Sage Journals: Discover world-class research

Abstract

Background:

The field of autism research in adults has not yet emerged in Arab countries, which is in part due to the lack of valid, reliable autism measures. The objectives of this study were to assess the psychometric properties of the abridged version of the autism-spectrum quotient (AQ-28) in a sample of non-clinical Arabic-speaking adults.

Methods:

A cross-sectional web-based study was carried-out from February to April 2024. Adults aged over 18 years from the general population of Lebanon (N = 1076, mean age of 27.90 ± 11.81 years, 63.8% females) were recruited using the snowball sampling technique. The AQ-28 was administered to participants, along with the relationship Questionnaire, the Patient Health Questionnaire-9, the Generalized Anxiety Disorder-7, and the Jong-Gierveld Loneliness Scale.

Results:

The original five-factor model of the AQ-28 had good fit to the data: RMSEA = 0.061 (90% CI: 0.059–0.064), SRMR = 0.060, CFI = 0.848, TLI = 0.831, and an excellent internal consistency (Cronbach’s α = 0.91). Measurement invariance has been established across sex. Mean sum scores observed in male participants were significantly higher than those displayed by females. Autistic traits as measured by the Arabic AQ-28 correlated positively with loneliness (r = 0.16; p < 0.001), depression (r = 0.17; p < 0.001) and anxiety (r = 0.15; p < 0.001).

Conclusion:

Findings suggest that Arabic-language version of the AQ-28 provided by this study is valid, reliable, and suitable for use among Arabic-speaking adults to measure core trait dimensions of autism and screen for autism spectrum disorder. It is our hope that the new scale will generate interest among researchers in the yet unexplored field of autism research in Arab countries.

Keywords

autism autistic traits autism-spectrum quotient AQ abridged version psychometric properties adults Arabic

Introduction

According to the dimensional approach, the defining prominent features of autism spectrum disorder (ASD), that is, the difficulties with social interaction and the restricted interests and repetitive behaviors, are common in the general population.¹ They present with varying degrees of severity along a continuum that extends from subclinical autistic traits to clinical ASD.¹ Over the last 2 decades, this dimensional approach has gained significant attention and has been extensively investigated because of its potential utility for both clinical practice and in the identification of endophenotypes to understand the genetics of ASD.^2,3 There has been a substantial amount of research describing the prevalence and nature of autistic traits, or the broad autism phenotype, in a wide variety of non-clinical populations, including unaffected first-degree relatives of people with ASD,⁴ children,⁵ adolescent twins,⁶ college students,⁷ and adult individuals without a diagnosis of an ASD.^8–13

Although they do not reach the threshold required for a clinical diagnosis of ASD, autistic traits may still be associated with a range of negative mental health consequences, including depression/anxiety,⁹ high levels of loneliness,^10,11 deficits in mentalizing abilities,¹⁴ decreased mental well-being,¹⁵ as well as more suicidal behavior¹⁶ and death by suicide.¹⁷ Beyond its impact on mental health, the broader autism phenotype has also attracted potential interest owing to its potential to help clinicians detect autism traits in clinical individuals with disorders other than ASD, such as mood disorders,¹⁸ obsessive compulsive disorder,¹⁹ schizophrenia,²⁰ anorexia nervosa,²¹ and borderline personality disorder.²² All these considerations highlight the potential benefits of studying autistic-trait groups, and the strong need for a psychometrically sound measure of the broader autism phenotype among adults from the general population.

Measurement instruments of autism traits

Multiple quantitative measures are available to assess the presence of Autism traits, which can be divided into informant-report and self-report measures. Among the informant-report measures are the Communication Checklist —Adult,²³ the Social Responsiveness Scale,²⁴ the Autism-Spectrum Screening Questionnaire,²⁵ the Social and Communication Disorders Checklist,²⁶ while self-report measures include the Broad Autism Phenotype Questionnaire,²⁷ the Communication Checklist—Self Report,²⁸ and the Autism-Spectrum Quotient (AQ).²⁹ The latter is one of the most extensively used measures quantifying autistic traits in the general population, in both clinical practice and research. The AQ was initially developed by Baron-Cohen et al.²⁹ to screen for ASD in people aged 16 years and older, and is originally composed of 50 items and 5 subscales (i.e., imagination, social skills, communication, attention switching, and attention to detail). The AQ has many advantages over other measures of autism traits, such as the inclusion of dimensions assessing both social and nonsocial aspects of cognition and behavior. Besides, its self-report nature and forced-choice format make it convenient to use.

In 2011, Hoekstra et al.³⁰ have shortened the AQ from 50 to 28 items as a rapid screening for ASD in a clinical setting and for easier implementation in large-scale studies. Indeed, a shorter version of the AQ enables saving of assessment time, respondents’ burden and related costs. In its initial validation, the AQ-28 demonstrated good psychometric qualities in different samples, including adults from the general population, university students, and individuals with a formal Asperger Syndrome diagnosis.³⁰ The AQ-28 has been adapted and validated in several languages, demonstrating adequate psychometric characteristics among various samples, including Persian-speaking university students,³¹ Spanish-speaking ASD first-degree relatives, patients with schizophrenia spectrum disorders and non-clinical adults,³² Malaysian- and Dutch-speaking adults from the general population,³³ as well as Scottish-speaking adults with ASD.³⁴ The AQ-28 has consistently demonstrated good psychometric properties, including measurement invariance to sex.^30,35,36 However, the scale has not yet been validated for the Arabic-speaking population.

Rationale

There is some evidence from cross-national studies that individuals from different countries/cultures may score differently on the AQ, which could be explained by variations in the interpretation and expression of autistic traits between cultures,³⁷ or even by differences in respondents’ primary spoken language.³⁸ Indeed, social norms on how people behave or process (facial) emotions might differ between Western societies (where the AQ was firstly developed) and other parts of the world.^37,39 Therefore, the translation, cultural adaptation and validation of the AQ for different languages and cultures is of utmost importance and should be encouraged. In this regard, the present study proposes to validate the Arabic version of the AQ-28. Making available an Arabic-language version of the Q-28 to measure autistic traits in the Arabic-speaking general populations may provide new insights from an unexplored region and culture, into the continuum nature of the distribution of autistic traits throughout the general population. This can foster research on autism in Arab countries, and bring clinicians and researchers closer to an enhanced understanding of the nature, etiology, and mechanisms underlying the emergence of ASD.⁴⁰

Therefore, the objectives of this study were to assess the factor structure, internal consistency reliability, measurement invariance across sex and concurrent validity of the AQ-28 in a sample of Arabic-speaking adults from the general population of Lebanon. It is hypothesized that the proposed five-factor structure will be supported and will be invariant across males and females. Based on previous literature, it is anticipated that the scale will show satisfactory reliability for all factors and that good concurrent validity will be evidenced by positive correlations with anxiety, depression,⁹ and loneliness scores.^10,11

Methods

Participants and procedure

A cross-sectional web-based study was carried-out from February to April 2024. The study’s inclusion criteria were the following: (1) Being aged 18 years and over, (2) Being of Lebanese nationality and residency, (3) Having access to the Internet, and (4) Willing to participate. Participants who did not fit inclusion criteria or those who did not fill the complete questionnaire were excluded. Non-probability sampling, by using the snowball sampling technique, was employed. Data was gathered using an online questionnaire in the Arabic language, disseminated via Google Forms and distributed to potential participants via popular social media platforms, including WhatsApp, Facebook, and Instagram. The study information was provided online via text in the first section of the questionnaire, and each participant was asked to read and give their informed consent before starting to fill out the survey. Confidentiality and anonymity were ensured, and no financial rewards were offered. The present study was performed according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.⁴¹

Measures

The questionnaire involved sociodemographic information, along with five measurement instruments (Appendix 1).

Sociodemographic information

Participants were asked to provide their age, sex (male, female), and level of education (elementary, middle, secondary, university).

The Autism-Spectrum Quotient-28 (AQ-28)

The AQ-28 is a self-administered scale consisting of 28 items. The scale is composed of two higher-order factors, and a total of five factors: (1) Difficulties with social skills (“Social skills”; e.g., “I find social situations easy”), (2) Preference for routine (“Routine”; e.g., “I prefer to do things the same way over and over again”), (3) Attention switching difficulties (“Switching”; e.g., “I frequently get strongly absorbed in one thing”), (4) Difficulties with imagination (“Imagination”; e.g., “Trying to imagine something, I find it easy to create a picture in my mind”), and (5) Fascination for numbers/patterns (“Numbers/patterns”; e.g., “I usually notice car number plates or similar strings of information”).³⁰ The four first factors are part of the higher-order “Social behavioral difficulties” factor. Items are scored on a 4-point Likert scale (“definitely agree”, “slightly agree”, “slightly disagree”, or “definitely disagree”) to specify the extent to which respondents agree/disagree with each statement. Total scores range from 0 to 28 points, with higher scores indicating greater autistic traits. After obtaining permission from the original author, Professor Simon Baron-Cohen, the AQ-28 underwent a rigorous translation and cultural adaptation for the Arabic language and environment. These processes ensured that the meaning of each item remains consistent with the original version, following international norms and guidelines.⁴² The forward-backward technique was adopted. Firstly, the AQ-28 was translated from English to Arabic by a Lebanese translator who was not involved in the study. Secondly, the Arabic translated version was back-translated to English by a Lebanese psychologist who is fluent in English. This approach was considered to guarantee a balance between contextual and literal translation. Afterwards, both the original and the back-translated English versions were compared by a panel of experts composed of the translators, the research team, two psychiatrists, and one psychologist, in order to confirm the accuracy of the translation, resolve any inconsistencies,⁴³ and ensure that the scale’s conceptual consistency in both the original and Arabic settings was maintained.⁴⁴ The Arabic version was examined to determinate any differences between source and Arab, Lebanese culture and maintain meaning equivalence. At this stage, no problems with adaptation into Arabic were identified in the 28 questions, and no adjustment to the translation was made upon review. Specifically, no changes in the wording of the items, the language of the item, or the instruction set were done. A pilot study on 30 persons was then conducted to ensure that items are clear and easily interpreted. No further adjustments were needed.

The Patient Health Questionnaire-9 (PHQ-9)

The PHQ-9 is a self-administered scale containing 9 items (e.g., “Little interest or pleasure in doing things”) to measure the severity of depressive symptoms.⁴⁵ Items are scored on a 4-point Likert scale ranging between 0 (not at all) and 3 (nearly every day), with total scores varying from 0 to 27. Greater scores reflect more severe depression. The Arabic validated version was utilized,⁴⁶ and had good internal consistency in this study (Cronbach α = 0.91).

The Generalized Anxiety Disorder 7-item scale (GAD-7)

This is a self-report measure composed of seven items (e.g., “Feeling nervous, anxious, or on edge”) assessing the severity of anxiety symptoms over the past 2 weeks.⁴⁷ Items are rated on a four-point Likert-type scale ranging from 0 (not at all) to 3 (nearly every day). Total scores can vary from 0 to 21. The Arabic validated version of the GAD-7 was adopted,⁴⁸ which yielded a Cronbach α of 0.92 in our sample.

The Jong-Gierveld Loneliness Scale (JGLS)

The degree of loneliness experienced by participants was assessed using the 5-item version of the JGLS (e.g., “I miss having people around”; “I experience a general sense of emptiness”).⁴⁹ Each item is scored following a yes/no question format; 0 point is awarded for a negative answer and 1 point for a positive answer. Higher scores indicate higher levels of loneliness. The Arabic validated version of the scale was used,⁵⁰ with a Cronbach α of 0.78 in this study.

We included PHQ-9 and GAD-7 to assess convergent validity, two constructs commonly associated with autistic traits, using the Arabic validated versions of those scales. To evaluate discriminant validity, we used the JGLS scale, as loneliness represents a related but conceptually distinct emotional experience.

Analytic strategy

Confirmatory factor analysis

There were no missing responses in the dataset. We used data from the total sample to conduct a CFA using the SPSS AMOS v.29 software. We aimed to enrol a minimum of 560 persons following the recommendations of Mundfrom et al.⁵¹ of 3 to 20 times the number of the scale’s variables. Parameter estimates were obtained using the maximum likelihood method. Multiple fit indices were calculated: root mean square error of approximation (RMSEA), standardized root mean square residual (SRMR), Tucker–Lewis Index (TLI), and Comparative Fit Index (CFI). Values ≤ 0.08 for RMSEA, ≤0.05 for SRMR and ≥0.90 for CFI and TLI indicate good fit of the model to the data.⁵² We also calculated the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC) for model comparisons; if the difference between the two values is <2, both models are considered adequate, whereas if the difference between the values is >10, the model with the lower value is preferred. Multivariate normality was not verified at first (Bollen–Stine bootstrap p = 0.002); therefore, we performed a bootstrapping procedure with 500 samples to adjust for the effects of non-normality on model fit estimation. Additionally, we computed bias-corrected 95% CI for parameter estimates (e.g., standardized loadings) using 500 bootstrap samples, providing more robust inference under non-normal data conditions.

Measurement invariance across sex was examined to ensure that the AQ-28 measures autistic traits similarly in males and females. This step is crucial given prior evidence of sex-related differences in the expression and recognition of autistic traits (Napolitano, 2022, #82),⁵³ which may impact scale functioning. Measurement invariance was assessed at the configural, metric, and scalar levels.⁵⁴ We accepted ΔCFI ≤ 0.010 and ΔRMSEA ≤ 0.015 or ΔSRMR ≤ 0.010 as evidence of invariance.⁵⁵ Mann–Whitney test was used to compare two means.

Internal consistency reliability was assessed using McDonald’s ω and Cronbach’s α, with values greater than 0.70 reflecting adequate Internal consistency reliability. Spearman test was used to correlate the AQ scores with the other scales in the survey.

Results

Participants

One thousand seventy-six persons participated in this study, with a mean age of 27.90 ± 11.81 years, 63.8% females and 82.3% with a university level of education.

Confirmatory factor analysis

The fit indices of all models are summarized in Table 1. The results showed that the five-factor model had the best fit, with the lowest AIC and BIC values. The standardized factor loadings for that model are summarized Table 2. Internal consistency reliability of scores were as follows: Factor 1 (ω = 0.71/α = 0.71), Factor 2 (ω = 0.57/α = 0.57), Factor 3 (ω = 0.67/α = 0.66), Factor 4 (ω = 0.77/α = 0.78), Factor 5 (ω = 0.82/α = 0.82), and the total score (ω = 0.91/ α = 0.91).

Table 1.

Fit indices of the different models of the AQ-28 scale.

Model	RMSEA (90% CI)	χ² (df)	SRMR	CFI	TLI	AIC	BIC
1 factor	0.078 (0.076; 0.081)	2662.63 (350)	0.067	0.746	0.726	2774.63	3053.56
2 factors	0.069 (0.066; 0.071)	2115.24 (349)	0.066	0.806	0.790	2229.24	2513.16
5 factors	0.061 (0.059; 0.064)	1722.41 (340)	0.060	0.848	0.831	1854.41	2183.16
Second order of the 5-factors	0.063 (0.060; 0.066)	1804.68 (345)	0.061	0.840	0.825	1926.68	2230.52

Table 2.

Standardized loading factors of the five-factor model of the Autism-Spectrum Questionnaire-28 in Arabic.

Item	Loading factor	Standard error	p
Factor 1—social skills
1. I prefer to do things with others rather than on my own	0.58	—	—
2. I find social situations easy	0.61	0.07	<0.001
3. I would rather go to a library than to a party	0.32	0.07	<0.001
4. I find myself drawn more strongly to people than to things	0.55	0.07	<0.001
5. I find it hard to make new friends	0.29	0.07	<0.001
6. I enjoy social occasions	0.68	0.07	<0.001
7. I enjoy meeting new people	0.66	0.07	<0.001
Factor 2—routine
8. I prefer to do things the same way over and over again	0.44	—
9. It does not upset my if my daily routine is disturbed	0.53	0.10	<0.001
10. I enjoy doing things spontaneously	0.61	0.10	<0.001
11. New situations make me anxious	0.42	0.09	<0.001
Factor 3—switching
12. I frequently get strongly absorbed in one thing	0.50	—	—
13. I can easily keep track of several different people’s conversations	0.61	0.09	<0.001
14. I find it easy to do more than one thing at once	0.61	0.09	<0.001
15. If there is an interruption, I can switch back very quickly	0.61	0.09	<0.001
Factor 4—imagination
16. Trying to imagine something, I find it easy to create a picture in my mind	0.61	—	—
17. Reading a story, I can easily imagine what the characters might look like	0.62	0.06	<0.001
18. I find making up stories easy	0.60	0.07	<0.001
19. Reading a story, I find it difficult to work out the character’s intentions	0.48	0.07	<0.001
20. I find it easy to work out what someone is thinking or feeling	0.60	0.07	<0.001
21. I find it difficult to imagine what it would be like to be someone else	0.50	0.07	<0.001
22. I find it difficult to work out people’s intentions	0.47	0.07	<0.001
23. I find it easy to play games with children that involve pretending	0.55	0.07	<0.001
Factor 5—numbers and patterns
24. I usually notice car number plates or similar strings of information	0.66	—	—
25. I am fascinated by dates	0.70	0.06	<0.001
26. I am fascinated by numbers	0.76	0.06	<0.001
27. I notice patterns in things all the time	0.66	0.06	<0.001
28. I like to collect information about categories of things	0.68	0.06	<0.001

Sex invariance

We were able to show the invariance across sex at the configural, metric, and scalar levels (Table 3). A significantly higher AQ-28 score was found in males compared to females (12.20 ± 3.30 vs 11.63 ± 3.55; t(1074) = 2.67; p = 0.005; Cohen’s d = 0.166).

Table 3.

Measurement invariance of the Autism Questionnaire across sex in the total sample.

Model	CFI	RMSEA	SRMR	Model comparison	ΔCFI	ΔRMSEA	ΔSRMR
Configural	0.835	0.046	0.066
Metric	0.833	0.045	0.074	Configural versus metric	0.002	0.001	0.008
Scalar	0.829	0.045	0.074	Metric versus scalar	0.004	<0.001	<0.001

CFI: comparative fit index; RMSEA: root mean square error of approximation; SRMR: standardized root mean square residual.

Concurrent and discriminant validity

The total AQ score was significantly associated with the social skills (rho = 0.68; p < 0.001), routine (rho = 0.44; p < 0.001), switching (rho = 0.43; p < 0.001), imagination (rho = 0.58; p < 0.001), and numbers and patterns (rho = 0.25; p < 0.001) subscales. Higher loneliness (rho = 0.13; p < 0.001), depression (rho = 0.14; p < 0.001) and anxiety (rho = 0.14; p < 0.001) were significantly associated with more autistic traits.

Discussion

The AQ is valuable as a quick screening instrument for core trait dimensions of autism, and has been widely used globally in clinical practice and autism research. To our knowledge, the present study is the first attempt to validate the abridged version of the AQ scale in the Arabic language. Our findings showed that the Arabic version of the AQ-28 exhibited a five-factor structure that was invariant across sex, and satisfactory internal consistency. In addition, the scale demonstrated good concurrent validity as evidenced by appropriate patterns of correlations with measures of loneliness, depression and anxiety.

The AQ-28 items were originally developed to reflect core features of autistic traits and have been widely used in both clinical and non-clinical populations. As such, the scale demonstrates adequate face validity, as its items are clearly and recognizably related to behaviors and traits associated with autism. Additionally, the Arabic translation underwent expert review to ensure that item content remained conceptually faithful and culturally understandable, further supporting face validity in our target population.

The current results were able to replicate the original five-factor model found by Hoekstra et al.³⁰ in an Arabic-speaking population and country. Other linguistic versions of the AQ-28 validated in different countries and settings also landed support to the proposed multifactorial model.^31–34 Furthermore, findings of the reliability analysis showed an acceptable internal consistency for the AQ-28 (Cronbach ranging from 0.57 to 0.91), thus supporting the results previously found in the original (α ranging from 0.77 to 0.86)³⁰ as well as other validation studies (e.g., Spanish, α between 0.79 and 0.88,³² Dutch, α between 0.72 and 0.89,³³ Malaysian, α between 0.50 and 0.88,³³ Persian, α between 0.48 and 0.95). While the total AQ-28 score demonstrated excellent internal consistency (ω = 0.91/α = 0.91), three of the five subscales exhibited Cronbach’s alpha values below the conventional threshold of 0.70. Although this indicates limited internal consistency for these specific dimensions, it is important to note that values between 0.60 and 0.70 are often considered acceptable for brief or exploratory subscales in psychological research.⁵⁶ Despite these facts, their inclusion was retained to preserve comparability with established factor structures and to reflect the multidimensional nature of autistic traits, despite the modest reliability coefficients.

As a multidimensional measure, the AQ-28 has not only the capacity to offer an overall evaluation of autism trait levels through a total scale score, but also the ability to evaluate trait dimensions independently at the subscale level. In particular, the scale measures five clearly defined domains, assessing attention switching difficulties (“Switching”), difficulties with imagination (“Imagination”), a preference for routine (“Routine”), difficulties with social skills (“Social skills”), and a fascination for numbers/patterns (“Numbers/patterns”). This multi-dimensional approach to the examination of the broader Autism Phenotype can be highly valuable in determining the differential association between some autism domains and certain individual characteristics. For instance, more social difficulties were distinctly and specifically linked to the presence of a local processing bias in non-clinical individuals.⁵⁷ Other research showed that imagination difficulties are specific predictors of difficulty in perceiving speech in noise in non-autistic individuals.⁵⁸ Therefore, as autistic traits are heterogeneous in nature, it is crucial to not treat all autistic individuals as a homogenous group, and rather explore the shared as well as distinct characteristics of particular domains of autism using a psychometrically-sound multifactorial measure such as the AQ-28. Some fit indices (e.g., CFI and TLI) fall below the commonly accepted thresholds for good model fit; while these values were interpreted as indicating an acceptable fit, we acknowledge that they reflect only a marginal fit and should be interpreted with caution. These cut-off values should not be interpreted rigidly^59,60 and values between 0.80 and 0.90 for CFI and TLI can indicate acceptable but mediocre fit to the data.⁶¹ This suggests that the model may not fully capture the underlying factor structure and warrants further investigation in future research. Although it has been suggested that each factor should ideally have at least four items with loadings of 0.60 or higher to ensure stability,⁶² this criterion was not met in our results. Nevertheless, the factor structure was retained given the theoretical relevance of the items and the overall pattern of loadings, in line with common practice in psychological scale validation where such ideal thresholds are not always attainable. Although our CFA results and internal consistency estimates provide preliminary support for the multidimensional structure of the AQ-28, we acknowledge that further psychometric evaluation is needed to justify the independent use of subscale scores. As Reise et al.⁶³ noted, multidimensionality alone does not guarantee that subscales function as psychometrically sound standalone constructs. Although the CFA supported a multidimensional structure of the AQ-28, some subscales exhibited relatively low internal consistency in this sample. Given the high internal consistency of the total score, we recommend that the AQ-28 total score be prioritized for interpretation. Subscale-level scores may still be useful for clinical purposes but should be interpreted with caution when used separately.

Measurement invariance was tested and established across male and female respondents at the configural, metric, and scalar levels. This psychometric property is a necessary step for the questionnaire validation to ensure meaningful group comparisons based on observed AQ-28 scores and guarantee that any observed differences are not caused by test bias.⁶⁴ Consistent with our findings, cross-sex measurement invariance has previously been reported for the AQ in both autistic and non-autistic individuals.^35,65 Of note, an emerging body of research noticed that the AQ had some shortcomings, including differential item functioning (DIF) between groups, leading researchers to question the validity of comparisons of AQ scores and to caution users against interpretational biases in responses to individual items.⁶⁶ For instance, Belcher et al.⁶⁷ found that the likelihood of endorsement of most of the AQ items was influenced by gender. One potential explanation could be a bias towards males in the diagnosis of autism,⁶⁸ as existing studies have predominantly involved male participants.⁶⁹ Interestingly, Agelink van Rentergem⁷⁰ observed that the set of items that showed DIF consistently across analyses was not included in the AQ-28, thus recommending this short form for use when comparing groups.

The total AQ score was significantly associated with all five subscales, with correlations ranging from 0.25 to 0.68. While moderate to strong associations were observed with most subscales- particularly social skills and imagination- the weaker correlation with numbers and patterns suggests partial divergence in trait dimensions. These findings support a partially unified but multidimensional structure of autistic traits, consistence with prior evidence,⁷¹ and indicate that the total score should be interpreted with caution, especially when subscale-level differentiation is theoretically important.

In our study, mean sum scores observed in male participants were significantly higher than those displayed by females. In the original validation study,²⁹ sex differences in AQ scores were observed, with male students scoring higher on all AQ subscales than their female counterparts, with the exception of attention to detail. However, as measurement invariance was not analyzed, it is unclear whether the differences seen reflect true differences or sex differential functioning of the scale’s items. Likewise, studies in Japanese⁷² and Polish⁷³ adult students found that males displayed greater severity of autistic traits than females, scoring higher in all AQ scores but attention to detail. A systematic review encompassing 73 articles and 6934 participants showed that males and female adults from the general population had significantly different mean AQ scores, with males scoring greater than females; whereas no significant sex difference was found in AQ scores in the clinical ASD population.⁴ These observations along with our findings might suggest that sex differences in AQ scores seem to be cross-culturally independent. However, mixed findings have also been previously reported in the literature, with studies indicating either the lack of sex differences in AQ scores,⁷⁴ or even slightly higher (but still non-significant) AQ scores exhibited by females compared to males.⁷⁵ Altogether, there remains a debate around the extent to which an apparent disparity exists between males and females in the incidence of autism.^76,77 More research in various cultural contexts and geographical locations are required to further elucidate the nature of sex differences. This research should consider using autism measures that are psychometrically invariant for males and females to rule out test bias as possible explanations for between-group differences in scores.

This study’s findings provided evidence for concurrent validity and indicated that autistic traits as measured by the Arabic AQ-28 correlated positively with loneliness, depression and anxiety. These results are in line with earlier studies consistently showing that sub-threshold Autistic Traits among adults from the general population are associated and increased susceptibility towards the development of different psychopathological trajectories.⁷⁸ In agreement with our findings, prior research reported that adults with more pronounced autistic traits are likely to experience loneliness more frequently and more intensely than their non-autistic counterparts,⁷⁹ and to develop depression and anxiety symptoms.⁸⁰ Taken together, our findings support the clinical utility of the AQ-28 for autistic traits assessment and ASD diagnosis screening, as well as for identifying co-occurring psychopathology. However, users of the AQ-28 should be cautioned that the measure is not intended to make a diagnosis, it is rather used as a screening tool to help in the diagnostic process of ASD and related conditions.

Study limitations

This study has some limitations that can guide future lines of research. First, the target population consisted of adults from the general population, which can limit the generalizability of results to other clinical and non-clinical populations, such as students or individuals diagnosed with ASD or with schizophrenia. Second, due to snowball sampling, the sample representativeness may be affected. Third, an internet survey was adopted in the current study. Although this recruitment method has several advantages, as it offers the possibility of reach a large group of potential respondents and the opportunity to cover a larger geographic area, in addition to being quicker to conduct and less expensive than other methods. However, online surveys are also prone to sampling biases and limited representativeness, as they may not reach certain populations (e.g., those who are less likely to have internet access and to respond to online questionnaires).⁸¹ Fourth, female and highly educated participants were slightly overrepresented, which could have influenced the results and limit their generalizability. Additionally, the sample had a relatively low mean age (27.90 years). At the same time, it can be suggested that the sample size was large enough for the psychometric testing. To uncover these limitations, future studies using probability sampling methods are needed to improve the robustness of our findings may and produce more representative results. Future work should also incorporate mixed-effects modelling to capture nuanced gender-based differences across domains of autistic traits. Fifth, the study involved Arabic-speaking adults from a single country. Future studies using a multinational sample could specifically examine cross-country differences in the factor structure of the AQ-28. In addition, other autism screening tools could not be used for convergent validity, due to the lack of validated measures assessing ATs in the Arabic language. Besides, information on personal history of physician-diagnosed autism spectrum disorder or related conditions and on gender identity were not gathered in the context of the present study. To address these limitations, future research should include non-binary people, and use standardized interview to diagnose autism among participants.

Furthermore, additional psychometric properties of the Arabic AQ-28, such as test-retest reliability, divergent validity and cut-off score, still need to be investigated, in order to respectively verify the stability of the measure over time and help predict respondents most likely to fulfil diagnostic criteria for ASD. There is also a need to examine the relationship between AQ scores and scores on measures that assess similar constructs (e.g., BAPQ, CATI). Finally, although measurement invariance across sex has been established in our sample, the AQ may have some issues related to DIF. An alternative could be the validation of other measures, such as the Comprehensive Autistic Trait Inventory,⁸² that was designed to capture broader experiences of autism in both sexes.

Implications and future perspectives

These limitations aside, the current study extends and makes a major contribution to the international literature on autism by validating, for the first time, the AQ-28 into Arabic. The Arabic version of AQ-28 consists of 28 statements referring to five functional domains, which is almost half the length of the original long-form. Screening for autistic traits in clinical and community settings using 50 items may be too demanding, particularly in the developing-Arab-countries’ settings, where time, resources and costs can be a strain. Due to its brevity and good psychometric qualities, the AQ-28 allows to reduce the burden of test administration, while preserving the validity and reliability of its scores. This would facilitate its routine application in clinical settings. Moreover, multi-group analyses provided evidence of the measurement equivalence of the scale across sex, making it possible to make accurate between-sex group comparisons on AQ-28 scores. The way how item interpretation could be affected by linguistic or cultural factors remains a potential research perspective that needs further development using different cross-national and cross-cultural data sets.

It needs to be emphasized that the AQ-28 is a screening but not a diagnostic instrument. Individuals endorsing high scores on this scale do not necessarily have ASD. To establish a clinical diagnosis and conduct detailed clinical research, more comprehensive assessments and interviews remain essential to determine the presence and clinical significance of key features and core symptoms of autism.²⁹

To date, the field of autism research in adults has not yet emerged in Arab countries. It is our hope that the Arabic-language version of the AQ-28 provided by this study will generate interest among researchers in this important area of research, and help to galvanize efforts to better understand the nature and etiology of the broader autism phenotype in the general population across cultures, as well as implications for practice and health policy.

Conclusion

The present findings together suggest that the Arabic translated version of the AQ-28 is a useful tool in quantifying autistic traits among Arabic-speaking adults from the general population. The availability of a brief, easy-to-administer, valid, and reliable self-report measure of autistic traits in Arabic will hopefully allow to timely recognize adults with mildest-forms of ASD, who have psychological problems and adaptive difficulties, and who seek care in clinical settings in the Arabic speaking countries. Additionally, the Arabic AQ-28 is expected to open broader perspectives for researchers who work in this field in the different Arab countries.

Supplemental Material

sj-docx-1-phj-10.1177_22799036251377610 – Supplemental material for Validation of the abridged version of the autism-spectrum quotient (AQ-28) in the Arabic-speaking adult general population

Supplemental material, sj-docx-1-phj-10.1177_22799036251377610 for Validation of the abridged version of the autism-spectrum quotient (AQ-28) in the Arabic-speaking adult general population by Feten Fekih-Romdhane, Fouad Sakr, Amthal Alhuwailah, Leila Sarra Chaibi, Mai Helmy, Hanaa Ahmed Mohamed Shuwiekh, Nedjem Eddine Boudouda, Btissame Zarrouq, Abdallah Y. Naser, Kamel Jebreen, Mohammed Lakhdar Roubi, Bassam Abdul Rasool Hassan, Nisma Merdad, Rizwana Amin, Inad Nawajah, Ali Haider Mohammed, Sinan Subhi Farhan, Omar Abdulwahid AlAni, Majda Cheour, Mariam Dabbous, Diana Malaeb, Sahar Obeid and Souheil Hallit in Journal of Public Health Research

Supplemental Material

sj-docx-2-phj-10.1177_22799036251377610 – Supplemental material for Validation of the abridged version of the autism-spectrum quotient (AQ-28) in the Arabic-speaking adult general population

Supplemental material, sj-docx-2-phj-10.1177_22799036251377610 for Validation of the abridged version of the autism-spectrum quotient (AQ-28) in the Arabic-speaking adult general population by Feten Fekih-Romdhane, Fouad Sakr, Amthal Alhuwailah, Leila Sarra Chaibi, Mai Helmy, Hanaa Ahmed Mohamed Shuwiekh, Nedjem Eddine Boudouda, Btissame Zarrouq, Abdallah Y. Naser, Kamel Jebreen, Mohammed Lakhdar Roubi, Bassam Abdul Rasool Hassan, Nisma Merdad, Rizwana Amin, Inad Nawajah, Ali Haider Mohammed, Sinan Subhi Farhan, Omar Abdulwahid AlAni, Majda Cheour, Mariam Dabbous, Diana Malaeb, Sahar Obeid and Souheil Hallit in Journal of Public Health Research

Footnotes

Acknowledgements

The authors would like to thank all participants.

ORCID iDs

Feten Fekih-Romdhane

Fouad Sakr

Souheil Hallit

Ethical considerations

The protocol was approved by the home institutions of the study’s principal investigators (FFR and SH), namely the ethics committee of Razi Psychiatric Hospital, Manouba, Tunisia (ECRPH-2024-032) and the Lebanese International University’s School of Pharmacy ethics committee (2024ERC-025-LIUSOP).

Consent to participate

When filling out the online form, each participant provided written informed consent. All methods were performed in accordance with the relevant guidelines and regulations (in accordance with the Declaration of Helsinki).

Consent for publication

Not applicable.

Author contributions

FF-R and SH designed the study; FF-R drafted the manuscript; SH carried out the analysis and interpreted the results; FS, AA, LSC, MH, HAMS, NEB, BZ, AYN, KJ, MLR, NM, RA, BARH, AHM, SSF, OAA, and MD collected the data; MC, DM, and SO reviewed the paper for intellectual content; all authors reviewed the final manuscript and gave their consent.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

Because of ethical committee constraints, none of the data collected or analyzed during this study are publicly available. However, the corresponding author (SH) may make the data available upon reasonable request.

Supplemental material

Supplemental material for this article is available online.

References

American Psychiatric Association. Diagnostic and statistical manual of mental disorders: DSM-5, vol. 5. Washington, DC: American Psychiatric Association, 2013.

Grzadzinski

Huerta

Lord

DSM-5 and autism spectrum disorders (ASDs): an opportunity for identifying ASD subtypes. Mol Autism 2013; 4: 12.

Mosconi

Stevens

Unruh

, et al Endophenotype trait domains for advancing gene discovery in autism spectrum disorder. J Neurodev Disord 2023; 15(1): 41.

Eyuboglu

Baykara

Eyuboglu

Broad autism phenotype: theory of mind and empathy skills in unaffected siblings of children with autism spectrum disorder. Psychiatr Clin Psychopharmacol 2018; 28(1): 36–42.

Jussila

Junttila

Kielinen

, et al Sensory abnormality and quantitative autism traits in children with and without autism spectrum disorder in an epidemiological population. J Autism Develop Disord 2020; 50: 180–188.

Constantino

Todd

RD.

Autistic traits in the general population: a twin study. Arch Gen Psychiatry 2003; 60(5): 524–530.

Stevenson

Hart

KR.

Psychometric properties of the autism-spectrum quotient for assessing low and high levels of autistic traits in college students. J Autism Develop Disord 2017; 47(6): 1838–1853.

Ruzich

Allison

Smith

, et al Measuring autistic traits in the general population: a systematic review of the Autism-Spectrum Quotient (AQ) in a nonclinical population sample of 6,900 typical adult males and females. Mol Autism 2015; 6(1): 2.

Gurbuz

Riby

South

, et al Associations between autistic traits, depression, social anxiety and social rejection in autistic and non-autistic adults. Sci Rep 2024; 14(1): 9065.

10.

Stice

Lavner

JA.

Social connectedness and loneliness mediate the association between autistic traits and internalizing symptoms among young adults. J Autism Develop Disord 2019; 49: 1096–1110.

11.

Suzuki

Inagaki

The relationships among autism spectrum disorder traits, loneliness, and social networking service use in college students. J Autism Develop Disord 2021; 51(6): 2047–2056.

12.

Gallitto

Leth-Steensen

Autistic traits and adult attachment styles. Pers Individ Differ 2015; 79: 63–67.

13.

Ameli

Ghaderi

Aslezaker

, et al The relationship between attachment styles and autistic traits: considering the mediating role of empathy. Res Bull Med Sci 2019; 24(1): e13.

14.

Gökçen

Frederickson

Petrides

Theory of mind and executive control deficits in typically developing adults and adolescents with high levels of autism traits. J Autism Develop Disord 2016; 46: 2072–2087.

15.

Stimpson

Hull

Mandy

The association between autistic traits and mental well-being. J Happiness Stud 2021; 22(1): 287–304.

16.

Richards

Kenny

Griffiths

, et al Autistic traits in adults who have attempted suicide. Mol Autism 2019; 10(1): 26.

17.

Cassidy

Au-Yeung

Robertson

, et al Autism and autistic traits in those who died by suicide in England. Br J Psychiatry 2022; 221(5): 683–691.

18.

Postorino

Vicari

Mazzone

Mood disorders and autism spectrum disorder. In: Vitiello

Mazzone

(eds.) Psychiatric symptoms and comorbidities in autism spectrum disorder. New York, NY: Springer International Publishing, 2016, pp. 1–19.

19.

Wikramanayake

WNM

Mandy

Shahper

, et al Autism spectrum disorders in adult outpatients with obsessive compulsive disorder in the UK. Int J Psychiatry Clin Pract 2018; 22(1): 54–62.

20.

Barlati

Deste

Gregorelli

, et al Autistic traits in a sample of adult patients with schizophrenia: prevalence and correlates. Psychol Med 2019; 49(1): 140–148.

21.

Rhind

Bonfioli

Hibbs

, et al An examination of autism spectrum traits in adolescents with anorexia nervosa and their parents. Mol Autism 2014; 5: 1–9.

22.

Dell’Osso

Cremone

Carpita

, et al Correlates of autistic traits among patients with borderline personality disorder. Compr Psychiatry 2018; 83: 7–11.

23.

Whitehouse

AJO and Bishop DVM.

The Communication Checklist–Adult version (CC-A), London: Pearson, 2009.

24.

Constantino

JN.

Social responsiveness scale. In: Volkmar

(ed.) Encyclopedia of autism spectrum disorders. Cham: Springer, 2021, pp. 4457–4467.

25.

Posserud

Lundervold

Gillberg

Autistic features in a total population of 7–9-year-old children assessed by the ASSQ (Autism Spectrum Screening Questionnaire). J Child Psychol Psychiatry 2006; 47(2): 167–175.

26.

Skuse

Mandy

Scourfield

Measuring autistic traits: heritability, reliability and validity of the Social and Communication Disorders Checklist. Br J Psychiatry 2005; 187(6): 568–572.

27.

Hurley

Losh

Parlier

, et al The broad autism phenotype questionnaire. J Autism Develop Disord 2007; 37: 1679–1690.

28.

Bishop

Whitehouse

Sharp

The communication checklist-self report (CC-SR). London, UK: Pearson, 2009.

29.

Baron-Cohen

Wheelwright

Skinner

, et al The autism-spectrum quotient (AQ): evidence from asperger syndrome/high-functioning autism, malesand females, scientists and mathematicians. J Autism Develop Disord 2001; 31: 5–17.

30.

Hoekstra

Vinkhuyzen

Wheelwright

, et al The construction and validation of an abridged version of the autism-spectrum quotient (AQ-Short). J Autism Develop Disord 2011; 41: 589–596.

31.

Ameli

Ghaderi

Aslzaker Lighvan

, et al Psychometric properties of the abridged version of the autism-spectrum quotient among Iranian students. Pract Clin Psychol 2019; 7(4): 303–318.

32.

Lugo-Marín

Díez-Villoria

Magán-Maganto

, et al Spanish validation of the Autism Quotient Short Form Questionnaire for adults with autism spectrum disorder. J Autism Develop Disord 2019; 49(11): 4375–4389.

33.

Chee

Scheeren

De Vries

The factor structure and measurement invariance of the Autism Spectrum Quotient-28: a cross-cultural comparison between Malaysia and the Netherlands. Autism 2024; 28(1): 32–42.

34.

Kuenssberg

Murray

Booth

, et al Structural validation of the abridged Autism Spectrum Quotient–Short Form in a clinical sample of people with autism spectrum disorders. Autism 2014; 18(2): 69–75.

35.

Grove

Hoekstra

Wierda

, et al Exploring sex differences in autistic traits: a factor analytic study of adults with autism. Autism 2017; 21(6): 760–768.

36.

Waldren

Livingston

Shah

Is there an optimal self-report measure to investigate autism-related sex differences?

Res Autism 2022; 125: 202617.

37.

Freeth

Sheppard

Ramachandran

, et al A cross-cultural comparison of autistic traits in the UK, India and Malaysia. J Autism Develop Disord 2013; 43: 2569–2583.

38.

Chee

de Vries

Language matters: the autism-spectrum quotient in English, Mandarin and Bahasa Malaysia. J Autism Develop Disord 2022; 52(9): 3814–3824.

39.

Masuda

Ellsworth

Mesquita

, et al Placing the face in context: cultural differences in the perception of facial emotion. J Pers Soc Psychol 2008; 94(3): 365.

40.

Robinson

Koenen

McCormick

, et al Evidence that autistic traits show the same etiology in the general population and at the quantitative extremes (5%, 2.5%, and 1%). Arch Gen Psychiatry 2011; 68(11): 1113–1121.

41.

von Elm

. STROBE initiative. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med 2007; 147: 573–577.

42.

Van Widenfelt

Treffers

De Beurs

, et al Translation and cross-cultural adaptation of assessment instruments used in psychological research with children and families. Clin Child Fam Psychol Rev 2005; 8(2): 135–147.

43.

Fenn

Tan

C-S

George

Development, validation and translation of psychological tests. BJPsych Adv 2020; 26(5): 306–315.

44.

Ambuehl

Inauen

Contextualized measurement scale adaptation: a 4-Step tutorial for health psychology research. Int J Environ Res Public Health 2022; 19(19): 12775.

45.

Kroenke

Spitzer

Williams

JB.

The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med 2001; 16(9): 606–613.

46.

Dagher

Samaha

Mhanna

, et al Depressive symptoms among a sample of Lebanese adolescents: scale validation and correlates with disordered eating. Arch Pediatr 2023; 30(6): 401–407.

47.

Spitzer

Kroenke

Williams

, et al A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med 2006; 166(10): 1092–1097.

48.

Sawaya

Atoui

Hamadeh

, et al Adaptation and initial validation of the Patient Health Questionnaire - 9 (PHQ-9) and the Generalized Anxiety Disorder - 7 Questionnaire (GAD-7) in an Arabic speaking Lebanese psychiatric outpatient sample. Psychiatry Res 2016; 239: 245–252.

49.

Wilson

Krueger

Arnold

, et al Loneliness and risk of Alzheimer disease. Arch Gen Psychiatry 2007; 64(2): 234–240.

50.

Awad

Rogoza

Gerges

, et al Association of social media use disorder and orthorexia nervosa among Lebanese university students: the indirect effect of loneliness and factor structure of the social media use disorder short form and the Jong-Gierveld Loneliness Scales. Psychol Rep 2022; 127(3): 1065–1084.

51.

Mundfrom

Shaw

TL.

Minimum sample size recommendations for conducting factor analyses. Int J Test 2005; 5(2): 159–168.

52.

Bentler

PM.

Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct Equ Model Multidiscip J 1999; 6(1): 1–55.

53.

Napolitano

Schiavi

La Rosa

, et al. Sex Differences in Autism Spectrum Disorder: Diagnostic, Neurobiological, and Behavioral Features. Front Psychiatry 2022; 13: 889636. https://doi.org/10.3389/fpsyt.2022.889636

54.

Vadenberg

Lance

A review and synthesis of the measurement in variance literature: suggestions, practices, and recommendations for organizational research. Organ Res Methods 2000; 3: 4–70.

55.

Swami

Todd

Azzi

, et al Psychometric properties of an Arabic translation of the Functionality Appreciation Scale (FAS) in Lebanese adults. Body Image 2022; 42: 361–369.

56.

Hair

Jr Hult

GTM

Ringle

, et al Evaluation of reflective measurement models. In: Partial least squares structural equation modeling (PLS-SEM) using R: A workbook. New York, NY: Springer International Publishing, 2021, pp. 75–90.

57.

Russell-Smith

Maybery

Bayliss

, et al Support for a link between the local processing bias and social deficits in autism: an investigation of embedded figures test performance in non-clinical individuals. J Autism Dev Disord 2012; 42(11): 2420–2430.

58.

Tsuji

Nishiguchi

Noda

, et al

Autistic traits and temporal integration of auditory and visual stimuli in the general population: The role of imagination. 2025. https://doi.org/10.31234/osf.io/38d6j_v3

59.

Heene

Hilbert

Draxler

, et al Masking misfit in confirmatory factor analysis by increasing unique variances: a cautionary note on the usefulness of cutoff values of fit indices. Psychol Methods 2011; 16(3): 319.

60.

Perry

Nicholls

Clough

, et al Assessing model fit: caveats and recommendations for confirmatory factor analysis and exploratory structural equation modeling. Meas Phys Educ Exerc Sci 2015; 19(1): 12–21.

61.

Hooper

Coughlan

Mullen

Evaluating model fit: a synthesis of the structural equation modelling literature. In: 7th European conference on research methodology for business and management studies: 2008, London, UK, 19–20 June 2008, pp. 195–200.

62.

Guadagnoli

Velicer

WF.

Relation of sample size to the stability of component patterns. Psychol Bull 1988; 103(2): 265.

63.

Reise

Bonifay

Haviland

MG.

Scoring and modeling psychological measures in the presence of multidimensionality. J Pers Assess 2013; 95(2): 129–140.

64.

Wicherts

Dolan

CV.

Measurement invariance in confirmatory factor analysis: an illustration using IQ test performance of minorities. Educ Meas Iss Pract 2010; 29(3): 39–47.

65.

English

Gignac

Visser

, et al A comprehensive psychometric analysis of autism-spectrum quotient factor models using two large samples: model recommendations and the influence of divergent traits on total-scale scores. Autism Res 2020; 13(1): 45–60.

66.

Murray

Allison

Smith

, et al Investigating diagnostic bias in autism spectrum conditions: an item response theory analysis of sex bias in the AQ-10. Autism Res 2017; 10(5): 790–800.

67.

Belcher

Uglik-Marucha

Vitoratou

, et al Gender bias in autism screening: measurement invariance of different model frameworks of the Autism Spectrum Quotient. BJPsych Open 2023; 9(5): e173.

68.

Cruz

Zubizarreta

Costa

, et al Is there a bias towards males in the diagnosis of autism? A systematic review and meta-analysis. Neuropsychol Rev 2025; 35(1): 153–176.

69.

D’Mello

Frosch

, et al Exclusion of females in autism research: empirical evidence for a “leaky” recruitment-to-research pipeline. Autism Res 2022; 15(10): 1929–1940.

70.

Agelink van Rentergem

Lever

Geurts

HM.

Negatively phrased items of the Autism Spectrum Quotient function differently for groups with and without autism. Autism 2019; 23(7): 1752–1764.

71.

English

MCW

Gignac

Visser

TAW

72.

Kunihira

Senju

Dairoku

, et al ‘Autistic’ traits in non-autistic Japanese populations: relationships with personality traits and cognitive ability. J Autism Develop Disord 2006; 36: 553–566.

73.

Pisula

Kawa

Szostakiewicz

, et al Autistic traits in male and female students and individuals with high functioning autism spectrum disorders measured by the Polish version of the Autism-Spectrum Quotient. PLoS One 2013; 8(9): e75236.

74.

Hurst

Nelson-Gray

Mitchell

, et al The relationship of Asperger’s characteristics and schizotypal personality traits in a non-clinical adult sample. J Autism Develop Disord 2007; 37: 1711–1720.

75.

Hurst

Mitchell

Kimbrel

, et al Examination of the reliability and factor structure of the Autism Spectrum Quotient (AQ) in a non-clinical sample. Pers Individ Differ 2007; 43(7): 1938–1949.

76.

Constantino

Charman

Gender bias, female resilience, and the sex ratio in autism. J Am Acad Child Adolesc Psychiatry 2012; 51(8): 756–758.

77.

Dworzynski

Ronald

Bolton

, et al

How different are girls and boys above and below the diagnostic threshold for autism spectrum disorders?

J Am Acad Child Adolesc Psychiatry 2012; 51(8): 788–797.

78.

Dell’Osso

Lorenzi

Carpita

Autistic traits and illness trajectories. Clin Pract Epidemiol Ment Health 2019; 15: 94.

79.

Grace

Remington

Lloyd-Evans

, et al Loneliness in autistic adults: a systematic review. Autism 2022; 26(8): 2117–2135.

80.

Hollocks

Lerh

Magiati

, et al Anxiety and depression in adults with autism spectrum disorder: a systematic review and meta-analysis. Psychol Med 2019; 49(4): 559–572.

81.

Greenacre

ZA.

The importance of selection bias in internet surveys. Open J Statist 2016; 6(3): 397–404.

82.

English

Gignac

Visser

, et al The Comprehensive Autistic Trait Inventory (CATI): development and validation of a new measure of autistic traits in the general population. Mol Autism 2021; 12(1): 37.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB

0.03 MB