Heredity, IVF and environment: the history before birth

Yes. Autism and ADHD commonly involve inherited susceptibility, often through many genetic variants with small effects. Some cases involve a rarer genetic condition; some relevant variants arise newly rather than being inherited. Genetic does not always mean hereditary, and an inherited susceptibility does not guarantee a particular diagnosis or level of difficulty. MedlinePlus Genetics: autism and ADHD

Dyslexia also often runs in families, with differing presentations and severity. We should examine each condition's evidence and the person's combined profile, rather than give every form of neurodivergence one inheritance rule. British Dyslexia Association: family and developmental history

What heritability does—and does not—mean

For scale, a large five-country autism study estimated heritability at about 81%. It estimates how much of the variation in susceptibility within the studied populations is associated with genetic differences between people. It does not mean an autistic parent has an 81% chance of having an autistic child, that 81% of one child's autism was caused by genes, or that the rest can be assigned to identifiable preventable events. Bai and colleagues, 2019

A recurring family pattern deserves concrete description. Longstanding difficulty with change, sensory discomfort, attention, reading or social situations may be useful history even without a previous diagnosis. A family resemblance in temperament, values or coping is not itself a diagnosis. Ask about when a difficulty began, the settings in which it appeared, its consequences and the support or effort that concealed it.

Include both biological parents' family histories, where available. A clinician or genetic counsellor can help decide what warrants investigation and whether any testing would answer a useful question. It is reasonable to raise these questions before pregnancy; an incomplete family history should not prevent appropriate care. ^[3]

A proposed discussion record would include the following. It is a preparation aid, not a diagnostic score or a checklist that can guarantee a child's future development.

• Confirmed diagnoses and any available clinical or genetic reports; separately, observations about relatives who have never been assessed. Record who supplied the information and what remains uncertain.
• Persistent developmental, language, literacy, coordination, attention, sensory or social difficulties, alongside strengths, compensating strategies and functional impact. Include relevant neurological and mental-health history without relabelling it as autism or ADHD.
• Known genetic conditions, intellectual disability, congenital differences, recurrent pregnancy losses or other reproductive concerns that may warrant a genetics discussion. Obtain relatives' records with their agreement; leave unknown details marked unknown.
• Relevant fertility and pregnancy records: the reason for treatment if known, IVF or ICSI, fresh or frozen embryo transfer, any frozen eggs or sperm, donor history where available, medicines and their timing, significant illnesses, gestational age at birth and neonatal care. These are contextual facts, not a list of presumed causes.
• A clinician's written interpretation: which observations matter, which do not currently suggest a specific concern, what additional information would help, and who will review the child and when. Parents can contribute the record; they should not have to perform the clinical interpretation themselves.

This deserves a direct answer, with the comparison made explicit. Studies do not give one consistent estimate that isolates the IVF procedure from the reasons it was needed, parental characteristics and pregnancy outcomes. IVF, ICSI, embryo freezing and multiple pregnancy should not be treated as interchangeable exposures.

In an Ontario cohort of about 1.37 million children, autism diagnoses occurred in approximately 1.6% after unassisted conception and 1.9% after IVF or ICSI during the study's follow-up. After adjustment, diagnoses were more frequent than in the general unassisted group, but not more frequent than among children of parents with infertility who conceived without treatment. These are cohort observations, not an individual's lifetime probabilities or proof of causation. Velez and colleagues, 2023

A US cohort published in June 2026, which relied on caregiver-reported diagnoses, found no significant association for IVF, while reporting associations involving underlying fertility difficulties and some non-IVF treatments. The IVF sample was much smaller than the full cohort. A finding without statistical significance does not prove that every possible effect is absent. Kahn and colleagues, 2026 — abstract

Frozen and fresh embryo transfers

A Taiwanese study published online in September 2026 compared 4,604 children after frozen embryo transfer with 16,877 after fresh transfer. It found no significant difference between those groups for autism, ADHD, intellectual disability or cerebral palsy. Both assisted-conception groups nevertheless showed higher autism and ADHD rates than controls conceived without assistance; the authors could not fully separate treatment from underlying subfertility. Chang and colleagues, 2026 — abstract

ICSI evidence also varies: an earlier Taiwanese cohort reported an autism association, whereas the Ontario and newer Taiwanese comparisons did not identify an additional ICSI association. Differences between populations, treatment indications and study methods matter. Lo and colleagues, 2022

Cell injury and developmental outcomes are different questions

Freezing can damage cells: not all embryos survive freezing and thawing. Embryos are frozen with protective substances, commonly using vitrification to limit ice-crystal damage. That is different from freezing isolated DNA. HFEA: embryo freezing

The human outcome studies reviewed here do not establish that embryo freezing causes autism or ADHD by producing DNA fractures or fragments. Cell injury, changes in gene regulation and DNA-sequence changes are different questions; a laboratory observation would still need to be connected to clinical outcomes. Frozen-versus-fresh comparisons cannot rule out effects shared by both IVF groups. The appropriate conclusion is neither a proven freezing mechanism nor a guarantee of zero risk.

Here, environment includes biological and physical exposures, not only parenting or the home. Evidence differs by exposure, timing and outcome. A finding about autism cannot automatically be applied to dyslexia, ADHD or every developmental difficulty.

Some prenatal exposures have established developmental risks. Valproate during pregnancy is an important medicine-specific example. Pregnancy planning should include review with the prescribing clinician; stopping treatment independently can itself cause harm. European Medicines Agency: valproate and pregnancy

Other research concerns associations involving air pollution, pesticides, maternal diabetes and inflammatory or infectious processes. Such findings deserve examination, but they do not all have the same evidence strength or show that an exposure caused a particular child's condition. NIEHS: autism and environmental research

Prematurity, birth complications and neonatal illness also belong in the developmental history. They may affect follow-up needs. A record of a difficult pregnancy, a medicine, a stressful period or assisted conception is not enough to assign causation. Review the specific exposure, indication, dose and timing with an appropriate clinician rather than treating all medication or all pregnancy events as equivalent.

The present environment still needs action

The present environment still requires action. Adjustments to sensory load, communication, learning demands, predictability and recovery can change what the child can manage. Finding a biological contribution does not remove adults' responsibility to improve those conditions.

Genetic counselling can connect the family and reproductive history to an assessment of which tests, if any, are informative. It can also explain uncertainty, recurrence questions and the meaning of a result for other family members. ^[8]

Tests for a known familial genetic change, carrier screening, chromosome testing and a polygenic risk score answer different questions. A normal result on one test does not establish that a child will not be autistic, have ADHD or experience learning difficulties. Conversely, a genetic finding does not by itself describe the child's future strengths, experience or support needs.

What embryo scores cannot establish

Commercial embryo scores should not be presented as a reliable way to predict or exclude these complex profiles. In December 2025, the American Society for Reproductive Medicine concluded that polygenic embryo screening was not ready for clinical use. ^[9]

The practical request is: tell us what can be learned, what cannot, how a result would change care, and what follow-up is appropriate even when no explanatory result is found. The availability, referral process and funding of services in Greece still require a separate local review.

A family history can justify more attentive observation and timely discussion. It should produce an explicit monitoring question rather than a label placed on a child in advance. The examples below are editorial illustrations, not validated screening rules.

Family reading difficulties → record language and literacy development as opportunities to learn arise → ask the relevant educator or clinician to review persistent difficulties and provide support. Do not wait for a genetic explanation before addressing an observed need.

Recurring sensory or social difficulties in the family → listen to the child's experience and track patterns across settings → discuss persistent concerns with the paediatrician. Familiarity within the family should not make distress invisible; resemblance alone should not be treated as diagnostic.

A documented pregnancy exposure or significant neonatal complication → bring the actual records → ask the clinical team which developmental follow-up is indicated. An exposure record should lead to an informed plan, not a presumed cause.

For each concern, record the observation and its source, the working interpretation, what would strengthen or weaken it, the next action, the responsible professional and a review point. Revise the interpretation as evidence changes. A parent should be able to see whether the information has affected the plan, rather than repeatedly supplying the same history to a new specialist.

This also connects to stigma: a history assembled only from formal diagnoses may miss experiences that relatives did not disclose or never had assessed. Make it possible to report a difficulty without requiring the family to diagnose, accuse or publicly identify a relative.

Evidence note · 26 October 2026. The links beside each claim lead to its source. Inheritance guidance, observational treatment studies, medicine-specific hazards and practical suggestions carry different kinds of evidence. The history record and monitoring examples are editorial synthesis.

The Ontario study was checked in its full text; the June 2026 US study and September 2026 frozen-versus-fresh study are used at abstract level. These recent findings require fuller appraisal, including treatment indications, technique and era, pregnancy outcomes, follow-up length and how diagnoses were recorded. This page does not establish a specific IVF-to-DNA-damage mechanism or an individual child's cause.

Detailed biological and epigenetic mechanisms, exposure-specific causal evidence, differences across developmental conditions, and verified genetics/referral routes and funding in Greece remain open research. Specialist review and final bilingual approval remain pending. Assessment preparation and current support responsibilities remain useful while those questions are investigated.