We have developed an unobtrusive technique that estimates how much smoke particulate matter and nicotine is obtained by smokers from their cigarettes in their everyday environment [2-5]. The technique, called part-filter analysis, involves analysing the retained smoke particulates and nicotine in the mouth-end filter section of a smoked cigarette. The part-filter technique has been used in a number of studies including a clinical study on candidate PREPs and a longitudinal study of smoking behaviour.
This page describes the following in detail:-
Part-filter analysis, along with data from the same cigarettes smoked in the laboratory, enables the measurement of nicotine-free dry particulate matter (NFDPM) or ‘tar’ and nicotine to give an estimation of mouth level exposure from the filters of smoked cigarettes. This is the maximum amount of NFDPM and nicotine available for smokers to take into the mouth - including any smoke that might spill from the mouth before and during inhalation and exhalation.
Part-filter analysis gives a direct estimate of the yields produced by cigarettes when smoked by humans [5,6]. Although it is often used for research purposes within the tobacco industry, part-filter analysis has yet to be endorsed as an alternative to biomarkers of exposure by scientists outside the industry. However, recently groups external to the tobacco industry have published filter analysis work and reviewed the current state of the art for these methodologies [7,8]. In addition, correlation studies have been carried out, comparing part-filter analysis with biomarkers of exposure. These studies have shown significant correlation between mouth level exposure measurements using the part filter analysis and relevant biomarkers of exposure for specific smoke constituents in 24-hour urine samples from smokers [9-12].
The part-filter analysis methodology estimates human mouth level exposure to NFDPM and nicotine by comparing the UV absorbance of filter extracts and the nicotine content of human smoked cigarettes respectively with those from machine smoked cigarettes for which the smoke yield is known. This is possible because the yield of nicotine and NFDPM in the smoke and that remaining in the filter are related to the filtration efficiency/flow relationship of the filter. The part-filter method, relies on the analysis of a section (10 or 7 mm) from the mouth-end of the filter where the filtration efficiency remains relatively constant irrespective of puff flow rates (as demonstarted below) and butt lengths. This method has also been found to be less susceptible to the effect of nicotine condensation .
The figure below describes the part-filter analysis methodology.
For the part-filter analysis method, calibration is required for each test product by machine smoking using a range of regimes that encompass the wide range of human smoking behaviours including puff volumes, durations, intervals and smoke flows. By varying these parameters we obtain the smoking regimes which comprise three flow rates, 20ml/s, 33.3ml/s and 46.7ml/s.
Smoke yield and part-filter analysis data from these calibrations are used to produce regression equations that are combined with human smoked part-filter data to provide human nicotine and NFDPM mouth level exposure estimates.
For each product to be calibrated, five cigarettes per smoking regime are smoked using smoking machines. The smoked cigarette filter tips are then cut to either 7mm or 10mm from the mouth-end downstream of any ventilation holes - depending on filter design - and stored in a dark location at ambient temperature for subsequent analysis of nicotine and filter tip extract UV absorbance alongside human smoked part-filters. This ensures that calibration and human smoked part-filters are equally aged.
Mouth end portions of freshly smoked cigarette filter tips are collected by smokers using a specially designed filter collector to cut, collect and protect the filter tips after smoking.
Each collector has a barcode to identify:-
Each collector also has an anti-tamper seal to prevent interference with the collector.
The smoker extinguishes a cigarette, inserts the mouth end of the filter into the collector hole labelled 'filter end' and pushes the handle 'A' to cut a 10mm or 7mm section depending on cigarette filter design.
Handle 'B' is then pulled to allow the cut filter section to drop into the storage box and the tobacco end section of the filter is discarded.
Generally for part-filter analysis studies, smokers are asked to collect a minimum of 15 filter tips from the test cigarettes they have smoked, as a representative sample allowing three measurements using groups of 5 filters. However, other studies have been carried out using part-filter analysis of single filters. One such study investigated variation in mouth level exposure in smokers over a 24 hour period and concluded that there was no significant difference in the mouth level exposure for single filters from individual smokers over this period of time .
The collectors are returned to the laboratory where 15 tips are separated into three random groups of five. The length (± 0.1mm) of each tip is recorded and the mean length for each batch of five calculated to allow for differences in the tip length cut.
Mainstream smoke nicotine and NFDPM yields obtained from calibration smoking are plotted against the calibration filter tip nicotine and filter tip extract UV absorbance, respectively, to obtain regression equations that can be used to estimate human nicotine and NFDPM mouth level exposure. Each batch of five filters, from either smokers or machine smoking, is extracted in 20mL methanol and analysed for nicotine and NFDPM content by gas chromatography and UV absorbance (310nm) respectively.
Human nicotine mouth level exposure is estimated for each smoker by using the filter tip extract nicotine concentrations and the linear regression equation derived from the calibration. Normally three values are averaged - one from each batch of five tips.
Similarly, human NFDPM mouth level exposure is estimated for each smoker by using the UV absorbance per tip values and the linear regression equation derived from calibration smoking.
Below are examples of the calibration graphs obtained using this method. Mainstream smoke nicotine is plotted against tip nicotine and mainstream NFDPM is plotted against extract UV absorbance. The linearity of these graphs is affected by filtration efficiency/flow of the given cigarette filter and can vary depending on the product design.