OBSERVATIONS Association Between Smoking , Hematological Parameters , and Metabolic Syndrome in Japanese Men

C igarette smoking increases the risk for metabolic syndrome (1), and it may also affect hematological parameters (2). Because certain hematological parameters may be associated with metabolic syndrome (3), we have investigated whether the mode of association between smoking and metabolic syndrome varies according to hematological parameters. Among individuals who had undergone a general heath screening test between 1994 and 2003, 27,972 subjects (9,729 never smokers [52.8 10.7 years], 7,242 former smokers [54.8 9.9 years], and 11,001 current smokers [50.4 9.8 years]) answered in full a questionnaire concerning their smoking habits and were enrolled in the current study. Metabolic syndrome was defined as the presence of three or more of the following: 1) fasting glucose 110 mg/dl, 2) blood pressure 130/85 mmHg, 3) triglycerides 150 mg/dl, 4) HDL cholesterol 40 mg/dl, and 5) BMI 25 kg/m. The interquartile cutoff points were 4,700, 5,500, and 6,600 cells/ l for white blood cell (WBC) count and 14.4, 15.1, and 15.7 g/dl for hemoglobin level. Compared with the never smokers, the WBC count and hemoglobin level were significantly higher in the current smokers (5,200 1,200 vs. 6,400 1,800 cells/ l, P 0.0001, and 14.8 1.0 vs. 15.2 1.0 mg/dl, P 0.0001, respectively). After adjusting for age and total cholesterol level, logistic regression analysis showed that current smokers had a higher incidence of metabolic syndrome with an odds ratio (OR) of 1.59 (95% CI 1.47–1.73) compared with never smokers. Compared with the lowest quartile (Q), the incidence of metabolic syndrome was significantly more frequent in the higher quartiles of the WBC count (Q2, OR 1.73 [95% CI 1.54–1.95]; Q3, 2.50 [2.23–2.80]; and Q4, 3.80 [3.41–4.24]) and in those of the hemoglobin level (Q2, 1.65 [1.47–1.86]; Q3, 2.41 [2.15– 2.70]; and Q4, 4.05 [3.63–4.53]). The association between current smoking and metabolic syndrome was found to be statistically significant in lower quartiles of the WBC count (Q1, OR 1.40 [95% CI 1.10–1.79] and Q2, 1.36 [1.13–1.64]) but not in the higher ones (Q3, 1.02 [0.87–1.18] and Q4, 1.04 [1.89–1.21]). By contrast, the association between current smoking and metabolic syndrome was statistically significant regardless of the hemoglobin level (Q1, 1.50 [1.19–1.88]; Q2, 1.53 [1.27–1.84]; Q3, 1.43 [1.21–1.67]; and Q4 1.25 [1.09– 1.43]). These results suggest that the association between smoking and metabolic syndrome may be heavily confounded by certain factors that increase the circulating WBC count.