In 12 fetal cord blood samples, Fryer, et al. found higher LINE-1 methylation levels among samples from higher birthweight infants but comparison groups were generated indirectly by hierarchical clustering [25].
http://plosone.org/article/info:doi/10.1371/journal.pone.0025254#article1.body1.sec3.p2

Fryer at al examined genome wide methylation quantitative interrogation of 27,578 CpG loci associated with 14,496 genes at single-nucleotide resolution across 12 fetal cord blood samples where unbiased analysis of methylation across samples revealed a significant correlation of methylation patterns with birth weight centile. Hence, these epigenetic changes were assessed at a gene specific level. Also, Fryer et al used hierarchial clustering to allow unbiased exploration of data coupled with univariate linear regression analysis.

Importantly, Fryer et al used birth weight centiles calculated on a customized growth chart rather than birth weight as incorrectly cited here. This is important as babies born small for gestational age are not necessarily growth restricted and some of these babies will be small but healthy. However, the use of customized birth weight standards increases identification of fetuses at risk of poor pregnancy outcomes and hence have a better diagnostic accuracy in identifying fetal growth restriction (e.g. Clausson B, Gardosi J, Francis A, Cnattingius S. Perinatal outcome in SGA births defined by customised versus population-based birthweight standards. BJOG 2001 108:830-834 and Figueras F, Gardosi J. Should We Customize Fetal Growth Standards? Fetal Diagn Ther 2009 25:297-303.)