What is specific about employment status, workplace experiences and requirements in individuals with autism in Germany?

The challenge of sustainably integrating highly educated individuals with ASD without intellectual disabilities in the first labor market is repeatedly described in literature. In a retrospective study, a group of 197 clinically late-diagnosed adults with ASD without intellectual disabilities was compared to a closely matched group of 501 individuals who did not meet the criteria for the diagnosis of ASD within a utilization population of the Cologne Autism Outpatient Clinic. Results indicated that the pronounced demand for reduction of social and interpersonal requirements at the workplace (including planned or limited contact with colleagues and customers) as well as the experience of difficulties following unexpected changes in the daily routine were specific for ASD. In addition, individuals with ASD reported greater difficulties in finding a suitable job and being able to live on their wages, taking age and educational qualification into account. Supported employment measures were provided significantly more frequently to individuals in the ASD group. In conclusion, impairments in social skills emerged as one of the main obstacles of workplace performance for individuals with ASD emphasizing the necessity to develop and apply ASD-specific support services.


Abstract
The challenge of sustainably integrating highly educated individuals with ASD without intellectual disabilities in the first labor market is repeatedly described in literature. In a retrospective study, a group of 197 clinically late-diagnosed adults with ASD without intellectual disabilities was compared to a closely matched group of 501 individuals who did not meet the criteria for the diagnosis of ASD within a utilization population of the Cologne Autism Outpatient Clinic. Results indicated that the pronounced demand for reduction of social and interpersonal requirements at the workplace (including planned or limited contact with colleagues and customers) as well as the experience of difficulties following unexpected changes in the daily routine were specific for ASD. In addition, individuals with ASD reported greater difficulties in finding a suitable job and being able to live on their wages, taking age and educational qualification into account. Supported employment measures were provided significantly more frequently to individuals in the ASD group. In conclusion, impairments in social skills emerged as one of the main obstacles of workplace performance for individuals with ASD emphasizing the necessity to develop and apply ASD-specific support services.

Lay Summary
Individuals with autism spectrum disorder (ASD) without intellectual disabilities show high unemployment rates despite high level of education. In order to achieve a cooperative interaction between people with and without autism, barriers that people with autism face in their working life should be spelled out. We studied 698 persons who visited the Adult Autism Outpatient Clinic, in 197 the diagnosis was confirmed, in 501 cases the diagnosis was not confirmed. Results indicated that, while all individuals in the sample had requirements for adjustment of the working environment, individuals with ASD reported a specific need for reduction of social and interpersonal requirements and for structured daily routines at work. Further, individuals with ASD reported difficulties in finding a suitable employment with sufficient salary also influenced by age and school education. In addition, individuals with ASD in the tested sample were more often supported by sheltered training measures ('supported employment'). In conclusion, difficulties in dealing with social skills demands were one of the main factors challenging CF and KV should be considered joint last authors.

INTRODUCTION
The clinical diagnosis of an Autism spectrum disorder (ASD) is characterized according to ICD-10 by impairments in social interaction and communication as well as restrictive, inflexible behavior patterns (American Psychiatric Association, 2013). Behavioral, cognitive, social, and sensory characteristics occur in varying degrees of severity, and the diversity and classification of the different manifestations is becoming increasingly important (Bottema-Beutel et al., 2020;Buijsman et al., 2023;Bury et al., 2020;Happé, 1999;Kapp et al., 2013;Johnson & Joshi, 2016;Monk et al., 2022;Sarrett, 2017;Tepest, 2021). Despite growing public awareness of autism, the psychosocial outcomes of people with ASD lag behind those of non-disabled adults and people with other disabilities (Shattuck et al., 2012). Unemployment rates are alarmingly high in autism (Espelöer et al., 2022) and it is time to draw society's attention to the needs of people with autism in the working environment in order to enable them to join the labor market, from which all parties could benefit, considering special strengths people with autism could bring to the workforce (Johnson & Joshi, 2016;Sarrett, 2017). However, in the labor market, social skills are often required, which presents a major challenge for individuals with ASD (Arora, 2017;Fazekas, 2020;Riggio, 2020). Already the job application process is a challenge for individuals with ASD, requiring flexibility and adaptation to new routines, overcoming miscommunication as well as social interaction deficits (Baldwin et al., 2014;Hurlbutt & Chalmers, 2004;Müller et al., 2003;Sarrett, 2017). Accordingly, asking about challenges in working life, individuals with ASD report difficulties in understanding expectations of employers or colleagues, recognizing implicit as well as explicit statements, and completing work assignments to the correct degree and period. Difficulties are reported in dealing with customer contact and in adapting quickly to unfamiliar situations and demands. Social requirements in particular represent a barrier that often cannot be overcome and interpersonal difficulties are more likely to cause termination than professional problems (Espelöer et al., 2022;Hurlbutt & Chalmers, 2004;Müller et al., 2003;Proft et al., 2016).
When asking people with ASD about their specific requirements for their work environments, the needs for limited social interaction, clear communication, reduced sensory input, support in prioritization and sufficient time to learn new tasks were expressed (Kirchner & Dziobek, 2014;Müller et al., 2003;Proft et al., 2016).
Furthermore, the possibility of flexibly structuring their daily working schedule according to individual needs as well as consistent, predictable and clearly defined work assignments appear to be important. Nevertheless, intellectual demands must not lose relevance (Baldwin et al., 2014;Müller et al., 2003). Hence, alarmingly increased unemployment rates of up to five-fold higher compared to the general population (Espelöer et al., 2022;Maslahati et al., 2022) or underemployment paired with over-education in individuals with ASD without intellectual disabilities are the results (Baldwin et al., 2014;Frank et al., 2018;Hurlbutt & Chalmers, 2004;Maslahati et al., 2022;Riedel et al., 2016). Roux et al. (2013) described lower wages for individuals with ASD compared to individuals with learning or language difficulties or other mental health problems. Additionally, individuals with ASD without intellectual disabilities often need specific support, but are at the same time at higher risk of being overlooked in support services compared to individuals with other disabilities because of preserved intellectual abilities (Baldwin et al., 2014;Chen et al., 2015;Espelöer et al., 2022;Kirchner & Dziobek, 2014;Lehnhardt et al., 2012;Roux et al., 2013;Shattuck et al., 2012;Vogeley et al., 2013). Especially in the transition period of the first two years after leaving school, support seems to be of great importance so that individuals with ASD can enter working life successfully (Chen et al., 2015;Shattuck et al., 2012).
The importance of social competences in the workplace poses a challenge for people with mental disorders, as various disorders are associated with impairments in social interaction and communication, which might cause challenges in finding and maintaining suitable workplaces (Arora, 2017;Brunello & Schlotter, 2011;Fazekas, 2020;Olesen et al., 2013;Riggio, 2020). Autism-specific social characteristics should receive special scrutiny because, in particular, the psychosocial outcomes of young adults with ASD lag behind compared to persons with other mental disorders or disabilities and comparable demographics (Chen et al., 2015;Howlin, 2013;Howlin & Moss, 2012;Roux et al., 2013;Shattuck et al., 2012).
To gain a better understanding of specific needs of late-diagnosed individuals with ASD, the current study includes a clinical comparison group of individuals referred for diagnostic assessment to our autism outpatient clinic, for whom a diagnosis of ASD was ruled out (see Bloch et al., 2021;Falter-Wagner et al., 2022). This group of persons, who share many features with persons with ASD, represents a particularly informative comparison group, as a comparison allows differentiation of autism-specific needs for the working environment over and above needs across F-diagnoses with social interaction challenges. Indeed, our results show that comparing these two groups of patients with social interaction challenges, either with or without ASD, there were very specific needs reported by study participants with ASD. Only a thorough understanding of these specificities can lead to alignment of therapeutic support services and development of specific employment services for individuals with ASD different from those for individuals with social interaction difficulties without ASD. The current study sought to approach an autism-specific profile in working life. The (i) first aim was to present the educational, vocational, and employment situation of individuals with ASD in comparison with those of a close clinical comparison group in order to identify indications of specific differences of persons with ASD. To investigate this research question, the first part of a vocational questionnaire (Proft et al., 2016; see section Instrument) was used referred to as 'professional development'. The (ii) second aim was to identify specific characteristics regarding workplace experiences and needs of the ASD group by investigating differences between these two groups. Here, the second part of the vocational questionnaire 'specific workplace characteristics' was used.

METHOD Participants
Data were obtained retrospectively from the clinical database of the Adult Autism Outpatient Clinic, Department of Psychiatry, of the University Hospital of Cologne, Germany, partially overlapping with the sample of a previous study that focused on unemployment rates (see Espelöer et al., 2022). 3520 individuals were referred for diagnostic assessment during the period of 2014 to 2021 and received a vocational questionnaire (Proft et al., 2016) before their first interview to assess their employment situation. 1877 questionnaires were returned corresponding to a response rate of 53.3%. 852 of 1877 individuals completed the diagnostic process. Data of persons presenting with questionnaires with more than five missing values were excluded from the current analysis. Further, we excluded data of individuals without completed school education, mentally disabled persons, and individuals diagnosed with other pervasive developmental disorders (F84.8, F84.9). In total, 698 individuals were included in the current analysis. The sample consists of two groups, 197 individuals were clinically diagnosed with ASD according to ICD-10 (WHO, 1992) criteria (ASD+ group), whereas in 501 individuals a diagnosis of ASD was ruled out (ASD-group). In the ASD group, 85.3% of individuals were diagnosed with F84.5 (n = 168), 9.6% with F84.0 (n = 19), and 5.1% with F84.1 (n = 10).
Individuals attended the outpatient clinic for diagnostic clarification, usually based on suspected ASD due to social emotional symptoms. Registration for diagnosis in the outpatient clinic is subject to referral and thus based on a prior assessment by a psychiatrist or clinical psychologist. All individuals included in the current study showed Autism-Spectrum Quotient (AQ, Baron-Cohen et al., 2001) values above the clinical cut-off of 32. Diagnostic procedures complied with the German guidelines on ASD (Association of the Scientific Medical Societies, 2016) and were based on consensus diagnoses, which involves the independent assessment of at least two experienced clinicians. The exclusive aim is to verify or reject the diagnosis of ASD, possible differential diagnoses cannot be systematically investigated in detail because of time limitations. Individuals in both groups reported social difficulties, individuals in the ASD-group, however, did not fulfill criteria for any diagnosis within the classification of pervasive developmental disorders (ICD-10 F84). IQ testing is not standard in the diagnostic process. Due to a high educational qualification level in our sample (see Table 2), we assume that cognitive disabilities did not influence our results (Ritchies et al. 2018).

Instrument
The vocational questionnaire (Proft et al., 2016) was completed before starting the diagnostic procedure and before communicating diagnostic results. The questionnaire was designed based on statements regarding personal workplace experiences and wishes for an ideal workplace of individuals with ASD in a prior qualitative study (Proft et al., 2016). The vocational questionnaire (Proft et al., 2016) consists of the two parts 'professional development' and 'specific workplace characteristics'. The second part again comprises the two categories 'Wishes and requirements for an ideal workplace' (W) and 'Workplace experiences' (E).

'Professional development'
The first part is referred to as 'professional development' and captures descriptive data about formal education level, occupational skill level, employment status, psychosocial situation, periods of unemployment, and frequency and reasons of terminations (The German translation of levels of formal qualification attached as Supplementary material and Espelöer et al., 2022).

'Specific workplace characteristics'
The second part 'specific workplace characteristics is composed of the two categories 'Wishes and requirements for an ideal workplace' (W) and 'Workplace experiences' (E). Items were reversely coded where required prior to calculations.
'Wishes and requirements for an ideal workplace' (W). Ratings were performed using a 5-point Likert scale ("very important", "important", "moderately important", slightly important", and "unimportant"). This category of the questionnaire originally consists of 44 items (e.g. "W_02 Few people in the working environment"; "W_35 Retreat possibilities during breaks and/or when overstrained in the daily work routine"). Initially, the questionnaire was designed to examine individuals with ASD. Two items that required an existing ASD diagnosis were excluded from the analyses ("Informing people in the working environment about the autistic condition"; "Increasing awareness of autistic disorders"). Forty-two items were included in the present analyses.
'Workplace experiences' (E). Ratings were performed using a 5-point Likert scale ("strongly agree", "agree", "undecided", "disagree", and "strongly disagree"). The category of the questionnaire originally consists of 39 items (e.g. "E_04 I do not have common topics of conversation with colleagues"; "E_13 Shifting appointments and changes in the daily schedule at short notice (e.g. due to unforeseen meetings) are problematic for me"). Initially, the questionnaire was designed to examine individuals with ASD. Seven items that required an existing ASD diagnosis were excluded from the analyses (e.g. "My employer mainly tries to meet my autistic needs through concrete interventions (e.g. individual office)"). Thirty-two items were included in the present analyses.

Basic statistics
'Professional development' For group comparisons, Chi square tests of association and Mann Whitney U tests were used.

'Specific workplace characteristics'
The distribution of the dataset was checked using Little's Missing Completely at Random Test (Little, 1988). For both parts of the questionnaire ('Wishes and requirements for an ideal workplace' (W) and 'Workplace experiences' (E)), we expected that values were missing by chance. Missing values were replaced by the procedure of multiple imputation if not more than five values were missing. Data did not meet the assumption of normality and thus non-parametric tests were used.
An exploratory factor analysis was performed in order to reduce the complexity of the data. The exploratory approach has to be taken into account when describing and interpreting results (Bender & Lange, 2001). The required assumptions for factor analysis were met: The Kaiser-Meyer-Olkin coefficient (Kaiser & Rice, 1974) showed high values greater than 0.8, respectively (E: KMO = 0.831, W: KMO = 0.883), all MSA-coefficients showed values greater 0.5, and results of the Bartlett's tests of sphericity (Bartlett, 1951) were significant (E: Χ 2 (496) = 6997, p < 0.001, W: Χ 2 (861) = 8914, p < 0.001). After generating factors, post hoc comparisons were performed using reliability analyses and non-parametric Mann-Whitney U tests. To account for influences of age and school education on the results we employed linear regressions.
The number of factors to be extracted was determined using Horn's parallel analysis. Here, the eigenvalue progression and the scree plot were examined (see Supplementary material). In addition to the statistical analysis, the practical content-related importance of the respective items were taken into account as well. Items with a main factor loading of at least 0.4 (Gaskin & Happell, 2014)   were included in the analysis. An item-rest correlation of ≥0.3 is suggested for an item to be included in a factor.

RESULTS 1
'Professional development' Education. Group comparisons show that individuals in the ASD-group were more highly educated compared to individuals in the ASD+ group (U = 42,914, p = 0.009).

'Specific workplace characteristics'
First, factor analyses of the categories 'Workplace experiences' and 'Wishes and requirements for an ideal workplace' were performed. In the following, group comparisons were presented for both categories, respectively (see section Group comparisons of the categories 'Workplace experiences' and 'Wishes and requirements for an ideal workplace').
'Workplace experiences'. Taking into account the result of the factor extraction by Horn's parallel analysis, the eigenvalue analysis, the scree plot, the content interpretation as well as the assumption of a minimum loading of 0.4 and a minimum number of three items per factor, a three-factor structure appeared to be most appropriate. A reduction from 32 to 20 items could be achieved by eliminating items with low main loadings. No relevant cross-loadings (>0.3) were obtained, so that the following three factors with a simple structure pattern were generated (see Supplementary material). E_Factor 1 'Social challenges' comprises difficulties in interpersonal interaction and communication, in understanding social rules, with teamwork, customer contact, flexibility, or lack of structure. E_Factor 2 'Job fit' comprises sufficiency of salary as well as difficulties in finding a suitable job. E_Factor 3 'Specific needs' refers to options to bring in one's own specific, individual interests, strengths, and requirements at the workplace.
Item E_03 was included in E_Factor 3 'Specific needs' despite a low item-rest correlation of 0.178. However, the item shows an acceptable factor loading of 0.406 and contributes content-related to the factor. The internal consistency of factors 1 and 2 were in the acceptable range of >0.7. Factor 3 did not meet sufficient but acceptable internal consistency with Cronbach's α of 0.576. The average inter-item correlation was in the acceptable range of 0.2 to 0.4 for factor 1 and factor 3. An increased value for factor 2 (r = 0.45) indicates a homogeneous factor containing items that measure the same characteristic (Table 3). For a factor with a low item count of three items, it is reasonable to tolerate mean inter-item correlations >0.4 in order to measure a specific domain (Bühner, 2011).
'Wishes and requirements for an ideal workplace'. When including the factor extraction by Horn's parallel analysis, the analysis of the eigenvalues, the scree plot, content related interpretations, the assumption of a minimum loading of 0.4 as well as a minimum number of three items per factor, a four-factor structure is suggested containing factors with a clear content based differentiation. A reduction from 42 to 25 items could be achieved by eliminating items with low main loadings so that the following four factors with a simple structure pattern were generated (see Supplementary material). W_Factor 1 'Social challenges' comprises specific requirements for social interaction and communication in the workplace such as reduced, specific, professional personal contact with colleagues, supervisors, and customers as well as structured daily schedules. W_Factor 2 'Job fit' comprises the need for sufficient salary as well as permanent employment. W_Factor 3 'Specific needs' comprises the need for individual support, strategies for dealing with excessive demands, and recognition of individual abilities. W_Factor 4 'Individual work setting' comprises the need for specific working conditions such as home office, individual office, or flexible work scheduling.
Only one relevant cross-loading (>0.3) was obtained for item W_27 on factor 1 (0.438) and factor 4 (0.502). With regard to content-related and statistical fit, the item was assigned to factor 4. Item W_08 was eliminated despite a factor loading >0.4 (0.416) due to a low itemrest correlation of <0.3 (0.249) as well as an increased internal consistency of the factor when the item is dropped (Cronbach's α = 0.598 vs. Cronbach's α 'if item dropped' = 0.646). Item W_20 was included in factor 4 despite a low item-rest correlation of 0.274. However, the item shows an acceptable factor loading of 0.448 and contributes content-related to the factor (see Supplementary material). The internal consistency of factor 1 and 3 were in the acceptable range of >0.7. Factor 2 with Cronbach's α = 0.646 and factor 4 with Cronbach's α = 0.557 did not meet sufficient internal consistency probably due to the small number of items. Both factors include only three items, which, however, show acceptable factor loadings >0.4 as well as content-related fit.
The homogeneity of all factors was in the acceptable range of 0.2 to 0.4 (Table 4).
Group comparisons of the categories 'Workplace experiences' and 'Wishes and requirements for an ideal workplace' 'Workplace experiences'. Although individuals in the ASD+ group indicated worse experiences related to social requirements (E_Factor 1 'Social challenges') than individuals in the ASD-group this difference was relevant only at a trend level: U = 44,734, p = 0.054, r = 0.094 (Table 5). Single item comparisons indicated greater impairment for individuals with ASD referred to item E_13 (Shifting appointments and changes in the daily schedule at short notice (e.g., due to unforeseen meetings) are problematic for me, U = 44,089, p = 0.021, r = 0.107), E_04 (I do not have common topics of conversation with colleagues, U = 43,927, p = 0.018, r = 0.110), and E_11 (I have problems with contacts to customers, U = 40,389, p < 0.001, r = 0.182) ( Table 6). The ASD+   (Table 7). Including age and school education as covariates, an influence on the group effect was revealed (R 2 = 0.0728, F(3,688) = 18.0, p < 0.001). No interaction effect was found. A marginal group difference remains (B = 0.159, t = 1.71, p = 0.087), but age (B = 0.017, t = 4.59, p < 0.001) and school education (B = À0.164, t = À4.30, p < 0.001) needed to be included in the interpretation of the group effect. Results did not show group differences for E_Factor 3 'Specific needs' (Table 5). 'Wishes and requirements for an ideal workplace'. Individuals with ASD show significantly higher specific requirements for social interaction and communication on the factor level (W_Factor 1), attributing higher importance to the reduction of communication and interaction at work than individuals in the ASD-group, U = 42,651, p = 0.005, r = 0.136 (Table 8). The ASD+ sample attributes significantly higher importance to the following requirements than individuals in the ASDgroup: W_15: Personal contact with colleagues only for a short time window during the day for factual communication, U = 43,408, p = 0.011, r = 0.120, W_02: Few people in the working environment, U = 41,450, p < 0.001, r = 160, W_12: No contact with customers, U = 41,073, p < 0.001, r = 0.168, W_34, No business trips and/or field assignments, U = 41,547, p < 0.001, r = 0.158, and W_10, As few contacts as possible in the company, U = 43,949, p = 0.018, r = 0.109 (Table 9).
Results did not show group differences for factor W_Factor 2 'Job fit', W_Factor 3 'Specific needs', and W_Factor 4 'Individual work setting' (Table 8). Only one single item comparison shows differences. Individuals with ASD indicated a greater importance for the following item: W_35 (item of W_Factor 3), Retreat possibilities during breaks and/or when overstrained in the daily work routine, U = 45,245, p = 0.049, r = 0.083 (Table 10).

DISCUSSION
Aim of the current study was to study both experiences on the one hand and requirements on the other hand concerning working life and employment status that are specific for individuals diagnosed with autism (ASD+) in comparison to individuals suffering from social difficulties to whom a diagnosis of ASD does not apply (ASD-). Presumably because of the shared social difficulties motivating referral for diagnostic evaluation of the whole population examined, we found in both groups comparable 'professional developments', including employment status, durations of unemployment, frequency of terminations including reasons thereof and vocational qualifications. Answering the first research aim, as an exception  of these many similarities, supported employment measures were utilized significantly more frequently by individuals with ASD. Specific differences were hypothesized for the increased requirement for reduced interpersonal interactions and communication at work and the tendency toward greater difficulties in dealing with social demands in the ASD group, representing the second aim related to 'specific workplace characteristics'. Greater difficulty in finding a suitable workplace with sufficient salary was found in persons with ASD, especially among younger individuals and persons with lower level of education.

'Specific workplace characteristics'
'Workplace experiences'. Difficulties in dealing with social requirements became evident in both groups on the factor level, with a tendency toward greater impairments for individuals with ASD (E_Factor 1 'Social challenges'). However, evidence indicated that challenges with flexible scheduling, customer contact, and identifying shared interests and conversation topics with colleagues were specifically pronounced among individuals with ASD. Here, the so-called hygiene factors become apparent (Herzberg, 1972) as one key component of his two-factor motivation theory. The theory includes two independent, non-complementary factors and states that so-called motivators (e.g., achievement, recognition, responsibility) might increase job satisfaction, while dissatisfaction might be increased by a lack of so-called hygiene factors (e.g., salary, job security, supervision, relationships at work). Hence, autism-specific impairments in social skills as well as inflexible routines and ritualized behaviors significantly affect the successful integration and maintenance of employment and according to Herzberg (1972) as a lack of hygiene factors might increase dissatisfaction in working life. Reports of workplace experiences indicate that individuals with ASD were more likely struggling to live on their wages than individuals with a ruled out ASD diagnosis. People in the ASD group also reported greater difficulties in finding suitable employment (E_Factor 2 'Job fit'). It turned out that age and educational qualifications were found to have an impact on salary and job fit with greater difficulties at younger ages and lower educational levels. It could be assumed that the level of school-leaving qualifications and thus the advanced age could affect the amount of salary and thus job security (Whittenburg et al. 2019). However, we find a lack of sufficient salary and job security as a dissatisfaction of hygiene factors might increase dissatisfaction in working life.
'Wishes and requirements for an ideal workplace'. Autism-specific qualitative impairments in processing social information intuitively as well as reduced competences in flexible planning and prioritization seem to require specific working conditions. In accordance with qualitative studies of Müller et al. (2003) as well as  T A B L E 9 Single item comparisons for W_Factor 1 'Social challenges' of the questionnaire 'Wishes and requirements for an ideal workplace'  Hurlbutt and Chalmers (2004), results of the current study suggested that the wish to reduce interpersonal demands for communication and interaction in the workplace represents a specific characteristic for individuals with ASD (W_Factor 1 'Social challenges'). More precisely, people with ASD, in contrast to non-autistic persons, attribute significantly greater importance on interactions that are brief and focus purely on content and on interactions with only a limited number of colleagues or preferably, if possible, only one specific contact person as well as not having customer contact. Again, it might be assumed that job dissatisfaction becomes apparent by the lack of hygiene factors.
The wish of individuals in the ASD group for reduction of social and sensory stimulation becomes evident in a variety of areas. The ASD group appears to place a higher importance on not traveling or doing fieldwork as part of their job duties than individuals with a ruled out ASD diagnosis (W_Factor 1 'Social challenges'). Overloads due to sensory overstimulation were frequently reported in individuals with ASD and have now been included as a diagnostic criterion for ASD in DSM-5 (American Psychiatric Association, 2013). This becomes also evident in the current study. The need for a retreat during breaks and in situations of excessive demands is suggested to be specific for people with ASD compared to people without ASD (W_Factor 3 'Specific needs'). The recognition of individual needs at the workplace, the importance of financial security, and the possibility of flexibly adapting framework conditions (home office, individual office, flexible work scheduling; comparable to hygiene factors) to one's own needs were similarly pronounced in both groups (W_Factor 2 'Job fit'). Both groups attribute similar importance to the general conditions of a secure job and a salary that is appropriate to the professional activity and ensures economic security. However, confirming previous research (Roux et al., 2013), individuals with ASD show greater difficulties in obtaining adequate salary and a suitable workplace, which was even aggravated with younger age and lower educational qualification.
It turns out that, in both groups, there is a marked discrepancy between individual wishes and requirements at work and experienced workplace conditions. Comparable to previous results (Hurlbutt & Chalmers, 2004;Müller et al., 2003), autism-specific needs and impairments in social competences and flexible behavior become apparent. Difficulties in dealing with customers and colleagues, the need for professional rather than personal exchange, and for minimizing continuously interactions with superiors and colleagues were found to be more pronounced in both factors ('Workplace experiences' and 'Wishes and requirements for an ideal workplace') in individuals with ASD than in individuals without ASD. Consistent with a qualitative study by Müller et al. (2003), both factors indicate that the opportunity to work alone and autonomously accommodates the need for stimulus reduction and retreat.
A more recent, empirically supported theoretical concept extending the two-factor motivation theory by Herzberg (1972) is the job demands-resources model (Demerouti et al., 2001;Demerouti & Bakker, 2011 ;Schaufeli, 2017), which differentiates the two components of job demands and job resources in order to understand the emergence of health impairments such as stress, and motivational processes such as work engagement, and the effect on job performance. Job demands refer to physical, psychological, social as well as organizational aspects of work (e.g., workload, conflicts with colleagues, financial insecurity) and may affect the employee's stress level, which in combination with reduced job resources may have a negative impact on mental health and job performance. Job resources (e.g., support from others, team climate, feedback, financial security, job control) have motivational qualities by enhancing job engagement and may protect against stress. It becomes apparent that the reduction of job demands (pronounced difficulties with social demands, need for reduced social and sensory stimulation, lack of financial security and job fit) does not automatically increase job engagement (Demerouti & Bakker, 2011;Schaufeli, 2017). Following the job demands-resources model, the current results suggest autism-specific employment support structures with attention to individual resources. Awareness of ASDspecific requirements at the workplace seems to be crucial. Employers may also benefit in this way if job performance can be increased by paying attention to the respective job resources and reducing job demands.

'Professional development'
Vocational qualification. What differentiates between groups is that individuals in the ASD group are supported significantly more often by supported employment programs suggesting a higher need of support services (Shattuck et al., 2012;Vogeley et al., 2013). Results might also indicate difficulties for individuals with ASD to successfully transit from school into vocational qualification (Müller et al. 2003). It can be assumed that the more structured environment of education, where social difficulties are met with higher acceptance and tolerance compared to the context of employment could be beneficial for people with ASD (Frank et al., 2018;Maslahati et al., 2022;Müller et al., 2008). Socially insecure or withdrawn behavior may lead to irritation in the workplace (Vogeley et al., 2013). Demands for social skills, adaptability, and flexible behavior seem to be increasing in working life. The likelihood of finding employment decreases due to impaired conversational skills (Roux et al., 2013). However, unfortunately there is a distinct lack of appropriate support services especially for individuals with ASD without intellectual disabilities (Shattuck et al., 2012) resulting in the need for ongoing parental engagement (Baldwin et al., 2014).
In contrast to the differences described above, both groups were comparable in terms of their employment status and vocational qualifications. Individuals in both groups are comparatively often in process of graduation or without completed vocational qualification at the time of data collection. Individuals with ASD show decreased employment rates and, respectively, increased unemployment rates compared to individuals without ASD, however, results did not differ significantly.
Gap between educational level and employment status. Unemployment rates of 24.5% in the ASD group and 19.9% in the group of individuals without ASD are alarmingly high in both groups compared to those of the general German population (5.2%, Federal Employment Agency, 2022; see Espelöer et al., 2022 for a full discussion). Increased unemployment rates are frequently described in international as well as national literature, with rates ranging from 24% to 73% (Baldwin et al., 2014;Howlin, 2013;Howlin & Moss, 2012;Taylor & Seltzer, 2011;Vogeley et al., 2013) even in above-average educated individuals with ASD (Espelöer et al., 2022;Frank et al., 2018;Maslahati et al., 2022;Riedel et al., 2016). Although over half of the ASD group achieved a university entrance-level qualification (53.3%), this number was even higher in the comparison group of individuals without ASD (65.5%). Complementary to these results, higher rates of basic secondary education were achieved by the ASD group (18.8%) compared to the group of individuals without ASD (10.1%). Compared to the general German population with university entrance-level qualifications of 32.5%, an overqualification in individuals with ASD without intellectual disability becomes evident (see Espelöer et al., 2022 for a full discussion). However, an increased need for supported employment among individuals with ASD is reported. The risk of unemployment persists in both groups despite high levels of educational and vocational qualifications (Baldwin et al., 2014;Frank et al., 2018;Riedel et al., 2016). This discrepancy could be associated with Herzberg's (1972) intrinsic motivators, which might lead to decreased job satisfaction due to reduced recognition, responsibility, and development opportunities.
Termination and periods of unemployment. Difficulties in successfully participating in the working life are also reflected in repeated terminations (Espelöer et al., 2022;Frank et al., 2018). Limited abilities in soft skills and inflexible and rigid adherence to ritualistic and stereotyped behaviors could arguably hinder successful job retention. The majority of individuals in both groups had already experienced unemployment (ASD+: 68.5%, ASD-: 70.4%), with nearly half of individuals in the ASD group (47.5%) and 43.6% in the group of individuals with a ruled out ASD diagnosis experiencing long-term unemployment. In the general German population, 39.3% of unemployed persons were affected by long-term unemployment (Federal Employment Agency, 2022). Comparable to previous research (Frank et al., 2018), average periods of unemployment of 24.9 months in both groups were reported in the current study, too. Likewise, both groups similarly reported to have experienced job termination most common due to interpersonal difficulties rather than professional difficulties.

Strengths and limitations
We present data of a new and comprehensive questionnaire based on statements of individuals with ASD, which were collected in a prior qualitative study (Proft et al., 2016). Limiting the study, our convenience samples happen to differ in age and educational qualification given the posthoc database analysis. We did not adjust groups with respect to age and educational qualifications, but included both factors as covariates to avoid an artificial distortion of the population that visited the autism outpatient clinic in Cologne. This is also recommended for future studies. In the present study, we focused on persons with late-diagnosed ASD without intellectual disability. It is important to emphasize that the results cannot be generalized to all people within the autism spectrum. A large sample size as well as a sufficient response rate of 53.3% were achieved based on data from people who were motivated and able to fully complete the questionnaire.
We did not ask about the level of income and therefore could not include this point in the discussion of the results that people with ASD report greater difficulties in being able to make a living on their wages than individuals without ASD. This point should be included in future studies in order to be able to make precise objective statements. Differential diagnoses and comorbid disorders are common in adults with ASD (Strunz et al. 2014;Chen et al., 2015;Hudson et al. 2019) and have been discussed as relevant factors in research on the employment situation and level of functioning in individuals with ASD (Chen et al., 2015;Croen et al., 2015;Fombonne et al., 2021;Riedel et al., 2016). In the current study, self-reported diagnoses during lifetime provided by patients themselves were collected (see table in Supplementary material). A detailed clinical examination of differential diagnoses or comorbidities was not performed in the present study because the aim of the autism outpatient clinic is exclusively to confirm or reject the diagnosis of ASD. Level of functioning of individuals with ASD was not explicitly recorded but a relatively high level of functioning could be assumed on the basis of high educational qualification levels and late diagnoses of mainly F84.5 (85.3%). A more specific characterization of samples should be investigated as an important further aspect in future prospective studies.

CONCLUSION
The results of the current analyses described an autismspecific profile of a mismatch between social skills and social requirements in the workplace both with respect to past experiences as well as future requirements. Our results show that there are very high unemployment rates and sometimes unsurmountable challenges by the workplace posed on individuals with social interaction difficulties. There is no doubt that there is a marked need for improved support structures across these individuals. The current results suggest though that support structures should not rely on one-fits-all approaches. Instead, the results demonstrate an autism-specific profile characterized by pronounced need for reduction of social and interpersonal demands at work, for structured working conditions as well as the autism-specific difficulties with finding a suitable job ensuring economic security. Future studies should consider the effects of age and educational qualification on employment outcomes. As aspects could be described based on the exploratory approach that may influence successful integration into working life for people with interactional difficulties and specifically with ASD, these characteristics should be further investigated and corroborated in future prospective studies. Autism-specific employment support is urgently needed and awareness and knowledge about the specific requirements of individuals with ASD without intellectual disability at the workplace is potentially very helpful. Employers in turn can potentially benefit from the diverse skills set people with ASD can bring to the workplace if challenges for them are minimized.

AUTHOR CONTRIBUTIONS
All authors contributed to the study conception and design. Material preparation and analysis were performed by JE. The first draft of the manuscript was written by JE and all authors commented on previous versions of the manuscript. All authors listed have made a substantial, direct, and intellectual contribution to the work and approved the final manuscript for publication. All authors had full access to data and CF verified the data.

ACKNOWLEDGMENTS
We thank all the participants who took part in this study. Open Access funding enabled and organized by Projekt DEAL.

FUNDING INFORMATION
The study was funded by a grant of the "Landschaftsverband Rheinland (LVR)" for a project that developed a supported employment program for adults with ASD in Germany. CF was supported by the DFG (Deutsche Forschungsgemeinschaft) [grant numbers FA 876/3-1, FA 876/5-1]. KV received funding from the Federal German Ministry of Education and Research (grant number: 16SV7244) and from the EC, Horizon 2020 Framework Program, FET Proactive (grant agreement id: 824128).

CONFLICT OF INTEREST STATEMENT
On behalf of all authors, the corresponding author states that there is no conflict of interest.
DATA AVAILABILITY STATEMENT Data availability The datasets generated and analyzed in the current study are not publicly available as they are retrieved from a clinical database of the Adult Autism Outpatient Clinic, Department of Psychiatry, of the University Hospital of Cologne, Germany. The authors will provide data and materials supporting the results or analyses presented in the current paper upon reasonable request.

ETHICS STATEMENT
The manuscript was submitted to the local Ethics Committee of the Medical Faculty, University of Cologne, which confirmed that the study is exempt from the requirement of ethical approval as under German law no separate ethics application to and statement of ethical approval by the local ethics committee are required for performing purely retrospective clinical database analyses.