Skip to main content

Table 1 Characteristics of the subcutaneous fat thickness data with different types of trimming

From: The effect of measurement error of phenotypes on genome wide association studies

Trait Sample n mean (mm) s.d. s.e.m. C.V. range (mm) r B
CHILLP8A GWAS 888 10.99 4.18 0.14 0.38 1-30 -
P8FAT A GWAS 940 11.89 5.21 0.17 0.44 0-37 0.72
P8MEAN A GWAS 888 11.55 4.34 0.15 0.38 2.5-29.5 0.91/0.94 C
CHILLP8 trim10% 799 10.80 3.22 0.11 0.30 5-18 -
P8FAT trim10% 846 11.63 4.10 0.14 0.35 5-21 0.61 D
CHILLP8 diff1 < 36 800 10.92 4.16 0.15 0.38 2-30 -
P8FAT diff1 < 36 800 11.56 4.62 0.16 0.40 1-30 0.85
CHILLP8 diff1 < 4 564 10.46 4.09 0.17 0.39 2-30 -
P8FAT diff1 < 4 564 10.68 4.23 0.18 0.40 2-30 0.95
CHILLP8 all 8,139 10.75 5.05 0.06 0.47 0-38 -
P8FAT all 8,653 11.63 5.65 0.06 0.49 0-42 0.81
  1. A Subcutaneous fat thickness measured in millimetres aligned with the crest of the 3rd sacral vertebra by two independent groups, with P8MEAN the average between the two values
  2. B Correlation between CHILLP8 and P8FAT for the same method of data trimming
  3. C Correlation between P8MEAN to CHILLP8 and P8FAT respectively
  4. D For samples in common for the two univariate trimmed data sets