TABLE 5.
OCM marker | Studies, n | GG, n | CC, n | Total, n | SMD ± SE (95% CI of SMD)2 | P2 | I2 (95% CI),3 % | P4 |
Holotranscobalamin | 7 | 256 | 527 | 783 | −0.445 ± 0.116 (−0.673, −0.217) | <0.001 | 48.16 (0.00, 78.10) | 0.07 |
EUR | 5 | 112 | 195 | 307 | −0.524 ± 0.194 (−0.906, −0.141) | 0.007 | 61.02 (0.00, 85.37) | 0.04 |
Vitamin B-12 | 11 | 2583 | 4278 | 6861 | −0.099 ± 0.144 (−0.381, 0.183) | 0.49 | 91.67 (87.09, 94.62) | <0.0001 |
EUR | 8 | 2408 | 3867 | 6275 | 0.103 ± 0.106 (−0.105, 0.311) | 0.33 | 75.78 (51.45, 87.91) | 0.0002 |
Homocysteine | 11 | 2595 | 4308 | 6903 | 0.112 ± 0.0585 (−0.020, 0.240)5 | 0.095 | 36.23 (0.00, 68.66) | 0.11 |
EUR | 8 | 2420 | 3897 | 6317 | 0.070 ± 0.0215 (0.020, 0.120)5 | 0.015 | 0.00 (0.00, 49.59) | 0.73 |
Folates | 5 | 2245 | 3663 | 5908 | −0.003 ± 0.052 (−0.106, 0.099) | 0.948 | 17.67 (0.00, 83.88) | 0.30 |
EUR | 4 | 2214 | 3584 | 5798 | −0.011 ± 0.057 (−0.124, 0.101) | 0.846 | 22.73 (0.00–90.02) | 0.27 |
Methylmalonic acid6 | 4 | 2137 | 3601 | 5738 | 0.255 ± 0.458 (−0.643, 1.153) | 0.58 | 97.70 (96.09, 98.65) | <0.0001 |
RBC folates6 | 4 | 132 | 339 | 471 | 0.230 ± 0.191 (−0.145, 0.605) | 0.23 | 63.23 (0.00, 87.60) | 0.04 |
EUR, European-ancestry subjects; OCM, one-carbon metabolism; RBC, red blood cell; SMD, standard mean difference.
Random-effect model according to the DerSimonian and Laird procedure (22).
I2 is the percentage of the observed total variation across studies that was due to real heterogeneity rather than to chance (24).
Significance for heterogeneity was assessed with the use of the χ2-based Q statistic (24).
When the P value that was associated with the DerSimonian and Laird model was between 0.01 and 0.05, the 95% CI of the pooled effect size and its associated P value were calculated using the Hartung-Knapp-Sidik-Jonkman procedure (23).
No analysis on EUR was performed because of the low number of studies.