Autism - Understanding Developmental trajectories, and Inequalities in diagnosis and Treatment in adults aged 50 and over (AUDIT-50)

Background/aims: About 240,000 UK adults aged 50-years and over (≥50) have an Autism Spectrum Disorder (hereafter called autism*). The sparse evidence suggests they have poor outcomes and lack support. Improvements in care and wellbeing are hindered by limited understanding of needs. Our three-year project addresses this.

It aims to: 1. provide the first estimate of UK autism diagnostic coverage among ≥50s, elucidating inequalities in diagnostic provision; 2. describe the mental and physical health needs of ≥50s diagnosed with autism, and their receipt of healthcare, identifying any autism-related inequalities in provision; 3. capture the experience of ≥50s with high autistic traits (including those with no diagnosis); investigating how high levels of autistic traits relate to cognitive and mental health in ageing, and mechanisms underpinning these relationships; 4. understand older autistic* people’s views on their healthcare needs, facilitators and barriers to diagnosis and support, and how to overcome these barriers.

Methods: Aims 1-2 will utilise The Health Improvement Network (THIN), a longitudinal, UK primary care database (n=~11,500,000; >12,000 ≥50s diagnosed with autism).

Aim 1: We will (i) calculate autism diagnosis prevalence in ≥50s who have records in THIN; (ii) compare this with expected true prevalence (modelled from epidemiological literature), in order to estimate the proportion of autistic people aged ≥50 who are undiagnosed; (iii) compare proportion undiagnosed across gender, socioeconomic status, and ethnic group to elucidate any inequalities in the likelihood of people from marginalised groups receiving a diagnosis.

Aim 2: We will (i) describe prevalence of mental and physical health comorbidities in ≥50s diagnosed with autism in THIN; (ii) test the hypothesis that prevalence exceeds that in controls matched on demographics and local geography; (iii) use established measures of receipt of healthcare to test the hypothesis that ≥50s diagnosed with autism have fewer health assessments and medication reviews, but more psychotropic prescriptions than matched controls.

Aim 3: We will add autism measurement (Broad Autism Phenotype-Questionnaire, Autism Quotient-10) to PROTECT (longitudinal cohort study of 20,700 ≥50s; estimated 1.4% autistic) and (i) use multivariate approaches to explore whether, in ≥50s, higher levels of autistic traits predict poorer subsequent cognition (e.g. working memory) and mental health (e.g. anxiety); (ii) investigate putative processes (e.g. sleep disturbance, loneliness) underlying these relationships.

Aim 4: We will: interview ~20 purposively-sampled autistic older people and ~5 informal caregivers; analyse transcribed data using Framework Analysis; and use findings to contextualise quantitative outputs. *Terms preferred by the autism-community (see PPI)