Table E3.

Ninety-five percent confidence intervals for additional parameters (i.e. in addition to the eight model C parameters) included in full model D: ${\sigma }_d^2,\hspace{0.17em}{p}_{d1}$, m_d (see Table 1)

Phenotype	${\mathit{\sigma }}_d^2$	${\mathit{p}}_{d1}$	md
SCZ 2014	[1.33e−4, 1.42e−4]	[4.2e−4, 4.5e−4]	[512, 525]
BIP	—–	—–	—–
CD	[6.96e−2, 8.24e−2]	[3.8e−4, 4.4e−4]	—–
UC	[6.86e−2, 9.20e−2]	[2.7e−5, 3.6e−5]	—–
CAD	[4.51e−3, 6.07e−3]	[3.8e−5, 5.1e−5]	[101, 104]
AD Chr19	—–	—–	—–
AD NoC19	—–	—–	—–
ALS Chr9	—–	—–	—–
Edu	[8.19e−5, 8.75e−5]	[3.4e−3, 3.7e−3]	[442, 448]
IQ 2018	[3.55e−5, 3.66e−5]	[5.8e−2, 6.0e−2]	[563, 568]
Height 2010	—–	—–	—–
Height 2014	—–	—–	—–
Height 2018	—–	—–	—–
HDL	[1.88e−2, 2.56e−2]	[1.8e−5, 2.2e−5]	[63, 69]
LDL	[6.32e−3, 8.29e−3]	[2.0e−6, 2.5e−6]	[339, 352]
BMI GIANT 2015	—–	—–	—–
BMI GIANT-UKB 2018	—–	—–	—–
TC	[2.67e−4, 3.14e−4]	[7.9e−4, 9.3e−4]	—

These confidence intervals are likely underestimates (too narrow), due to linkage disequilibrium and a consequent over-counting of the number of degrees of freedom (independent z-scores).