Differences in mitochondrial DNA copy number between major depressive disorder, bipolar disorder and schizophrenia: a meta-analysis

Background: Growing body of evidence indicates the mitochondrial involvement in the pathophysiology of psychiatric disorders such as major depressive disorder (MDD), bipolar disorder (BD) and schizophrenia (Sch) [1]. Mitochondrial DNA copy number (mtDNAcn), the ratio of mitochondrial and nuclear DNA, has been proposed as a marker of mitochondrial health [2]. Previous studies have implicated that mtDNAcn alterations may play a role in the pathogenesis of various psychiatric disorders but with conflicting results [3]. The current meta-analysis was conducted to show mtDNAcn differences between individuals with MDD,BD or Sch in comparison with healthy controls (HC).Methods: A literature search was conducted on the electronic databases PubMed and Scopus to identify the relevant studies published until 21 June 2022. The search was performed using the following keywords (mitoch* DNA copy number OR mitoch* copy number OR mtDNA copy) AND (bipolar OR depress* OR psychosis OR schizop* OR affective OR mood) and limited to English language. Screening, data extraction, and risk of bias assessment were independently conducted using Covidence, a software program, and discrepancies were resolved by consensus. The analysis was conducted with R version 3.1.2. and a random-effects model was used to calculate effect sizes.Results: After screening 377 publications, a total of 25 studies that met inclusion criteria were included, comprising 6,369 individuals (n = 1,229 BD with n= 492 BD Type 1; n = 846 MDD; n = 426 Sch; n = 3,798 HC). The included studies reported the following findings:(i) Individuals with BD Type 1 had lower mtDNAcn compared to HC, with a significant and moderate effect size after excluding a possible outlier study (d = -0.57, p < 0.0001), while individuals with BD, regardless of type, had lower mtDNAcn compared to HC but with an insignificant and small effect size (d = -0.27, p =0.07).(ii) Individuals with MDD had higher levels of mtDNAcn compared to HC, but with an insignificant and small effect size (d = 0.35, p = 0.07) and had considerably different heterogeneity in Q test (Q = 8, p < 0.0001).(iii) No significant differences in mtDNAcn were found between individuals with Sch and HC (d = -0.05, p= 0.85). However, moderator analysis revealed significantly lower mtDNAcn in medicated individuals with Sch (Qb= 22.1, Qbp<0.001).Conclusion: Our findings present a significantly lower mtDNAcn in BD type I with a moderate effect size when compared to HC. Despite a trend of increase in MDD, the change was not statistically significant, possibly due to the heterogeneity of studies. Future studies with larger sample sizes and standardized quantification methods may yield more consistent results in MDD. Individuals with Sch presented comparable mtDNAcn with HC, but medicated individuals present lower mtDNAcn. The impacts of medications deserve further investigation. In addition, since the biological significance of changes in mtDNAcn in peripheral blood is still limited, future studies elucidating the correlations between blood and brain mtDNAcn may enhance our understanding.