Differential predictors of well-being versus mental health among parents of preschoolers with autism

Abstract and purpose

• Large body of research shows elevated mental health difficulties among parents of children with autism compared to other parents.
• Well-being is a distinct construct from mental health; predictors of well-being are less well characterized.
• Focus on parents of young, newly diagnosed autistic children (ages 13–48 months) to identify child-, parent-, and family/socioeconomic factors associated with concurrent mental health and well-being.
• Key findings:
  • Parental mental health was predicted by both trait negative emotionality and parent-reported child autism symptoms.
  • Parental well-being was predicted by parent factors alone, specifically trait extraversion and mindfulness.
  • Broader child characteristics and family/socioeconomic context did not add significant predictive value in regression models.
• Practical implication: well-being has unique, modifiable predictors (mindfulness) suggesting potential early supports that directly benefit parents in the autism context.

Lay summary ( Lay abstract )

• In a sample of 136 parents of preschoolers with autism, well-being was reduced and mental health problems were common relative to norms.
• Mental health problems were linked to a combination of child factors (e.g., autism symptom severity) and parent factors (e.g., higher education level in Salomone et al., 2018 note), whereas well-being was predicted mainly by parent factors (excluding child characteristics).
• Mindfulness emerged as a key, modifiable predictor of well-being, alongside extraversion.
• Implication: Mindfulness-based interventions may directly support parental well-being when parenting a young child with autism.

Key concepts and definitions

• Mental health difficulties: Depression, anxiety, and stress levels (as measured by DASS-21).
• Well-being: Self-reported happiness and quality of life (as measured by WEMWBS); not merely the absence of mental health symptoms.
• Distinction between constructs:
  • Mental health and well-being are related but distinct; interventions may differentially target them.
• Modifiable vs non-modifiable predictors:
  • Mindfulness and extraversion are modifiable to some extent (mindfulness more so via training); other factors like child severity are not easily modifiable.

Theoretical and contextual background

• Prior evidence shows higher depression, anxiety, and stress in parents of children with autism and parenting stress in daily life.
• Well-being may predict resilience against adverse outcomes (physical health, academic achievement).
• Shared and distinct predictors for mental health and well-being have been found in population studies and autism-focused studies (e.g., gratitude, hope, self-esteem, coping styles).
• Prior study (Salomone et al., 2018) found well-being linked to child factors in school-age parents, whereas mental health was linked to a combination of child and parent factors.
• The current study aims to replicate and extend: examine whether child-, parent-, and family/context factors concurrently relate to both mental health and well-being in a pre-school sample and identify unique predictors for each construct.
• Hypotheses:
  • Well-being would be lower and mental health problems higher than normative levels.
  • Common and unique predictors would emerge for mental health vs well-being across child-, parent-, and family/context factors.
  • Mental health would be predicted by a combination of child, parent, and family factors; well-being would be more strongly predicted by child characteristics (as per Salomone et al., 2018) but also by parent factors.

Methods

• Design: Cross-sectional analysis of baseline data from a service-linked cohort; data collected at intake prior to intervention.
• Participants: 136 parents of preschoolers with autism; service intake context within a university-based early intervention program (2016–2019).
• Ethics: La Trobe University HREC approval; informed consent obtained.
• Sample characteristics (Table 1):
  • 136 parents; 124 mothers (91%), 11 fathers (8%), 1 other (1%).
  • Education: 30 (25%) ≤Secondary; 88 (75%) tertiary.
  • Household income: 23 low-income (20%); 95 not low income (80%).
  • Primary language: 99 English (72%); 38 Other (28%).
  • Number of children: 58 with 1 child (50%); 57 with >1 child (50%).
  • Children with autism: 84 simplex (73%); 31 multiplex (27%).
• Measures: (detailed below)
• Procedure: Measures collected at service intake; child diagnoses/assessments conducted as part of standard intake; parent questionnaires completed.
• Attrition for outcomes: Mental health data available for all 136 parents; well-being data available for 121.

Measures (key instruments and what they capture)

• Mental health and well-being outcomes
  • DASS-21 ( Depression Anxiety Stress Scales, 21 items): measures mental health difficulties across three subscales; higher scores indicate greater problems. Raw scores were multiplied by two to align with original DASS scoring; used as a total score for analyses after transformation.
  • WEMWBS (Warwick-Edinburgh Mental Well-being Scale): 14 items; higher scores indicate greater well-being.
• Child characteristics
  • ADOS-2 CSS (Autism Diagnostic Observation Schedule, 2nd edition, Calibrated Severity Scores): examiner-rated autism severity, range 0–10; higher means greater severity.
  • SCQ (Social Communication Questionnaire): parent-reported autism severity; range 0–39; higher means greater severity.
  • MSEL (Mullen Scales of Early Learning): developmental quotient (DQ) computed from average age-equivalent scores; higher indicates skills near or above age expectations; standard interpretation: 100 is age-level; below indicates delay.
  • VABS-II ABC (Vineland Adaptive Behavior Scales – 2nd edition, Adaptive Behavior Composite): standardized score; higher indicates greater adaptive ability (M=100, SD=15).
  • CBCL (Child Behavior Checklist, Preschool version): parent-reported child behavioral and emotional difficulties; Internalising and Externalising domain scores; higher scores indicate more problems.
• Parent and family context predictors
  • BFI-2 (Big Five Inventory – 2nd edition): five domains with 15 facets; Extraversion, Agreeableness, Conscientiousness, Negative Emotionality, Open-Mindedness; facet scores retained for analysis.
  • MAAS (Mindful Attention Awareness Scale): mindfulness; higher scores indicate greater mindfulness.
  • Demographics: education level, primary home language, household income, number of children.
• Notes on data handling
  • Some SES variables dichotomised for analyses; robustness checked with full-range analyses; no substantive changes observed.

Statistical approach

• Preliminary analyses: descriptive statistics; normality checks; transformations:
  • Square-root transformation for skewed: DASS-21 total, MSEL DQ, VABS ABC SS, CBCL Internalising.
  • Outliers winsorized to within 2.5 SD.
• Correlations (Table 3): examined correlations between potential predictors and parental mental health (DASS-21 total) and well-being (WEMWBS total).
• Group comparisons (Table 4): t-tests across family/socioeconomic categories for DASS-21 and WEMWBS.
• Regression analyses (Table 5): standard multiple regression to identify unique predictors for each outcome, using factors significantly correlated with that outcome:
  • DASS-21 model predictors pooled from child, parent, and family domains that showed significant bivariate associations.
  • WEMWBS model predictors similarly drawn from significant associations.
• Model interpretation: report R^2, F, and significant predictors with unstandardized (B) and standardized (β) effects, along with p-values.

Results: descriptive and initial associations

• Mental health and well-being levels (Table 2, descriptive, pre-transform):
  • WEMWBS mean = 45.20 (SD = 8.57); range 27–64; sample size n = 121 for well-being.
  • DASS-21 total mean = 31.63 (SD not specified here for transformed data); n = 136 for mental health difficulty total.
  • 38%–50% of parents reported DASS-21 subscale scores in mild–extremely severe ranges across Depression, Anxiety, and Stress subscales (Figure 1).
  • WEMWBS scores were significantly below population norms (norms: M = 51.61, SD = 8.71).
• Correlations (Table 3): key associations with DASS-21 total and WEMWBS total:
  • DASS-21 total: positive associations with Negative Emotionality (r = 0.60, p < 0.001); negative associations with Extraversion (r = -0.14, p = .174 not significant), Agreeableness (r = -0.28, p = .004), MAAS mindfulness (r = -0.42, p < .001); positive association with SCQ Autism Symptoms (r = 0.29, p = .001); CBCL Internalising (r = 0.32, p = .001); CBCL Externalising (r = 0.31, p = .001).
  • WEMWBS total: positive associations with Extraversion (r = 0.45, p < .001); Agreeableness (r = 0.26, p = .006); Conscientiousness (r = 0.23, p = .019); Open-Mindedness (r = 0.26, p = .006); MAAS Mindfulness (r = 0.50, p < .001); negative association with Negative Emotionality (r = -0.53, p < .001); with less robust links to child measures (ADOS CSS: r = 0.04, p = .668; SCQ: r = -0.10, p = .284; CBCL Externalising: r = -0.32, p = .001).
• Group differences (Table 4): SES and language factors showed some differences:
  • Primary language non-English was associated with lower DASS-21 total (better mental health) and higher well-being (p-values: DASS-21 p = .003; WEMWBS p = .044).
  • No robust, consistent differences across education level or household income for either mental health or well-being in the full dataset.
• Child factors: mixed associations with parental mental health; significant correlation of parental mental health with parent-reported SCQ and CBCL externalising/internalising behaviors; little consistent association between child ADOS CSS or MSEL DQ and parental well-being.

Results: regression models (unique predictors)

• Model for DASS-21 total (mental health difficulties) – predictors jointly explain $R^2 = 0.484$
  • Significant unique predictors:
  • BFI-2 Negative Emotionality: eta = 0.483, ext{ SE}_{eta} = 0.282, t = 4.696, p < 0.001
  • SCQ Child Autism Symptoms (parent-reported): $\beta = 0.194, ext{ SE}_{\beta} = 0.033, t = 2.155, p = 0.034$
  • Other predictors (not uniquely significant): BFI-2 Agreeableness, MAAS Mindfulness, CBCL Internalising, CBCL Externalising, Language category, Number of children with autism.
  • Overall model statistics: F(8, 86) = 9.145, p < 0.001.
• Model for WEMWBS total (well-being) – predictors jointly explain $R^2 = 0.529$
  • Significant unique predictors:
  • DASS-21 total (mental health difficulties): $\beta = -0.226, ext{ SE}_{\beta} = 0.354, t = -2.438, p = 0.017$
  • BFI-2 Extraversion: $\beta = 0.317, ext{ SE}_{\beta} = 1.091, t = 3.219, p = 0.002$
  • MAAS Mindfulness: $\beta = 0.256, ext{ SE}_{\beta} = 0.837, t = 2.817, p = 0.007$
  • Other predictors (not uniquely significant): BFI-2 Agreeableness, Conscientiousness, Negative Emotionality, Open-Mindedness, CBCL Externalising, English vs Other Home Language.
  • Overall model statistics: F(9, 99) = 11.234, p < 0.001.
• Note on interpretation of unique effects:
  • Mental health difficulties are best predicted by a combination of a parent’s negative emotionality and the child’s autism symptoms as perceived by the parent.
  • Well-being, beyond its association with mental health, is uniquely predicted by higher parent extraversion and greater mindfulness, even after accounting for mental health level.

Discussion: interpretation and integration with prior work

• Distinct predictors for mental health vs well-being support the notion that these are not simply opposite ends of a single continuum.
• Child factors (e.g., autism severity via SCQ) contribute to parental mental health when considered alongside parent personality traits, but child factors do not uniquely predict well-being when controlling for parent traits and mindfulness.
• Parent personality traits that predict well-being align with broader literature: higher Extraversion (sociability) and higher Mindfulness contribute to greater well-being.
• Mindfulness as a modifiable factor aligns with mindfulness-based interventions showing reductions in depression, anxiety, and stress among parents of children with autism and developmental disabilities, as well as improvements in well-being in other contexts.
• The observed link between well-being and mindfulness suggests direct targets for early support in families with preschoolers with autism.
• The difference from Salomone et al. (2018) may relate to developmental stage (pre-school vs school-age), sample SES, and different parent/child characteristics across samples.
• The current findings imply that early parent-focused supports that cultivate mindfulness and promote sociable, outgoing behaviors could bolster parental well-being in parallel with addressing mental health needs.

Limitations and future directions

• Limitations:
  • Cross-sectional design limits causal inferences; directionality between parent factors and child behavior is bidirectional and complex.
  • Self-report measures introduce potential shared-method variance and measurement bias.
  • Predominance of mothers limits generalizability to fathers/siblings and potential sex differences in associations.
  • Dichotomization of SES variables may reduce sensitivity; however, robustness checks with full ranges showed no substantive changes.
  • Cultural/contextual diversity present (including a sizable South-East Asian subset), which could influence findings; replication in other cultural contexts needed.
• Future research recommendations:
  • Longitudinal designs to examine directional effects and mediation/moderation pathways (e.g., does mindfulness reduce mental health symptoms, thereby improving well-being? or do improvements in well-being feedback to mental health and child outcomes?).
  • Intervention trials targeting mindfulness and mindful behavior in parents of young children with autism to test causal impact on well-being.
  • Exploration of moderation by parent gender, child characteristics, and family context to tailor supports.
  • Investigation of how perceptions of child symptoms (parent-reported) relate to parental adaptation versus objective clinician ratings.

Implications for practice and policy

• Mental health and well-being are interconnected but anchored in different predictor profiles; interventions should be targeted accordingly.
• Mindfulness-based interventions (MBIs) for parents of preschoolers with autism may directly enhance well-being and potentially buffer mental health difficulties.
• Programs should consider supporting parent extraversion-related behaviors (enhancing social engagement) and mindfulness practices early in the diagnostic/service-entry period.
• Clinicians and service providers should assess both parental mental health and well-being separately to identify distinct needs and tailor supports.
• The findings underscore the value of early family-centered supports that acknowledge parent well-being as a critical component of family functioning and child development in autism contexts.

Supplemental and contextual references (selected highlights)

• Measurement tools and validation: DASS-21, WEMWBS, SCQ, ADOS-2 CSS, MSEL, VABS-II, CBCL, MAAS, BFI-2.
• Key related literature: Salomone et al. (2018) on school-age parents; Rayan & Ahmad (2016, 2017, 2018) on mindfulness with autism; Anglim et al. (2020) meta-analysis on personality and well-being; various studies on broader autism phenotype and parent well-being across populations.
• Important constants and concepts used in analyses: population norms for WEMWBS; transformation of DASS-21 scores; norms and interpretation of BFI-2 and MAAS scales.

Quick reference: key results to remember

• Mental health model: R^2 = 0.484; significant unique predictors: eta = 0.483, ext{ p }<0.001 for Negative Emotionality; $\beta = 0.194, ext{ p }=0.034$ for SCQ Autism Symptoms.
• Well-being model: R^2 = 0.529; significant unique predictors: $\beta = -0.226, ext{ p }=0.017$ for DASS-21 total; $\beta = 0.317, ext{ p }=0.002$ for Extraversion; $\beta = 0.256, ext{ p }=0.007$ for MAAS Mindfulness.
• Cross-construct relation: well-being moderately related to mental health (negative association with DASS-21), but also with parent Extraversion and Mindfulness independent of mental health level.
• Child factors showed limited unique predictive value for well-being; the parent-focused traits (mindfulness and extraversion) held stronger predictive relevance for well-being in this preschool-age sample.