The use of psychoactive substances, including
methamphetamine (MA) may cause changes in DNA methylation. The aim of this study was to examine the effects of MA use on long interspersed element-1 (LINE-1) methylation patterns in association with MA-induced
paranoia. This study recruited 123 normal controls and 974 MA users, 302 with and 672 without MA-induced
paranoia. The Semi-Structured Assessment for
Drug Dependence and
Alcoholism was used to assess demographic and
substance use variables. Patterns of LINE-1 methylation were assessed in peripheral blood mononuclear cells and a combined
bisulfite restriction analysis (COBRA) was used to estimate overall LINE-1 methylation (mC) while COBRA classified LINE-alleles into four patterns based on the methylation status of two CpG dinucleotides on each strand from 5' to 3', namely two methylated (mCmC) and two unmethylated (uCuC) CpGs and two types of partially methylated loci (mCuC that is 5'm with 3'u and uCmC that is 5'u with 3'm CpGs). MA users showed higher % mCuC and % mCuC + uCmC levels than controls. Use of
solvents and
opioids, but not cannabis and
alcohol dependence, significantly lowered % uCmC levels, while current smoking significantly increased % uCuC levels. MA-induced
paranoia was strongly associated with changes in LINE-1 partial methylation patterns (lowered % uCmC), heavy MA use, lower age at onset of MA use, and
alcohol dependence. Women who took
contraceptives showed significantly lower LINE-1 % mC and % mCmC and higher % uCuC levels than women without
contraceptive use and men. The results show that MA-induced changes in LINE-1 partial methylation patterns are associated with MA-induced
paranoia and could explain in part the pathophysiology of this type of
psychosis. It is argued that MA-induced neuro-oxidative pathways may have altered LINE-1 partial methylation patterns, which in turn may regulate neuro-oxidative and immune pathways, which may increase risk to develop MA-induced
paranoia.