The question
Many people picture mental health conditions as arriving in midlife, if at all. The age-of-onset research says first episodes cluster much earlier. This piece lets a reader enter an age, their own or a child's, and see for twelve conditions how much of the eventual first-onset is already behind people that age and how much is still ahead. It describes populations, not individuals: it is not a risk score and not a diagnosis. Our World in Data published its own charts of this research on 2026-10-19, which is why the question is live now.
The data
Every figure comes from Solmi et al. (2022), a meta-analysis of 192 studies and 708,561 people. For each condition the paper gives the share of people whose onset came before age 14, 18 and 25, the peak onset age, and the median with its interquartile range. We kept those and nothing else. Between them we draw the cumulative curve by straight-line interpolation through (0, 0%), the 25%, 50% and 75% quartile ages, and the 14, 18 and 25 shares; where two anchors share an age we keep the published share. Past the last published point the curve runs straight to 100% at age 70. That tail is our assumption, not data. Each of the 100 dots is one person, placed at the age where the curve passes that person's rank.
The form
A line chart of cumulative onset holds the same numbers, but it asks the reader to read a percentage off a slope. Here a hundred people stand for each condition and light up as the age passes their first episode, so "half by 18" is something you watch happen, and twelve rows of dots can be compared by eye at the age that matters to the reader. The reader's age is the way in, and the dots are ordered so the lit region grows from the left. It borrows the one-dot-one-person move from A Day in the Life of Americans.
What it leaves out
Studies defined onset differently (symptoms, diagnosis or first hospitalisation) and are weighted to high-income countries in North America, Europe and Australia, so the curves may not describe every place. The dots count people who will ever have the condition, not everyone, so they say nothing about how common a condition is. A first episode is not a lifetime sentence and many people recover. Age 70 as the end point is ours. Group rows pool several diagnoses. Nothing here predicts any one person, and anyone worried about their own or someone else's mental health should speak to a doctor or a local health service.
Sources