Two people can absorb the exact same radiation dose and face very different lifetime cancer risk, because dose alone doesn't determine harm — age at exposure does too. Infants and young children have tissue that is still rapidly dividing (more cells mid-cycle when radiation strikes), and they have decades of remaining life for a damaged cell to progress into a tumor. Real dosimetry (following the BEIR VII / ICRP approach) does not model this as a separate curve per age — it applies an age-dependent scaling factor on top of the same base dose-response used for adults.
risk(D, age) = base(D) × scale(age)
base(D) = k · D (linear no-threshold, k ≈ 5.5%/Sv)
scale(age) = exp(θ · (age − 30) / 10) (θ ≈ −0.3, BEIR VII-style)
- Claimed dose — the absorbed dose in millisieverts, identical for every age; this is the "exposure" input, not the risk.
- Age at exposure — where on the curve the same dose lands. Age 30 is the reference (scale = 1×); an infant (age 0) sits at roughly 2.5× that baseline risk for the same dose, while a 70-year-old sits at roughly 0.3×.
- Age scaling factor —
exp(θ·(age−30)/10) — the multiplier that turns a raw dose into an age-appropriate risk estimate. It falls monotonically with age because remaining lifespan (time for a radiation-induced mutation to become a cancer) falls too.
- vs. adult — the ratio of the selected age's scaled risk to the age-30 reference risk at the same dose, i.e. how much more (or less) that dose matters at this age.
Real-world relevance: this is why pediatric CT protocols use reduced dose settings, why radiation workers have age-based exposure guidance, and why regulatory dose limits for the public are set using age-averaged risk rather than a single adult-only coefficient.