Pitfalls in the assessment of smoking status detected in a cohort of South African RA patients.

This study was conceived in an attempt to explain the unexpectedly high frequency of elevated levels of serum cotinine measured retrospectively in a cohort of predominantly black South African females with rheumatoid arthritis (RA), findings that were inconsistent with the smoking histories derived from health questionnaires. The discrepant findings suggested either a greater tendency towards underreporting of smoking status in the study cohort, or possible confounding effects of the use of smokeless tobacco products. In addition to the cohort of RA patients (n = 138, of whom 115 (83 %) were female), blood samples were also taken from a second cohort consisting of 29 declared smokers, 18 (62 %) of whom where females, 29 smokeless tobacco (SLT) users (all female), and 22 non-users of any tobacco products, 18 (82 %) of whom were females. Serum cotinine levels were determined using an ELISA procedure. Cotinine levels of >10.0 ng/ml were detected in serum specimens from 43 (31 %), RA patients of whom 35 (81 %) were female, with a median value of 50.1 ng/ml and interquartile range (iqr) of 68.6. Only 18 of the 35 females indicated that they smoked. The groups of declared smokers and SLT users had equivalent median serum cotinine levels of 88.0 ng/ml (iqr = 10.8 ng/ml) and 87.0 ng/ml (iqr = 15.6 ng/ml), respectively, while cotinine was undetectable in specimens from non-tobacco product users (<0.2 ng/ml). Users of SLT products in South Africa are predominantly female and have serum cotinine levels which are comparable with those of current smokers, raising concerns about the validity of measurement of cotinine as the sole objective marker of smoking status in populations with high usage of SLTs. This situation can be rectified by ensuring that usage of SLT products is accurately recorded in health questionnaires, while inclusion of measurement of one or more additional, objective biomarkers of smoking in combination with cotinine may enable reliable distinction between smoking and usage of SLTs which, given the associated risks, is a strategy of particular relevance in RA.