Table 1 Two-sample Mendelian randomisation results using inverse-variance weighted analyses of genetically predicted cardiorespiratory fitness on type 2 diabetes and related intermediate traits

Outcome	Radial-filtered	n_SNPs	beta	s.e.	p	CochQp	EGGER intercept p-value
Type 2 diabetes^a	No	157	−0.0284	0.0165	0.086	<0.001	0.330
Type 2 diabetes^a	Yes	126	−0.0171	0.0060	0.005	0.69	0.780
fasting insulin	No	156	−0.0073	0.0052	0.160	<0.001	0.543
fasting insulin	Yes	134	−0.0112	0.0032	0.001	0.79	0.562
fasting glucose	No	156	0.0011	0.0057	0.854	<0.001	0.702
fasting glucose	Yes	140	−0.0002	0.0027	0.926	0.98	0.423
2-hr glucose	No	156	−0.0157	0.0072	0.029	<0.001	0.152
2-hr glucose	Yes	141	−0.0099	0.0051	0.052	0.86	0.595
HbA1c	No	155	−0.0011	0.0048	0.821	<0.001	0.426
HbA1c	Yes	137	−0.0009	0.0031	0.775	0.74	0.325
BMI	No	157	−0.0033	0.0038	0.388	0.00	0.778
BMI	Yes	108	−0.0024	0.0015	0.103	0.11	0.470

^a The effect sizes (betas) for type 2 diabetes MR analyses are logOdds.
2-hr glucose = 2-h post-load plasma glucose, HbA1c = glycated haemoglobin, BMI = BMI.
The effect estimates of the optimised genetic instruments on type 2 diabetes were extracted from the meta-analysis of GWAS summary statistics from the DIAMANTE consortium excluding UK Biobank participants (55,005 cases, 400,308 controls)⁵⁵. The effect estimates on FI, FPG, 2hrPG and HbA1c were extracted from meta-analysis summary statistics for these traits among European population acquired from the MAGIC investigators working group⁵⁶. The effect estimates on BMI were obtained from the publicly available meta-analysis results of BMI GWAS from Genetic Investigation of Anthropometric Traits (GIANT) consortium and UK Biobank⁵⁷.

Quick links

Search