Table 3.

Cancer mortality 1961–2015 in G2 men by paternal grandparents’ harvest exposures in SGP: hazard ratios with 95% confidence limits (in brackets) based on Cox regression

	All cancers
Access to food	Model 1		Model 2
Paternal grandmother
Good	1.09	[0.39, 3.00]	1.20	[0.40, 3.62]
Intermediate	1.00	ref.	1.00	ref.
Poor	1.37	[0.59, 3.18]	1.45	[0.63, 3.34]
Paternal grandfather
Good	3.35 a	[1.95, 5.76]	3.44 b	[1.87, 6.34]
Intermediate	1.00	ref.	1.00	ref.
Poor	0.65	[0.20, 2.06]	0.63	[0.20, 1.99]
Observations	3224		3224
Number of deaths	117		117
Cancers not related to smoking
Paternal grandmother
Good	0.00	[0.00]	0.00	[0.00,]
Intermediate	1.00	ref.	1.00	ref.
Poor	1.01	[0.31, 3.28]	1.13	[0.35, 3.65]
Paternal grandfather
Good	3.51 c	[1.77, 6.97]	4.39 d	[2.02, 9.53]
Intermediate	1.00	ref.	1.00	ref.
Poor	0.78	[0.19, 3.26]	0.86	[0.21, 3.52]
Observations	3224		3224
Number of deaths	70		70

Statistically significant estimates (95% CI) in bold type

Model 1: Adjusted for G2 birth year, sibship size and sibling order, father’s harvest exposure in SGP, social class, income and education, and any parental death before age 18

Model 2: + linear trends for grandparents birth years, with confidence limits based on sibling cluster robust standard errors

^aInteraction by gender: p = 0.006

^bInteraction by gender: p = 0.005

^cInteraction by gender: p = 0.013

^dInteraction by gender: p = 0.009