Longitudinal association between inflammatory markers and specific symptoms of depression in a prospective birth cohort

BACKGROUND: Low-grade inflammation is associated with depression, but studies of specific symptoms are relatively scarce. Association between inflammatory markers and specific symptoms may provide insights into potential mechanism of inflammation-related depression. Using longitudinal data, we have tested whether childhood serum interleukin 6 (IL-6) and C-reactive protein (CRP) levels are associated with specific depressive symptoms in early adulthood. METHODS: In the ALSPAC birth cohort, serum IL-6 and CRP levels were assessed at age 9 years and 19 depressive symptoms were assessed at age 18 years. We used modified Poisson generalised linear regression with robust error variance to estimate the risk ratio (RR) and 95% confidence interval (95% CI) for each depressive symptom. In addition, we used confirmatory factor analysis to create two continuous latent variables representing somatic/neurovegetative and psychological dimension scores. Structural equation modelling was used to test the associations between IL-6 and these dimension scores. RESULTS: Based on data from 2731 participants, IL-6 was associated with diurnal mood variation, concentration difficulties, fatigue and sleep disturbances. The adjusted RRs for these symptoms at age 18 years for participants in top, compared with bottom, third of IL-6 at age 9 years were 1.75 (95% CI, 1.13–2.69) for diurnal mood variation, 1.50 (95% CI, 1.11–2.02) for concentration difficulties, 1.31 (95% CI, 1.12–1.54) for fatigue, and 1.24 (95% CI, 1.01–1.52) for sleep disturbances. At dimension level, IL-6 was associated with both somatic/neurovegetative (β = 0.059, SE = 0.024, P = 0.013) and psychological (β = 0.056, SE = 0.023, P = 0.016) scores. CONCLUSIONS: Inflammation is associated with specific symptoms of depression. Associations with so-called somatic/neurovegetative symptoms of depression such as fatigue, sleep disturbances and diurnal mood variation indicate that these symptoms could be useful treatment targets and markers of treatment response in clinical trials of anti-inflammatory treatment for depression.