Intellectual disability fell across South America. Autism diagnoses barely moved.
A single-author analysis of Global Burden of Disease estimates finds the region's neurodevelopmental picture split in two — and argues the flat autism curve reflects diagnostic capacity, not risk.
An analysis published in Arquivos de Neuro-Psiquiatria on 12 May pulls Global Burden of Disease 2023 estimates for 12 South American countries and tracks three neurodevelopmental conditions from 1990 to 2023 [s1]. The result splits cleanly. Idiopathic developmental intellectual disability declined in every country. Autism spectrum disorder rose slightly everywhere. ADHD went essentially nowhere [s1].
The 2023 numbers
Age-standardised prevalence rates in 2023 ranged from 0.57% in Guyana to 1.06% in Chile for autism; from 1.00% in Colombia to 2.55% in Suriname for ADHD; and from 0.42% in Colombia to 0.54% in Uruguay for idiopathic developmental intellectual disability, abbreviated IDID [s1].
The trend lines matter more than the levels. Autism prevalence rose in all 12 countries, with average annual percent changes running from +0.35% per year in Paraguay to +0.79% per year in Chile [s1]. ADHD showed almost no net movement, ranging from −0.10% per year in Argentina to +0.39% per year in Brazil [s1]. IDID fell consistently everywhere, from −0.79% per year in Bolivia to −0.30% per year in Suriname [s1].
For scale: the analysis cites global GBD 2021 age-standardised prevalence of 1.11% for ADHD, 0.79% for autism and 1.16% for IDID [s1]. Most of the South American autism estimates in the study's country table sit below that global figure; several of the ADHD estimates sit well above it [s1].
The argument about what the flat lines mean
The interesting claim is not that autism prevalence rose but that it rose so little. Compared with high-income settings where much larger relative increases have been documented over comparable periods, the changes here are small [s1]. The author's reading is that this reflects persistent diagnostic constraints — a limited number of trained specialists, uneven access to standardised screening and diagnostic services, and health-system inequities — with diagnostic capacity concentrated in major urban centres and likely under-ascertainment among rural and socioeconomically disadvantaged populations [s1]. On that account, broadened diagnostic criteria and rising awareness had a muted effect in the region because the machinery to act on them is thinly distributed.
The ADHD result gets a parallel treatment. Recent GBD-based analyses report slight global declines; South America instead shows stability [s1]. The analysis suggests diagnostic practices may not have changed uniformly across the region, and that differences between school-based screening and clinic-based evaluation, alongside documented under-recognition in underserved populations, could obscure underlying shifts in risk [s1].
The IDID decline is the one the analysis treats as most likely to be real. It mirrors global reductions reported in GBD 2019 analyses and is described as compatible with sustained improvements in maternal, perinatal and neonatal care — increased skilled birth attendance and institutional delivery, strengthened obstetric and newborn care, and expanded coverage of essential maternal and child health interventions across the region [s1]. The author immediately qualifies this: primary epidemiological studies on IDID remain scarce in South America, which limits direct characterisation of causes and increases reliance on modelled inputs [s1].
The sex gap, and where it is widening
Sex differences were large and persistent for autism and ADHD, and small for IDID [s1]. The male-to-female ratio trajectories suggest a slight widening of the autism gap over the study period, largely stable ADHD disparities, and minimal variability for IDID [s1].
The analysis connects the autism finding to existing evidence that girls may present with subtler or more compensatory social-communication profiles and different behavioural manifestations, which can reduce referral and diagnosis rates [s1]. It concludes that girls and women in the region may still be under-recognised or diagnosed late [s1]. That is an inference about detection drawn from a prevalence ratio, not a direct measurement of missed diagnoses.
What this study cannot tell you
The limitations are stated by the author and are substantial. First, the analysis relies on modelled estimates in settings with limited primary data, and variation in diagnostic infrastructure and service availability may influence the underlying inputs — particularly for autism and ADHD [s1]. This is the central caveat: GBD prevalence in data-sparse regions is partly an output of modelling assumptions, so a flat national curve can reflect flat inputs as much as flat reality.
Second, the study examines prevalence rather than incidence, and did not analyse age-specific trends, which limits inferences about changes in early identification and cohort effects [s1]. Third, joinpoint regression was not applied, so the analysis summarises overall direction and magnitude rather than identifying years in which trends changed [s1].
The gap between modelled regional figures and local primary studies is visible in the paper's own literature summary. Cited primary estimates include ADHD at 5.8% among Brazilian adolescents aged 12 to 14 and 10.5% among Chileans under 19; autism at 0.17% among Venezuelan children aged three to nine and 0.11% among Ecuadorian children aged five to 15; and, in an Argentinian study, intellectual disability of all forms at 0.93% among those under 19, with IDID at 0.66% [s1]. These are heterogeneous in method and population and are not directly comparable to age-standardised national rates — which is precisely the author's point about why regional surveillance is thin.
What it means for readers in the region
The practical reading is about services, not risk. If the IDID decline is driven by better perinatal care while autism and ADHD curves are held flat by diagnostic bottlenecks, then the two halves of the finding call for different responses: continuing what worked in maternal and newborn care, and building assessment capacity outside major cities. The analysis explicitly frames the region's structural barriers — segmented health systems, unequal specialist distribution, substantial out-of-pocket costs, limited coordination between health and education sectors, and a lack of culturally adapted assessment tools — as contributors to under-ascertainment [s1].
What to watch
Whether any South American country builds population-based neurodevelopmental surveillance capable of testing these modelled curves against directly measured prevalence. Until that exists, a flat line on a GBD chart remains ambiguous between two very different explanations.
Sources
- Trends in the age-standardized prevalence of neurodevelopmental disorders in South America, 1990-2023 — Arquivos de Neuro-Psiquiatria, 12 May 2026
Sources
- Trends in the age-standardized prevalence of neurodevelopmental disorders in South America, 1990-2023 — Arquivos de Neuro-Psiquiatria , May 12, 2026
More on
Stimulants beat behaviour therapy for ADHD at 14 months. By 8 years the gap was gone.
The MTA trial randomised 579 children and found medication management clearly superior short-term. Its own long-term follow-up found the treatment groups no longer differed by 6 to 8 years.
New ADHD prescriptions rose 157% in nine years, fastest in women aged 25 to 44
A population study covering all 15 million residents of Ontario found the rate of new ADHD medication prescriptions accelerated sharply after 2019 — and shifted decisively toward adults.
Two prenatal chemical studies, opposite answers, and why both count
One found specific phenanthrene metabolites associated with fetal growth restriction and chased the mechanism. The other tested a 24-metabolite mixture against ADHD in 3,962 children and found nothing.
Athletes with ADHD face higher concussion odds and longer recovery, a review finds
The paper is a literature review, not a new trial — it synthesizes existing research rather than testing anything new. The authors argue return-to-play protocols may need to treat ADHD as a distinct variable.