Background: Elevated levels of plasma C-reactive protein (CRP) have been associated with many diseases including depression, but it remains unclear whether this association is causal. We tested the hypothesis that CRP is causally associated with depression, and compared these results to those for cancer, ischemic heart disease, chronic obstructive pulmonary disease, and all-cause mortality.
Methods: We performed prospective and instrumental variable analyses using plasma CRP levels and four CRP genotypes on 78,809 randomly selected 20- to 100-year-old men and women from the Danish general population. End points included hospitalization or death with depression and somatic diseases, prescription antidepressant medication use, and all-cause mortality.
Results: A doubling in plasma CRP yielded an observed odds ratio (OR) of 1.28 (95% confidence interval [CI]: 1.23-1.33) for hospitalization or death with depression, whereas for genetically elevated CRP, the causal OR was .79 (95% CI: .51-1.22; observed vs. causal estimate, p = .03). For prescription antidepressant medication use, corresponding ORs were 1.12 (1.11-1.15) and .98 (.83-1.15; p = .08). These results were similar to those for risk of cancer (p = .002), ischemic heart disease (p = 4 × 10(-99)), chronic obstructive pulmonary disease (p = 6 × 10(-86)), and all-cause mortality (p = .001) examined in the same individuals.
Conclusions: Elevated CRP was associated with increased risk of depression in individuals in the general population, but genetically elevated CRP was not. This indicates that CRP per se is not a causal risk factor for depression.
Keywords: All-cause mortality; CRP; depression; genetic variants; inflammation; mendelian randomization; somatic disease.
© 2013 Society of Biological Psychiatry Published by Society of Biological Psychiatry All rights reserved.