Measuring subjective quality of life in autistic adults with the PROMIS global–10: Psychometric study and development of an autism-specific scoring method

Abstract

Quality of life is widely acknowledged as one of the most important outcomes in autism research, but few measures of this construct have been validated for use in autistic people. The goal of the current study was to examine the psychometric properties of the Patient-Reported Outcomes Measurement Information System Global–10, an established self-report measure of health-related quality of life, in a sample of 901 autistic adults (predominantly female and adult-diagnosed) recruited from the Simons Foundation Powering Autism Research for Knowledge cohort. Using an item response theory framework, we performed a comprehensive psychometric evaluation of the Patient-Reported Outcomes Measurement Information System Global–10 in this sample, examining its latent structure, differential item functioning, reliability, and construct validity. After developing an autism-specific measurement model, the Patient-Reported Outcomes Measurement Information System Global–10 demonstrated excellent psychometric properties in the current sample, including excellent model-data fit, high reliability, minimal differential item functioning across subgroups of autistic adults, and an expected pattern of correlations with external variables. Exploratory analyses indicated that lower quality of life was associated with female sex and identified as a sexual/gender minority. A free online score calculator has been created to facilitate the use and interpretation of normed Patient-Reported Outcomes Measurement Information System Global–10 general quality of life latent trait scores for clinical and research applications (available at https://asdmeasures.shinyapps.io/promis_qol).

Lay Abstract

Quality of Life an outcome that both researchers and autistic advocates agree is extremely important to consider when implementing services, interventions, and supports for autistic people. However, there has been little research on the topic of how quality of life can best be measured in autistic people or whether existing quality of life questionnaires are appropriate for use in the autistic population. This study aimed to validate an established quality of life measure, the Patient-Reported Outcomes Measurement Information System Global–10, in a large sample of autistic adults recruited online. We created a new way to score the Patient-Reported Outcomes Measurement Information System Global–10 scale and generate a “General quality of life” score specific to autistic adults. This new score performed very well in this sample, showing very little measurement error and relating in expected ways to similar constructs, such as physical health and emotional distress. Exploratory analyses found that lower quality of life was associated with female sex and self-identification as a sexual or gender minority (i.e. LGBTQ + identity). These findings suggest that the new Patient-Reported Outcomes Measurement Information System Global–10 quality of life score is a reliable and valid measure of quality of life in autistic adults, although additional studies are necessary to further explore its measurement properties in other subsets of the autistic population, such as individuals with intellectual disabilities. This measure is freely available for use as an outcome in both research and clinical practice, and an online score calculator is available to support the use of this measure in real-world applications.

Keywords

adults behavioral measurement item response theory mental health physical health PROMIS psychometrics quality of life reliability validity

Quality of Life (QoL) is a multidimensional construct defined as “an overall general well-being that comprises objective descriptors and subjective evaluations of physical, material, social, and emotional well-being together with the extent of personal development and purposeful activity, all weighted by a personal set of values” (Felce & Perry, 1995; Karimi & Brazier, 2016). QoL is a particularly salient outcome for autistic people, as increased subjective QoL is often considered within the neurodiversity paradigm to be the end goal of any intervention, accommodation, or service provided to autistic people (Brown et al., 2022; Burgess & Gutstein, 2007; Kapp, 2018; Robertson, 2009). Although QoL is frequently measured in the studies of autistic adults (Ayres et al., 2018; Lam et al., 2021), few studies have examined the psychometric properties of self-reported QoL measures in this population (Holmes et al., 2020; McConachie et al., 2018; Williams & Gotham, 2021a). Such work is particularly important, as it can uncover previously unknown sources of systematic measurement error (i.e. differential item functioning (DIF); Penfield & Camilli, 2006; Tay et al., 2015) that could significantly bias within- and between-group comparisons of QoL scores (Williams & Gotham, 2021a). As interest in the topic of QoL in autism continues to grow, additional psychometric studies are desperately needed to determine whether existing measures used to quantify QoL in autistic adults are adequate for this purpose and free of significant measurement bias.

The body of research examining the psychometric properties of QoL in autistic adults is relatively recent, as prior to 2018, no studies had systematically attempted to validate existing QoL measures in this population (Ayres et al., 2018). A landmark study by McConachie et al. (2018) was the first to do so, examining the psychometric properties of the World Health Organization Quality of Life–Brief Version (WHOQOL–BREF; The WHOQOL Group, 1998) in a sample of over 300 autistic adults. In this sample, the original WHOQOL–BREF factor structure exhibited poor fit to the data, requiring multiple post hoc model modifications to achieve even minimally acceptable fit indices. However, the authors reported adequate reliability and convergent/divergent validity for the WHOQOL–BREF subscale scores, providing some psychometric support for the use of this measure in the autistic population. Based on input from autistic adults regarding the content of the WHOQOL–BREF and the aspects of QoL they found to be most salient (McConachie et al., 2018, 2019), the same group of researchers developed the Autism Spectrum Quality of Life measure (ASQoL), a supplementary QoL item pool to be used in conjunction with the WHOQOL–BREF in autistic adults. Although the authors reported preliminary psychometric support for an 8-item ASQoL composite score, an independent psychometric assessment of this measure found several ASQoL items to exhibit problematically large amounts of DIF across multiple demographic categories, including sex/gender (Williams & Gotham, 2021a). The recent study by Williams and Gotham (2021a) also validated a composite score derived from four WHOQOL–BREF items (termed the WHOQOL-4) as a brief QoL measure in their autistic adult sample. However, the WHOQOL-4 does not contain any items from the “physical,” “social,” or “environment” domains of the WHOQOL–BREF, and it therefore may not sufficiently represent the relevant facets of QoL captured by the longer form (Williams and Gotham, 2021a). Based on these few psychometric studies, all validated scales for measuring QoL in autistic adults have notable limitations, and additional studies are warranted to explore other putative measures of QoL in this population.

One QoL measure of particular interest is the 10-item “Global Health” scale created as a part of the Patient-Reported Outcomes Measurement Information System (PROMIS) initiative (Hays et al., 2009), hereafter referred to as the “PROMIS Global–10.” The PROMIS Global–10 contains high-level items that ask respondents to evaluate their overall physical, mental, and social well-being, and their overall QoL (e.g. “In general, how would you rate your satisfaction with your social activities and relationships?”). Items such as those on the PROMIS Global–10 allow respondents to individually weigh different aspects of their experiences to arrive at an overall judgment of their QoL in each domain. The PROMIS Global–10 has not been frequently used in studies of QoL in autistic adults, and no studies to date have assessed its reliability, latent structure, or DIF in the autistic population. However, it is notable that this measure is included within the recently proposed PROMIS Autism Battery–Lifespan (PAB-L; Holmes et al., 2020), a battery of PROMIS scales designed specifically to comprehensively measure QoL in autistic adults. Moreover, the PROMIS Global–10 items are written in such a way that individuals may report a “good QoL” without conforming to normative standards of a “good life” (e.g. an individual who has no friends but does not desire any friends could nevertheless report a high level of satisfaction with their social relationships). The relative brevity of the questionnaire is also notable, as a 10-item QoL assessment is substantially shorter than the somewhat more comprehensive 26-item WHOQOL-BREF. For these reasons, we believe that the PROMIS Global–10 has potential utility as a QoL measure in the autistic adult population, and thus, that further investigation of its psychometric properties in this population is warranted.

In this study, we addressed the present gaps in the literature by evaluating the psychometric properties of the PROMIS Global–10 in a large sample of independent autistic adults, assessing the measure’s latent structure, reliability, and DIF across multiple demographic and clinical subgroups. As a secondary aim, we sought to use item response theory (IRT) to improve the interpretation of PROMIS Global–10 scores by (a) generating autism-specific normative scores for within-group comparison and (b) linking global QoL scores from the PROMIS Global–10 to qualitative descriptors. It is notable that the goal of this investigation was not to preserve the original structure of the PROMIS Global–10 or present scores that could be compared with general population samples. Rather, we sought to model the structure of the PROMIS Global–10 items in the population of autistic adults, specifically deriving a novel scoring method that could generate valid scores for this particular population (e.g. for use as outcome measures in clinical care or research studies evaluating services or interventions). Moreover, to assess the nomological validity of the newly derived latent trait scores in the autistic population, we examined the relationships between these scores and theoretically related clinical variables. As the PROMIS Global–10 is predominantly an index of health-related QoL (comprising both physical and mental health domains), we hypothesized that perceived levels of physical health problems and mental health problems would both predict overall QoL in the current sample, even after controlling for the other problem domain. Finally, although a comprehensive examination of predictors of QoL was outside the scope of the current study, we performed exploratory analyses to assess whether QoL was related to demographic variables (age, sex, sexual/gender minority (SGM) identities, race/ethnicity, level of education, number of psychiatric conditions, and prior receipt of special education services) in this sample of autistic adults.

Methods

Participants

Independent autistic adults (i.e. those who are their own legal guardians) above the age of 18 years were recruited in spring 2021 from the Simons Foundation Powering Autism Research for Knowledge (SPARK) cohort (Feliciano et al., 2018) through the SPARK Research Match service (Project No. RM0111Woynaroski_DST). All individuals self-reported a professional diagnosis of autism spectrum disorder or equivalent condition (e.g. Asperger syndrome, Pervasive Developmental Disorder–Not Otherwise Specified), and while these diagnoses were not independently confirmed, prior research has supported their validity in most cases (Fombonne et al., 2021). These participants completed a series of online surveys assessing demographics, medical/psychiatric history, core features of autism, co-occurring psychopathology, somatic symptom burden, and QoL. Participants were compensated with US$10 in Amazon gift cards for the completion of these surveys. A total of 1271 individuals consented to participate in the Research Match study, 901 of whom were included in the current analyses. Data underwent rigorous quality checks, and survey participants were excluded from analyses if they (a) met the SPARK definition of a possibly invalid autism diagnosis (e.g. age of diagnosis is below 1 year of age; diagnosis rescinded by a professional; n = 134), (b) did not self-report a professional diagnosis of autism on the study-specific demographics form (n = 45), (c) reported demographic variables (e.g. age, sex at birth, receipt of special education services in childhood) that were inconsistent with those originally reported to SPARK (n = 34), (d) reported the use of a cochlear implant (an exclusionary criterion for the larger study; n = 3), or (e) endorsed a professional diagnosis of either Alzheimer’s disease or dissociative identity disorder (indicating either careless/random responding or a true diagnosis that could compromise the validity of self-report; n = 23). In addition, individuals who dropped out of the study before completing the PROMIS Global–10 questionnaire (n = 223) were not included in the current analyses. All participants gave informed consent, and all study procedures were approved by the institutional review board at Vanderbilt University Medical Center.

Measures

PROMIS 10-item global health

The PROMIS Global–10 (short form version 1.2; Hays et al., 2009) is a 10-item measure of general QoL developed as a part of the National Institutes of Health PROMIS initiative (Broderick et al., 2013; Cella et al., 2010). The measure consists of items assessing general health, general QoL, physical health (two items), mental health, and social functioning (two items), and three additional items that respectively assess the degree of emotional distress, fatigue, and pain the respondent experienced in the past 7 days. All items are scored on a 1–5 scale with varying response options, except for the pain item, which is scored on a 0–10 scale and converted to scores of 1–5. Based on these global items, the PROMIS group found evidence for two 4-item summary scores assessing physical and mental health, respectively (Hays et al., 2009). However, these two constructs are highly interrelated in prior work (Hays et al., 2009), and in the current study, we opted to model all 10 items of the PROMIS Global–10 simultaneously using a bifactor structure with a single “General QoL” factor and specific factors representing “Physical Health” and “Mental Health.” In addition, due to a small number of individuals endorsing extreme pain in our sample (i.e. < 10 responses of “9” or “10”), we opted to recode the pain item to a 5-point scale (0/1 = “0,” 2/3 = “1,” 4/5 = “3,” 6/7 = “4,” and 8/9/10 = “5”) that differs from the scale used by the original PROMIS investigators (Hays et al., 2009). Latent trait scores (used in further analyses of the measure’s nomological validity) were calculated from individual PROMIS Global–10 item response patterns using autism-specific IRT model parameters derived from the current sample.

Additional measures

In addition to the PROMIS Global–10, individuals in the current study completed the measures of past-week negative affect (PROMIS Emotional Distress (PROMIS-ED) composite; Pilkonis et al., 2011) and somatic symptom burden (Somatic Symptom Scale-8 (SSS-8); Gierk et al., 2014) to measure perceived mental and physical health problems, respectively. Additional details on these measures and their psychometric properties in the current sample can be found in the Supplemental Information.

Statistical analyses

Assessment of latent structure

All statistical analyses were performed in the R statistical computing environment (R Core Team, 2021). Initially, we iteratively checked for redundant items using unique variable analysis (UVA; Christensen et al., 2020), with the weighted topological overlap (wTO) metric used to indicate pairs of items with local dependence and potentially redundant content (i.e. wTO > 0.3 signals significant overlap; Gysi et al., 2018). After removing or combining any redundant items, we fit the remaining PROMIS Global–10 items to a confirmatory bifactor model (equivalent to a bifactor graded response IRT model; Cai et al., 2011; Gibbons et al., 2007; Toland et al., 2017), with the “general QoL” item loading only onto the general factor and all other items permitted to load on physical and/or mental health factors based on the network structure generated by the UVA (Golino & Epskamp, 2017). We fit this model using full-information marginal maximum likelihood estimation (quasi-Monte Carlo Expectation Maximization (QMCEM) algorithm), as implemented in the mirt R package (Chalmers, 2012). Missing data were addressed using full-information maximum likelihood estimation. Model fit was assessed using the limited-information C₂ statistic (Cai & Monroe, 2014) and C₂-based approximate fit indices (i.e. the Comparative Fit Index (CFI_C2) and Root Mean Square Error of Approximation (RMSEA_C2)) and the standardized root mean square residual (SRMR). Individual standardized residuals were also examined to determine the areas of local dependence, with values greater than ± 0.1 indicative of local model misfit.

DIF was also analyzed according to age (< 40 versus ⩾ 40 years), race/ethnicity (non-Hispanic White vs all others), sex assigned at birth (male vs female), SGM (Alexander et al., 2016) status (cisgender-heterosexual vs all others), level of education (4-year college degree or higher vs all others), co-occurring lifetime psychiatric diagnoses (mood disorder, anxiety disorder, and attention-deficit/hyperactivity disorder (ADHD)), receipt of any special education services in primary/secondary school, and age of autism diagnosis (< 18 years vs ⩾ 18 years) using iterative Wald tests (Cao et al., 2017; Williams, 2021; Williams et al., 2021) with the Benjamini–Hochberg correction (Benjamini & Hochberg, 1995). Notably, the categories of “sexual minority” (non-heterosexual) and “gender minority” (transgender and/or non-binary gender) were combined for the purposes of DIF testing, as too few participants were identified as a gender minority to perform adequately powered DIF analyses on this subgroup. Standardized effect sizes (ESSD, interpretable on the metric of Cohen’s d; Meade, 2010) were used to evaluate whether statistically significant DIF was large enough to be practically meaningful, with the values of ESSD > 0.5 (“medium effects”) used to flag an item as problematic.

Scoring, reliability, and validity

After establishing a final structural model for the PROMIS Global–10, we assessed the reliability and validity of this final model in our current sample of autistic adults. Sum score reliability and general factor saturation were quantified using coefficients omega total (ω_T) and omega hierarchical (ω_H), respectively (Green & Yang, 2009; Revelle & Condon, 2019; Rodriguez et al., 2016a, 2016b). In addition, model-based latent trait scores on the general QoL factor, normed on our autistic sample, were generated using the estimated a-posteriori (EAP) estimator (Bock & Mislevy, 1982) and converted to T-scores (M = 50, SD = 10) for the ease of interpretability. Marginal reliabilities for these latent trait scores (ρ_xx) were additionally calculated using the mirt package.

To examine the nomological validity of the newly generated PROMIS Global–10 scores, we examined the zero-order correlations of these scores with several clinical variables, including PROMIS-ED mean scores, SSS-8 total scores, and number of lifetime psychiatric diagnoses, all of which we expected to be inversely related to QoL. As we expected the general QoL score derived from the PROMIS Global–10 to reflect both physical and mental well-being, we further hypothesized that both PROMIS-ED mean scores and SSS-8 total scores would explain meaningful amounts of variance in PROMIS Global–10 scores (i.e. r_p > 0.1) when regressing QoL scores on both predictors simultaneously. Correlations and partial correlations were calculated using robust Bayesian estimators (Kurz, 2019; Williams & Gotham, 2021b; see Supplemental Information for additional details).

We also explored the degree to which QoL scores were predicted by demographic variables, including age, sex, race/ethnicity, age of autism diagnosis (assessed continuously and dichotomized as < 18 and < 4 years to indicate pediatric and early childhood diagnoses, respectively), educational attainment, receipt of special education services, sexual minority status, and gender minority status (using Bayesian t-tests or correlations as appropriate). Effect sizes were tested against interval null hypotheses of d = [−0.2, 0.2] or r = [−0.1, 0.1] (i.e. the null hypothesis states that the true effect is smaller than Cohen’s (1992) definition of a “small” effect) using the region of practical equivalence (ROPE) Bayes factor (BFROPE; Makowski, Ben-Shachar, Chen, & Lüdecke, 2019; Makowski, Ben-Shachar, & Lüdecke, 2019). As tests of QoL according to demographics were exploratory, we did not have a priori hypotheses regarding these effects.

Community involvement

The current study was conducted as a collaboration between autistic and non-autistic researchers. Author Z.J.W., who is autistic, conceptualized and designed the study (including the larger SPARK project from which the current sample was drawn), cleaned and processed the data, performed all statistical analyses, and drafted and revised the article.

Results

Autistic adults in the current study (N = 901, mean age = 37.41 years, age range 18–79 years) were predominantly non-Hispanic White (81.2%), assigned female at birth (63.2%), college educated (58.2% completed a 4-year college degree or more), and diagnosed with autism in adulthood (median age of diagnosis = 23.08 years; 39.3% diagnosed before 18 years of age; 6.7% diagnosed before 4 years of age). A large proportion of individuals in this sample (37.5%) also identified as sexual or gender minorities. Many participants reported lifetime psychiatric or neurodevelopmental diagnoses other than autism (mean 2.5 additional diagnoses, range 0–8), with anxiety disorders (70.8%), mood disorders (56.4% unipolar depression; 11.8% bipolar disorder), and ADHD (42.3%) most frequently endorsed. Additional demographic and clinical information can be found in Table 1.

Table 1.

Sociodemographic and clinical characteristics of the sample.

	Males (n = 332)	Females (n = 569)	Total (N = 901)
Age (years)	38.91 (14.24)	36.54 (12.49)	37.41 (13.21)
Gender identity
Cisgender	305 (91.9%)	146 (82.0%)	782 (86.8%)
Transgender	10 (3.0%)	18 (4.5%)	28 (2.8%)
Non-binary/other	17 (5.1%)	74 (13.0%)	91 (10.1%)
Sexual minority status^a	75 (27.6%)	240 (50.0%)	315 (41.9%)
Race
American Indian/Alaska Native	13 (3.9%)	42 (7.4%)	55 (6.1%)
Asian	13 (3.9%)	23 (4.0%)	36 (4.0%)
Black/African American	14 (4.2%)	22 (3.9%)	36 (4.0%)
Middle Eastern/North African	1 (0.3%)	7 (1.2%)	8 (0.9%)
Native Hawaiian/Pacific Islander	1 (0.3%)	3 (0.5%)	4 (0.4%)
White	301 (90.7%)	525 (92.3%)	826 (91.7%)
Other race	6 (1.8%)	11 (1.9%)	17 (1.9%)
Hispanic/Latino ethnicity	25 (7.2%)	41 (7.5%)	66 (7.3%)
Education
No high school diploma	8 (0.9%)	6 (0.7%)	14 (1.6%)
High school diploma/general education development	52 (5.8%)	67 (7.4%)	119 (13.2%)
Vocational certificate	11 (1.2%)	22 (2.4%)	33 (3.7%)
Some college	70 (7.8%)	141 (15.6%)	211 (23.4%)
Associate degree	31 (3.4%)	59 (6.5%)	90 (10.0%)
Bachelor’s degree	76 (8.4%)	137 (15.2%)	213 (23.6%)
Some graduate/professional school	15 (1.7%)	36 (4.0%)	51 (5.7%)
Graduate/professional degree	69 (20.8%)	101 (17.8%)	170 (18.9%)
Age of autism diagnosis (years)	25.09 (18.81)	25.34 (14.82)	25.25 (16.39)
Received any special education services	165 (50.3%)	214 (37.7%)	379 (42.1%)
Number of other psychiatric diagnoses	2.02 (1.63)	2.78 (1.68)	2.50 (1.70)
PROMIS Anger T-score	52.51 (11.47)	56.57 (10.41)	55.07 (10.98)
PROMIS Anxiety T-score	55.63 (10.84)	60.90 (9.52)	58.96 (10.34)
PROMIS Depression T-score	54.76 (11.28)	57.20 (9.72)	56.30 (10.38)
PROMIS-ED mean score (1–5)	2.35 (0.94)	2.69 (0.87)	2.56 (0.91)
SSS-8 total score (0–32)	9.27 (6.77)	13.15 (6.43)	11.72 (0.85)
Global health (PROMIS Global–10 item 1)
Excellent (5)	45 (13.6%)	28 (4.9%)	73 (8.1%)
Very good (4)	98 (29.5%)	133 (23.4%)	231 (25.6%)
Good (3)	113 (34.0%)	215 (37.8%)	328 (36.4%)
Fair (2)	64 (19.3%)	146 (25.7%)	210 (23.3%)
Poor (1)	12 (3.6%)	47 (8.3%)	59 (6.5%)
Global QoL (PROMIS Global–10 item 2)
Excellent (5)	40 (12.0%)	47 (8.3%)	87 (9.7%)
Very good (4)	111 (33.4%)	140 (24.6%)	251 (27.9%)
Good (3)	105 (31.6%)	207 (36.4%)	312 (34.6%)
Fair (2)	57 (17.2%)	134 (23.6%)	191 (21.2%)
Poor (1)	19 (5.7%)	41 (7.2%)	60 (6.7%)

Note. “Males” and “Females” refer to biological sex as assigned at birth. While the survey question on sex allowed for an “Intersex/Undetermined” response, no participants in the current sample selected this response. Values are presented as M (SD) for continuous variables and n (%) for categorical variables. PROMIS: Patient-Reported Outcomes Measurement Information System; PROMIS-ED: PROMIS Emotional Distress Short Form Composite; SSS-8: Somatic Symptom Scale-8; QoL: Quality of Life.

Sexual orientation data only reported by 752 participants (272 Males, 480 Females).

The PROMIS Global–10 items were subjected to UVA, which indicated that Items 1 and 3 (“global health” and “global physical health”) were highly redundant (wTO = 0.563), a finding supported by a very high polychoric correlation between the two variables (r_poly = 0.925). As these two items correlated so highly, we opted to remove the “global health” item from the analysis entirely, retaining the “global physical health” item due to it being more specific. A subsequent UVA was conducted on the remaining nine items, demonstrating one additional redundant item pair (the two social items, 5 and 6; wTO = 0.305). As these items had a greater degree of unique variance (r_poly = 0.679), we opted to not remove either item, instead altering the measurement model to address the local item dependence. Although an additional “social” factor could feasibly have been added to the model, this factor would have been underidentified without additional constraints on loadings. Therefore, we chose instead to use the sum of Items 5 and 6 as a single nine-point “super-item” in subsequent analyses, thereby eliminating all local dependence.

The eight remaining PROMIS Global–10 items were then fit to a bifactor graded response model with secondary loadings on physical/mental health factors dictated by the results from the UVA community structure (i.e. Items 3, 7, 9, and 10 loading on the physical health factor and Items 4, 5/6, and 8 loading on the mental health factor). This initial model did not fit the data well (C₂(13) = 205.1, CFI_C2 = 0.963, RMSEA_C2 = 0.128, CI_90% [0.113, 0.144], SRMR = 0.057), and examination of the residual correlation matrix indicated unmodeled local dependency between Item 8 (“emotional problems”) and both Items 9 (“fatigue”; r_res = 0.212) and 10 (“pain”; r_res = 0.107). We then re-fit the model, allowing Items 9 and 10 to cross-load onto the mental health factor (Table 2), and the resulting model demonstrated much improved fit (C₂(11) = 56.0, CFI_C2 = 0.990, RMSEA_C2 = 0.071, CI_90% [0.054, 0.089], SRMR = 0.030) and no local dependency (all|r_res|< 0.078). Loadings onto the general factor in this model were moderate to very strong (λ_G = 0.405–0.912; Table 2) and highest for the “general QoL” item, supporting the construct validity of our operationalization of “general QoL.” We therefore decided to retain this structural model for the remainder of the analyses.

Table 2.

PROMIS Global–10 graded response model parameters and equivalent factor loadings for full sample.

Item	Item content	a₁ _(G)	a₂ _(MH)	a₃ _(PH)	d₁	d₂	d₃	d₄	d₅	d₆	d₇	d₈	λ_G	λ_MH	λ_PH
2	In general, how would you say your quality of life is?	3.788	—	—	6.156	2.540	−1.248	−5.467	—	—	—	—	0.912	—	—
3	In general, how would you rate your physical health?	1.969	—	1.244	3.928	1.229	−1.446	−4.200	—	—	—	—	0.682	—	0.431
4	In general, how would you rate your mental health, including your mood and your ability to think?	2.252	1.179	—	3.516	0.305	−2.218	−5.009	—	—	—	—	0.736	0.385	—
7	To what extent are you able to carry out your everyday physical activities, such as walking, climbing stairs, carrying groceries, or moving a chair?	1.822	—	1.820	4.885	2.316	0.445	−1.270	—	—	—	—	0.590	—	0.590
8	How often have you been bothered by emotional problems, such as feeling anxious, depressed, or irritable?	1.995	2.733	—	3.717	0.052	−3.344	−5.525	—	—	—	—	0.527	0.722	—
9	How would you rate your fatigue on average?	1.384	1.098	1.048	3.502	1.383	−1.289	−3.590	—	—	—	—	0.519	0.412	0.393
10	How would you rate your pain on average?	1.119	0.675	1.745	4.872	1.998	0.273	−1.643	—	—	—	—	0.405	0.244	0.631
5/6	5. In general, how would you rate your satisfaction with your social activities and relationships? 6. In general, please rate how well you carry out your usual social activities and roles. (This includes activities at home, at work and in your community, and responsibilities as a parent, child, spouse, employee, friend, and so on.)	2.068	0.512	—	4.048	2.748	1.277	−0.053	−1.483	−2.592	−3.761	−5.234	0.758	0.188	—

Note. Parameters estimated using maximum marginal likelihood based on the QMCEM algorithm. PROMIS: Patient-Reported Outcomes Measurement Information System; a₁–a₃ = slope parameters; G: general factor; MH: mental health factor; PH: physical health factor; d₁–d₈ = item intercept parameters (more positive values indicate “easier” items, meaning they are more likely endorsed at lower latent trait levels); λ = factor loading.

DIF testing indicated that the PROMIS Global–10 items were invariant across most of the tested comparisons, including sex (all p_FDR > 0.636; all|ESSD|< 0.175), race/ethnicity (all p_FDR > 0.167; all|ESSD|< 0.154), SGM status (all p_FDR > 0.528; all|ESSD|< 0.114), education status (all p_FDR > 0.308; all|ESSD|< 0.099), lifetime mood disorder (all p_FDR > 0.343; all|ESSD|< 0.176), lifetime anxiety disorder (all p_FDR > 0.151; all|ESSD|< 0.338), lifetime ADHD (all p_FDR > 0.700; all|ESSD|< 0.155), and age of autism diagnosis (all p_FDR > 0.086; all|ESSD|< 0.173). However, statistically significant DIF was detected between age groups for the combined item 5/6 (χ²(10) = 43.13, p_FDR < 0.001, UIDS = 0.623, ESSD = 0.409). Post hoc Wald tests indicated that both slopes (χ²(2)= 8.97, p = 0.011) and intercepts (χ²(8)= 41.08, p < 0.001) significantly differed between groups, with higher slopes and lower intercept parameters (more item “difficulty,” i.e., being endorsed less frequently at the same latent trait score) in individuals older than 40 years of age. Nevertheless, this DIF did not meet our a priori criterion for practical significance (i.e. |ESSD|> 0.5). Statistically but not practically significant DIF was also found for Item 7 (physical function) between adults with and without prior special education services (χ²(6) = 23.86, p_FDR = 0.001, UIDS = 0.117, ESSD = −0.049). DIF in this item was driven by item intercepts (χ²(4)= 21.36, p < 0.001), with lower item intercepts (higher item difficulty) in participants who had received special education services. As no items on the PROMIS Global–10 demonstrated practically significant DIF in the current sample, we did not modify the structural model any further based on the results of these analyses.

Reliability of the PROMIS Global–10 total score (excluding Item 1) was found to be very high (ω_T = 0.933), with a sizable portion of score variance attributable to the general QoL factor (ω_H = 0.747). The EAP latent trait scores for the general QoL factor also demonstrated high marginal reliability (ρ_xx(G) = 0.901), although reliabilities were sub-optimal for the residualized mental health (ρ_xx(MH) = 0.706) and physical health (ρ_xx(PH) = 0.677) factors. Thus, only the general factor score was analyzed in relation to the criterion variables and exploratory demographic predictors. QoL latent trait scores demonstrated moderate to large correlations with PROMIS-ED scores (r = −0.617, CrI_95% [−0.657, −0.574]), SSS-8 scores (r = −0.543, CrI_95% [−0.589, −0.494]), and the number of lifetime psychiatric diagnoses (r_poly = −0.354, CrI_95% [−0.410, −0.297]). As hypothesized, the partial correlations for PROMIS-ED scores (r_p = −0.435, CrI_95% [−0.489, −0.382]) and SSS-8 scores (r_p = −0.275, CrI_95% [−0.337, −0.214]) indicated that both physical and mental health complaints have incremental validity over the other health domain in predicting general QoL. Finally, although most demographic predictors had non-zero relationships with general QoL, only sex, sexual minority status, gender minority status, and autism diagnosis before 4 years of age predicted QoL to a practically meaningful extent (BF_ROPE > 6.79). Associations of QoL with clinical and demographic variables are presented in Table 3.

Table 3.

Relations between PROMIS Global–10 general QoL trait scores and other clinical and demographic variables.

Predictor	Effect size (95% CrI)	ROPE	BF _ROPE	$P (ROPE \| Data)$
SSS-8 total score	r = −0.543 [−0.590, −0.495]	[−0.1, 0.1]	>1 × 10 ¹⁰	<0.001
PROMIS-ED mean score	r = −0.617 [−0.657, −0.574]	[−0.1, 0.1]	> 1 × 10¹⁰	<0.001
Age	r = −0.045 [−0.110, 0.024]	[−0.1, 0.1]	0.010	0.906
Age of autism diagnosis (continuous)	r = −0.089 [−0.158, −0.024]	[−0.1, 0.1]	0.106	0.619
Number of additional psychiatric diagnoses	r_poly = −0.354 [−0.410, −0.297]	[−0.1, 0.1]	3.75 × 10 ⁸	<0.001
Education	r_poly = 0.097 [0.360, 0.481]	[−0.1, 0.1]	0.082	0.536
Sex	d = −0.314 [−0.452, −0.169]	[−0.2, 0.2]	5.93	0.061
Race/ethnicity (Non-Hispanic White vs other)	d = 0.111 [−0.058, 0.275]	[−0.2, 0.2]	0.065	0.852
Autism diagnosis (< 18 vs ⩾ 18 years)	d = 0.277 [0.138, 0.418]	[−0.2, 0.2]	2.38	0.137
Autism diagnosis (< 4 vs ⩾ 4 years)	d = 0.555 [0.281, 0.820]	[−0.2, 0.2]	63.8	0.006
Sexual minority status	d = −0.332 [−0.484, −0.181]	[−0.2, 0.2]	6.79	0.043
Gender minority status	d = −0.426 [−0.616, −0.221]	[−0.2, 0.2]	23.6	0.013
Received special education (Y/N)	d = 0.266 [0.129, 0.382]	[−0.2, 0.2]	1.78	0.175

Note. Bayes factors indicating substantial evidence against the interval null hypothesis (i.e. d lies within (−0.2, 0.2) or r lies within (−0.1, 0.1)) are presented in bold, whereas Bayes factors indicating substantial evidence for the interval null hypothesis are presented in italics. Effect sizes are estimated using Bayesian methods and are presented along with 95% highest density CrI. PROMIS: Patient-Reported Outcomes Measurement Information System; CrI: credible intervals; ROPE: region of practical equivalence; BF_ROPE = Bayes factor assessing interval null hypothesis that the effect falls within the ROPE; $P (ROPE | Data)$ = proportion of the d/r posterior distribution falling within the ROPE, conditioned on the observed data (i.e. probability that the interval null hypothesis is true); SSS-8: Somatic Symptom Scale-8; ED: Emotional Distress.

As an exploratory post hoc analysis, we additionally investigated whether a subset of three PROMIS Global–10 items (Items 2, 3, and 4, measuring global QoL, global physical health, and global mental health, respectively) could provide reliable estimates of an individual’s standing on the latent “general QoL” factor. In the current sample, EAP latent trait scores based on three rather than nine items demonstrated good marginal reliability (ρ_xx(G) = 0.840) and a very high correlation with EAP scores derived from the full measure (r = 0.981, CrI_95% [0.979, 0.984]). Notably, despite the strong concordance between two general QoL scores, the nine-item EAP score was substantially more reliable at the extremes of the trait distributions (r_xx > 0.7 for θ_G in (−3.21, 3.39)) than the three-item EAP score (r_xx > 0.7 for θ_G in (−2.35, 2.48)).

Discussion

The PROMIS Global–10 is a measure of health-related QoL developed for clinical and research purposes using the psychometrically rigorous methods of the NIH PROMIS initiative (PROMIS Health Organization & PROMIS Cooperative Group, 2013; Reeve et al., 2007). Despite the inclusion of this measure in a newly established QoL battery for autistic adults (Holmes et al., 2020), there have been no prior studies examining its psychometric properties in this population. The current study sought to remedy this lack of research by investigating the latent structure, reliability, and nomological validity of the PROMIS Global–10 in a large sample of autistic adults. Using state-of-the-art psychometric methods, we developed a novel autism-specific general QoL score that can be derived from the PROMIS Global–10, additionally demonstrating that this index has excellent model-data fit, high reliability, good construct validity, and negligible DIF across a wide range of demographic and clinical variables. Despite not containing items that tap all aspects of QoL found to be relevant to the autistic experience (McConachie et al., 2019), the near-unity correlation of PROMIS Global–10 latent trait scores and the item “In general, would you say your quality of life is: [Excellent / Very good / Good / Fair / Poor]” (r = 0.912) clearly demonstrates that the composite score validly captures the construct of subjective QoL as (implicitly) defined by autistic adults themselves. Although the PROMIS Global–10 is a generic health-related QoL measure with no autism-specific content, our findings provide strong preliminary support for the use of the novel “general QoL” latent trait score as a measure of general QoL in independent autistic adults who are capable of self-report. To help autism researchers and clinicians utilize this score in applied settings (e.g. to track global QoL improvement within a system of measurement-based psychiatric care; McFayden et al., 2021), we have created a free online score calculator (available at https://asdmeasures.shinyapps.io/promis_qol/) that can be used to generate QoL latent trait scores and corresponding T-scores based on the item parameters in the current study.

The general QoL factor reported in the current study demonstrated moderate to high loadings from all included PROMIS Global–10 items, particularly those tapping global QoL, social relationships and activities, global mental health, and global physical health. Notably, an EAP-based latent trait score calculated using data from a subset of three items (global QoL, global physical health, and global mental health) demonstrated adequate reliability and correlated at near-unity with its nine-item counterpart. However, as the abbreviated score sacrifices a substantial amount of precision at the tails of the latent trait distribution, we recommend that the three-item composite only be used in situations where it is impractical or burdensome to administer the full PROMIS Global–10.

Within the factor structure of the PROMIS Global–10 items, loading magnitudes indicated that mental health appears to contribute somewhat more to the overall variance in self-reported QoL, a finding that was further supported by stronger partial correlations between QoL and mental health than between QoL and physical health. However, both physical and mental health accounted for unique variance in overall QoL ratings, suggesting that the PROMIS Global–10 scores are likely sensitive to changes in both physical and mental health status. It is also notable that items on the PROMIS Global–10 assessing pain and fatigue appeared to be indicative of both physical and mental health concerns, with significant loadings on both specific factors in the present sample of autistic adults. Though a cross-loading for the fatigue item was noted in the original PROMIS validation study (Hays et al., 2009), the item assessing pain severity did not load meaningfully onto that factor in general population samples. Combined with the high levels of multi-system somatic complaints reported by autistic adults (Williams & Gotham, 2022), including those in our current sample, an elevated loading of pain onto the mental health factor indicates that variance in pain ratings in autistic adults may be disproportionately affected by the presence of so-called Central Sensitivity Syndromes (CSS), such as fibromyalgia and irritable bowel syndrome (Yunus, 2015), the severity of which can be greatly influenced by psychosocial stressors (Alok et al., 2014). Although infrequently assessed, CSS and associated symptoms appear highly prevalent in autistic adults (Grant et al., 2022), and it is possible that these conditions play an outsized role in predicting health-related QoL in this population, particularly in samples, such as ours with relatively good physical health. Notably, these claims are purely speculative at this time, and additional research is warranted to specifically assess whether somatic symptom burden in autism is primarily driven by CSS diagnoses and to determine whether biopsychosocial interventions targeting central sensitization (Adams & Turk, 2018; Gatchel et al., 2007) meaningfully increase QoL for autistic adults.

Although the current study did not seek to comprehensively investigate the predictors of QoL in autistic adults, exploratory analyses also demonstrated small yet practically significant differences in self-reported QoL according to sex assigned at birth (M > F), sexual minority status (sexual minority < heterosexual), and gender minority status (transgender/non-binary <cisgender). The mediators of these demographic effects were not explored in the current study, but we suspect that known group differences in both physical and mental health conditions (DaWalt et al., 2021; Grant et al., 2022; Kassee et al., 2020; Lai et al., 2019; Murphy et al., 2020; Simantov et al., 2022; Strauss et al., 2021; Weir, Allison, & Baron-Cohen, 2021; Weir, Allison, Warrier, & Baron-Cohen, 2021; Williams & Gotham, 2022) contribute substantially. Additional research is needed to further interrogate these relationships and determine whether increased rates of co-occurring mental and physical health conditions/symptoms fully explain, at least in part, the associations between female sex and SGM status and reduced QoL in autistic adults. In addition, individuals reporting a diagnosis of autism in early childhood demonstrated moderately higher QoL ratings than those diagnosed later, replicating previous findings (Kamio et al., 2013). Although correlational in nature, this finding supports the claim that early detection of autism is associated with improved outcomes (Elder et al., 2017), potentially due to early access to autism-specific services or early childhood behavioral interventions. Further research is therefore needed to assess the mechanism whereby early diagnosis increases QoL in this population, specifically determining which services or interventions appear most influential in this regard.

Though this study has provided strong support for the use of a new PROMIS Global–10 score in autistic adults, this recommendation is qualified by multiple limitations. Perhaps most importantly, the sample was predominantly White, female, highly educated, and diagnosed with autism in adulthood, thereby being highly unrepresentative of the broader population of autistic adults. This appears to be a standard occurrence in independent adult samples recruited from SPARK (e.g. Williams et al., 2021; Yerys et al., 2021; Zheng et al., 2021) and other registries of autistic adults (Cassidy et al., 2018; Grove et al., 2021; McConachie et al., 2018; Weir, Allison, Warrier, & Baron-Cohen, 2021). Although the specific reasons for this demographic imbalance in adult autism samples have not been formally explored, we hypothesize that many of the individuals most likely to participate in autism research have been diagnosed with autism relatively recently (i.e. most often as adults). As adult females are among the fastest-growing demographic groups to be newly diagnosed with autism (Posserud et al., 2021; Russell et al., 2021), we speculate that the female predominance in our sample may reflect the high numbers of newly diagnosed autistic women (and to a lesser extent gender minorities) who wish to take part in research as a way of learning more about their new diagnosis. It is also notable that women are on average more likely than men to participate in online surveys (e.g. Smith, 2008), potentially contributing to the gender imbalance in the SPARK cohort and other adult autistic samples. Though the unrepresentative sample may limit the generalizability of normative score percentiles to the broader autistic population, no practically meaningful DIF was found across any demographic category, and thus, it is likely that the item parameters generated from the current study are relatively unbiased by the sample composition (Embretson, 1996). Moreover, while intelligence was not formally assessed, it is unlikely that the current sample of independent adults contained a large portion of individuals with intellectual disability. Thus, it remains unknown how appropriate the PROMIS Global–10 is for use in the “dependent” subset of the autistic population with a greater degree of cognitive impairments (cf. Fombonne et al., 2020), and future work is necessary to investigate such properties as readability and DIF according to intellectual disability status. The measurement model of the PROMIS Global–10 was also modified post hoc to accommodate local misfit, and thus validation of this model in independent samples will further support its use. Other psychometric properties of the PROMIS Global–10 General QoL Score, such as test–retest reliability, DIF across repeated administrations, and sensitivity to change, were also not assessed in the current study and should be investigated in future psychometric work. Notably, measure validation is an ongoing process (Chan, 2014; Williams, 2022), and it is paramount that other researchers interested in QoL in autistic adults seek to independently replicate and extend the current psychometric findings in future studies.

Supplemental Material

sj-docx-1-aut-10.1177_13623613221085364 – Supplemental material for Measuring subjective quality of life in autistic adults with the PROMIS global–10: Psychometric study and development of an autism-specific scoring method

Supplemental material, sj-docx-1-aut-10.1177_13623613221085364 for Measuring subjective quality of life in autistic adults with the PROMIS global–10: Psychometric study and development of an autism-specific scoring method by Zachary J Williams, Carissa J Cascio and Tiffany G Woynaroski in Autism

Supplemental Material

sj-docx-2-aut-10.1177_13623613221085364 – Supplemental material for Measuring subjective quality of life in autistic adults with the PROMIS global–10: Psychometric study and development of an autism-specific scoring method

Supplemental material, sj-docx-2-aut-10.1177_13623613221085364 for Measuring subjective quality of life in autistic adults with the PROMIS global–10: Psychometric study and development of an autism-specific scoring method by Zachary J Williams, Carissa J Cascio and Tiffany G Woynaroski in Autism

Footnotes

Acknowledgements

The authors thank all of the individuals and families enrolled in SPARK, the SPARK clinical sites, and SPARK staff, with special thanks to Brianna Vernoia and Casey White for their continued support. They thank for granting access to demographic and phenotypic data on SFARI Base. Approved researchers can obtain the SPARK population dataset described in this study by applying at https://base.sfari.org after a pre-specified embargo period. They also thank the developers of the PROMIS instruments for their contributions to the psychometric literature on patient-reported outcome measurement. The PROMIS Global–10 measure described in this study is freely available for download at .

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Z.J.W. has received consulting fees from Roche and Autism Speaks; he also serves as a member of the autistic researcher review board of the Autism Intervention Network for Physical Health (AIR-P) and as a community partner for the Autism Care Network Vanderbilt site. T.G.W. is the parent of an autistic adult. C.J.C. reports no competing interests.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the grants from the National Institute on Deafness and Other Communication Disorders (Award No. F30-DC019510); National Institute of General Medical Sciences (Award No. T32-GM007347); National Center for Advancing Translational Sciences (CTSA Award No. UL1-TR002243); Nancy Lurie Marks Family Foundation; and the Family S Endowed Graduate Scholarship in Autism Research at Vanderbilt University. Content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or any other funding organization. No funding body or source of support had a role in the study design, data collection, analysis, or interpretation, decision to publish, or preparation of this manuscript.

ORCID iDs

Zachary J Williams

Carissa J Cascio

Supplemental material

Supplemental material for this article is available online.

References

Adams

L. M.

Turk

D. C.

(2018). Central sensitization and the biopsychosocial approach to understanding pain. Journal of Applied Biobehavioral Research, 23(2), e12125. https://doi.org/10.1111/jabr.12125

Alexander

Parker

Schwetz

(2016). Sexual and gender minority health research at the National Institutes of Health. LGBT Health, 3(1), 7–10. https://doi.org/10.1089/lgbt.2015.0107

Alok

Das

S. K.

Agarwal

G. G.

Tiwari

S. C.

Salwahan

Srivastava

(2014). Problem-focused coping and self-efficacy as correlates of quality of life and severity of fibromyalgia in primary fibromyalgia patients. JCR: Journal of Clinical Rheumatology, 20(6), 314–316. https://doi.org/10.1097/RHU.0000000000000130

Ayres

Parr

J. R.

Rodgers

Mason

Avery

Flynn

(2018). A systematic review of quality of life of adults on the autism spectrum. Autism, 22(7), 774–783. https://doi.org/10.1177/1362361317714988

Benjamini

Hochberg

(1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 57(1), 289–300. https://doi.org/10.2307/2346101

Bock

R. D.

Mislevy

R. J.

(1982). Adaptive EAP estimation of ability in a microcomputer environment. Applied Psychological Measurement, 6(4), 431–444. https://doi.org/10.1177/014662168200600405

Broderick

J. E.

DeWitt

E. M.

Rothrock

Crane

P. K.

Forrest

C. B.

(2013). Advances in patient-reported outcomes: The NIH PROMIS® measures. EGEMS, 1(1), 1015. https://doi.org/10.13063/2327-9214.1015

Brown

H. M.

Dwyer

P. S. R.

Gassner

D. L.

Kapp

S. K.

Ne’eman

Ryan

J. G.

Waisman

Williams

Z. J.

(2022). Introducing the autistic researcher review board of the autism intervention research network on physical health. Pediatrics, 149(s4), Article e2020049437F. https://doi.org/10.1542/peds.2020-049437

Burgess

A. F.

Gutstein

S. E.

(2007). Quality of life for people with autism: Raising the standard for evaluating successful outcomes. Child and Adolescent Mental Health, 12(2), 80–86. https://doi.org/10.1111/j.1475-3588.2006.00432.x

10.

Cai

Monroe

(2014). A new statistic for evaluating item response theory models for ordinal data (CRESST Report 839). National Center for Research on Evaluation, Standards, and Student Testing, University of California. https://eric.ed.gov/?id=ED555726

11.

Cai

Yang

J. S.

Hansen

(2011). Generalized full-information item bifactor analysis. Psychological Methods, 16(3), 221–248. https://doi.org/10.1037/a0023350

12.

Cao

Tay

Liu

(2017). A Monte Carlo study of an iterative Wald test procedure for DIF analysis. Educational and Psychological Measurement, 77(1), 104–118. https://doi.org/10.1177/0013164416637104

13.

Cassidy

Bradley

Shaw

Baron-Cohen

(2018). Risk markers for suicidality in autistic adults. Molecular Autism, 9(1), 42. https://doi.org/10.1186/s13229-018-0226-4

14.

Cella

Riley

Stone

Rothrock

Reeve

Yount

Amtmann

Bode

Buysse

Choi

Cook

DeVellis

DeWalt

Fries

J. F.

Gershon

Hahn

E. A.

Lai

J.-S.

Pilkonis

Revicki

. . . Hays

(2010). The Patient-Reported Outcomes Measurement Information System (PROMIS) developed and tested its first wave of adult self-reported health outcome item banks: 2005–2008. Journal of Clinical Epidemiology, 63(11), 1179–1194. https://doi.org/10.1016/j.jclinepi.2010.04.011

15.

Chalmers

R. P.

(2012). mirt: A multidimensional item response theory package for the R environment. Journal of Statistical Software, 48(6), 1–29. https://doi.org/10.18637/jss.v048.i06

16.

Chan

E. K. H.

(2014). Standards and guidelines for validation practices: Development and evaluation of measurement instruments. In Zumbo

B. D.

Chan

E. K. H.

(Eds.), Validity and validation in social, behavioral, and health sciences (pp. 9–24). Springer. https://doi.org/10.1007/978-3-319-07794-9_2

17.

Christensen

A. P.

Garrido

L. E.

Golino

(2020). Unique variable analysis: A novel approach for detecting redundant variables in multivariate data. https://doi.org/10.31234/osf.io/4kra2

18.

Cohen

(1992). A power primer. Psychological Bulletin, 112(1), 155–159. https://doi.org/10.1037/0033-2909.112.1.155

19.

DaWalt

L. S.

Taylor

J. L.

Movaghar

Hong

Kim

Brilliant

Mailick

M. R.

(2021). Health profiles of adults with autism spectrum disorder: Differences between women and men. Autism Research, 14, 1896–1904. https://doi.org/10.1002/aur.2563

20.

Elder

J. H.

Kreider

C. M.

Brasher

S. N.

Ansell

(2017). Clinical impact of early diagnosis of autism on the prognosis and parent–child relationships. Psychology Research and Behavior Management, 2017(10), 283–292. https://doi.org/10.2147/PRBM.S117499

21.

Embretson

S. E.

(1996). The new rules of measurement. Psychological Assessment, 8(4), 341–349. https://doi.org/10.1037/1040-3590.8.4.341

22.

Felce

Perry

(1995). Quality of life: Its definition and measurement. Research in Developmental Disabilities, 16(1), 51–74. https://doi.org/10.1016/0891-4222(94)00028-8

23.

Feliciano

Daniels

A. M.

Snyder

L. G.

Beaumont

Camba

Esler

Gulsrud

A. G.

Mason

Gutierrez

Nicholson

Paolicelli

A. M.

McKenzie

A. P.

Rachubinski

A. L.

Stephens

A. N.

Simon

A. R.

Stedman

Shocklee

A. D.

Swanson

Finucane

. . . Chung

W. K.

(2018). SPARK: A US cohort of 50,000 families to accelerate autism research. Neuron, 97(3), 488–493. https://doi.org/10.1016/j.neuron.2018.01.015

24.

Fombonne

Coppola

Mastel

O’Roak

B. J.

(2021). Validation of autism diagnosis and clinical data in the SPARK cohort. Journal of Autism and Developmental Disorders. Advance online publication. https://doi.org/10.1007/s10803-021-05218-y

25.

Fombonne

Green Snyder

Daniels

Feliciano

Chung

Abbeduto

Aberbach

Acampado

Ace

A. J.

Albright

Alessandri

Amaral

D. G.

Amatya

Anglo

Annett

R. D.

Arriaga

Ashley

Astrovskaya

Baalman

, . . . The SPARK Consortium. (2020). Psychiatric and medical profiles of autistic adults in the SPARK cohort. Journal of Autism and Developmental Disorders, 50(10), 3679–3698. https://doi.org/10.1007/s10803-020-04414-6

26.

Gatchel

R. J.

Peng

Y. B.

Peters

M. L.

Fuchs

P. N.

Turk

D. C.

(2007). The biopsychosocial approach to chronic pain: Scientific advances and future directions. Psychological Bulletin, 133(4), 581–624. https://doi.org/10.1037/0033-2909.133.4.581

27.

Gibbons

R. D.

Bock

R. D.

Hedeker

Weiss

D. J.

Segawa

Bhaumik

D. K.

Kupfer

D. J.

Frank

Grochocinski

V. J.

Stover

(2007). Full-information item bifactor analysis of graded response data. Applied Psychological Measurement, 31(1), 4–19. https://doi.org/10.1177/0146621606289485

28.

Gierk

Kohlmann

Kroenke

Spangenberg

Zenger

Brähler

Löwe

(2014). The Somatic Symptom Scale–8 (SSS-8): A brief measure of somatic symptom burden. JAMA Internal Medicine, 174(3), 399–407. https://doi.org/10.1001/jamainternmed.2013.12179

29.

Golino

H. F.

Epskamp

(2017). Exploratory graph analysis: A new approach for estimating the number of dimensions in psychological research. PLOS ONE, 12(6), Article e0174035. https://doi.org/10.1371/journal.pone.0174035

30.

Grant

Norton

Weiland

R. F.

Scheeren

A. M.

Begeer

Hoekstra

R. A.

(2022). Autism and chronic ill health: An observational study of symptoms and diagnoses of central sensitivity syndromes in autistic adults. Molecular Autism, 13, Article 7. https://doi.org/10.1186/s13229-022-00486-6

31.

Green

S. B.

Yang

(2009). Reliability of summed item scores using structural equation modeling: An alternative to coefficient alpha. Psychometrika, 74(1), 155–167. https://doi.org/10.1007/s11336-008-9099-3

32.

Grove

Begeer

Scheeren

A. M.

Weiland

R. F.

Hoekstra

R. A.

(2021). Evaluating the latent structure of the non-social domain of autism in autistic adults. Molecular Autism, 12(1), 22. https://doi.org/10.1186/s13229-020-00401-x

33.

Gysi

D. M.

Voigt

Fragoso

T. M.

Almaas

Nowick

(2018). wTO: An R package for computing weighted topological overlap and a consensus network with integrated visualization tool. BMC Bioinformatics, 19(1), 392. https://doi.org/10.1186/s12859-018-2351-7

34.

Hays

R. D.

Bjorner

J. B.

Revicki

D. A.

Spritzer

K. L.

Cella

(2009). Development of physical and mental health summary scores from the patient-reported outcomes measurement information system (PROMIS) global items. Quality of Life Research, 18(7), 873–880. https://doi.org/10.1007/s11136-009-9496-9

35.

Holmes

L. G.

Zampella

C. J.

Clements

McCleery

J. P.

Maddox

B. B.

Parish-Morris

Udhnani

M. D.

Schultz

R. T.

Miller

J. S.

(2020). A lifespan approach to patient-reported outcomes and quality of life for people on the autism spectrum. Autism Research, 13(6), 970–987. https://doi.org/10.1002/aur.2275

36.

Kamio

Inada

Koyama

(2013). A nationwide survey on quality of life and associated factors of adults with high-functioning autism spectrum disorders. Autism, 17(1), 15–26. https://doi.org/10.1177/1362361312436848

37.

Kapp

S. K.

(2018). Social support, well-being, and quality of life among individuals on the autism spectrum. Pediatrics, 141(Suppl. 4), S362–S368. https://doi.org/10.1542/peds.2016-4300N

38.

Karimi

Brazier

(2016). Health, health-related quality of life, and quality of life: What is the difference? PharmacoEconomics, 34(7), 645–649. https://doi.org/10.1007/s40273-016-0389-9

39.

Kassee

Babinski

Tint

Lunsky

Brown

H. K.

Ameis

S. H.

Szatmari

Lai

M.-C.

Einstein

(2020). Physical health of autistic girls and women: A scoping review. Molecular Autism, 11(1), 84. https://doi.org/10.1186/s13229-020-00380-z

40.

Kurz

A. S.

(2019, February 10). Bayesian robust correlations with brms (and why you should love Student’s t). A. Solomon Kurz. https://solomonkurz.netlify.app/post/2019-02-10-bayesian-robust-correlations-with-brms-and-why-you-should-love-student-s-t/

41.

Lai

M.-C.

Kassee

Besney

Bonato

Hull

Mandy

Szatmari

Ameis

S. H.

(2019). Prevalence of co-occurring mental health diagnoses in the autism population: A systematic review and meta-analysis. The Lancet Psychiatry, 6(10), 819–829. https://doi.org/10.1016/s2215-0366(19)30289-5

42.

Lam

G. Y. H.

Sabnis

Migueliz Valcarlos

Wolgemuth

J. R.

(2021). A critical review of academic literature constructing well-being in autistic adults. Autism in Adulthood, 3(1), 61–71. https://doi.org/10.1089/aut.2020.0053

43.

Makowski

Ben-Shachar

M. S.

Chen

S. H. A.

Lüdecke

(2019). Indices of effect existence and significance in the Bayesian framework. Frontiers in Psychology, 10, Article 2767. https://doi.org/10.3389/fpsyg.2019.02767

44.

Makowski

Ben-Shachar

M. S.

Lüdecke

(2019). BayestestR: Describing effects and their uncertainty, existence and significance within the Bayesian framework. The Journal of Open Source Software, 4(40), 1541. https://doi.org/10.21105/joss.01541

45.

McConachie

Mason

Parr

J. R.

Garland

Wilson

Rodgers

(2018). Enhancing the validity of a quality of life measure for autistic people. Journal of Autism and Developmental Disorders, 48(5), 1596–1611. https://doi.org/10.1007/s10803-017-3402-z

46.

McConachie

Wilson

Mason

Garland

Parr

J. R.

Rattazzi

Rodgers

Skevington

Uljarevic

Magiati

(2019). What Is important in measuring quality of life? Reflections by autistic adults in four countries. Autism in Adulthood, 2(1), 4–12. https://doi.org/10.1089/aut.2019.0008

47.

McFayden

T. C.

Gatto

A. J.

Dahiya

A. V.

Antezana

Miyazaki

Cooper

L. D.

(2021). Integrating measurement-based care into treatment for autism spectrum disorder: Insights from a community clinic. Journal of Autism and Developmental Disorders, 51(10), 3651–3661. https://doi.org/10.1007/s10803-020-04824-6

48.

Meade

A. W.

(2010). A taxonomy of effect size measures for the differential functioning of items and scales. The Journal of Applied Psychology, 95(4), 728–743. https://doi.org/10.1037/a0018966

49.

Murphy

Prentice

Walsh

Catmur

Bird

(2020). Autism and transgender identity: Implications for depression and anxiety. Research in Autism Spectrum Disorders, 69, Article 101466. https://doi.org/10.1016/j.rasd.2019.101466

50.

Penfield

R. D.

Camilli

(2006). Differential item functioning and item bias. In Rao

C. R.

Sinharay

(Eds.), Handbook of statistics (Vol. 26, pp. 125–167). Elsevier. https://doi.org/10.1016/S0169-7161(06)26005-X

51.

Pilkonis

P. A.

Choi

S. W.

Reise

S. P.

Stover

A. M.

Riley

W. T.

Cella

, & PROMIS Cooperative Group. (2011). Item banks for measuring emotional distress from the patient-reported outcomes measurement information system (PROMIS®): Depression, anxiety, and anger. Assessment, 18(3), 263–283. https://doi.org/10.1177/1073191111411667

52.

Posserud

M.-B.

Solberg

B. S.

Engeland

Haavik

Klungsøyr

(2021). Male to female ratios in autism spectrum disorders by age, intellectual disability and attention-deficit/hyperactivity disorder. Acta Psychiatrica Scandinavica, 144(6), 635–646. https://doi.org/10.1111/acps.13368

53.

PROMIS Health Organization & PROMIS Cooperative Group. (2013). PROMIS® instrument development and validation scientific standards, Version 2.0. http://www.nihpromis.org/Documents/PROMISStandards_Vers2.0_Final.pdf

54.

R Core Team. (2021). R: A language and environment for statistical computing (4.1.0) [Computer software]. R Foundation for Statistical Computing. https://www.R-project.org/

55.

Reeve

B. B.

Hays

R. D.

Bjorner

J. B.

Cook

K. F.

Crane

P. K.

Teresi

J. A.

Thissen

Revicki

D. A.

Weiss

D. J.

Hambleton

R. K.

Liu

Gershon

Reise

S. P.

Lai

J.-S.

Cella

(2007). Psychometric evaluation and calibration of health-related quality of life item banks: Plans for the Patient-Reported Outcomes Measurement Information System (PROMIS). Medical Care, 45(5), S22–S31.

56.

Revelle

Condon

D. M.

(2019). Reliability from α to ω: A tutorial. Psychological Assessment, 31(12), 1395–1411. https://doi.org/10.1037/pas0000754

57.

Robertson

S. M.

(2009). Neurodiversity, quality of life, and autistic adults: Shifting research and professional focuses onto real-life challenges. Disability Studies Quarterly, 30(1). https://doi.org/10.18061/dsq.v30i1.1069

58.

Rodriguez

Reise

S. P.

Haviland

M. G.

(2016a). Applying bifactor statistical indices in the evaluation of psychological measures. Journal of Personality Assessment, 98(3), 223–237. https://doi.org/10.1080/00223891.2015.1089249

59.

Rodriguez

Reise

S. P.

Haviland

M. G.

(2016b). Evaluating bifactor models: Calculating and interpreting statistical indices. Psychological Methods, 21(2), 137–150. https://doi.org/10.1037/met0000045

60.

Russell

Stapley

Newlove-Delgado

Salmon

White

Warren

Pearson

Ford

(2021). Time trends in autism diagnosis over 20 years: A UK population-based cohort study. Journal of Child Psychology and Psychiatry. Advance online publication. https://doi.org/10.1111/jcpp.13505

61.

Simantov

Pohl

Tsompanidis

Weir

Lombardo

M. V.

Ruigrok

Smith

Allison

Baron-Cohen

Uzefovsky

(2022). Medical symptoms and conditions in autistic women. Autism, 26, 373–388. https://doi.org/10.1177/13623613211022091

62.

Smith

W. G.

(2008). Does gender influence online survey participation? A record-linkage analysis of university faculty online survey response behavior. In Online submission (ED 501717). https://eric.ed.gov/?id=ED501717

63.

Strauss

Cook

Watson

Winter

Whitehouse

Albrecht

Wright Toussaint

Lin

(2021). Mental health difficulties among trans and gender diverse young people with an autism spectrum disorder (ASD): Findings from Trans Pathways. Journal of Psychiatric Research, 137, 360–367. https://doi.org/10.1016/j.jpsychires.2021.03.005

64.

Tay

Meade

A. W.

Cao

(2015). An overview and practical guide to IRT measurement equivalence analysis. Organizational Research Methods, 18(1), 3–46. https://doi.org/10.1177/1094428114553062

65.

Toland

M. D.

Sulis

Giambona

Porcu

Campbell

J. M.

(2017). Introduction to bifactor polytomous item response theory analysis. Journal of School Psychology, 60, 41–63. https://doi.org/10.1016/j.jsp.2016.11.001

66.

Weir

Allison

Baron-Cohen

(2021). Understanding the substance use of autistic adolescents and adults: A mixed-methods approach. The Lancet Psychiatry, 8, 673–685. https://doi.org/10.1016/S2215-0366(21)00160-7

67.

Weir

Allison

Warrier

Baron-Cohen

(2021). Increased prevalence of non-communicable physical health conditions among autistic adults. Autism, 25(3), 681–694. https://doi.org/10.1177/1362361320953652

68.

The WHOQOL Group. (1998). Development of the World Health Organization WHOQOL-BREF quality of life assessment. Psychological Medicine, 28(3), 551–558. https://doi.org/10.1017/s0033291798006667

69.

Williams

Z. J.

(2021). irt_extra: Additional functions to supplement the mirt R package [R]. https://www.researchgate.net/publication/340846037_irt_extra_Additional_Functions_to_Supplement_the_mirt_R_Package

70.

Williams

Z. J.

(2022). Commentary: The construct validity of “camouflaging” in autism: Psychometric considerations and recommendations for future research—Reflection on Lai et al. (2020). Journal of Child Psychology and Psychiatry, 63, 118–121. https://doi.org/10.1111/jcpp.13468

71.

Williams

Z. J.

Everaert

Gotham

K. O.

(2021). Measuring depression in autistic adults: Psychometric validation of the beck depression inventory–II. Assessment, 28(3), 858–876. https://doi.org/10.1177/1073191120952889

72.

Williams

Z. J.

Gotham

K. O.

(2021a). Assessing general and autism-relevant quality of life in autistic adults: A psychometric investigation using item response theory. Autism Research, 14(8), 1633–1644. https://doi.org/10.1002/aur.2519

73.

Williams

Z. J.

Gotham

K. O.

(2021b). Improving the measurement of alexithymia in autistic adults: A psychometric investigation of the 20-item Toronto Alexithymia Scale and generation of a general alexithymia factor score using item response theory. Molecular Autism, 12(1), Article 56. https://doi.org/10.1186/s13229-021-00463-5

74.

Williams

Z. J.

Gotham

K. O.

(2022). Current and lifetime somatic symptom burden among transition-aged autistic young adults. Autism Research. Advance online publication. https://doi.org/10.1002/aur.2671

75.

Yerys

B. E.

McQuaid

G. A.

Lee

N. R.

Wallace

G. L.

(2021). Impacts of co-occurring ADHD symptoms on critical markers of outcome for autistic adults. PsyArXiv. https://doi.org/10.31234/osf.io/zwjp7

76.

Yunus

M. B.

(2015). An update on central sensitivity syndromes and the issues of nosology and psychobiology. Current Rheumatology Reviews, 11(2), 70–85. https://doi.org/10.2174/157339711102150702112236

77.

Zheng

Adams

Taylor

J. L.

Pezzimenti

Bishop

S. L.

(2021). Depression in independent young adults on the autism spectrum: Demographic characteristics, service use, and barriers. Autism, 25(7), 1960–1972. https://doi.org/10.1177/13623613211008276

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

0.06 MB