Skip to main content
. 2015 Apr 16;233:1669–1678. doi: 10.1007/s00213-015-3919-9

Table 2.

The association between inflammatory proteins and depressive symptoms after 5 years

Inflammatory proteins No. of cases B (95 % CI) P value
Log IL-6 (per 1 SD increment)α
 Model 1 656 0.107 (1.32, 5.38) 0.001
 Model 2 650 0.105 (1.16, 5.33) 0.002
 Model 3 650 0.084 (0.48, 4.73) 0.016
Log ACT (per 1 SD increment)α
 Model 1 655 0.057 (−0.67, 10.76) 0.083
 Model 2 650 0.048 (−1.57, 10.04) 0.153
 Model 3 650 0.020 (−4.20, 7.75) 0.561
Log CRP (per 1 SD increment)α
 Model 1 656 0.090 (0.55, 3.19) 0.006
 Model 2 650 0.086 (0.38, 3.16) 0.013

The category “depressive symptoms” includes all subjects who were screened positive. α = levels of inflammatory biomarkers from measurements of 1997–1999: new onset of depressive symptoms at 5-year follow-up in participants with no depressive symptoms at baseline. Subjects on antidepressant medications were excluded from the analysis. Model 1 linear regression analysis adjusted for age and gender. Model 2 as model 1 and additionally adjusted for body mass index (BMI), smoking, physical illness (including stroke, history of cardiovascular disease and diabetes) and mini-mental state examination (MMSE). Model 3 as model 2 and additionally adjusted for acute inflammation. No model 3 was created for CRP as acute inflammation was calculated as CRP >10 mg/mL; therefore, no adjustment for acute inflammation could be done when CRP was used as a predictor

IL-6 interleukin-6, ACT α1-antichymotrypsin, CRP C-reactive protein, SD standard deviation, B standardized beta, CI confidence interval