Background Associations of socioeconomic status (SES) and smoking-related diseases depend on uniform validity of self-reported smoking habits in different SES groups. We investigated the influence of SES on validity of self-reported smoking status by means of urinary cotinine.

Methods We determined total urinary cotinine in the baseline population of the Heinz Nixdorf Recall Study. Participants with cotinine>200 µg/L were potential current smokers. We defined upper and lower 20% of the gender-specific distribution of the International Socio-Economic Index (ISEI) as high and low SES, respectively, else as intermediate. We analysed the association of self-reported smoking status and cotinine by ISEI and additional SES measures, stratified by gender. In self-reported non-smokers, we estimated age-adjusted ORs with 95% CI to detect differences by SES in the validity of self-reported smoking status.

Results In 2004 men and 1887 women, 78% and 80%, respectively, reported to be non-smokers. Median cotinine concentrations were 2 µg/L in non-smokers, and 3651 µg/L in male and 3127 µg/L in female smokers. Based on cotinine in non-smokers, 2.0 % of men (n = 32) and 1.8 % of women (n = 27) were potential smokers, with lower proportions in the subgroup of never-smokers (men: 0.7%, women: 0.5%). The validity of self-reported smoking status did not substantially differ by SES. Tendencies for increased underreporting were indicated for women with low ISEI (OR 1.35; 95% CI 0.54 to 3.39) and men in blue-collar jobs (OR 1.39; 95% CI 0.67 to 2.87).

Conclusion Validity of self-reported smoking status in this elderly German cohort was high and did not depend on SES.