The Association between Inflammatory Cytokines-Chemokines and Telomere Length.

Separate linear mixed models were used for each cytokine/chemokine. Models controlled for white blood cell count, neutrophil %, lymphocyte %, monocyte %, eosinophil %, current smoking intensity (cigarettes per day), age at baseline blood draw (years), BMI (log kg/m2), and years as a boilermaker (log years). Main effects were log-transformed to achieve a normal distribution. *p-values below the Bonferroni-corrected α-level of 0.007 were considered statistically significant. The estimate for the main effect represents the difference in telomere length per incremental pg/mL increase in the cytokine/chemokine at any follow-up time, controlling for other predictors. The estimate for follow-up time represents the rate of telomeric change; the difference in telomere length per day of follow-up when all other predictors are at zero. The estimate for the main effect x follow-up time interaction represents effect modification of the rate of telomeric change by the cytokine/chemokine; the change in telomere length per pg/mL increase in the cytokine/chemokine over each day of follow-up time, holding other predictors constant. 95% Confidence Intervals (95%CI) are the bounds in which with more independent samplings, the true estimate will fall within this range in 95% of those samplings.