Secondhand Smoke Exposure and Markers of Inflammation: The Multi-Ethnic Study of Atherosclerosis

Few studies have evaluated the association between secondhand smoke exposure (SHS) and markers of subclinical cardiovascular disease (CVD) in ethnically diverse populations. We assess the impact of SHS on subclinical markers of inflammation (C-reactive protein, interleukin-6 (IL-6), and fibrinogen).

We included 5,032 non-smoking adults aged 45-84 years without prior CVD participating in the Multi-Ethnic Study of Atherosclerosis (MESA) from 2000 to 2002. SHS exposure and smoking status were determined by self-report and urinary cotinine in a subset. A total of 3,035, 682, 428, 480, and 407 participants reported 0, 1, 2-3, 4-11, and 12 or more hours of SHS exposure per week, respectively.  Urinary cotinine, available in 2,983 participants, was detectable in 4.6% of the unexposed participants compared to 36.9% of the participants reporting 12 or more hours of exposure per week.  We used multivariable linear regression models to estimate geometric mean ratios comparing hsCRP≥2, IL-6, and fibrinogen levels by SHS exposure categories compared to unexposed. For dichotomous outcomes (hsCRP≥2), we used multivariable logistic regression.

The median (IQR) of hsCRP was 1.75 (0.80-3.80) mg/L and 2.38 (1.09, 5.04) mg/L in participants with 0 versus 12 or more hours of SHS exposure per week. For IL-6, the corresponding medians (IQR) were 1.09 (0.69-1.68) and 1.26 (0.84-2.03) pg/ml. After adjustment for age, gender, race/ethnicity, clinic site, education, income, hypertension, diabetes, LDL-C, treatment for dyslipidemia, physical activity, and cigarette smoking status, the geometric mean ratios for hscRP and IL-6 were 1.25 (95% CI: 1.11, 1.39) and 1.10 (95% CI: 1.03, 1.18) comparing 12 or more to 0 hours of SHS exposure per week. After further adjustment for body mass index the geometric mean ratios were 1.13 (95% CI: 1.02, 1.26) for CRP and 1.05 (95% CI: 0.98, 1.11) for IL-6.  The odds of hsCRP≥2 mg/L comparing 12 or more to 0 hours of SHS per week were 1.55 (95% CI: 1.24, 1.93) and 1.30 (1.04, 1.68), before and after adjustment for body mass index. SHS exposure was not associated with fibrinogen levels.

Self-reported SHS exposure was associated with increased hsCRP levels in serum. The association was attenuated, but remained statistically significant after accounting for BMI. For IL-6, the association with SHS disappeared after adjustment for body mass index.  Despite limited SHS exposure assessment, our findings support the association between SHS and hsCRP levels.