Circulating Serotonin Levels in COPD Patients

A Pilot Study

Pietro Pirina; Elisabetta Zinellu; Panagiotis Paliogiannis; Alessandro G. Fois; Viviana Marras; Salvatore Sotgia; Ciriaco Carru; Angelo Zinellu


BMC Pulm Med. 2018;18(167) 

In This Article


Table 1 reports the demographic and clinical characteristics in controls and COPD patients. As expected, COPD patients showed a reduced FEV1 and FEV1/FVC ratio. There were no between-group differences in smoking status or BMI. Serotonin blood levels were significantly higher in COPD patients (median 0.81 μmol/L, IQR: 0.61–4.02) than in controls (median 0.65 μmol/L, IQR: 0.53–1.39), p = 0.02 (Figure 1). As previously reported TBARS concentrations significantly increased, and PSH concentrations significantly decreased, according to COPD presence.[18] However, no significant correlations were observed between serotonin blood levels and oxidative stress indices. As reported in Figure 2, Spearman's correlations in the whole study population indicated that serotonin blood values are inversely associated with FEV1 (rho = − 0.25, p = 0.023) and FVC (rho = − 0.26, p = 0.017). Table 2 summarizes the results of the univariate logistic regression analysis, which evidenced that serotonin levels were independently associated with presence of COPD (crude OR = 7.29, 95% CI: 1.296–41.05, p = 0.003). This association remained significant also after adjusting for age, gender, BMI, smoking status, and oxidative stress indices (OR 21.92, 95% CI 2.02–237.83; p = 0.011).

Figure 1.

Blood levels of serotonin in healthy subjects (n = 43) and in the totality of COPD patients (n = 43). The central horizontal line on each box represents the median, the ends of the boxes are 25 and 75 percentiles and the error bars 5 and 95%. P-values derived from Student's t-test

Figure 2.

Correlation between FEV1 and FVC with blood serotonin in COPD patients

ROC curve analysis was performed to evaluate the sensitivity, specificity, and diagnostic accuracy of serum serotonin levels alone, or in combination with PSH and TBARS, in distinguishing COPD from healthy subjects (Figure 3 and Table 3). Serotonin alone, with a cut-off of 0.78 μmol/L discriminated COPD from controls with 53.5% sensitivity and 74.4% specificity (AUC = 0.647, 95% CI 0.537–0.747, p = 0.014). Serotonin in combination with PSH and TBARS produced the best result, with an AUC of 0.830 (95% CI 0.733–0.902, p < 0.0001), sensitivity 76.7% and specificity 74.4%. Pairwise comparison of ROC curves indicated that the combination of serotonin, PSH and TBARS yield a significant increase in AUC (+ 0.183, p = 0.0035) compared to AUC obtained with serotonin alone.

Figure 3.

The area under receiver operating characteristic curves of serotonin