Limited Value of Urinary Cotinine to Creatinine Ratio as an Indicator for Tobacco Smoke Exposure

Usefulness of urinary cotinine (COT)/creatinine (Cr) ratio, a widely accept practical indicator of tobacco smoke uptake, was evaluated in 4 groups of healthy male volunteers (21 in total) with different cigarette doses of 0, 10, 20 and 30 cigarettes per day. 24-hour urine samples were collected on the 7th day of successive controlled smoking to determine urinary COT, thiocyanate (SCN) and Cr concentrations. Although urinary COT concentration exhibited close correlation (r=0.91, p<0.01) with cigarette dose, this correlation was weakened to r=0.82 (p<0.01) when it was standardized by urinary Cr concentration. Urinary SCN concentration was, in contrast, unrelated to cigarette dose (r=0.09), which however become correlated (r=0.71, p<0.05) after Cr standardization, in part due to the variation of urinary Cr concentration which was found inversely linked to cigarette dose (r= -0.47, p<0.05). Multiple regression analyses revealed that extent of urine volume-dependent decrease in urinary COT concentrationwas of several orders smaller in magnitude than that of urinary SCN or Cr concentrations as estimated on the basis of standardized partial regression coefficients. These results suggested that urinary COT/Cr ratio, compared to urinary COT concentration, tends to exaggerate the actual smoke uptake in the individuals with increased urine flow as well as in the smokers with smoking-associated decrease in urinary Cr concentration. J Epidemiol, 1992; 2 : 91-95.

Usefulness of urinary cotinine (COT)/creatinine (Cr) ratio, a widely accepted practical indicator of tobacco smoke uptake, was evaluated in 4 groups of healthy male volunteers (21 in total) with different cigarette doses of 0, 10, 20 and 30 cigarettes per day. 24-hour urine samples were collected on the 7th day of successive controlled smoking to determine urinary COT, thiocyanate (SCN) and Cr concentrations. Although urinary COT concentration exhibited close correlation (r=0.91, p<0.01) with cigarette dose, this correlation was weakened to r=0.82 (p<0.01) when it was standardized by urinary Cr concentration. Urinary SCN concentration was, in contrast, unrelated to cigarette dose (r=0.09), which however become correlated (r= 0.71, p<0.05) after Cr standardization, in part due to the variation of urinary Cr concentration which was found inversely linked to cigarette dose (r= -0.47, p<0.05).
Multiple regression analyses revealed that extent of urine volume-dependent decrease in urinary COT concentration was of several orders smaller in magnitude than that of urinary SCN or Cr concentrations as estimated on the basis of standardized partial regression coefficients. These results suggested that urinary COT/Cr ratio, compared to urinary COT concentration, tends to exaggerate the actual smoke uptake in the individuals with increased urine flow as well as in the smokers with smoking-associated decrease in urinary Cr concentration. J Epidemiol, 1992; 2 : 91-95.
tobacco smoke, cotinine, thiocyanate, creatinine ratio Accurate assessment of tobacco smoke uptake is important since tobacco smoke has been incriminated to various hazardous health conditions not only of active smokers but of individuals involuntarily exposed to environmental tobacco smoke1-3).Among a variety of biological indicators of smoke uptake, cotinine (COT)4), a major metabolite of nicotine, has been regarded most reliable because of its specificity to tobacco smoke, longer circulating half life5) compared to its parental compound, nicotine, and because it is readily measurable by a sensitive radio immunoassay6) or gas chromatography7), inspite of a great interlaboratory variation in the measurement8). COT is particularly useful in assessing the extent of environmental tobacco smoke uptake since it is concentrated in urine and well correlates with self-reported degree of tobacco smoke exposure9-11) As a biological sample for COT determination, single-spot urine instead of 24-hour urine has been used for practical reasons and urinary COT excretion is sometimes expressed in the form of creatinine (Cr) ratio, expecting to adjust for the variable COT dilution by urine flow12). Irrespective of such a practical importance of COT/Cr ratio as an indicator of smoke uptake, validation for its use has as yet been, to our knowledge, fully established.
We here present the results of a controlled smoking experiment in men and suggest that COT/Cr ratio is useful only on a limited basis. The present data have been highly suggestive of environmental tobacco smoke exposure to our non-smoking subjects since smoker/non-smoker ratios of urinary COT concentration and COT/Cr were at the largest about 20 and 30, respectively, which were much smaller than the reported values in the literature11,15).

MATERIALS AND METHODS
Results of simple regression analyses are summarised in Table 2. In COT, urinary concentration and total urinary output were closely correlated with cigarette dose, while the correlation was slightly diminished in COT/Cr ratio. In contrast, although urinary concentration and total urinary output of SCN were little or poorly correlated with cigarette consumption, correlation was much improved in SCN/Cr ratio. Such an increase in correlation by expressing urinary SCN excretion in Cr ratio is in part or completely explained by a confounding effect of urinary Cr concentration which was found in a significant inverse relationship with cigarette dose. Decreased correlation by Cr ratio of COT, on the other hand, appeared further confounded by the effect of urine volume since urine   was however only slight, approximately 1 : 4-5 of urinary SCN and Cr concentrations on the basis of standard partial regression coefficients. It was also noted that urinary SCN concentration was in a positive correlation with cigarette dose, which was unable to detect in the simple regression analysis because of the confounding effect of urine volume. These results altogether suggested that the use of urinary COT/Cr ratio is limited compared to urinary COT concentration since it could be confounded by the cigarette dose-dependent decrease in urinary Cr concentration as well as by the urine volumedependency of urinary COT concentration which was unproportionally small compared to that of urinary Cr concentration.

DISCUSSION
The rationale for using Cr ratio as an index for estimating urinary excretion of micronutrients or their metabolic end products is based upon the assumptions (1) that daily Cr excretion is constant and (2) that urinary component is volume-dependently diluted by urine. Regarding the first assumption, it has been well recognized that there is a definite day-to-day variation in Cr excretion as well as the diurnal variation where the Cr excretion is lowest during early morning and highest in the afternoon 16). Body size-or age-dependency of urinary Cr excretion has been well established as one of the causes of inter-individual variation of Cr excretion 16,17) Our present results demonstrated that urinary Cr concentration correlated inversely with cigarette consumption in accordance with the observation of Adlkofer et al.18), suggesting a further potential source of variation in using Cr ratio in clinical or populational surveys. Difference in mean urinary Cr concentrations between non-or light smokers and heavy smokers was as great as more than 30% in the present study, similar to the values reported by Adlkofer et al.18) which was about 40% between the light smokers (smoking less than 10 cigarettes/day) and heavy smokers (smoking more than 30 cigarettes/ day). Our finding is unlikely confounded by the influence of body size or blood pressure17) between the study groups since neither body mass index nor blood pressure was related to cigarette consumption. From our results decreased urinary Cr concentration may be explained by a combined effect of decreased total urinary Cr output and the greater urine volume particularly in heavy smokers consuming 30 cigarettes per day. Although slightly reduced plasma Cr levels for smokers as compared with non-smokers were reported by some investigators19, 20), which could be a potential cause of lower urinary Cr concentration in smokers, such was not confirmed in the present study consistent with other authors21). Greater urine volume observed in the group smoking 30 cigarettes per day is difficult to simply explain by an effect of heavy smoke exposure since intake of fluid or diet was not controlled in the present study, which however might be an effect of smoking since accelerated urine flow by smoking was reported by other authors22).
Regarding the second assumption, multiple regression analysis in the present study revealed that urinary COT concentration was much less diluted by urine volume compared to urinary SCN and Cr concentrations. This finding suggests that expressing urinary COT excretion by means of Cr ratio could introduce rather than adjust the variation due to the inter-or intra-individual difference in urine volume which could be influenced by a variety of factors in daily life. In a occasion where urinary Cr concentration is decreased due to the increased urine flow, COT/Cr ratio may lead to an exaggerated estimation of actual tobacco smoke exposure.
From the present results it is concluded that urinary COT concentration may be more appropriate than COT to Cr ratio as a urinary index for tobacco smoke exposure contrary to what had been suggested 12). This would be so even for the assessment of environmental tobacco smoke exposures since variability of COT/Cr in terms of CV was twice greater than that of urinary COT concentrations in non-smoker group (27.2% vs 11.2%) and would be particularly so in the studies employing random spot urine samples where accurate informations on factors which influence the urine flow of individuals are hardly collectable.
Finally, the present findings might be extended to more generalized conclusion that possible influence of tobacco smoking on urinary Cr concentration could be one of the potential confounding factors in the investigations which use Cr ratio of urinary components and involve both non-smokers and smokers and that magnitude of urinary dilution require to be evaluated for any urinary constituent before it is standardized by urinary Cr concentration in clinical or epidemiological surveys.