Elevated plasma B and betaine levels in women with anorexia nervosa: possible role in illness pathophysiology and epigenetic regulation.

Background: Phenomenology in anorexia nervosa (AN) appears to be subject to epigenetic regulation via DNA methylation. The micronutrients B and betaine contribute directly to DNA methylation and have been shown to be abnormally elevated in blood samples from people with AN.

Methods: We measured plasma B and betaine levels, as well as leukocyte DNA methylation levels, among women with active AN (AN-active group), those in 1-year remission from AN (AN-remitted group), and those who had never experienced an eating disorder (NED group). We compared the groups on micronutrient levels and on the strength of association between micronutrients and methylation.

Results: We included 64 women in the AN-active group, 49 in the AN-remitted group, and 49 in the NED group. Relative to those with NED (B: mean 339.6 [standard deviation (SD) 224.3] μmol/L; betaine: mean 33.74 [SD 17.10] μmol/L), participants with active AN showed high B and betaine (B: mean 571.0 [SD 505.2] μmol/L; betaine: mean 43.73 [SD 22.50] μmol/L); AN-remitted participants had elevated B alone (B: mean 588.2 [SD 379.9] μmol/L; betaine: mean 33.50 [SD 19.20] μmol/L). There were also group-based differences in the strength of association between B and site-specific DNA methylation at genes regulating insulin function, glucose metabolism, cell regulation, and neurotransmitter function. These associations between B and methylation levels were generally stronger among those without an ED than among those with either active or remitted AN.

Limitations: The extent to which plasma nutrient levels provide a meaningful proxy to cellular processes affecting DNA methylation is uncertain and the sample size limits the stability of results. We included only biological females in this investigation.

Conclusion: Elevated B levels in AN resemble elevations reported among people with autoimmune, neoplastic, or other disorders. Such elevations imply that plasma B levels may misrepresent nutritional status among people with AN. Observed associations between levels of B and methylation at certain gene regions have ambiguous importance, but may indicate an influence of nutritional status on epigenetic mechanisms or may be the coincidence of separate processes that independently affect levels of micronutrients and DNA methylation.