Summary
This study focused on the variation in the yields of constituents in smoke from commercial cigarette brands available on the Japanese market. Nineteen commercial cigarette brands were sampled five times every two months from 2009 to 2010. The target constituents were benzo[a]-pyrene, 1,3-butadiene, benzene, formaldehyde, acetaldehyde, acrolein, N-nitrosonornicotine (NNN), 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), carbon monoxide, “tar”, and nicotine. The results of this study showed that the coefficient of variation (CV) values varied greatly by brands, constituents, and smoking regimes. The yields of NNN and NNK in the smoke were strongly correlated to their yields in the tobacco filler blend for most brands. The yields of benzo[a]pyrene under the International Organization for Standardization (ISO) and the Health Canada Intense (HCI) smoking regimes and 1,3-butadiene under the HCI smoking regime were found to be influenced by the measurement. It was shown that factors for variation were highly varied among constituents. The grand mean of CV values for NNN and formaldehyde associated with cigarette manufacturing over ten months and measurement at the JT laboratory under the HCI smoking regimes were 17.1% and 6.6% respectively. The grand mean of CV values for NNN and formaldehyde associated with both cigarette manufacturing over ten months and measurement at different laboratories under the HCI smoking regimes were 23.7% and 22.9% respectively. This is due to the fact that formaldehyde showed the highest CV values for reproducibility among the constituents. Thus, in order to set realistic and robust confidence intervals, it is very important to take into account the variations associated with cigarette manufacturing and measurement within and between laboratories.
Peer Review #2 of "Confidence intervals for the common coefficient of variation of rainfall in Thailand (v0.1)"