A systematic review of quality of life of adults on the autism spectrum

Abstract

Autism spectrum disorder is associated with co-existing conditions that may adversely affect an individual’s quality of life. No systematic review of quality of life of adults on the autism spectrum has been conducted. Our objectives were as follows: (1) review the evidence about quality of life for adults on the autism spectrum; (2) critically appraise current practice in assessing quality of life of adults on the autism spectrum. We searched bibliographic databases and other literature to identify studies using a direct measure of quality of life of adults on the autism spectrum. Hand searching of reference lists, citation searching and personal communication with field experts were also undertaken. In total, 827 studies were identified; 14 were included. Only one quality of life measure designed for use with the general autism spectrum population was identified. Quality of life of adults on the autism spectrum is lower than that of typically developing adults, when measured with tools designed for the general population. There are no comprehensive autism spectrum disorder–specific quality of life measurement tools validated for use with representative samples of adults on the autism spectrum. There is a pressing need to develop robust measures of quality of life of autistic adults.

Keywords

adults autism spectrum disorders measurement quality of life

Background

Autism spectrum disorder (ASD) is a neurodevelopmental disorder with an impact throughout the lifespan. Children and adults on the autism spectrum have an increased susceptibility to a range of comorbid psychiatric and neurological disorders such as anxiety disorders, depression and epilepsy (Levisohn, 2007; Simonoff et al., 2008; Stahlberg et al., 2004). While approximately 1% of the UK child and adult population have ASD (Baird et al., 2006; Brugha et al., 2011), the majority of ASD research concerns children and adolescents, with little research evidence about the lives of adults on the autism spectrum (Mukaetova-Ladinska et al., 2012) and their quality of life (QoL) (Van Heijst and Geurts, 2015).

The World Health Organization (WHO) defines QoL as ‘an individual’s perception of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards and concerns’ (Harper, 1998). The UK National Institute for Health and Care Excellence (NICE) recognised that ASD may have a significant effect on QoL and stated that improving QoL should be one aim of any psychosocial intervention for people on the autism spectrum (NICE, 2012, 2013). Most current QoL measurement tools were designed for use with people from the general population. In recent years, QoL measures have been designed for people with a range of specific health and mental health conditions in order to capture aspects of life that are in some way ‘different’ from that of the general population due to their condition. For instance, QoL measures often have a focus on social integration and social relationships – this emphasis might be less appropriate for people with a social communication disorder than those from the neurotypical population.

Several studies have investigated the QoL of individuals on the autism spectrum – for example, changes in QoL across the lifespan (Van Heijst and Geurts, 2015) – and have identified factors that influence QoL (Chiang and Wineman, 2014). However, there is currently no systematic review evidence of the QoL of adults on the autism spectrum or the measures used in studies. Our systematic review aims were two-fold: first, to systematically identify and review the evidence about QoL for adults on the autism spectrum, and second, to critically appraise the tools that have been used to assess QoL of adults on the autism spectrum.

Review methods

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for conducting systematic reviews were adhered to throughout the review process (Moher et al., 2009).

Review criteria

Randomised controlled trials (RCTs), observational, qualitative and mixed method studies published in English from any country of origin were eligible for inclusion in the review. Eligible studies were required to include adult participants (aged ⩾18 years) with a clinical diagnosis of ASD (confirmed by an appropriately qualified medical professional either for the purpose of the study or prior to recruitment) and use a direct measure of QoL (proxy, non-proxy and mixed measures of QoL). Studies using indirect measures of QoL, such as employment status, educational level, accommodation and perceived loneliness among others, including measures of psychological well-being, were excluded.

Search strategy

With specific guidance and expertise from an information scientist, and informed by scoping searches, a structured search strategy was developed using a combination of keywords and MeSH (Medical Subject Headings) terms.

The following bibliographic databases were searched: MEDLINE (1966–October 2016), EMBASE (1980–October 2016), PsycINFO (1880–October 2016), Scopus (1960–October 2016) and Web of Science (1864–October 2016) and the Cochrane Library. Key search terms used were autis*, Asperger* and Quality of Life. Citation and hand searching of studies that fulfilled the eligibility criteria were also conducted. Experts in the field were consulted in May 2015 to identify any work in progress or grey literature.

Study selection process

Studies were selected through screening of titles and abstracts according to the inclusion criteria by one reviewer (M.A.). Studies identified by the primary selection process were independently assessed for inclusion using a study selection form. Disagreements were resolved by discussion (occurred twice), or when agreement could not be reached, other co-authors were asked to adjudicate (occurred on three occasions).

Data extraction of included studies was undertaken using a data extraction form. When necessary, study authors were contacted with a request for additional information.

Assessment of methodological quality

Included studies were subjected to independent assessment of methodological quality by two reviewers (L.A. and D.F.). The Cochrane Collaboration tool for RCTs (Higgins and Green, 2009) was used to assess the risk of bias in six domains for the two experimental studies. A 5-item checklist (Flynn et al., 2012) adapted from the Newcastle–Ottawa Quality Assessment Scale for Case-Control Studies (Wells et al., 2014) was used to assess risk of bias in cross-sectional studies across five domains.

The 16-item Quality Assessment Tool for Studies with Diverse Designs (Sirriyeh et al., 2012) was used to assess the methodological quality of the remaining two pre-test/post-test studies. The 14 items relevant to quantitative designs were used, which are scored from 0 to 3 (range of total scores from 0 to 42), with total scores converted into percentages for reporting.

Data synthesis

Due to the expected heterogeneity of study designs and QoL measurement tools, and the small number of participants from individual studies, a narrative approach was used to synthesise the findings of the included studies. In synthesising the data from the methodological review tools, emphasis was placed on the review criteria relevant to the objectives of this current review.

Results

In total, 14 of 827 studies identified by the search strategy fulfilled the review criteria (Figure 1). Two were randomised studies (Garcia-Villamisar, 2010; Hesselmark et al., 2014). Nine studies were cross-sectional designs that assessed QoL of adults on the autism spectrum at a specific point in time (Hong et al., 2016; Jennes-Coussens et al., 2006; Kamio et al., 2013; Kamp-Becker et al., 2010; Katz et al., 2015; Khanna et al., 2014; Lin, 2014; Saldaña et al., 2009; Van Heijst and Geurts, 2015) and three were pre-test/post-test intervention studies (Gal et al., 2015; Garcia-Villamisar, 2010; Hesselmark et al., 2014). Sample sizes ranged from 12 (Jennes-Coussens et al., 2006) to 291 adults (Khanna et al., 2014). The combined sample size from the 14 included studies was 959 adults. Collectively, these studies focussed on individuals aged from 18 to 55 years with one study including participants up to 83 years (Van Heijst and Geurts, 2015). As expected, there was a predominance of male participants.

Figure 1.

PRISMA diagram showing the process used to identify studies to show study search and selection.

Tools used to assess QoL of adults on the autism spectrum

Seven tools were used to assess QoL of adults on the autism spectrum. The psychometric properties of the tools and the populations in which the tools were used are described in the following section.

World Health Organization Quality of Life–BREF

The World Health Organization Quality of Life (WHOQOL)-BREF tool is an abbreviated 26-item version of the original 100-item WHOQOL tool designed for use in subjective self-assessment of QoL across a range of population types. The tool assesses QoL across four domains: Physical Health, Psychological Health, Social Relationships and Environment. Reference values for different versions of the WHOQOL-BREF are available, and QoL is scored out of a maximum total of 100. The WHOQOL-BREF has not been validated in research with autistic adults. Research using the tool with the general population indicated internal consistency coefficients (Cronbach’s alpha) between 0.66 for the social domain and 0.84 for the physical domain. Test-retest reliability ranges from 0.66 for the physical domain to 0.87 for the environmental domain. Data relating to construct validity and sensitivity to change are not available.

Quality of Life Questionnaire

The Quality of Life Questionnaire (QoL-Q) tool is a 40-item questionnaire designed for assessing QoL of individuals with intellectual disability (ID). It was designed to be administered in an interview format. The QoL-Q assesses QoL across four domains: Satisfaction, Competence/Productivity, Empowerment/Independence and Social Belonging/Community Integration. No normative reference values for QoL scores are available. A theoretical minimum total QoL score is 40 and a theoretical maximum score would be 120. Internal consistency coefficient (Cronbach’s alpha) has been reported for the total score as 0.9. Test-retest reliability was 0.87. Data relating to construct validity and sensitivity to change are not available.

No guidance is available on categorisation of scores, for example, as low, average or good. Findings from studies using the QoL-Q cannot be easily compared to scores obtained using other QoL measurement tools because the domains of QoL assessed differ from other measures. The QoL-Q has not been validated in research with autistic adults.

Quality of Life Inventory

The Quality of Life Inventory (QOLI) is a 32-item tool that measures QoL across four domains: Health, Relationships, Employment and Living Conditions. Internal consistency coefficient (Cronbach’s alpha) for the domain total scores ranges from 0.77 to 0.89. Test-retest reliability ranges from 0.80 to 0.91. Data relating to construct validity and sensitivity to change are not available. The QOLI has not been validated in research with autistic adults.

Comprehensive Quality of Life Questionnaire

The Comprehensive Quality of Life Questionnaire (ComQOL) is a 35-item tool that assesses QoL over seven domains: Material Well-being, Health, Productivity, Intimacy, Safety, Place in Community and Emotional Well-being. No normative or control values for the ComQOL are available. Data relating to internal consistency, test-retest reliability, construct validity and sensitivity to change are not available. The ComQoL has not been validated in research with autistic adults.

Short Form Health Survey (SF-36)

The SF-36 is a 36-item tool that assesses QoL over eight domains: Bodily pain, General health, Vitality, Social Functioning, Physical Functioning, Mental Health, Emotional well-being, Limitations to role caused by mental health problems, and Limitations to role caused by physical health problems. It has not been validated in research with autistic adults. Research with the general population indicates internal consistency coefficients (Cronbach’s alpha) between 0.71 for the social domain and 0.92 for individual domains. Data relating to test-retest reliability, construct validity and sensitivity to change are not available.

Medical outcomes study short form health survey version 2 (SF-12 v.2)

The SF-12 v.2 (Medical Outcomes Study Short-Form Health Survey version 2) is an abbreviated 12-item version of the SF-36 where items are grouped into Physical Component Summary and Mental Component Summary. The SF-12 v.2 has not been validated in research with autistic adults. Data relating to internal consistency, construct validity and sensitivity to change are not reported. Test-retest reliability ranges from 0.89 for the physical component score to 0.76 for the mental component score.

Novel QoL measures (QOL1 and QOL2)

The QOL1 is a proxy assessment of ‘autism-friendly environment’ and includes assessment of the autism-specific knowledge held by the family, the provision of a structured education programme, the implementation of a personal treatment and care plan, the extent to which everyday life/employment is appropriate to the individual’s ability and an overall rating of QoL. Each domain is scored on a scale of 1–5, with a score of 1 described as ‘very good’, 2 as ‘good’, 3 as ‘average’, 4 as ‘poor’ and 5 as ‘very poor’. The QOL2 consists of a parent/carer rating of the well-being of an adult on the autism spectrum. This is scored on a scale of 1–5 and provides a marker of overall well-being in a residential setting. Cronbach’s alpha for QOL1 is reported as 0.96 (Billstedt et al., 2011). Data relating to test-retest reliability, construct validity and sensitivity to change are not available.

In summary, seven QoL measures were used across the 14 studies reviewed. A number of domains are assessed using the individual tools, the most common being physical health and mental health/well-being. Where they were reported, internal consistency and test-retest reliability were within acceptable levels, although it is important to note within the context of the current review that these metrics are derived from non-autistic samples, with no validation of these tools specifically for use with autistic adults (with the exception of QOL1 and QOL2). Only one study reported on the use of a QoL measure designed for use with adults on the autism spectrum (Billstedt et al., 2011). Billstedt et al. (2011) noted that their novel measure was brief and more work was required before it could be considered to be a valid QoL measure for autistic adults.

Results from studies using self-report QoL measures with autistic adults

The following section describes the results from eight identified papers reporting self-assessment of QoL by autistic participants.

Studies utilising the WHOQOL-BREF

Four studies included in the review utilised the WHOQOL-BREF (Jennes-Coussens et al., 2006; Kamio et al., 2013; Kamp-Becker et al., 2010; Lin, 2014). Jennes-Coussens et al. (2006) assessed the QoL of twelve 18- to 21-year-old males with a diagnosis of Asperger syndrome and compared this to a typically developing control group. Questionnaires were posted to participants. Results showed that participants with Asperger syndrome rated their QoL lower than controls (mean overall score of 3.7 (standard deviation (SD) = 0.7) vs 4.1 (SD = 0.6), respectively). Of the four domains, only differences in the scores for Social Relationships reached statistical significance with participants with Asperger syndrome reporting a mean score of 49.5 (SD = 23.0) compared to 78.4 (SD = 18.2) for typically developing controls.

In a substantially larger sample, Kamio et al. (2013) reported on the QoL of 154 adults on the autism spectrum. Questionnaires were posted to participants. The authors only report on two domains of the tool, Psychological Health and Social Relationships. Autistic participants scored significantly lower than a typically developing reference group (2.71 vs 3.19–3.25, respectively, for Social Relationships and 2.78 vs 3.26–3.32, respectively, for Psychological Health).

Kamp-Becker et al. (2010) assessed QoL of an all male sample of 26 adults on the autism spectrum compared to two age-specific reference samples (healthy individuals and individuals with schizophrenia spectrum disorder) using the German version of the WHOQOL-BREF. QoL scores were significantly lower for autistic individuals than in the typically developing sample (mean 60.6 (SD = 26.1) compared to 75.5 (SD = 16.4), respectively). In three of four domains, mean scores for autistic participants were significantly lower than for the reference population; Physical Health (70.1, SD = 19.1 vs 87.2, SD = 13.6), Psychological Health (61.5, SD = 21.9 vs 79.1, SD = 14.0) and Social Relationships (53.8, SD = 23.5 vs 74.9, SD = 17.5). Autistic participants rated QoL in the Social Relationships domain lowest.

In Taiwan, Lin (2014) conducted a cross-sectional study with 41 autistic adults (mean age 26.9 years) and compared to a control group of 41 adults without ASD selected from a community sample of 122 adults without ASD and matched to the autistic sample on age and gender. The autistic participants scored consistently lower than matched controls. Further descriptive comparisons were made without significance testing to a Taiwanese health population reference group of 9107 individuals for the Taiwanese version of the WHOQOL-BREF. Analysis showed autistic adults reported lower QoL in the Social Relationships domain (mean 1.5 SD below the reference group) and in the domains of Physical and Psychological Health (mean 1.0 SD below the reference group).

Studies utilising the QoL-Q

Two included studies used the QoL-Q, both of which provided data at more than one time point. Gal et al. (2015) sought to assess the impact of engaging in an army training and service programme on QoL of 25 autistic adults in Israel (autistic adults in Israel are exempt from national military service, which is usually undertaken at age 18). The authors sought to examine the effect of enrolment in the 3-month training programmes followed by a 6-month placement in a profession in the armed forces that would match the ‘special characteristics of people with ASD’, including aerial photography interpretation. QoL was measured immediately before and after the 3-month aerial photography interpretation training programme and after 6 months of integration in an army unit. No significant difference in overall QoL or individual domains was found between the beginning of the training period and afterwards. However, a statistically significant self-reported increase in QoL was noted after 6 months of integration in the army unit across all domains.

Katz et al. (2015) investigated the influence of taking part in a 9-month work placement on the self-reported QoL of 26 adults with ‘High Functioning’ ASD. Participants were either about to transfer from education to employment or were already seeking employment. QoL was measured every 3 months throughout their employment period. After 9 months, a non-significant increase from baseline in self-reported overall QoL was found, particularly in the self-competence domain.

Studies utilising QOLI

Only one included study utilised the QOLI. Hesselmark et al. (2014) assessed QoL of 60 autistic adults in an intervention study that evaluated the impact of Cognitive Behavioural Therapy (CBT) versus recreational activity on QoL and Self-Esteem. The CBT group received a programme aimed at introducing participants to concepts of acceptance and change, whereas those in the Recreational Activity intervention group received structured activity that promoted social interaction. The mean QOLI scores in the CBT and Recreational Activity groups were −0.11 and −0.28 at baseline, 0.64 and −0.01 at post-intervention and 0.64 and 0.30 at follow-up, respectively (negative scores indicate a response of ‘dissatisfied’). The authors also quoted a mean QOLI score for non-clinical controls of 2.8 and note that the total participants’ mean score was −0.19 in the autistic adult at baseline which ‘indicated a general dissatisfaction with life (and) is similar to scores of untreated post-traumatic stress disorder (PTSD) patients (mean = 0.3, SD = 3.2) observed in other studies’.

Studies utilising the SF-36

One study used the SF-36. In a cross-sectional study, Van Heijst and Geurts (2015) found QoL to be significantly lower in 24 autistic adults compared with controls (mean total score 66.7 (SD = 16.6) vs 82.0 (SD = 13.2)). Age and intelligence quotient (IQ) were not found to predict QoL but autism severity (as measured by the social responsiveness scale) was a significant predictor of QoL.

Studies utilising the SF-12 v.2

One study used a derivative of the SF-36, the SF 12 v.2. Khanna et al. (2014) conducted a cross-sectional study of QoL of 291 autistic adults registered with the Interactive Autism Network (IAN). The results indicated that the mental summary component of the SF-12 tool was lower for autistic adults compared to US norms. This statistically significant difference was present across both genders and all age groups. Scores for the physical summary component were only significantly lower in autistic adults aged 18–24 years compared to US norms. Severity of autism symptoms was assessed for each participant using the Adult Short Autism-Spectrum Quotient-10 Items (AQ-10). Physical component scores were higher in the high autism severity group (50.9) than the low autism severity group (46.9); mental component scores were higher for the low autism severity group (42.6) than the high autism severity group (39.7).

Results from studies using proxy reports of QoL of autistic adults

Studies utilising the ComQOL

Saldaña et al. (2009) assessed the QoL of 25 autistic adults via parent/carer report. Informants completed an interview which included aspects of the ComQOL. The domains ‘Material Well-being’ and ‘Intimacy with Family’ were reported to be the domains with the highest QoL (83.6% and 81.4%, respectively), followed by Place in Community (67.8%), Productivity (66.4%), Health (63.3%), Safety (61.3%) and Intimacy with Friends (55.8%). Assessment of the ‘Emotional Well-being’ domain was excluded from the study as the authors concluded that this domain would be too complex for proxy assessment.

Studies utilising the QOL1 and QOL2

Billstedt et al. (2011) conducted a cross-sectional study of QoL with 108 parents/carers of autistic individuals using two novel parent/carer proxy report tools designed for the study, the QOL1 and QOL2. Findings indicated a mean score of average/good across participants utilising the QOL1 tool; using the QOL2 tool, parents and carers rated their family member’s QoL as very good or good in 91/100 cases.

Results from studies using a combination of proxy and self-reported QoL

Studies utilising the WHOQOL-BREF

Hong et al. (2016) conducted a cross-sectional study investigating the relationship between autistic adult self-report, maternal proxy report and maternal report of QoL of a sample of 60 mother and adult offspring dyads. The mean age of the participating autistic adults was 32.0 (SD = 6.8). Correlation between autistic adult self-report and maternal proxy report was significantly higher than the correlation between autistic adult self-report and maternal self-report across three domains (Physical Health, Psychological Health and Social Relationships). For autistic adults self-report and maternal report, mean scores were not significantly different for three of the domains (Physical Health, Psychological Health and Environment) but autistic adult self-reported significantly higher QoL in the Social Relationships domain (M = 71.2, SD = 26.9) than on maternal report (M = 60.2, SD = 24.1). There were no significant differences in mean domain scores between adult self-report and maternal proxy report.

Studies utilising the QoL-Q

Garcia-Villamisar (2010) evaluated the effect of a 12-month leisure programme intervention on the QoL of 37 adults on the autism spectrum. The intervention group was engaged in exercise groups, craft workshops, media-based activities and organised events and outings, with a wait-list control group of 34 autistic individuals. The authors assessed receptive and expressive language skills in autistic participants and administered the QoL-Q to those deemed capable of self-reporting. For those who could not self-report, at least two individuals who knew the participant well completed the QoL-Q on the individual’s behalf (the number in this group is not reported). Compared with baseline scores, a statistically significant increase in QoL was reported in the intervention group after 12-month follow-up (improved from 50.6 to 63.6), but not in the control group (scores changed from 54.2 to 55.3).

Garcia-Villamisar et al. (2013) conducted a cross-sectional study to investigate the association between QoL, challenging behaviours and autistic traits in 70 autistic adults with Intellectual Disability (ID). As with Garcia-Villamisar et al. (2010), participants’ verbal skills were assessed and people considered able to do so completed the measure themselves (6 participants; therefore, 64 had proxy report). The mean overall baseline QoL score was 72.5; however, the authors do not provide interpretation of the QoL scores.

Synthesis of findings across studies

The majority of included studies used self-assessment to assess QoL of participants. All studies where a comparison was made to a typically developing sample reported QoL to be lower in autistic adults. It is of note that 7 of the 14 (50%) included studies did not include reference values to enable comparison of QoL scores from their samples with typically developing individuals (Billstedt et al., 2011; Gal et al., 2015; Garcia-Villamisar, 2010; Garcia-Villamisar et al., 2013; Hesselmark et al., 2014; Katz et al., 2015; Saldaña et al., 2009). It is therefore difficult to draw any firm conclusions regarding the comparison of the QoL of autistic individuals participating in these studies to that of the general population.

In studies that reported individual domain scores, QoL tended to be reported as pervasively lower for autistic adults and to be notably low in domains capturing social relationships and integration. For example, all four studies using the WHOQOL-BREF reported that QoL scores were lower for autistic adults overall and across all four domains, when compared to neurotypical comparison groups (Jennes-Coussens et al., 2006; Kamio et al., 2013; Kamp-Becker et al., 2010). Across all four studies, adults on the autism spectrum reported lowest QoL on the Social Relationships domain. Saldaña et al. (2009) found that scores were lowest in the ‘Intimacy with Friends’ domain of the ComQOL.

Evidence for the impact of autism severity on QoL has not been comprehensively examined. The two studies (Khanna et al., 2014; Van Heijst and Geurts, 2015) that do examine the associations between autism severity and QoL report a significant correlation between greater severity of autistic symptoms and lower QoL. Furthermore, Saldaña et al. (2009) attribute an inability for family members to act as proxies in assessing the QoL for an autistic relative to increased symptom severity.

Methodological quality of included studies

The methods by which participants had been diagnosed with ASD were comprehensively reported in most of the included studies. Nine studies included evidence that the ASD diagnosis had been made by a Clinical Psychologist, Psychiatrist or multidisciplinary clinical team (Billstedt et al., 2011; Garcia-Villamisar, 2010; Garcia-Villamisar et al., 2013; Hesselmark et al., 2014; Jennes-Coussens et al., 2006; Kamp-Becker et al., 2010; Lin, 2014; Saldaña et al., 2009; Van Heijst and Geurts, 2015). Further clarification about diagnosis was sought from the authors of two studies (Kamio et al., 2013; Khanna et al., 2014). Khanna et al. stated that they recruited through the IAN database and that most of the individuals included on IAN had received a diagnosis by a clinician or their diagnosis had been clinically validated (29 July 2015, personal communication); however, some individuals had volunteered to be included without a clinician-given ASD diagnosis, but this applied to a minority of participants (Daniels et al., 2012). Kamio et al. (2013) did not offer sufficient information about the method of ASD diagnosis in their study for us to comment further.

The methodological quality assessment using structured checklists can be found in the web appendix. Blinding of outcome assessment was inadequately addressed in the RCT by Hesselmark et al. (2014). The potential for other sources of bias was considered low risk for both experimental studies (Garcia-Villamisar, 2010; Hesselmark et al., 2014).

All nine cross-sectional studies included an adequate description of the presenting condition, including a definition of the exposure and ascertainment of outcomes. However, six of these studies were regarded as unclear risk of bias from selection, and five from non-response in accordance with the quality checklists used.

The two pre-test/post-test studies were of acceptable methodological quality. The total score percentages for Gal et al. (2015) and Katz et al. (2015) were 60% and 55%, respectively.

Discussion

We identified 14 studies that reported on QoL of autistic adults using either self-report or proxy report or a combination of the two. Narrative synthesis of data from these studies reveals a trend for QoL to be lower in autistic adults compared to typically developing adults. In total, 13 studies used QoL measures that have previously been shown to be reliable and valid indicators of QoL in general population samples but importantly these measures have not been validated for use with autistic adults. One measure of QoL, composed of the QOL1 and QOL2 tools, had been designed for use with individuals on the autism spectrum (Billstedt et al., 2011). The identification of widespread use of tools without established reliability and validity for measuring QoL of autistic adults is an important finding of this review because this practice may lead to misleadingly low or high scores.

A synthesis of the findings from across the included studies reveals that autistic adults are likely to score lower than their typically developing counterparts on QoL domains assessing social inclusion and interaction. Without a clear understanding of what specific aspects are important to assess in relation to QoL for autistic people, it is difficult to make sense of these data. Two potential explanations are possible. It is possible autistic people’s inherent difficulties regarding social interaction and communication are a risk factor for lower QoL reported in these domains. However, we need to also be mindful that autistic individuals may value some experiences and activities more/less than people without ASD (Tavernor et al., 2013). For example, time spent on repetitive behaviours and on circumscribed or particular interests may be considered more desirable than some activities undertaken by typically developing adults; social and group activities may be valued less by autistic adults. Therefore, the items included in these specific domains of assessment tools developed for and validated with the general population may lack validity when used with autistic individuals. Of course, these two explanations are not mutually exclusive. One firm conclusion that can be drawn from this review is that it is imperative we develop a greater understanding of what aspects of life matter most and affect the lives of autistic people and that we develop tools to measure these when assessing their QoL.

Seven of the 14 included studies showed a bias towards recruiting adults on the autism spectrum without intellectual (learning) disability (Hesselmark et al., 2014; Jennes-Coussens et al., 2006; Kamio et al., 2013; Kamp-Becker et al., 2010; Khanna et al., 2014; Lin, 2014; Van Heijst and Geurts, 2015). Thus, the review does not provide a comprehensive picture of QoL for those 30%–40% of autistic adults with co-occurring ID (Totsika et al., 2010) – and this review’s interpretation of study results may or may not be applicable to this population. Similarly, most studies included autistic adults aged 18–55 years. Studies including autistic adults from across the whole age range are needed to investigate whether younger adults and older people conceptualise QoL differently and report different findings.

The findings highlight the conundrum of how best to disentangle the likely complex interactions between autism symptom severity and QoL. Two studies suggest QoL is lower when autism severity is greater (Khanna et al., 2014; Van Heijst and Geurts, 2015). Saldaña et al. (2009) importantly remind us that a greater degree of autism severity may impact people’s capacity to self-report and further may also impact proxy raters accurately determining QoL for the person they are reporting for. Hong et al. present the first data directly comparing proxy and self-report of QoL data from autistic adults. Their results point to a correlation between maternal proxy report and self-report by autistic adults; however, the majority of autistic adults who participated in the study were described as high functioning (Hong et al., 2016). The ability level of the autistic participants may have made it easier for proxy reporters to make judgements about QoL for the autistic person they were reporting on. Furthermore, despite overall general agreement between self-report and proxy report, Hong and colleagues report a significant difference between adult self-report and maternal proxy report in the Social Relationships domain, where maternal ratings were lower. They conclude that although mothers are generally accurate reporters of autistic adult offspring QoL, there may be a tendency for proxy reporters to consider QoL in social relationships as poorer than autistic adult self-report. Interestingly, Hong et al. report strongest agreement in the environmental domain and least agreement in the psychological health domain. This may suggest proxy reporters have more difficulty interpreting the internal or intrapersonal aspects of QoL of autistic adults – or may provide further evidence for lack of construct validity of particular neurotypically derived and validated domains of QoL for autistic adults.

Taken together, the reviewed evidence highlights a clear take home message. To address the conceptual and methodological challenges, a measure of QoL either designed specifically for or validated with autistic adults is required to accurately assess their QoL. Improving measures already used with autistic adults rather than designing new tools from first principles is most likely to lead to the timely production of acceptable measures. Only one study identified in this review used a measure designed specifically for autistic people (Billstedt et al., 2011); interestingly, this was the only study to report better QoL for autistic adults, compared with control populations. Billstedt et al. (2011) stated that the measure would benefit from further development to provide a more in-depth assessment across the multiple domains that constitute QoL.

Strengths and weaknesses of the current review

A key strength of this review is inclusion of only studies that reported on participants for whom a valid ASD diagnosis had been ascertained (although of course this could not be verified as part of this study). The review of the psychometric properties of the QoL measures was undertaken according to defined criteria. This further highlighted the shortcomings and methodological/design issues that should be addressed when investigating QoL of autistic adults. A potential weakness of this review is the exclusion of studies published in languages other than English, as these may contain relevant information. Similarly, the exclusion of case studies, although limited in the reliability of their conclusions, may have excluded the results of studies using QoL measurement tools not identified in this review.

Implications for future research

The studies reviewed had several important methodological shortcomings, which should be addressed in future research. Bishop-Fitzpatrick et al. (2016) outline a need to address the conceptualisation of QoL among autistic people. They identify ‘normative’ outcomes (employment, independent living and social engagement) and ‘objective QoL’ (physical health, mental health, neighbourhood quality and family contact) as relevant concepts. The first step, therefore, is to reach consensus regarding what constitutes ‘good’ QoL for autistic individuals, across the lifespan. This cannot be achieved without active involvement of autistic people to define QoL – either as part of the research team or as research consultants. Without the involvement of autistic adults in design, measures are unlikely to have content validity or have acceptable wording and formatting (Parr, 2016). The participation of autistic adults and relatives in discussion groups is required to ensure a tool takes account of the views of people with ID or those who represent adults who are unable to participate themselves. It is possible that more than one QoL measurement tool will be required – for example, for adults on the autism spectrum with and without ID, and potentially for younger and older adults. There are two possible approaches to optimal development of these tools; either the adaptation of an existing QoL measurement tool or the development of an entirely novel tool. The former would allow access to an existing framework and have the significant benefit of control data from a neurotypical population being available, while the latter allows greater freedom to fully incorporate the views and experiences of adults on the autism spectrum and those people who know them well. Either approach will require the investigation of the psychometric properties of any new or adapted tool and their acceptability to adults on the autism spectrum, including to people who might complete the measure on their behalf. Following a large study of QoL at Newcastle University, UK, McConachie, Rodgers and Parr are working on developing items additional to the WHOQOL-BREF (with permission), with the aim of creating a reliable and valid way of measuring the QoL of autistic adults.

Conclusion

Adults on the autism spectrum typically report lower QoL than typically developing adults when measured with tools designed for the general population. There is a pressing need to develop a reliable and valid measure of QoL of adults on the autism spectrum. Large studies of representative populations of people with ASD are needed using measurement tools that are reliable and valid for use with autistic adults. Parallel collection of data about autism characteristics and conditions that often accompany autism would allow investigation of how individual factors (such as mental health conditions) affect QoL. As we know from childhood autism research, planned multisite studies, data sharing and joint publication by groups working in collaboration would be desirable to ensure timely new knowledge is available to the autism community and those planning services.

Supplemental Material

AUT714988_Lay_Abstract – Supplemental material for A systematic review of quality of life of adults on the autism spectrum

Supplemental material, AUT714988_Lay_Abstract for A systematic review of quality of life of adults on the autism spectrum by Michael Ayres, Jeremy R Parr, Jacqui Rodgers, David Mason, Leah Avery and Darren Flynn in Autism

Supplemental Material

AUT714988_Supplementary_Material – Supplemental material for A systematic review of quality of life of adults on the autism spectrum

Supplemental material, AUT714988_Supplementary_Material for A systematic review of quality of life of adults on the autism spectrum by Michael Ayres, Jeremy R Parr, Jacqui Rodgers, David Mason, Leah Avery and Darren Flynn in Autism

Footnotes

Acknowledgements

The authors thank the information scientist team for their help in devising and carrying out an appropriate search protocol.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: David Mason’s research post was funded by the Research Autism and the Newcastle University research programme on autism lifecourse and ageing (http://research.ncl.ac.uk/adultautismspectrum/) was funded by the autism research charity Autistica ().

References

Baird

Simonoff

Pickles

(2006) Prevalence of disorders of the autism spectrum in a population cohort of children in South Thames: the Special Needs and Autism Project (SNAP). The Lancet 368: 210–215.

Billstedt

Gillberg

(2011) Aspects of quality of life in adults diagnosed with autism in childhood: a population-based study. Autism 15: 7–20.

Bishop-Fitzpatrick

Hong

Smith

(2016) Characterizing objective quality of life and normative outcomes in adults with autism spectrum disorder: an exploratory latent class analysis. Journal of Autism and Developmental Disorders 46: 2707–2719.

Brugha

Mcmanus

Bankart

(2011) Epidemiology of autism spectrum disorders in adults in the community in England. Archives of General Psychiatry 68: 459–465.

Chiang

H-M

Wineman

(2014) Factors associated with quality of life in individuals with autism spectrum disorders: a review of literature. Research in Autism Spectrum Disorders 8: 974–986.

Daniels

Rosenberg

Anderson

(2012) Verification of parent-report of child autism spectrum disorder diagnosis to a web-based autism registry. Journal of Autism and Developmental Disorders 42: 257–265.

Flynn

Knoedler

Hess

(2012) Engaging patients in health care decisions in the emergency department through shared decision-making: a systematic review. Academic Emergency Medicine 19: 959–967.

Gal

Selanikyo

Erez

AB-H

(2015) Integration in the vocational world: how does it affect quality of life and subjective well-being of young adults with ASD. International Journal of Environmental Research and Public Health 12: 10820–10832.

Garcia-Villamisar

(2010) Effects of a leisure programme on quality of life and stress of individuals with ASD. Journal of Intellectual Disability Research 54: 611–619.

10.

Garcia-Villamisar

Dattilo

Matson

(2013) Quality of life as a mediator between behavioral challenges and autistic traits for adults with intellectual disabilities. Research in Autism Spectrum Disorders 7: 624–629.

11.

Harper

(1998) Development of the World Health Organization WHOQOL-BREF quality of life assessment. Psychological Medicine 28: 551–558.

12.

Hesselmark

Plenty

Bejerot

(2014) Group cognitive behavioural therapy and group recreational activity for adults with autism spectrum disorders: a preliminary randomized controlled trial. Autism 18: 672–683.

13.

Higgins

JPT

Green

(eds) (2009) Cochrane Handbook for Systematic Reviews of Interventions, version 5.0.2. The Cochrane Collaboration. Available at: www.cochrane-handbook.org (accessed 3 March 2011).

14.

Hong

Bishop-Fitzpatrick

Smith

(2016) Factors associated with subjective quality of life of adults with autism spectrum disorder: self-report versus maternal reports. Journal of Autism and Developmental Disorders 46: 1368–1378.

15.

Jennes-Coussens

Magill-Evans

Koning

(2006) The quality of life of young men with Asperger syndrome – a brief report. Autism 10: 403–414.

16.

Kamio

Inada

Koyama

(2013) A nationwide survey on quality of life and associated factors of adults with high-functioning autism spectrum disorders. Autism 17: 15–26.

17.

Kamp-Becker

Schroder

Remschmidt

(2010) Health-related quality of life in adolescents and young adults with high functioning autism-spectrum disorder. Psycho-Social Medicine 7: Doc03.

18.

Katz

Dejak

Gal

(2015) Work performance evaluation and QoL of adults with High Functioning Autism Spectrum Disorders (HFASD). Work 51: 887–892.

19.

Khanna

Jariwala-Parikh

West-Strum

(2014) Health-related quality of life and its determinants among adults with autism. Research in Autism Spectrum Disorders 8: 157–167.

20.

Levisohn

(2007) The autism-epilepsy connection. Epilepsia 48: 33–35.

21.

Lin

L-Y

(2014) Quality of life of Taiwanese adults with autism spectrum disorder. PLoS ONE 9: e109567.

22.

Moher

Liberati

Tetzlaff

(2009) Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Annals of Internal Medicine 151: 264–269.

23.

Mukaetova-Ladinska

Perry

Baron

(2012) Ageing in people with autistic spectrum disorder. International Journal of Geriatric Psychiatry 27: 109–118.

24.

National Institute for Health and Care Excellence (NICE) (2012) Recognition, referral, diagnosis and management of adults on the autism spectrum. Available at: https://www.nice.org.uk/guidance/cg142

25.

National Institute for Health and Care Excellence (NICE) (2013) Autism spectrum disorder in under 19s: Support and management. Available at: https://www.nice.org.uk/guidance/cg170

26.

Parr

(2016) How can we learn more about the lives of adults on the autism spectrum from across the age range, and their relatives? In: Wright

(ed.) Autism Spectrum Disorder in Mid and Later Life. London: Jessica Kingsley Publishers, pp.288–296.

27.

Saldaña

Álvarez

Lobatón

(2009) Objective and subjective quality of life in adults with autism spectrum disorders in southern Spain. Autism 13: 303–316.

28.

Simonoff

Pickles

Charman

(2008) Psychiatric disorders in children with autism spectrum disorders: prevalence, comorbidity, and associated factors in a population-derived sample. Journal of the American Academy of Child and Adolescent Psychiatry 47: 921–929.

29.

Sirriyeh

Lawton

Gardner

(2012) Reviewing studies with diverse designs: the development and evaluation of a new tool. Journal of Evaluation in Clinical Practice 18: 746–752.

30.

Stahlberg

Soderstrom

Rastam

(2004) Bipolar disorder, schizophrenia, and other psychotic disorders in adults with childhood onset AD/HD and/or autism spectrum disorders. Journal of Neural Transmission 111: 891–902.

31.

Tavernor

Barron

Rodgers

(2013) Finding out what matters: validity of quality of life measurement in young people with ASD. Child: Care, Health and Development 39: 592–601.

32.

Totsika

Felce

Kerr

(2010) Behavior problems, psychiatric symptoms, and quality of life for older adults with intellectual disability with and without autism. Journal of Autism and Developmental Disorders 40: 1171–1178.

33.

Van Heijst

BFC

Geurts

(2015) Quality of life in autism across the lifespan: a meta-analysis. Autism 19: 158–167.

34.

Wells

Shea

O’Connell

(2014) The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomized studies in meta-analyses. Ottawa Hospital Research Institute. Available at: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp

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

0.43 MB

0.31 MB