Functional Outcome in Late Adolescence/ Early Adulthood of Patients With Autism Spectrum Disorder and Its Relationships With Parental Burnout and Depression: Multi-center, Cross-sectional Study

The aim of this study is to determine functioning of adults with Autism Spectrum Disorders (ASDs) diagnosed in childhood and also depression and burnout levels among their parents. 261 adults with ASDs and their parents were recruited for the study. Both parents completed the Beck Depression and Maslach Burnout Inventories and reported the functioning of their adult offspring with ASDs. Only 5.4 % of our sample reported “good” or “very good” outcome. The most common psychiatric comorbidities were intellectual disabilities and attention-decit/ hyperactivity disorder. Maternal burnout and depression scores were signicantly elevated compared to those of fathers. There is an undeniable urgent need for more research to identify the needs of adults and families suffering from ASD. This multi-center, cross-sectional study aimed to evaluate functional outcomes among adults with ASD, parental depressive symptoms and burnout levels. Majority of adults with ASD in our sample had “poor” or “very poor” outcome according Howlin while only 5.4% were classied or good” 38 . The diagnoses of included patients at the time of the study were ascertained through clinical interviews by psychiatrists following them at the time as per DSM-5 1 . Informed consents were obtained from parents of early adults/ late adolescents with ASD as well as verbal individuals with ASD to be included in the study. Inclusion criteria for ASD group were a diagnosis of simplex ASD without known genetic syndromes (i.e. Fragile X syndrome, Down Syndrome, Rett syndrome or variants, etc.) at childhood (i.e. ≤ 12 years old). The patients in this study involved those who were diagnosed with “infantile autism” or “atypical pervasive developmental disorder” as per DSM-III, “autistic disorder” and “pervasive developmental disorder- not otherwise specied” (PDD-NOS) as per DSM-III-R and “autistic disorder”, “Asperger’s disorder” and PDD-NOS as per DSM-IV and IV-TR criteria. To this end, patients with ICD-10 codes F84.0 (“Childhood Autism”), F84.1 (“Atypical Autism”), F84.5 (“Asperger’s Syndrome”) and ICD-9 codes 299.0 (“infantile autism”) and 299.8 (“other”/ “atypical child psychosis”) at the study centers were screened for. From 1057 families, 429 could not be contacted, 297 refused participation in the study and 69 withdrew their consent once enrolled. One patient was found to be diagnosed at 17 years of age and his data were removed from the analysis leaving 261 patients and their parents. Patients with epilepsy and psychiatric or other neurodevelopmental comorbidities were included, and the effects of those confounders were controlled for in analyses.


Introduction
Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder that presents with communication and social impairments, limited interests and repetitive actions, and problems with sensory integration 1 . Although genetic factors are important, the symptoms are thought to arise from a complex interplay of genetic, epigenetic, and environmental factors 2,3 . Latest estimates of prevalence suggest that ASD may affect one out of every forty children in the USA 4 with a male-to-female ratio of 2.5:1, while the average worldwide frequency ranges from 1.8-7.2 per ten thousand 5,6 .
Based on frequency of children with ASD, it is expected that approximately 700,000 adults aged 65 and older in the next 20 years will have ASD in the US. Similar percentages will be expected all over the world, which is not ready based on the needs of the disorder 7 .
The transition from adolescence to adulthood is also di cult for neurotypical youth and their family members.
However, for individuals with ASD and their families this transition may be more complicated and stressful 8, 9 . As children age; parents' concerns focus on the future and issues of guardianship, education, day services/employment, independent living, transition planning, health and safety, social outlets and nancial security come to the forefront [10][11][12] . Blacher, Kraemer and Howell 13 suggested that parents of young adults with ASD were more concerned about the transitional processes compared to others whose children had Down syndrome or Cerebral Palsy. Also, mothers with lower education, male children, and older children may be signi cantly more anxious about the future 14 . Speci cally, parents' concerns on independent living of their children with ASD and their future well-being increase as they grow into adulthood 12,15 affecting their mental health. Studies addressing parents of adults with ASD found that behavior problems, psychiatric morbidity, gender of the child with diagnosis, poor physical health, and social impairments were associated with reduced quality of life, problems in intra-familial relationships, and elevated symptoms of depression and anxiety among parents as well as children [16][17][18][19] . A recent study by Hoyle and colleagues 20 found that mothers were under elevated risk of psychopathology even when their adult child with developmental disability had achieved independent living.
Studies on adults with ASD from Turkey are limited, although interest in their functional outcomes and comorbidities are increasing 21 . In the rst study, 22,23 , 21 patients with ASDs with an average age of 20.9 years (range = 16. 0-36.0) and with IQ ≥ 50 were evaluated. In this study; 20.0% of the patients were reported to have symptomatic improvement during puberty, while rate of comorbid psychotic disorders was 4.8%. Gormez 24 evaluated parenting burden among 101 Turkish adults with intellectual disabilities (IDs); of which 31.7 percent also had ASDs. In the study, elevated caregiving burden was associated with ASD, behavioral problems, and caregivers' higher education level. Gormez and colleagues 25 also evaluated a large group (n = 771) of adults with IDs and reported that 50.1% had psychiatric symptoms, while rate of challenging behaviors was 36.4%. Challenging behaviors were predicted by the presence of with operationally de ned criteria and all were limited to single centers.
Specialized Child and Adolescent Psychiatry Departments in Turkey have been in operation since 1990 26 , 27 and children initially diagnosed with ASD have been getting older. However, no study until now has concentrated on how Turkish late adolescents/ early adults diagnosed with ASD in their childhood function and the problems experienced by their parents. Therefore, the goals of this study were; a. to evaluate the sociodemographic features and functional outcomes of Turkish late adolescents/ early adults with ASD diagnosed in childhood, b. to determine predictors of favourable functional outcomes, c. to assess self-reported burnout and depression levels among their parents and to determine factors associated with clinicially signi cant depression levels among parents.

Sociodemographic characteristics of adults with ASDs and their families
Two hundred and sixty-one adults with ASDs (86.2% male) with a mean age of 21.  Table 1.  psychiatric comorbidity was present.
The majority of adults with ASD (n = 189, 72.7%) were receiving psychopharmacological treatment while enrolled in the study. Twenty-eight (14.8%) continued the medication prescribed by their child/adolescent psychiatrist without being followed up by any psychiatrist.

Language abilities, social skills and legal problems of adults with ASDs
At the study enrollment, almost half of the adults with ASD (n = 112, 42.9%) were able to construct ≥ 3-word sentences while almost one-third (n = 80, 30.7%) was mutistic. Improper behaviors such as touching/masturbation in community were reported by parents of 45 adults (17.2%). Ninety-nine participants (37.9%) harmed another at least once due to rage attacks and nine were previously involved in legal proceedings (3.4%) due to social communication problems and narrowly focused interests.
More than one fourth of adults with ASD (n = 74, 28.4%) could communicate with phone and 21.8% (n = 57) had social media accounts that they managed with some help from their parents. Thirty-seven adults with ASD (14.2%) had friends which they socialized with ≥ 3 hours/week and 3.4% (n= 9) had experienced romantic relationships lasting a few months with a member of the opposite sex. None had stable, long-term relationships and none were married at the time of the study.

Independent living, self-care, transportation and employment of adults with ASDs
Most of the participants with ASD (n = 257, 98.5%) were still living with their parents while four (1.5%) were living in nursing homes. Majority were continent (n = 233, 89.3%) while less than half (n = 120, 46.0%) could carry out self-care by themselves. Fifty-six (21.5%) could go shopping for themselves at nearby malls, while twenty-nine (11.1%) could manage travelling alone in their city by public transport. Less than one tenth (n = 16, 6.1%) could manage inter-city travel by bus/train or airplane, while only three (1.1%) passed the examination for their drivers' license. Less than onetenth (n = 24,9.2%) of adults with ASD worked at a job for at least a month. The median duration of work was 12.0 (IQR = 21.0, Range: 1.0-48.0) months. None of the adults with ASD could partake in the obligatory military service in Turkey. 5. Classi cation and predictors of "good"/ "very good" functional outcome of adults with ASD A composite measure of overall social functioning in our sample was calculated according to Howlin et al. (2004) 38 . As a result, the majority of ASD subjects in our sample was classi ed as having "very poor" (n= 196, 75.1%) or "poor" (n= 31, 11.9%) outcome. Only twenty participants (7.7%) were classi ed as having "fair" outcome. Rates for "good" and "very good" outcome were 2.7% each (n = 7). Logistic regression was used to evaluate predictors of "good" or "very good outcome" in the sample and age at diagnosis, duration of special education, age at speech onset, presence of IDs, comorbid psychopathology, epilepsy, maternal and paternal higher education (high-school or above) were entered as predictors. The overall model was signi cant (AIC = 93.8, BIC = 123.0, χ2 (8) = 12.5, p = 0.13) and may explain 17.3% variability in the outcome (Nagelkerke R2 = 0.173, Table 3). Duration of special education was the only signi cant predictor in the analysis, although offspring of fathers with higher education also tended to have more favourable outcomes. The model could correctly classify all of the participants without "good" or "very good" outcomes and none of those with "good" and "very good" outcomes (Accuracy = 0.93, AUC = 0.80).

Depression levels of parents of adults with ASDs and associated factors
The mean BDI scores of mothers and fathers were 11.7 (SD = 8.9) and 8.8 (SD = 9.

Discussion
This multi-center, cross-sectional study aimed to evaluate functional outcomes among adults with ASD, parental depressive symptoms and burnout levels. Majority of adults with ASD in our sample had "poor" or "very poor" outcome according to Howlin et al (2004) criteria while only 5.4% were classi ed as having "good" or "very good" outcome 38 .
Duration of special education was the only signi cant predictor of favourable outcome while offspring of fathers with higher levels of education also tended to fare better. Mothers were signi cantly more depressed compared to fathers.
More than one fth of mothers and one tenth of fathers reported signi cant levels of depression. Signi cant maternal depression was associated with greater severity of ASD symptoms, lack of spare time, and paternal depression.
Clinically signi cant depression symptoms among fathers were associated with lack of formal education for their children, a history of hospitalization among children and maternal depression. Mothers especially those with offspring who also had IDs reported elevated levels burnout compared to fathers.
Initially described disorders of infancy and often conceptualized as childhood disorders, ASD are now known to be lifelong conditions 38,39,18 . Children, who were diagnosed with autism in the late 1990s and early 2000, began to be adults.
This transition is challenging not only for individuals with ASD but also for their parents 8,40−42 . The majority of available studies on adults with ASD and their parents were conducted among western samples and no study up to now evaluated functioning of adults with ASD and depression and burnout levels of their parents in Turkey.
Caring for individuals with ASD from the time of diagnosis onward is a demanding process and parents must balance parenting roles within their marriage. They also act as co-therapists emphasizing positive behavior, while educating relatives, schools, and teachers on the nature of ASD, advocate for their children's rights and with the transition to adulthood must address problems such as guardianship, higher education, social/ romantic relationships of their children, nancial security, independent living and housing 10 . These lead them to experience elevated levels of stress and burnout.
Mothers of children with ASD are the primary caregivers experiencing elevated stress and this may contribute to the development of maternal psychopathology 36 , 43 . Most of the past studies supported that mothers of children with ASD experience elevated burden and depression in comparison to fathers 44 . Previous studies from Turkey have shown that both parents of children with disabilities experiences elevated stress, depression and burnout, but this was greater in mothers compared to fathers 36, 45 . In a study conducted among parents of Turkish adults with IDs, Gormez 24 reported that elevated caregiving burden was associated with comorbid ASD, behavioral problems and caregivers' higher education levels. Supporting those views self-reported depression and burnout levels in our study were also signi cantly elevated in mothers compared to fathers although we could not nd an effect of parental education. This difference may be due to heterogenous nature of patients in the Gormez study 24 , use of differing instruments (e.g. Zarit Caregiving Burden Interview vs. MBI) or sampling methods. Effects of parental education on burnout and depression may be evaluated with further studies.
Previous studies suggested that mothers with lower education, with male and/ or older children with ASD, and with children having behavior problems as well as psychiatric comorbidity, physical disorders and social impairments may be more depressed and anxious 14,16−19 . The presence of functional speech in children, higher paternal/ maternal education levels, access to specialized schooling for children, greater family income, lack of comorbidities among children, and using medications to manage ASD symptoms may also reduce parenting stress 30,32, 46-48 . Depression and anxiety in one parent, especially for mothers may spill-over to other family members 49 . Previous reports suggest that the rate of clinically signi cant depression in parents raising children with ASD ranged from 12.5 -34.2% [50][51][52][53][54] . A recent meta-analysis reported the depression rate in parents of children with ASD as 31.0% 53 . Supporting those views, we found that more than one fth of mothers and one tenth of fathers in our sample reported signi cant levels of depression. Signi cant depression in mothers was related to incontinence in children, lack of spare time, and depression in their husbands. Mothers of adults with ASD and comorbid IDs also tended to report increased depression. Clinically signi cant depression symptoms in fathers were related to lack of formal education for their children, a history of hospitalization among children, and signi cant depression symptoms among mothers. Mothers described increased emotional fatigue and total burnout compared to fathers. Presence of IDs and epilepsy in adult offspring with ASD also increased maternal burnout while there was a signal that paternal burnout may also be affected by presence of IDs. Those results support and extend the results of previous studies and may suggest that parenting stress, burnout and depression may continue over long term for parents of children with ASD. Depression in parent dyads may also spillover from mothers to fathers or vice versa 32 . Further studies on Turkish samples with ASD and their parents as they age may be required to evaluate the temporal changes in those constructs.
ASD is also associated with elevated rates of comorbid physical disorders, psychopathology as well as early mortality [55][56][57] . Lifetime rates for comorbid psychopathology among adults with ASD may vary between 69.0 -79. In another study, 64.1% of adults were prescribed at least one psychotropic medication, 46.4% took two or more psychotropic medications. Furthermore, twenty-two percent of these polypharmacy samples were not being followed by a psychiatrist 59 . Similar to those studies, more than half of adults with ASD in our sample also had a comorbid psychopathology. The majority were given psychopharmacological treatment, mostly with antipsychotics. However, differing from those studies most of the patients in our sample had comorbid ADHD and rates of mood and anxiety disorders were lower. This discrepancy may re ect the fact that they were being followed by psychiatrists who tend to evaluate symptoms with unstructured clinical interviews rather than using speci c instruments (e.g., PAS-ADD Checklist). In addition, the parents reported that their psychiatrists, most of whom lacked expertise in neurodevelopmental disorders consulted with the parents rather than observing the behavioral changes among their offspring with ASD, especially those without functional speech. The diagnoses of clinicians may also be affected by diagnostic overshadowing 60 . Further studies using structured interviews modi ed for ASD may be required to evaluate comorbid diagnoses among Turkish adults with ASD. Also, almost one-third of our patients was mutistic and with a similar rate of comorbid MR/ID that may limit their ability to express and desribe inner experience. On the other hand, Supporting to these results, in our sample, eighty-four participants were found to have chronic organic diseases requiring treatment. The most common of these diseases were epilepsy, obesity, and autoimmune thyroiditis. These results demonstrate that there is a need to develop improved treatment options and quality of care in this population 41,77 .
After high school, enrollment in tertiary education or nding employment are the most common steps. Those steps into adulthood can be challenging for those with ASD 88,89 . Previous studies reported limited access to tertiary education (28.0%) 88 and frequent problems in adapting to environment in those with access to tertiary education 90 . In Canada, Eaves and Ho 78 analyzed adults with ASD and reported that 45.0% had never been employed. In an older study from the US 91 only 27.0% of adults with ASD displayed some level of work/employment. In another study with adults in Australia, it was found that only a minority were engaged in paid employment 92 . Our results are consistent with those studies. The majority of our sample were still receiving special education at the time when they took part in the study, less than half completed high school and less than two percent could complete tertiary education. Additionally, less than one-tenth of Turkish adults with ASD in our sample could work at a job for at least a month. Our results underline the importance of life-long educational and vocational supports for adults with ASD. Further studies should evaluate predictors of gainful employment among this sample. Counseling services, vocational training and supports and coordination of services across schools, families, and potential employers for young adults and adults with ASD should be established as soon as possible in developing countries, including Turkey [93][94][95][96] .
An additional indicator of successful functioning in adulthood is independent living which includes separation from the family of origin and becoming socially active in the community. However, studies show that the most adults with ASD continue to live with their parents and never leave home 8, 73,78,92,97−103 . Consistent with previous studies, ve adults in our sample resided in nursing homes and the majority lived with their families. Transportation is another major problem for adults with ASD. Mobility provides a certain level of independence for young people to travel, access to health care, and attend group activities 104,105 . Those factors have been linked to enhanced quality of life and mental well-being 106-108 ; unfortunately, many adults with ASD continue to depend on family members for transportation. The situation may be improving in some countries and in a prior survey from the US, the majority of parents of adolescents with ASD without intellectual disability reported that their children were currently driving (29.0%) or planning to drive (34.0%) 109 .
Contrasting those results, approximately one percent of our sample could obtain their driver's licence and less than onetenth could travel between cities by using bus/ plane. This discrepancy may re ect the need to address transportation needs of Turkish adults with ASD. Driving instructors and courses may be informed of neurodevelopmental disorders including ASD and appropriate modi cations to the licensing process may be required. Staff working at intercity bus/ train/ airplane services may also be educated on the needs of adults with ASD as well as the importance of sensory problems in adapting to travel.
Social relations including friendships protect mental health among people, including adults with ASD. However, adults with autism experience signi cant di culties in establishing and maintaining friendships and romantic relationships due to their symptoms 110,111 . Previous studies found that only a small number of adults with ASD could form enduring friendships (romantic/ platonic) 38,97,99,112−116 . Our results are in accordance with those previous reports and we found that parents of just over one-tenth our sample reported their children as having friends and less than ve percent experienced romantic relationships.
Previous studies reported that 50.0 to 60.0% of adults with ASD may be classi ed as "poor" or "very poor" functioning, while rates of "good" and "very good" functioning are more variable (i.e. 3.0-48.0%; 18,38, 39,103,117 ). Rates of favorable outcome may change according to criteria used to de ne functioning, sampling methods, study centers, enrollment of participants with IDs and presence of social supports. Favorable outcome in those studies was associated with greater IQ levels, early functional language, social reciprocity, better initial functioning and lower levels of ASD symptoms 18,38,39,103,117 . The rate of "good"/ "very good" outcome in our sample was at the lower end of the ranges those previously reported. And differing from those studies, we found that the only predictor of favorable outcome in our sample was the duration of special education received. We could not evaluate the effects of IQ, adaptive functioning at baseline and severity of ASD symptoms due to lack of information in participants' records. Further studies may evaluate these variables on the functionality of adults with ASD in Turkey.

Limitations
Our study is not without limitations. First, most of the participants could not be contacted/ enrolled in the study, and their medical records were incomplete preventing us from evaluating the effects of baseline IQ, adaptive functioning, ASD symptom severity and effects of speci c symptoms. Parents of more impaired adults may refuse participation outright, withdraw their consent or may have moved to other addresses introducing selection bias to our results.
Second, we did not evaluate adults with ASD with clinical interviews and depended on the reports of psychiatrists following them for comorbid diagnoses. Further studies may evaluate adults with ASD via specialized instruments (e.g.
PAS-ADD etc.). Third, we evaluated depression and caregiving burden among parents with self-reports and there is a possibility of recall and bias in reporting as well as shared method variance. Structured interviews and scales (e.g. Zarit Caregiving Burden Interview, Hamilton Depression Rating Scale etc.) may provide greater objectivity in evaluating those constructs. Fourth, we depended on parents' own reports for their psychopathologies and medical disorders requiring treatment and those reports may also be subject to bias. Last, rather than using discrete questions tapping social participation, we could have used a psychometric scale to evaluate social competence among adults with ASD (e.g. McConkey and Walsh 118 ). Nevertheless, our study is the rst one to evaluate functioning among a large sample of adults with ASD from Turkey along with depression and burnout levels among their parents. Our data may contribute to the literature on adult outcome of ASD as most of the available data are from Western countries.

Conclusion
The transition to adulthood is very di cult for individuals with ASD and their families. Similar to the status of education services in other countries, many services that were offered as children are limited or no longer available for adults with ASD in Turkey and there are no structured supports easing transition to adult services 10 . Even in those countries with structured supports to ease transition to adult services, young adults with ASD may be socially isolated and lack participation in the community 39,114 . Lack of education among Health Care Workers (HcWs) for ASD in adulthood may present barriers to receiving help. This may lead to embitterment among parents and adults with ASD 8,20,119 . Our results support the urgent need of a structured transition plan for families of adults with ASD as well as those adults for adult mental health services such as those in other countries (e.g., STEPS) 90 .
Apart from mental health services, governmental and non-governmental (NGO) agencies may facilitate those families in using services for their general medical and social needs, education/ vocational training, housing, transportation, and employment. Turkish HcWs in general settings should be trained in recognition and management of medical comorbidities among adults with ASD especially those with limited communication and the importance of sensory problems limiting compliance with procedures. The majority of the mental health professionals serving adults in Turkey lack training in presentation and management of neurodevelopmental disorders, especially ASD among adults. There is only one clinic dedicated to adults with IDs/ ASD with two additional clinics dedicated to adults with ADHD and all those centers are located only in Istanbul. Those services should be introduced to other cities and regions of Turkey with additional coverage for other neurodevelopmental disorders (e.g., learning disorders, tic disorders, etc.) 97,120 .
Psychiatrists and mental health professionals practicing in Turkey and providing care for adults may also be trained to evaluate mutistic patients with ASD as well as those with limited communication skills. Behavioral signs and symptoms of internalizing disorders as well as issues of diagnostic overshadowing may be addressed in those trainings 121 . Special education services for ASD in Turkey are currently focused on academic, social, language and sensory integration skills as well as managing behavior problems and there are limited opportunities to receive vocational training for adolescents/ adults with ASD. Recognizing this gap, the Istanbul Municipality as well as some NGOs (e.g., Tohum Foundation for Autism, Autism Foundation) and isolated companies (e.g., Eker TM Dairy Products) introduced vocational trainings for youth with ASD. A grassroots organization started by Turkish youth with ASD started an online site focused on neurodiversity and reports the problems they face in their daily lives (https://merhabaspektrum.com). However, those attempts lack coordination and need governmental support and planning to achieve nation-wide effects.
As for academic education, the nationwide exams for Turkish high schools and universities may also provide accomodations for youth with ASD. Currently, accomodations are provided only for youth with ADHD and learning disorders. Universities, both private and government a liated may provide accomodations in classes as well as guidance and counselling on managing life at university especially for freshman classes. Senior students with ASD may be assigned as mentors to rst year students and this peer relationship may be bene cial for both parties. Driving licence examinations may be modi ed for youth with ASD and accomodations to intra-and intercity travel methods may be required. Trainings for train, bus and airplane staff may address problems experienced by those youth. Community participation and increased functioning for adults with ASD required multi-faceted interventions addressing various problems. Our study may be a tentative rst step in this direction and the results of our study group should be replicated with a nation-wide survey.

Study Centers
This research was performed in the Child and Adolescent Psychiatry Departments from different geographical regions of Turkey. The diagnoses of included patients at the time of the study were ascertained through clinical interviews by psychiatrists following them at the time as per DSM-5 1 .

Sampling
Informed consents were obtained from parents of early adults/ late adolescents with ASD as well as verbal individuals with ASD to be included in the study. Inclusion criteria for ASD group were a diagnosis of simplex ASD without known genetic syndromes (i.e. Fragile X syndrome, Down Syndrome, Rett syndrome or variants, etc.) at childhood (i.e. ≤ 12 years old). The patients in this study involved those who were diagnosed with "infantile autism" or "atypical pervasive developmental disorder" as per DSM-III, "autistic disorder" and "pervasive developmental disorder-not otherwise speci ed" (PDD-NOS) as per DSM-III-R and "autistic disorder", "Asperger's disorder" and PDD-NOS as per DSM-IV and IV-TR criteria. To this end, patients with ICD-10 codes F84.0 ("Childhood Autism"), F84.1 ("Atypical Autism"), F84.5 ("Asperger's Syndrome") and ICD-9 codes 299.0 ("infantile autism") and 299.8 ("other"/ "atypical child psychosis") at the study centers were screened for. From 1057 families, 429 could not be contacted, 297 refused participation in the study and 69 withdrew their consent once enrolled. One patient was found to be diagnosed at 17 years of age and his data were removed from the analysis leaving 261 patients and their parents. Patients with epilepsy and psychiatric or other neurodevelopmental comorbidities were included, and the effects of those confounders were controlled for in analyses.

Ethics
The study was conducted in accordance with the Declaration of Helsinki and local laws and regulations.

Beck Depression Inventory (BDI)
The BDI was developed by Beck in 1961 to evaluate symptoms of depression 28 , and a reliability and validity study for the inventory in Turkish was conducted by 29 . This scale includes 21 Likert-type items evaluating motivational, cognitive, somatic, social, and vegetative symptoms of depression. Previous studies on parents of children with disabilities including ASD supported its reliability and validity [30][31][32][33] Parents of adults with ASD completed the BDI in the current study. Cronbach alpha for the present sample was found as 0.79. Seventeen was taken as the cut-off score for clinically signi cant depressive symptoms, as suggested in the Turkish reliability and validity study of BDI 29 .

Maslach Burnout Inventory (MBI)
The MBI, a 22-item Likert-type scale, was developed to evaluate self-reported symptoms of burnout by Maslach and Jackson 34 . It consists of three subscales (emotional exhaustion, depersonalization, and personal accomplishment).
MBI was initially translated into Turkish and validated by Ergin 35 . It was later used to evaluate burnout among parents of children with disabilities 36 . In the study by Duygun and Sezgin 36 , factor analysis revealed that a two-factor model was more appropriate for parents and explained 35.6% of the variance. Accordingly, the items from the depersonalization subscale of the MBI were subsumed under emotional exhaustion, with the personal accomplishment items remaining the same 36 . However a recent study suggested a three factor solution was appropriate 37 among parents of children with ASD and we used the three factor solution. Mothers and fathers of adults with ASD completed the MBI in the current study. Cronbach alpha for the present sample was found as 0.74.

Study Procedure
Early adults/ late adolescents (≥ 18 years, henceforth termed "adults" for ease of use) diagnosed with ASD were identi ed by hospital charts. The clinicians completed the sociodemographic form while mothers and fathers of individuals with ASD completed BDI and MBI for themselves.

Statistics
Data were entered into a database prepared with Statistical Program for Social Sciences (SPSS™, IBM Inc., Armonk, NY) Version 22.0. Analyses were conducted with SPSS. Nominal data were summarized as counts and frequencies while quantitative data were summarized as means and standard deviations or medians and inter-quartile ranges (IQR) depending on normality and outliers. Comparisons between groups were conducted with chi-square tests for nominal variables. Yates', Fisher's, and Likelihood Ratio corrections were used as appropriate. Bivariate correlations were evaluated using Spearman correlation analyses. Bivariate comparisons of quantitative variables between independent and dependent groups were conducted with t tests. Predictors of "good"/ "very good" functioning in adulthood were evaluated with logistic regression (enter method). P was set at 0.05 (two-tailed). Bonferroni-Holm correction for controlling family-wise error was used in bivariate analyses of associated factors for parental depression and burnout Retrieved from: https://www.researchgate.net/publication/322568540_HolmBonferroni_sequential_correction_An_Excel_calculator_13).
Effect sizes for signi cant ndings were also reported.