and D for one or more of the phobic situations. Since the youths were below 18 years old the C-criteria, realizing that the fear is excessive or unreasonable, did not have to be fulfilled. The E-criterion was assessed with three yes/no questions, i.e. the student was asked whether the social fear was of such nature that it severely interfered with or severely interfered with his/her activities in school, during leisure-time or when being with peers. The eighth and last question covered the F-criterion of 6-month duration (yes/no question).

Diagnostic issues

For a probable diagnosis of social phobia, the student had to rate at least one potentially phobic situation as marked fear on the social fear scale. This particular situation had to be consistently endorsed in the diagnostic questions covering social phobia criteria A, B and D. The E-criterion had to be met, i.e the report of impairment in at least one of the three life domains assessed. Lastly, the F-criterion, concerning persistence of symptoms for more than 6 months, also had to be fulfilled. Four cut-off levels were chosen for measuring sub-clinical levels of social phobia. The first level concerned marked fear in at least one social situation on the first item of SPSQ-C. The second level concerned marked fear in at least one social situation and impairment in one life-domain. On the third level all criteria for social phobia except the 6-month duration (criteria A–E) were to be fulfilled and on the highest level, all criteria (A–F) for social phobia according to DSM-IV (criteria A–F) had to be fulfilled.

Statistical analyses

Prevalence rates are displayed as point prevalence with 95% confidence intervals (CI). For categorical variables, group differences were evaluated by use of χ2 tests. Logistic regression was used in order to assess the increased odds of presenting with social phobia in relation to gender. Grade was used as a covariate when calculating odds ratios for gender. All analyses were performed in SPSS version 14.0 (SPSS, Inc., Chicago, IL, USA). Grade was used as variable instead of age due to small numbers in the youngest age-group (aged 11) and the oldest (aged 15). This was also motivated by the notion that students from the same grade might be more similar with regard to psychological maturation and environmental demands as compared to students of the same age but in different grades.


Prevalence of social phobia

The proportions of respondents fulfilling the DSM-IV criteria of social phobia (probable cases) are presented in Table 1 for gender and for grade separately. In the total group, a point-prevalence of 4.4% (95%CI = 3.5–5.2) was found.
Table 1

Point-prevalence rates (%) of DSM-IV social phobia (probable cases) by gender among Swedish adolescents and summary results of logistic regression presented with odds ratios (OR) and P-values

Grade (age)

Gender (n)

Social phobia


n (%)


OR (95%CI)


6 (11–12)a

Boys (226)

3 (1.3)


4.20 (1.43–16.91)b


Girls (290)

18 (6.2)


Total (516)

21 (4.1)


7 (12–13)a

Boys (401)

8 (2.0)


2.94 (1.30–6.66)b


Girls (407)

23 (5.7)


Total (808)

31 (3.8)


8 (13–14)a

Boys (365)

7 (1.9)


4.30 (1.88–9.81)b


Girls (439)

34 (7.7)


Total (804)

41 (5.1)



Boys (992)

18 (1.8)


3.83 (2.27–6.50)c


Girls (1,136)

75 (6.6)


Grand total


93 (4.4)


aAge ranges are presented for the regular age ranges in each grade. In grade 6 eight children (1.6%) were 13 years old, in grade 7 thirteen students (1.6%) were 14 years old and in grade 8 two students (0.2%) were 12 years old and eleven students (1.4%) were fifteen years old

bWithin grade

cAdjusted for grade

A significant gender difference was found in the total group, and for each grade, with higher prevalence of social phobia in girls than in boys (see Table 1). There was no significant difference in prevalence between the grades neither in the total group (χ2 = 1.69, df  = 2, NS) nor when analysed separately for girls (χ2 = 1.60, df  = 2, NS) and for boys (χ2 = 0.40, df  = 2, NS). As is seen in Table 1, a more than three-fold increase in risk for girls to present with social phobia was found in the total group (OR = 3.83, 95%CI = 2.27–6.50, P < 0.001).

Data were further analysed according to urban status and the estimated prevalence was 4.0% for medium-sized municipalities as compared to 4.6% for large-sized municipalities (χ2 = 0.32, df=1, NS). Because this was not statistically significant, urban status was not further considered.

Sub-threshold social phobia

Prevalence estimates of sub-threshold social phobia, across gender and age, were based on four cut-off levels (see Table 2). Prevalence rates decreased from 13.8% at the most liberal level to 4.4% at the most stringent level (all criteria for social phobia fulfilled). Significantly higher rates for girls were noted on all levels (Table 2).
Table 2

Prevalence rates (%) of probable cases of social phobia and sub-threshold social phobia, with 95%CI, among Swedish adolescents according to four cut-off levels

Cut-off levela


Total group (n = 2,128)

Girls (n = 1,136)

Boys (n = 992)

ORb, 95%CI


n (%)


n (%)


n (%)



293 (13.8)


181 (15.9)


112 (11.3)


1.49 (1.16–1.92)



151 (7.1)


105 (9.2)


46 (4.6)


2.10 (1.46–3.00)



134 (6.3)


100 (8.8)


34 (3.4)


2.72 (1.82–4.05)



93 (4.4)


75 (6.6)


18 (1.8)


3.83 (2.27–6.45)


Summary results of logistic regression presented with odds ratios (OR) and P-values

aDefinition of cut-off levels: 1 = Marked fear in at least one social situation, 2 = Marked fear in at least one social situation+ impairment in 1 life-domain, 3 = All criteria for self-reported social phobia except 6-month duration, 4 = All criteria for self-reported social phobia according to DSM-IV

bBoys = reference group

Feared situations

Table 3 shows the proportion of students reporting marked fear in each of the eight situations in the SPSQ-C. “Speaking in front of class” was the most common fear followed by “making a phone-call to someone unfamiliar”. Grade did not have a significant impact on the prevalence of marked social fears (χ2 = 1.40, df  = 2, NS).
Table 3

Frequency of social fears, i.e. the proportion of respondents reporting “marked fear” for each of the eight social situations on the first item of the SPSQ-C


Prevalence n (%)

SP (n = 93)

No SP (n = 2,035)

Total group (n = 2,128)

1. Speaking in front of the class

59 (63.4)

73 (3.6)

132 (6.2)

2. Making a phone-call to someone unfamiliar

30 (32.3)

72 (3.5)

102 (4.8)

3. Initiating a conversation with someone unfamiliar

27 (29.0)

41 (2.0)

68 (3.2)

4. Looking someone in the eyes during a conversation

16 (17.2)

45 (2.2)

61 (2.9)

5. Raising your hand during a lesson

14 (15.1)

34 (1.7)

48 (2.3)

6. Eating together with others during lunch-break

8 (8.6)

38 (1.9)

46 (2.2)

7. Going to a party

7 (7.5)

27 (1.3)

34 (1.6)

8. Being together with others during the breaks

3 (3.2)

46 (2.3)

49 (2.3)


In the total group, impairment of school activities was more commonly reported (17.2%) than of activities during leisure time (4.1%) and when being with peers (3.1%). Significantly more girls than boys in the total group reported impairment in school (χ2 = 21.42, df  = 1, P < 0.01).

School impairment was reported by 91.4% of the social phobia group. Significantly more students in grade seven (aged 12–13) and grade eight (aged 13–14) reported impairment of school activities compared with students in grade six (aged 11–12) (χ2 = 8.45, df  = 2, P < 0.05).


In the present study a point-prevalence rate of 4.4% was found for self-reported DSM-IV social phobia in a community sample of Swedish adolescents in junior high school. The rate is in concordance with those found in European community samples, showing prevalence rates of DSM-II-R and DSM-IV social phobia of 0.9–13.1% in children and adolescents [26]. The finding implies that social phobia is a common condition among Swedish adolescents.

Our initial hypothesis of higher prevalence in girls than in boys was confirmed as more than a three-fold increase in risk for girls to present with self-reported social phobia was found. Higher prevalence of phobias and fears in general is reported in girls [25] and an increased risk for females compared to males to develop anxiety disorders and mood disorders [23, 28]. Especially, increasing female excess in self-reported general psychological distress is reported from the age of eleven to fifteen [38]. Despite reports of higher incidence of anxiety and depression in females than in males [33] and that the increasing female excess of internalizing symptoms is seen during early to mid-adolescence, the causes underlying this phenomenon remain incompletely understood. Biological processes such as greater fluxes in reproductive hormones in females, evident during puberty, might underlie increased proneness to anxiety and depression [2]. Hormonal changes during puberty is reported to partially explain higher rates of depression in adolescent girls than in boys [4]. The interaction of puberty and gender to predict symptoms of social anxiety has been studied by Deardorff and colleagues [12]. They found advanced pubertal development to be associated with heightened symptoms of social anxiety in girls but not in boys. The aetiology of social phobia is thought to be multi-factorial [30] and presumably a variety of biological, psychological and cultural factors interact. Early physical development can lead to unwanted sexual attention and body dissatisfaction where girls are thought to be more prone to negative self-evaluation and others’ opinions regarding appearance and behaviour [12, 28]. Taken together, a few hypotheses have been tested in explaining gender differences in social phobia and other internalizing disorders. It would be of interest to find a model within a developmental psychopathology framework for understanding possible gender-specific pathways in developing social phobia as well as in other common internalizing disorders [21].

Contrary to what was expected, significant age differences were not found in the present study. The prevalence of social phobia and of social anxiety has been reported to increase in mid-teenage [14, 44] with a peak in the age of 14–15 years [15, 29]. The highest prevalence rates of mental disorders is noted in the age-group of 15–24 years compared to older age-groups [24]. The present sample was drawn from a population of students in grade 6–8 (aged 12–14). There were a few students in the age of eleven (2.3%) and in the age of fifteen (0.5%). The possibility to detect age differences might have been limited by this relatively narrow range and by a small proportion of individuals in the age of 15 years.

Another objective of the present study was to investigate the prevalence of sub-threshold social phobia as this disorder is argued to exist on a continuum, varying in severity and symptoms. Prevalence rates increased from 4.4% for cases fulfilling all criteria for social phobia according to self-report, to 6.3% when removing the criteria of 6-month duration. A prevalence of 13.8% at the most liberal level, i.e. the report of marked fear of at least one social situation, was found. Studies of social fears vary in methodology, like type and number of situations assessed, and therefore a summary of findings is not easily captured. Recent results from a replication of the National Comorbidity Survey showed that 24.1% of the US adult population reported at least one social fear [32]. In a community study of German adolescents a high proportion, 47.2% of the total group, reported fear of any of the social situations investigated [14]. Thus, the prevalence of social fear was lower in the present study than in the studies mentioned.

Concerning type of fear, “speaking in front of the class” was most commonly reported. This finding is in line with other in other epidemiological studies of adults as well as children and adolescents, showing that public speaking is the most common fear [5, 42, 43]. However, Wittchen and colleagues [44] noted that fear of tests and examinations was the most prevalent social fear in German adolescents.

The above findings imply that the most commonly feared situations for children are encountered in schoolsettings [37]. Social phobia has previously been associated with academic underachievement, school refusal or school drop-out [40]. We noted that 17.2% of all respondents and 91.4% of probable cases of social phobia reported impairment of school activities due to social fear. Significantly more girls than boys were reporting social phobia and also, significantly more girls than boys reported school-impairment. Despite the facts that social phobia is highly prevalent and associated with negative consequences for school life, the recognition of social anxiety in students has been reported to be poor in school personnel as well as in parents [22]. Thus, to detect students with excessive social anxiety and to limit its negative effects on academic achievement and mental well-being in general, there is a need for increased awareness and knowledge about social phobia in school settings, with extra attention to female students on this issue.


Some methodological limitations should be addressed in the present study. Firstly, in this initial study, the prevalence of social phobia was established by use of a screening questionnaire only. Therefore, cases of social phobia should be regarded as probable cases rather than as a formal diagnosis of social phobia.

Secondly, in the present study, students were included only if they attended school at the time for the investigation and if they agreed to participate. Social phobia is associated with school-absenteeism and school-refusal and students with disabling levels of social anxiety might have been absent due to social fear or chose not to participate due to fear of negative attention, possibly lowering the prevalence estimate. Still, there was a relatively high response rate and the prevalence rate in the present study was similar to rates found in other studies.

Lastly, no control was made for socio-demographic factors other than gender, age and urban status and the generalizability of the results is limited to students in grade 6–8 (aged 12–14) in Swedish municipalities of 30.000–135.000 inhabitants. The schools were recruited from municipalities that were representative for the population regarding major socio-demographic variables but data on the sample, like ethnicity and parental characteristics would have been of interest.


Social phobia tends to be a common condition among Swedish adolescents, especially in girls. The results from this study call for longitudinal studies on the development and course of social phobia due to gender. There is a need for theoretical models to explain gender differences found in social anxiety proneness to achieve better guidance for prevention as well as clinical interventions. The high proportion of students reporting impairment in the school domain due to social anxiety, calls for reliable and valid methods and developmentally sensitive instruments to screen for these conditions. School-health professionals and teachers need to have knowledge about social phobia, and efficacious interventions to help students overcome social anxiety need to be made available.


This study was supported by the Sven Jerring Foundation, the Mayflower Foundation, the Research Council of South-Eastern Sweden (FORSS), the Swedish Psychiatry Foundation and the Organon Foundation.

Copyright information

© Springer-Verlag 2008