Multi-locus models of genetic risk of disease

Table 1 Recurrence risk (λ_R) to relatives (of type R) for several common complex genetic diseases ordered by prevalence (K)

Disease	Reference	K	λ _MZ ^a	λ _Sib ^b	λ _OP		^d	^e	^f	^g
Major depression (population cohort)	[27]	0.24	2	1.3		0.32		3.3	1.2	0.34
Age related macular degeneration	[28, 29]	0.12	4.7	2.1		0.50		3.4	1.1	0.64
Myocardial infarction	[30]	0.056	4.6	3.2		0.21		1.6	0.4	0.72
Breast cancer	[31]	0.036	4.1	2.2	1.9	0.12	1.3	2.6	0.8	0.37
Type II diabetes	[32]	0.028	10.4	3.5		0.27		3.8	0.8	0.58
Asthma	[33]	0.019	6.6	3.4		0.11		2.3	0.6	0.49
Rheumatoid arthritis	[34]	0.01	12.2	3.6		0.11		4.3	0.9	0.42
Bipolar disorder	[5]	0.01	60	7	7	0.60	1.0	10	1.2	0.70
Schizophrenia	[3]	0.0085	52.1	8.6	10	0.44	0.8	6.7	0.7	0.76
Type I diabetes	[35]	0.005	79	14		0.39		6.0	0.4	0.85
Multiple sclerosis	[36]	0.001	190	20		0.19	~1	9.9	0.5	0.68
Crohn's disease	[37]	0.001	600	64		0.60		10	0.1	1.00
Ankylosis spondylitis	[6]	0.001	630	82	79	0.63	1.0	7.8	0.1	1.00
Systemic lupus erythematosus	[38]	0.001		29	27		1.1			0.80
	[39, 40]	0.0003	774	65		0.24		12	0.2	0.84

^aThe maximum prevalence for K_MZis 1, so λ_MZ= K_MZ/K is constrained to be ≤1/K. λ_MZwas calculated from probandwise concordance rates K_MZand prevalence rates if λ_MZwas not directly reported. ^bEstimated from either sibling, dizygotic twin or first degree relative risks. ^c broad sense heritability on the risk scale (Equation 1). ^dThis ratio is expected to be 1 in the absence of dominance effects on the risk scale. ^eThis ratio is expected to be 2 under an additive model on the risk scale. ^fThis ratio is expected to be 1 under the unconstrained Risch model. ^gCalculated from the estimates of K and λ_Sib[41, 42], constrained to a maximum of 1.