Early childhood patterns of picky eating can ripple through development for some

At the University of Oslo, psychologists and collaborators following Norwegian families identified a sizable group of children whose eating patterns centered on avoidant and restrictive intake and whose difficulties stretched across development with some genetic associations.

Children whose avoidant or restrictive eating persisted from age 3 to 8 years showed more developmental problems and higher rates of conditions such as autism, ADHD, and epilepsy. Common genetic variants explained 8% to 16% of variation in avoidant or restrictive food intake phenotypes and overlapped with mental health, cognitive, growth, and gastrointestinal traits.

Feeding challenges beyond picky eating

Many young children push plates away, refuse new foods, or eat only a narrow set of familiar meals. Previous work describes "picky eating" in roughly 10% to 15% of children at different ages in early childhood and about 5.5% when picky eating persists across childhood. Broader eating problems can include low interest in food, sensory sensitivities to taste or texture, or fear of discomfort and choking.

Avoidant/restrictive food intake disorder (ARFID) entered the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders and the 11th revision of the International Classification of Diseases to capture clinically significant avoidant or restricted eating that is not driven by weight or shape concerns.

ARFID involves a limited variety or volume of food with consequences such as nutritional deficiencies or failure to meet energy needs. Prevalence estimates for ARFID and ARFID-like symptoms span from 1.5% to 64% across clinical and community samples, ages, and definitions, with a recent meta-analysis placing nonclinical prevalence around 4.5%.

Studies have linked ARFID symptoms and persistent picky eating to physical complaints, emotional and attention problems, social difficulties, and co-occurring neurodevelopmental and gastrointestinal conditions.

Twin work has suggested high heritability (79%) for ARFID, as well as for related traits such as food fussiness, fruit and vegetable liking, and early appetitive characteristics.

Even so, knowledge about how avoidant and restrictive eating emerges, persists, or resolves across development remains limited, and most existing studies come from North American settings. Few investigations have used genome-wide methods to examine a genetic background.

Norwegian children over time

In the study, "Prevalence, Characteristics, and Genetic Architecture of Avoidant/Restrictive Food Intake Phenotypes," published in JAMA Pediatrics, researchers used data from the Norwegian Mother, Father, and Child Cohort Study to estimate the prevalence of avoidant and restrictive food intake phenotypes, to track associated developmental characteristics from infancy to adolescence, and to investigate genetic architecture using genome-wide association analyses.

MoBa, the national pregnancy cohort at the center of the work, recruited approximately 114,500 children, 95,200 mothers, and 75,200 fathers. Genotype data are available for about 80% of participants, drawn from blood collected during pregnancy for parents and at birth for mothers and infants.

For the main analyses, investigators focused on 35,751 children with reported avoidant/restrictive food intake (ARFI) data at both 3 and 8 years of age. That group included 17,515 girls (49%) and 18,236 boys (51%). Questionnaires provided information about ARFI-related behaviors at 3 and 8 years, and national health registries supplied diagnostic codes from primary care and specialist services, including eating disorders, nutritional problems, growth concerns, and several medical conditions.

Defining avoidant and restrictive patterns

Study measures captured ARFI in two nested ways. ARFI-broad identified children with at least one reported symptom of avoidant or restrictive intake at age 3 and/or 8 years.

Items differed by age and included, for example, "does not eat well," "fussy," care devoted to ensuring the child eats enough, low enjoyment of new foods, getting full easily, eating slowly, leaving food on the plate, or eating less when upset or angry. For each item, the highest response option, such as "very true" or "always," served as the threshold indicating that a symptom was present.

Children meeting this symptom threshold at either age 3 or 8 years were then grouped into three ARFI-broad patterns. A persistent pattern meant symptoms at both ages. A transient pattern meant symptoms only at age 3 years. An emergent pattern meant symptoms only at age 8 years.

ARFI-clinical requires ARFI-broad plus at least one clinical significance indicator related to nutrition or eating. Indicators included too little weight gain, body mass index below the 5th percentile at ages 3 or 8 years, delayed physical growth, abnormal weight loss, severe or unspecified protein-energy malnutrition, nutrition problems, vitamin deficiencies, and childhood eating or feeding disorder diagnoses.

Patterns of picky eating

From the 35,751 children with data at both ages, 11,468, or 32.1%, met criteria for ARFI-broad at age 3, age 8, or both, leaving 24,283 children, or 67.9%, with no ARFI-broad symptoms.

Within the ARFI-broad group, 2,129 children, or 6.0% of the entire sample, showed persistent ARFI-broad with symptoms at both 3 and 8 years. Another 6,338 children, or 17.7%, showed transient ARFI-broad at 3 years only, and 3,001 children, or 8.4%, showed emergent ARFI-broad at 8 years only.

Patterns shifted further when clinical indicators were taken into account. Across the full sample, 2,265 children, or 6.3%, met ARFI-clinical criteria. That group included 624 children with persistent ARFI-clinical, representing 1.8% of the total sample, 1,157 with transient ARFI-clinical, representing 3.2%, and 484 with emergent ARFI-clinical, representing 1.4%.

Weight loss and growth problems formed the largest category of clinical indicators within ARFI-broad groups. Among children with persistent ARFI-broad, 22.59% showed weight loss or failure to grow, compared with 14.88% of the transient group and 12.10% of the emergent group.

Nutritional deficiency diagnoses affected 6.48% of the persistent group, 3.66% of the transient group, and 3.67% of the emergent group. Eating disorder or feeding disorder diagnoses associated with clinical significance indicators appeared in 7.14% of children with persistent ARFI-broad, 1.66% of those with transient ARFI-broad, and 2.03% of those with emergent ARFI-broad.

Symptom patterns at each age also carried distinctive features. At age 3, 60.17% of children with persistent ARFI-broad were described as fussy, compared with 37.90% of children with transient ARFI-broad.

Within the persistent group at that age, 28.56% did not eat well, 10.05% were not happy eating food, 26.87% had caregivers who were very careful to ensure enough intake, 36.03% had caregivers who tried to get them to eat even when they said they were not hungry, and 35.60% needed guidance or regulation to eat enough.

At age 8, children with persistent ARFI-broad often displayed patterns that limited exposure to new foods and reduced meal volume. In that group, 28.18% did not enjoy tasting new foods, 20.29% got full easily, 28.89% ate slowly, 4.13% took more than 30 minutes to finish a meal, 14.61% got full before finishing, 30.81% did not enjoy a variety of foods, 38.09% were not interested in tasting new food, 8.92% ate less when upset, 13.95% left food on the plate at the end of a meal, and 10.85% ate less when angry. Children with emergent ARFI-broad at age 8 showed similar behaviors at lower frequencies.

Developmental and diagnostic connections

Children with persistent ARFI-broad did not differ from peers only at mealtimes. Standardized scores across reported measures of eating-related difficulties, language development, motor skills, emotional problems, attention and hyperactivity, restricted and repetitive behaviors, and aggressive or uncooperative behaviors were higher for this group from infancy through age 14 years, indicating more difficulties.

Statistical tests comparing children with persistent ARFI-broad to those with no ARFI-broad showed significant group differences at all ages for reported measures.

By age 14, adolescents with persistent ARFI-broad also reported less prosocial behavior than peers without ARFI-broad. Children with transient or emergent ARFI-broad generally showed intermediate levels of difficulty in descriptive trends, falling between the persistent and no-ARFI groups. Children with persistent ARFI-clinical showed the highest levels of difficulties across developmental domains in descriptive comparisons.

Later diagnoses reinforced the depth of difficulties among children with persistent ARFI-broad. Intellectual disability diagnoses appeared in 0.36% of children with no ARFI-broad and 1.50% of those with persistent ARFI-broad.

Global developmental delay diagnoses affected 0.37% of children with no ARFI-broad and 0.99% of those with persistent ARFI-broad. Autism diagnoses appeared in 1.70% of the no-ARFI group and 5.50% of the persistent ARFI-broad group.

ADHD diagnoses occurred in 5.32% of children with no ARFI-broad and 9.49% of those with persistent ARFI-broad. Epilepsy diagnoses appeared in 1.33% of the no-ARFI group and 2.25% of the persistent ARFI-broad group.

Genetic signals and shared risks

Broad cases for genome-wide association included 10,219 children with at least one ARFI symptom at age 3 years, 4,430 with at least one ARFI symptom at age 8 years, 13,128 with symptoms at age 3 or 8 years, and 1,521 with symptoms at both ages.

Clinical cases included 2,336 children with symptoms at age 3 or 8 years plus a clinical indicator and 452 with both symptoms at ages 3 and 8 years plus a clinical indicator. Controls included 26,107 children with no ARFI-broad at either age and no eating disorder or relevant general medical condition diagnoses.

Genetic differences at common DNA markers explained a modest but real share of which children showed avoidant and restrictive eating patterns. At age 3, common DNA variants together accounted for about 8% of the differences between children in broad ARFI. At age 8, that share rose to roughly 12%.

When eating problems came with clear clinical consequences, common genetic variants explained about 16% of the differences between children. Checks that used stricter ways of defining who counted as a case produced very similar results, with strong statistical support.

Analyses that compared genetic patterns across traits pointed to small to moderate links between ARFI and several other outcomes. Children who carried more genetic risk for avoidant or restrictive eating tended to carry less genetic loading for higher educational attainment and for stronger cognitive performance.

Genetic patterns also tied higher ARFI risk to lower adult and childhood body mass index and to a lower inherited tendency to prefer low-calorie and savory foods. The same variants lend an inherited tendency toward consuming caffeinated sweet drinks.

Shared genetic influences extended into gastrointestinal and neurodevelopmental conditions. Higher genetic risk for ARFI went along with higher genetic risk for inflammatory bowel disease, ulcerative colitis, and celiac disease, as well as ADHD. Genetic overlap between ARFI at ages 3 and 8 years appeared substantial, indicating that many of the same common variants contributed at both ages.

Implications for children

Common variants accounted for 8% to 16% of variation in ARFI phenotypes, leaving substantial unexplained variance and limiting power to detect small-effect variants, especially in persistent ARFI-broad and ARFI-clinical groups.

Patterns in this Norwegian cohort show that avoidant and restrictive eating in childhood can be common and, for a substantial minority, persistent and clinically significant.

Only a few children with persistent avoidant or restrictive intake experienced developmental challenges, higher rates of neurodevelopmental and neurological diagnoses, and connections to growth and gastrointestinal problems.

Genetic signals, including two genome-wide loci and an association with ADCY3, along with broad genetic correlations across psychological, cognitive, anthropometric, food-related, and gastrointestinal traits, mark avoidant and restrictive intake as part of a wider biological and developmental landscape.

Awareness and early recognition of persistent patterns along with patient, broad-based support for affected children emerges as a central need within the picture drawn by the analysis.