Table 3.

Mortality difference between family cases and controls and their spouses

	A			B
	N (mean)	HR (95% CI)	p‐value	N (mean)	HR (95% CI)	p‐value
Family‐based case/control group
Control group (ref)	3,714 (0.62)			3,714 (0.50)
Case group	2,282 (0.38)	0.75 (0.69–0.82)	1.75e−10	2,282 (0.30)	0.74 (0.68–0.80)	4.08e−12
Spouses of cases				541 (0.07)	0.94 (0.82–1.07)	3.44e−01
Spouses of controls				937 (0.13)	1.12 (1.00–1.25)	4.07e−02
Birth year	5,996 (1933)	0.99 (0.98–0.99)	1.99e−05	7,474 (1932)	0.98 (0.98–0.99)	1.39e−12
Sex
Males (ref)	3,133 (0.52)			3,364 (0.45)
Females	2,863 (0.48)	0.56 (0.52–0.61)	<1.00e−15	4,110 (0.55)	0.49 (0.46–0.53)	<1.00e−15
Sibship size
Small—1–2 sibs (ref)	1,531 (0.26)
Medium—3–5 sibs	1,770 (0.30)	1.17 (1.04–1.32)	8.51e−03
Large—6–8 sibs	927 (0.15)	1.22 (1.04–1.43)	1.21e−02
Exceptional—9–15 sibs	441 (0.07)	1.36 (1.09–1.68)	5.84e−03
Single child—0 sibs	1,327 (0.22)	1.81 (1.62–2.02)	<1.00e−15

“A” corresponds to the CH curves of panel A of Figure 3. Means represent a mean for a continuous variable and a proportion for a categorical variable. When the p‐value was lower than 1.00e−15, we indicated the p‐value as <1.00*10–15. CI = confidence interval. F3 descendants with relatives who were still alive and had no last moment of observation ≥100 years were removed to assure an equal comparison between cases and controls. In “B”, the spouses of cases and controls are adjusted for the fact that they could not die before the birth of at least their first child (left truncation). We adjusted for this left truncation by entering the spouses of cases and controls in the model based on the first observed death in the groups (cases: 30 years and controls: 25 years). In model A, no adjustment for left truncation was necessary. In both models, we adjusted for right censoring by including a censoring indicator in the Cox model.