Abstract MP074: Pathway Linking Depression & Inflammation: A 5-year Longitudinal Twin Difference Study

Abstract

Background: Depression and inflammation are risk factors for cardiovascular disease. Depression and inflammation are associated, but the causal direction remains unclear. Genetic background (i.e., genes common to both conditions) may confound this association. We sought to examine the temporal relationship between depression and inflammation, after controlling for genetic confounding, in a 5-year longitudinal monozygotic (MZ) and dizygotic (DZ) twin difference study. Methods: This analysis is based on a longitudinal in-person follow-up of a male middle-aged sample of twin pairs from the Vietnam Era Twin Registry. Inflammation was measured by log-transformed plasma levels of interleukin-6 (IL-6), and depression was evaluated using the Beck Depression Inventory-II (BDI). Inflammation and depression were measured at baseline (visit 1) and after 5 years (visit 2). A longitudinal cross-lagged model was run on signed sibling difference scores in IL-6 and BDI (i.e. twin A - twin B). The path coefficient was examined by the robust maximum likelihood test. We also tested the interaction between zygosity and within-pair difference in IL-6 or BDI score. Results: A total of 166 male twins (83 pairs, mean age of 54) were examined at both visits, including 94 MZ and 72 DZ twins. There was a significant association between sibling differences in IL-6 and BDI score across two visits. At visit 1, the cross-sectional association between sibling difference in IL-6 and BDI score was not significant, but it became significant at visit 2 in MZ twins. The primary focus of our analysis was the longitudinal cross-lagged bidirectional association of BDI and IL-6. There was no significant association between baseline within-pair difference in BDI and their difference in IL-6 at follow-up for both MZ and DZ twins. However, for MZ twins there was a positive association between baseline within-pair difference in IL-6 and their difference in BDI at follow-up, which suggests that members of the MZ twin pair who had higher levels of IL-6 than their siblings at baseline had significantly higher BDI scores at follow-up. This association was only significant in MZ twins (r=0.36, p=0.001) and not in DZ twins (r=0.09, p=0.49), with a significant interaction (p=0.03) between zygosity and within-pair difference in IL-6. Conclusion: Higher inflammation is linked to higher depressive symptoms over time and not vice versa, which suggests that inflammation may be a cause of depression, rather than a consequence. This association is stronger in MZ than DZ twins, thus genetic confounding likely does not play a role.