Table 2.
Traits | Pop | k1/k0 | Sample size (case/control) | Data source |
---|---|---|---|---|
ALS | EUR | 20,806/59,804 | AVS [49] | |
T2D | EUR | 67/139 | 62,892/596,424 | PCTG [38] |
BMI | EUR | 91/95 | 339,224 | GIANT [60] |
Fasting glucose | EUR | 35/35 | 133,010 | MAGIC [61] |
Fasting insulin | EUR | 14/14 | 108,557 | MAGIC [61] |
HbA1c | EUR | 36/37 | 123,665 | MAGIC [62] |
ALS | EAS | 1234/2850 | Benyamin [63] | |
T2D | EAS | 34/72 | 36,614/155,150 | JENGER [64] |
Here k1 is the final number of instruments employed in the analysis while k0 is the number of candidate instruments. For T2D, we conducted stringent procedures (Fig. 1) to carefully choose instruments. For the remaining exposures, we performed the standard B1 and B2 selection procedures shown in Fig. 1 to select the final set of instrumental variables
T2D type 2 diabetes, BMI body mass index, ALS amyotrophic lateral sclerosis, HbA1c hemoglobin A1c, GIANT Genetic Investigation of ANthropometric Traits Consortium, MAGIC Meta-Analyses of Glucose and Insulin-related traits Consortium, JENGER Japanese ENcyclopedia of GEnetic associations by Riken, CHN China Health and Nutrition Survey, AVS ALS Variant Server, PCTG Program in Complex Trait Genomics, EUR European, EAS East Asian