The review, led by Gunhild Hodt of Innlandet Hospital Trust and the University of Oslo, together with colleagues from the University of Bergen, Haukeland University Hospital, Sahlgrenska University Hospital and the University of Gothenburg, was conducted in accordance with the PRISMA reporting guidelines, the international standard for systematic reviews in health research. The team searched five major databases — MEDLINE, Embase, the Cochrane Library, CINAHL and PsycINFO — with coverage extending up to January 12, 2026. Their search strategy initially surfaced 8,678 potentially eligible articles, a figure that underscores both the breadth of literature on autism and feeding and, by contrast, how few studies actually address formal eating disorder diagnoses in this population.

The inclusion criteria were deliberately strict. To qualify, studies had to be peer reviewed and published in English, include participants younger than 18 years, involve samples of at least three individuals, and apply recognized diagnostic frameworks — either the DSM-IV or DSM-5, the ICD-9 or ICD-10, or the forthcoming ICD-11. Importantly, the researchers excluded studies that described ARFID-like symptoms or phenotypes without a formal diagnosis, a decision that sharpened the review’s focus on clinically confirmed conditions but also revealed how rarely such diagnoses are formally made in autistic children. After screening, just 23 studies survived: ten examined the combination of autism and ARFID, twelve examined autism and pica, and a single study addressed autism and rumination disorder.

Pica, the most heavily studied of the three conditions in this population, involves the persistent ingestion of non-nutritive, non-food substances such as soil, hair, paper, plastic or metal objects. In typically developing children, the behavior is common in infancy but is considered pathological when it persists beyond a developmental window. In autistic children, particularly those with intellectual disability, pica can persist for years and carries serious medical consequences: lead poisoning from ingested paint chips, gastrointestinal obstruction or perforation, dental damage and parasitic infection. The studies included in the review indicated an increased risk of pica among children with autism, and the authors highlight that potentially severe adverse outcomes appear responsive to early detection and intervention — a point with immediate clinical implications, since a behavior that might be dismissed as a quirk of autism can instead signal a diagnosable, treatable and medically dangerous disorder.

Avoidant restrictive food intake disorder, or ARFID, presents a very different clinical picture. Introduced as a formal diagnosis in DSM-5, ARFID describes a restrictive eating pattern driven not by concerns about weight or body shape — the hallmark of anorexia nervosa and bulimia — but by sensory aversion to food textures, tastes and smells, fear of aversive consequences such as choking or vomiting, or a simple lack of interest in eating. The result can be weight loss, nutritional deficiency, dependence on supplements or tube feeding, and profound psychosocial impairment. The ten studies in the review that examined autism and ARFID together support a substantially increased co-occurrence of the two conditions, a finding that aligns with the broader clinical observation that sensory sensitivities and rigid, repetitive behaviors characteristic of autism frequently shape what, how and how much an autistic child will eat.

One of the review’s most important conceptual contributions is its confirmation that the eating disorders co-occurring with autism are fundamentally different in motivation from the classic eating disorders of adolescence. The authors state that the substantial co-occurrence of autism and these eating disorders involves concerns in which weight and body shape are not the primary drivers. This distinction matters enormously for clinical practice. Screening tools and treatment protocols developed for body-image-driven eating disorders may be poorly suited to an autistic child who restricts intake because of texture aversion, or who eats inedible objects because of sensory seeking. Misclassification can delay appropriate care, and the review’s authors emphasize that early detection of these eating disorders may improve clinical outcome.

Rumination disorder, the third condition examined, received by far the least scientific attention. The condition involves the effortless regurgitation of recently eaten food, which may then be re-chewed, re-swallowed or spit out, and it is not explained by gastrointestinal illness or another medical condition. In the entire body of literature screened by the review team, only one study met the criteria for examining rumination disorder in autistic children and adolescents. The authors conclude that rumination disorder may still be underrecognized and understudied in autism, a gap with real consequences: chronic regurgitation can cause dental erosion, esophageal injury, weight loss and malnutrition, and the behavior is frequently mistaken for reflux or vomiting, leading to ineffective medical workups rather than behavioral treatment.

Beyond the headline findings on occurrence, the review delivers a sobering assessment of the methodological quality of the existing evidence. The included studies, the authors report, have small cohorts and lack standardized diagnostic assessment and treatment protocols. In practical terms, this means that estimates of how common pica, ARFID and rumination disorder truly are among autistic children remain imprecise, that comparisons across studies are difficult, and that no evidence-based treatment pathway has been firmly established for any of the three conditions in this population. The abbreviations listed in the article — from the Brief Autism Mealtime Behaviour Inventory to the Food Frequency Questionnaire and the Gastrointestinal Symptom Inventory for ASD — hint at the patchwork of instruments researchers have used, none of which constitutes a gold-standard diagnostic tool for these disorders in autistic youth.

The clinical pathway from recognition to treatment emerged as another weak link. Because pica, ARFID and rumination disorder sit at the intersection of pediatrics, child and adolescent psychiatry, clinical psychology and gastroenterology, autistic children with these conditions may be seen — or missed — in any of several settings. A child with pica may present first to a pediatrician with lead poisoning; a child with ARFID may be referred to an eating disorder clinic built around adolescent anorexia; a child with rumination disorder may undergo repeated gastrointestinal investigations. Without standardized diagnostic criteria applied consistently across these pathways, the review suggests, the same underlying disorder can be interpreted very differently depending on which door the family walks through first.

The research team — which also included Gro Janne Henningsen Wergeland, Ola Skjeldal, Jørn Isaksen, Mads Holten-Andersen, Vesna Bryn and senior author Maj-Britt Posserud — calls implicitly for a new generation of studies: larger cohorts, validated diagnostic instruments adapted for autistic children, and controlled trials of interventions ranging from behavioral treatment for pica and rumination to graded food exposure for ARFID. Published open access in the Journal of Eating Disorders and supported by the Innlandet Hospital Trust, the review arrives at a moment of growing scientific interest in the intersection of autism and eating behavior. Its message to clinicians and researchers alike is clear: eating disorders in autistic children are real, common enough to matter, dangerous when missed, and responsive to early detection — but the field must first build the diagnostic and therapeutic foundations that other areas of eating disorder research have long taken for granted.

Subject of Research: Co-occurring feeding and eating disorders in children and adolescents with autism spectrum disorder

Article Title: Pica, avoidant restrictive food intake disorder and rumination disorder in children and adolescents with autism spectrum disorder: A systematic review

Article References: Hodt, G., Wergeland, G. J. H., Skjeldal, O., Isaksen, J., Holten-Andersen, M., Bryn, V., & Posserud, M.-B. (2026). Pica, avoidant restrictive food intake disorder and rumination disorder in children and adolescents with autism spectrum disorder: A systematic review. Journal of Eating Disorders, 14(1), Article 226. https://doi.org/10.1186/s40337-026-01772-x

Image Credits: AI Generated

DOI: 10.1186/s40337-026-01772-x

Keywords: autism spectrum disorder, pica, ARFID, rumination disorder, eating disorders, systematic review, children, adolescents, feeding behaviour, diagnostic assessment, early intervention, Journal of Eating Disorders

The research, published in BMC Biology by Andrew Y. Morozov of the University of Leicester, Boris W. Berkhout of the University of Amsterdam, and Donald DeAngelis of the University of Miami, combines laboratory experiments on a striking freshwater predator with a new mathematical framework that explicitly represents the internal structure of predator populations. The team’s central claim is provocative: if individual foragers develop stable preferences for particular food resources, then feeding cannot be described using traditional functional responses based on total predator density alone.

The empirical anchor of the study is the assassin snail, Anentome helena, a freshwater gastropod that feeds on other, non-predatory snails. On paper, this species looks like a textbook generalist: across its range it attacks a variety of prey snails, and the population as a whole exploits a broad feeding niche. But when the researchers ran controlled feeding experiments, offering individual assassin snails a menu of prey species including ramshorn snails, trumpet snails, pond snails, and quilted melania snails, a very different picture emerged. Individual predators displayed strong and persistent preferences for particular prey types, and the feeding niche of each individual was far narrower than that of the predator population as a whole.

This pattern, in which a generalist population is effectively composed of cohorts of specialists, is not merely a curiosity of snail behaviour. The authors argue it is likely widespread in nature, arising whenever individual predators learn to handle one prey type efficiently, imprint on a particular foraging strategy, or simply differ in their innate tendencies. In such populations, the aggregate functional response measured at the population level is a statistical artefact, a blend of many narrow individual responses, and using it in models can mask the true dynamics of predation.

To address this, the researchers developed a generic modelling framework in which the predator population is explicitly divided into specialist cohorts, each dedicated to a particular prey species. Crucially, individuals are not locked into their cohorts forever. The framework allows predators to switch between specialist strategies, with the switching governed by the relative profitability of each foraging option. When a particular prey species becomes abundant and profitable, more predators drift toward specialising on it; when that prey declines, individuals gradually abandon the strategy and adopt alternatives. This dynamic reallocation of foraging effort within the predator population is the mathematical heart of the new approach.

The team embedded this structured predator population into a tri-trophic food web model, complete with a basal resource, multiple competing prey species, and the cohort-structured predator. They then compared its behaviour against two classical alternatives: a model in which the generalist predator feeds on all prey according to a multi-prey Holling type III functional response with frequency-dependent food selectivity, and a model in which the prey are each attacked by their own dedicated specialist predator species. The comparison reveals just how much hinges on the assumption of homogeneity within predator populations.

The most striking result concerns coexistence. In the classical framework, a generalist predator feeding on several competing prey tends to destabilise the system or eliminate inferior competitors, because the predator concentrates its attack on whichever prey is currently most abundant, driving boom-and-bust cycles that often end in extinctions. In the new cohort-structured model, by contrast, the internal division of labour within the predator population promotes the coexistence of competing prey species. Because only a fraction of the predator population specialises on any given prey at a time, no single prey species faces the full brunt of predation, and inferior competitors can persist in the shadow of their dominant rivals.

Yet coexistence comes with a caveat that ecologists may find unsettling: the outcome depends on the initial configuration of specialist cohorts within the predator population. In other words, the same community, with the same species and the same environmental conditions, can arrive at different long-term states depending on how the predators’ dietary specialisations were distributed at the start. This sensitivity to initial conditions challenges the classical assumption that ecological communities converge on a predictable equilibrium determined solely by their parameters.

The model also generates a dynamical pattern that, according to the authors, has not been reported in previous predator-prey models: pronounced oscillations in prey densities while the total predator density remains approximately constant. In this regime, the predator population acts as a kind of steady regulatory backdrop, its overall numbers barely changing, while the composition of its specialist cohorts shifts continuously in response to the fluctuating prey. Individual prey species rise and crash in succession, but the predator community as a whole absorbs these swings through internal reallocation rather than demographic change. This decoupling of prey fluctuations from predator abundance is invisible to any model that treats the predator as a homogeneous mass.

Beyond these specific findings, the framework points to broader ecological implications. The authors highlight the potential for long-term ecological transients, extended periods in which community composition keeps shifting for very long times before settling, if it settles at all. Such transients could help explain why some ecosystems appear to be in perpetual flux even under stable environmental conditions. The structured-predator perspective also offers a mechanistic route to the high biodiversity observed in many natural communities, suggesting that the hidden dietary structure within predator populations may be an underappreciated engine of species coexistence.

The study amounts to a critical reappraisal of one of ecology’s oldest modelling conventions. The functional response has served the field well, but the authors argue it rests on an implicit assumption of homogeneity that frequently fails in nature. Their alternative does not discard the functional response entirely; rather, it embeds individual feeding preferences and strategy switching into the population-level description, producing a richer and, they argue, more realistic account of how generalist predators regulate prey. For ecologists modelling pest control, conservation, or food web dynamics, the message is that who eats what within a predator population matters just as much as how much the population eats in total.

The work also illustrates the value of pairing simple experimental systems with abstract theory. The assassin snail, a popular species in the aquarium trade, provided a tractable window into individual-level feeding decisions that would be difficult to observe in large vertebrate predators. By translating those observations into a general mathematical structure, the researchers have produced a tool that can, in principle, be applied wherever individual predators show stable dietary specialisation, from insects parasitising specific host species to fish specialising on particular foraging grounds.

As food web ecology grapples with predicting how communities will respond to environmental change, models that capture within-population structure may prove essential. If generalist predators everywhere are really assemblies of hidden specialists, then the regulation of prey populations, and the biodiversity those populations support, may depend on dynamics that classical theory has never been able to see.

Subject of Research: Modelling the regulation of prey populations by generalist predators composed of individual feeding specialists

Article Title: A novel framework to modelling regulation of prey populations by a generalist predator

Article References: Morozov, A. Y., Berkhout, B. W., & DeAngelis, D. (2026). A novel framework to modelling regulation of prey populations by a generalist predator. BMC Biology. https://doi.org/10.1186/s12915-026-02732-2

Image Credits: AI Generated

DOI: 10.1186/s12915-026-02732-2

Keywords: predator-prey interaction, functional response, generalist predator, food webs, Anentome helena, ecological modelling, species coexistence, ecological transients, feeding behaviour, specialist cohorts, population structure, theoretical ecology