Table 2.

Linear excess relative risk and effect modification estimates for all solid cancer mortality and incidence: LSS, 1958–2009

	ERR/Gy				Age at exposure, % change per decade	Attained age, power
	Sex-averaged	Males	Females	F:M ratio		Males	Females
	95% CI	95% CI	95% CI	95% CI	95% CI	95% CI	95% CI
Mortalitya	0.44b	0.28	0.60	2.10	−16.76	−2.47	−1.33
	0.35 to 0.54	0.18 to 0.40	0.46 to 0.74	1.41 to 3.33	−27.41 to −5.62	−3.55 to −1.34	−2.08 to −0.54
Incidencea	0.46	0.28	0.64	2.31	−20.66	−2.54	−1.37
	0.38 to 0.54	0.19 to 0.38	0.52 to 0.77	1.66 to 3.37	−28.90 to −12.02	−3.39 to −1.68	−1.88 to −0.86

Note. LSS, Life Span Study. ERR/Gy, excess relative risk per 1 Gy. 95% CI, confidence interval.

^aData are for the same individuals (i.e., NIC residents included, individuals with missing doses excluded, those who died or were diagnosed with cancer prior to Jan 1, 1958 excluded). Mortality case series are for entire Japan while incidence case series include cases diagnosed in Hiroshima and Nagasaki prefectures and exclude “autopsy only” cases.

^bAll estimates are from linear ERR model with parametric background, multiplicative adjustment for smoking, independent effect modification by sex, sex-specific attained age, age at exposure, and high dose. The ERR estimates per 1 Gy are for survivors at age 70 following exposure at age 30.