- Top of page
Aim: There is some evidence that age and gender influence expression of schizotypal traits in the normal population. One important limitation of previous studies, however, is the restricted age range of subjects. The aim of the present study was therefore to investigate the effect of gender and age on schizotypal personality features.
Methods: A total of 1024 participants aged between 16 and 90 were assessed using the schizotypal personality questionnaire. The effect of gender and age on schizotypal features was investigated. Confirmatory factor analysis was used to examine the factor structure of the scale in different age groups.
Results: Male subjects scored higher than women in negative and disorganized-like symptoms. Female subjects scored significantly in social anxiety and odd belief subscales. Younger participants significantlyscored higher in reference, odd belief, unusual perceptual experiences, odd behavior and odd speech subscales. While decreases in schizotypy scores were gradual for most of the scales, increased disorganization in female subjects was restricted to late adolescence. Confirmatory factor analysis of the data supported modified versions of the Raine three-factor model across all age groups.
Conclusion: Psychological changes related to adolescence and better social adaptation as a result of life experience might contribute to the age- and gender-related differences in schizotypy.
ALTHOUGH CATEGORICAL VIEWS have dominated clinical practice regarding psychotic disorders, there is increasing evidence supporting the validity of dimensional views of psychosis. Psychotic symptoms observed in schizophrenia and other mental illnesses were suggested to lie in the upper end of a continuum between schizophrenia and the normal population.1,2 Subclinical psychotic-like experiences and some personality traits commonly observed in non-clinical subjects seem to be associated with the same genetic susceptibility factors causing schizophrenia.2,3 These characteristics were defined under the term ‘schizotypy’.4 One of the most common strategies for studying schizotypy in the normal population is the investigation of schizotypal personality disorder (SPD)-like features in normal population.
Age and gender are important variables affecting presentation of psychosis in clinical cases. Age is also one of the important confounders of schizotypal traits. In general, schizotypal traits were found to be negatively associated with age in adults and adolescents,5–7 although a recent study did not support these findings in adolescents.8 Unfortunately, most of these studies investigated the effect of age on a sample restricted to young adults or adolescents.
Gender is also an important determinant of schizotypal traits in non-clinical populations. Negative and disorganization symptoms were reported to be higher in male subjects.6,7 Female subjects were also reported to score higher on social anxiety, magical thinking and paranoid-like symptoms.6,7,9 These findings are interesting considering the fact that a similar pattern of gender and symptom profiles is also observed in schizophrenia. Interaction between age and gender in schizotypal traits might also be relevant, because age of onset of psychotic-like experiences and florid psychosis was reported to be younger in male subjects.10,11
Schizotypal Personality Questionnaire (SPQ) is a very commonly used self-rated instrument that was constructed to assess all nine DSM criteria for SPD.12 Many of the findings related to effects of age and gender were studied using the SPQ. There is some debate over factor structure of the schizotypal traits. According to the hypothesis that schizotypal features are an attenuated form of clinical psychosis, both conditions should have a similar factor structure. Most of the studies supported the three-factor model of Raine et al..13,14 These factors are cognitive–perceptual, interpersonal and disorganization, which are in parallel with a three-dimensional symptom pattern observed in schizophrenia. Recent work, however, suggested that a model with a separate fourth paranoid factor could better explain the factor structure of schizotypal traits.15,16 Previous work also supported the same four-factor model in a sample consisting of 1059 university students.17 Two of these studies, however, also suggested that a modified version of the Raine model could satisfactorily explain the factorial structure of schizotypal traits in normal population.16,17 Gender and age are also among the factors that can influence the factor structure of schizotypal traits. Unfortunately only a few studies reported data regarding consistency of the factor structure of schizotypy in different age groups and gender. Fossatti et al. supported Raine's model in both adolescents and young adults.6 There is also some evidence of replicability of three-factor structure in mature adults.18,19 Those studies, however, did not take the model of Stefanis et al. and the proposed modifications in Raine's model into account.15–17
The aim of the present study was therefore to study the effect of age in both sexes on schizotypal personality features. Another goal of the study was to compare the factor structure of schizotypal traits in adolescents, young adults, mature adults and old age.
- Top of page
Because there was a modest but significant difference for age between groups, age was accepted as a covariate in male–female comparisons for SPQ and its subscales. In these ancova analyses, male subjects had a significantly higher schizotypy score than female subjects in constricted affect (F = 19.1, P < 0.001), no close friend (F = 16.6, P < 0.001), odd behavior (F = 37.0, P < 0.001), and odd speech (F = 9.8, P = 0.002). Female subjects had higher scores in social anxiety (F = 14.2, P < 0.001), and odd belief (F = 11.7, P = 0.001).
In five of nine subscales (Reference, Odd belief, Unusual perceptual experiences, Odd behavior, Odd speech), anova indicated significant effect of age (Table 1). For all of these five subscales, schizotypy scores were higher in younger subjects. For female subjects, the effect of age was highly significant in all five subscales (P < 0.01). For male subjects, significance of the effect of age was less pronounced for odd beliefs (P < 0.05) and there was only a tendency for odd speech (P = 0.07). For male subjects and female subjects, higher scores in these subscales decreased gradually towards the later age groups. There was a significantly different pattern, however, for odd behavior and odd speech for female subjects. Although female subjects had relatively high scores in these two subscales in the late adolescence group, they scored significantly lower after 19 years of age.
Table 1. Effect of gender and age on SPQ subscales
| ||16–18 years n = 113||19–20 years n = 152||21–24 years n = 415||25–35 years n = 98||36–45 years n = 77||46–65 years n = 72||66–90 years n = 97||Full sample|
|Mean ± SD||Mean ± SD||Mean ± SD||Mean ± SD||Mean ± SD||Mean ± SD||Mean ± SD||Mean ± SD|
|SPQ total|| || || || || || || || |
| Female**||31.7 ± 13.2||29.6 ± 11.2||26.6 ± 12.5||23.6 ± 11.0||23.6 ± 10.7||23.6 ± 15.6||24.7 ± 12.7||26.9 ± 12.6|
| Male**||31.9 ± 9.0||31.7 ± 13.0||28.4 ± 12.9||25.5 ± 14.2||26.5 ± 13.8||25.6 ± 9.5||23.2 ± 13.5||28.0 ± 12.9|
|Reference|| || || || || || || || |
| Female**||4.5 ± 2.6||4.6 ± 2.3||4.0 ± 2.3||3.1 ± 2.1||2.4 ± 1.9||2.1 ± 2.0||3.2 ± 2.0||3.7 ± 2.4|
| Male**||4.5 ± 2.0||4.0 ± 2.2||3.6 ± 2.1||2.9 ± 2.6||3.2 ± 2.5||2.3 ± 1.7||2.4 ± 1.9||3.4 ± 2.2|
|Social anxiety|| || || || || || || || |
| Female||4.3 ± 2.8||4.7 ± 2.7||4.3 ± 2.4||4.0 ± 2.6||4.0 ± 2.0||4.3 ± 2.9||4.1 ± 2.6||4.3 ± 2.5|
| Male||3.5 ± 2.1||4.0 ± 2.5||3.8 ± 2.3||3.3 ± 2.8||3.8 ± 2.6||3.7 ± 1.9||2.9 ± 2.5||3.6 ± 2.4|
|Odd belief|| || || || || || || || |
| Female**||2.9 ± 1.8||3.0 ± 1.8||2.7 ± 1.7||2.3 ± 1.5||2.2 ± 1.7||1.7 ± 1.5||1.9 ± 1.6||2.6 ± 1.7|
| Male*||2.2 ± 1.4||2.7 ± 1.8||2.3 ± 1.8||2.0 ± 2.0||1.7 ± 1.4||1.6 ± 1.3||1.7 ± 1.5||2.2 ± 1.7|
|Unusual perceptual experiences|
| Female**||4.2 ± 2.1||3.9 ± 2.3||3.3 ± 2.2||2.7 ± 2.0||2.9 ± 2.0||2.2 ± 2.2||2.5 ± 1.9||3.3 ± 2.2|
| Male**||4.5 ± 2.0||3.7 ± 2.3||3.4 ± 2.3||3.2 ± 2.0||2.6 ± 2.0||2.4 ± 1.7||2.1 ± 1.8||3.2 ± 2.2|
|Odd behavior|| || || || || || || || |
| Female**||2.7 ± 2.6||1.6 ± 1.8||1.5 ± 2.0||1.1 ± 1.7||1.3 ± 1.9||1.5 ± 2.1||1.4 ± 2.0||1.6 ± 2.1|
| Male**||3.5 ± 2.3||3.0 ± 2.3||2.5 ± 2.2||2.3 ± 2.3||1.5 ± 1.9||1.4 ± 1.6||1.9 ± 2.1||2.4 ± 2.2|
|No close friends|| || || || || || || || |
| Female||3.3 ± 2.0||3.4 ± 2.1||3.0 ± 2.1||3.2 ± 2.1||3.1 ± 2.3||3.5 ± 2.8||3.7 ± 2.1||3.2 ± 2.2|
| Male||3.8 ± 2.1||4.1 ± 2.4||3.7 ± 2.1||3.4 ± 2.1||3.9 ± 2.3||4.4 ± 1.9||3.5 ± 2.5||3.8 ± 2.2|
|Odd speech|| || || || || || || || |
| Female**||3.8 ± 2.7||2.7 ± 2.0||2.8 ± 2.2||2.3 ± 2.1||2.6 ± 2.3||2.8 ± 2.2||2.4 ± 2.3||2.8 ± 2.3|
| Male||3.7 ± 2.4||3.8 ± 2.4||3.3 ± 2.6||2.9 ± 2.3||3.1 ± 2.5||3.1 ± 1.9||2.4 ± 2.0||3.3 ± 2.4|
|Constricted affect|| || || || || || || || |
| Female||2.5 ± 1.8||2.5 ± 1.9||2.1 ± 1.9||1.8 ± 1.4||2.4 ± 2.0||2.7 ± 2.3||2.5 ± 1.9||2.3 ± 1.9|
| Male||2.8 ± 1.7||3.1 ± 2.2||2.7 ± 1.8||2.6 ± 1.7||3.5 ± 2.5||2.7 ± 1.9||2.7 ± 2.0||2.8 ± 1.9|
|Paranoia|| || || || || || || || |
| Female||3.5 ± 2.0||3.3 ± 2.1||3.0 ± 2.0||3.1 ± 1.9||2.8 ± 2.0||2.8 ± 2.0||3.1 ± 2.2||3.1 ± 2.0|
| Male||3.4 ± 1.5||3.5 ± 1.8||3.2 ± 2.0||2.8 ± 2.2||3.4 ± 1.9||3.9 ± 1.8||3.3 ± 2.2||3.3 ± 2.0|
CFA in the combined sample and effect of gender on factor structure of the SPQ
The Raine model did not fit the data adequately (RMSEA > 0.08). The Stefanis et al. four-factor model fitted the data satisfactorily (Table 2), but modifications suggested by Wuthrich and Bates16 and Bora and Arabaci17 resulted in better fit than the Stefanis four-factor model.15 Gender did not influence these results significantly; modified versions of the Raine model and the Stefanis et al. four-factor model provided an adequate fit in both sexes.
Table 2. CFA in the full study sample and effect of gender on factor structure
| ||χ2||d.f.||AIC||CFI||RMSEA||RMSEA 90%CI|
|Stefanis et al.15||130.3||19||182.3||0.98||0.076||0.064–0.088|
|Wuthrich and Bates16†||119.4||20||169.4||0.98||0.070||0.058–0.082|
|Bora and Arabaci17†||97.6||20||147.6||0.99||0.062||0.050–0.074|
CFA in different age groups
The Raine and the Stefanis et al. models did not adequately fit the data except for the age group 21–24 years (Table 3). Both modified versions of the Raine model provided satisfactory fit for all age groups. RMSEA was <0.08 and CFI was >0.97 for these models in all age groups. χ2 test and AIC also supported the superiority of the modified Raine models over the Raine and Stefanis et al. models for all age groups.
Table 3. CFA in different age groups
| ||χ2||df||AIC||CFI||RMSEA||RMSEA 90%CI|
|16–20 years|| || || || || || |
| Stefanis et al.15||55.8||19||107.8||0.96||0.086||0.60–0.112|
| Wuthrich and Bates16†||53.0||20||103.0||0.97||0.079||0.054–0.104|
| Bora and Arabaci17†||40.7||20||90.7||0.98||0.063||0.034–0.090|
|21–24 years|| || || || || || |
| Stefanis et al.15||69.6||20||119.6||0.98||0.077||0.058–0.098|
| Wuthrich and Bates16†||54.5||20||104.5||0.98||0.065||0.044–0.085|
| Bora and Arabaci17†||42.6||20||92.6||0.99||0.052||0.030–0.074|
|25–45 years|| || || || || || |
| Stefanis et al.15||50.0||19||102.0||0.97||0.097||0.065–0.129|
| Wuthrich and Bates16†||27.7||20||79.7||0.99||0.051||0.012–0.090|
| Bora and Arabaci17†||31.7||20||83.7||0.99||0.062||0.017–0.099|
|46–90 years|| || || || || || |
| Stefanis et al.15||45.1||20||95.1||0.98||0.086||0.053–0.120|
| Wuthrich and Bates16†||31.1||20||81.1||0.99||0.057||0.019–0.095|
| Bora and Arabaci17†||30.90||20||80.9||0.99||0.057||0.019–0.095|
- Top of page
The present study investigated schizotypal features and the factor structure in different age groups ranging from 16 to 90 years. The present results suggested that age and gender had a significant impact on severity of schizotypal features in the normal population, but the factor structure of schizotypal characteristics seems to be similar in all age and gender groups. In five of nine scales of the SPQ, subjects in younger age groups scored significantly higher than those in older groups. Male subjects had more disorganized and negative symptom-like features than female subjects. Female subjects scored higher in social anxiety and odd belief items. While most of the female subjects had low scores in disorganized dimension, adolescent female subjects had relatively high scores. Modified versions of the Raine three-factor model seem to be the best model to explain the structure of schizotypal features because they adequately fitted the data in all age groups and in both sexes.
The present study adds valuable information to previous findings regarding the impact of age on schizotypal features. Most of the previous studies reported data in more restricted age samples of adults.6,18 Interpretation of the meaning of the relationship between age and schizotypal features is important because age is an important factor for the onset of psychosis in schizophrenia spectrum disorders. There is also a differential effect of age for gender. Full-blown psychosis tends to arise earlier in male subjects than female subjects in early adulthood. Psychosis onset in adolescence and childhood is also more frequent in male subjects. Although psychotic-like features, which are similar to schizotypal features, arose earlier than full-blown psychosis in both sexes, they might also be observed earlier in male subjects.10 Psychotic symptom-like features and disorganization seems to decrease from late adolescence to older ages in both sexes. Although this reduction in schizotypal features was gradual in most cases, there was a different pattern for disorganization for female subjects. Female subjects seemed to have high levels of disorganization only in late adolescence. If schizotypal features are an attenuated form of psychosis, one should expect earlier (and more severe) onset of schizotypal features in male subjects. But the relatively high disorganization in late adolescence in female subjects and abrupt reduction of these symptoms afterwards suggest that psychological changes related to adolescence rather than proneness to psychosis can contribute to increased disorganization in adolescents and in young adults. There may be several explanations for gradual decrease in other aspects of schizotypal features with increasing age. It may be that the reduction may be related to the impact of psychosis conversion in high-risk individuals; that is, although high-risk individuals might be included in younger groups, they would be less likely to be included in older groups because a substantial number of them would have undergone conversion to the psychosis group, and therefore be excluded. Conversion to psychosis, however, cannot explain most of this reduction because only 1% of the population develops schizophrenia. Better social adaptation as a result of life experience and social learning might be more likely to explain age-related decreases in schizotypal features in the normal population.
The present findings about the impact of gender on schizotypy were in accordance with the published findings. Increased social anxiety and higher scores in the odd beliefs/magical thinking scale have been reported previously.6,9 Some studies reported higher scores for reference in female subjects,6 but the present results did not support this finding. Higher scores for negative and disorganized dimensions in male subjects are also compatible with previous work on schizotypal traits.6,7
One of the relatively new aspects of the present study was the application of CFA to different age groups. The results of the factor analysis support the validity of the SPQ because factor structure of the scale seems to be stable across the different age groups and sexes. But the present findings did not support previous studies that reported good fit to the data for the Raine three-factor model in different age groups.13,14 We believe that this discrepancy is related to some characteristics of the previous work. First, in some cases the goodness of fit of the Raine model was not very satisfactory according to our criteria.17 Second, these studies did not take the Stefanis et al. model into account and did not consider modifications in the Raine model. Finally, these studies did not investigate the Raine model across the whole adult age range. Some studies in younger adults reported better fit for the Stefanis et al. four-factor structure and modified version of the Raine model.15,16 The present work supports these studies and extends these findings to other age groups. We think that this is a major contribution of the present study because it shows that the Raine model has similar problems across all adult age groups. We believe that these findings are related to inadequacy of some of the items of the SPQ for clearly loading on one of the underlying three or four separable factors. Improvements in the SPQ and other measures can contribute to better definition of the dimensions of schizotypal personality features.
One of the limitations of the present study was its cross-sectional nature. Longitudinal studies might provide more reliable information regarding onset and prognosis of schizotypal features. Another limitation may be related to the fact that the SPQ is a self-report measure of schizotypal features. Also because there was no diagnostic interviews in the present study, it is possible that some individuals with schizophrenia spectrum disorders and schizotypal personality disorder were included in the study. But we do not think that this would have had a significant impact on the results because the sample size was large. And we may also expect that people with schizotypal personality disorder could be reluctant to participate in this kind of study. Finally, inclusion of children and younger adolescents would be desirable.