Table 1.

Temporal directionality and connectivity significance of selected disease pairs unique to each race cohort

Pop.	Disease 1	Disease 2	P-val	β
EA	Thyroid cancer	Postsurgical hypothyroidism	<6.4E−324	5.25
EA	Lymphosarcoma	Aplastic anemia	<6.4E−324	3.42
EA	Ulcerative colitis	Intestinal obstruction	<6.4E−324	3.27
EA	Toxic diffuse goiter	Postsurgical hypothyroidism	1.3E−153	3.11
EA	Familial hypercholesterolemia	Acute cystitis	<6.4E−324	3.09
AA	Diabetes mellitus, type 2	Diabetic cataract	2.1E−16	5.73
AA	Hyperthyroidism	Toxic diffuse goiter	<6.4E−324	5.10
AA	Chronic ulcer of skin	Osteomyelitis	1.4E−235	4.96
AA	Hypertension	IgA glomerulonephritis	6.4E−75	4.09
AA	HIV disease	Esophageal candidiasis	<6.4E−324	3.87
HL	Diabetes mellitus, type 1	Clostridium difficile colitis	3.3E−73	2.51
HL	Benign essential hypertension	Phobic disorder	5.1E−28	2.25
HL	Coronary artery disease	ARDS	1.7E−61	1.89
HL	Generalized anxiety disorder	Anemia	3.1E−64	1.72
HL	Major depressive disorder	Decubitus ulcer	2.1E−42	1.67

For each population, we determined which temporally related disease pairs had Bonferroni-corrected significant connectivity (P < 1.42 × 10⁻⁰⁶). We present particular disease pairs of interest from among the top-25 associations for each population, ranked by effect size. Effect size, or β, can be interpreted as the odds ratio of disease 2 occurring given disease 1, holding age and sex constant.