The animation first compares height predictions for Eric’s identical twin and for Lisa based on her parents’ heights. A shaded range expands four centimetres above and below Eric’s height. A dotted bridge joins the tops of Lisa’s parents’ heads, marks their 175-centimetre midpoint, and extends to Lisa’s prediction. Her shaded range expands nine centimetres above and below that midpoint while the twin’s narrower range remains beside it as a faint outline labeled “Twin”. The animation then shows an illustrative continuous population distribution of ADHD symptoms. Lisa’s parents appear at different symptom levels, joined by a dotted line with their midpoint marked. A guide drops from that midpoint as a broad bracket highlights a wide part of the curve crossing the diagnosis threshold. Three translucent Lisas show different places where she might land: below threshold, near threshold, and in the diagnosed tail. The range is schematic, not a calculated confidence interval or clinical prediction.
ADHD is a highly heritable trait. But heritability is not always as simple as it seems.
Knowing family members’ ADHD traits still tells us surprisingly little about one child’s symptoms.
Children receive a different mix of genes from each parent.
So children can be quite different from their parents.
Twin studies commonly estimate that genetic differences account for around 70–80% of the variation in childhood ADHD traits. This does not mean that 70–80% of one child’s traits are genetic, or that we can predict a child’s traits with 70–80% accuracy (Faraone & Larsson, 2019; Merwood et al., 2013).
Adult height is also highly heritable. Identical twins are usually close in height, yet even both parents’ heights leave a wider range for their child’s adult height. In one recent family study, mid-parental height explained 36–40% of the differences in adult children’s heights (Segal et al., 2009; Zeevi et al., 2024).
In a study of 11,566 eight-year-olds, children’s ADHD-trait scores correlated .24 with their mothers’ scores and .10 with their fathers’ scores. The study estimated that this resemblance was mainly due to genetic transmission. Even so, parents’ scores left most of the differences between children unexplained (Kleppestø et al., 2024).
The height whiskers show rounded prediction ranges: ±4 cm around an identical twin’s height and ±9 cm around an estimate based on both parents. The ADHD distribution shows an illustrative continuous population curve. Its shaded span and three translucent Lisas are deliberately schematic: they show that parents’ ADHD symptom scores leave a broad range of plausible scores for their child, not exact endpoints, a calculated confidence interval, or a clinical prediction. The diagnosis threshold is schematic, and symptom level alone does not determine diagnosis. Exact relationships vary by age, questionnaire and reporter.