Psychopharmacology (2003) 170:434–442 DOI 10.1007/s00213-003-1541-8
ORIGINAL INVESTIGATION
Michael Dixon · Neena Kochhar · Krishna Prasad · Jim Shepperd · David M. Warburton
The influence of changing nicotine to tar ratios on human puffing behaviour and perceived sensory response Received: 1 April 2003 / Accepted: 22 May 2003 / Published online: 7 August 2003 Springer-Verlag 2003
Abstract Rationale: Smokers modify their smoking behaviour when switching from their usual product to higher or lower tar and nicotine-yield cigarettes. Objective: The aims of the current study were to assess the influence of varying nicotine yields at constant tar yield on human puffing measures, nicotine deliveries under human smoking conditions and the sensory response to mainstream cigarette smoke. These assessments would allow an evaluation of the degree of compensation and the various possible causes of changes, if any. Methods: The participants were 13 regular smokers of commercial or hand-rolled cigarettes. They were tested with four cigarettes, which exhibited a wide range of nicotine to ‘tar’ ratios at a relatively constant ‘tar’ yield. Their smoking behaviour was monitored by placing the test cigarettes into an orifice-type holder/flowmeter attached to a custom-built smoker behaviour analyser. In addition, a comprehensive sensory evaluation of the products was carried out. Results: The differences in the nicotine to tar ratios of the samples did not significantly influence the puffing behaviour patterns, i.e. puff number and interval, total and average puff volume, integrated pressure and puff duration. Additionally the pre- to post-exhaled CO boosts were not significantly influenced by the experimental samples used in the study. However, the nicotine yields obtained by the smokers were significantly influenced by the machine-smoked nicotine yields or the M. Dixon ()) British American Tobacco, 4 Temple Place, London, WC2R 2PG, UK e-mail:
[email protected] Tel.: +44-20-78451649 N. Kochhar · D. M. Warburton Department of Psychology, University of Reading, Whiteknights, Reading, RG6 6AL, UK K. Prasad · J. Shepperd British American Tobacco, R&D Centre, Regents Park Road, Southampton, SO15 8TL, UK
nicotine to tar ratios of the samples. The machine-smoked nicotine yields were highly correlated with the nicotine yields obtained under human smoking conditions. For the sensory evaluation, there was only a significant difference between the samples in the intensity of the impact. Conclusion: These observations imply that these puffing variables are not controlled by the nicotine yield of the cigarette. Keywords Smoking behaviour · Nicotine yield · Tar yield · Sensory response · Compensatory smoking
Introduction It is recognised that smokers may modify their smoking behaviour when switching from higher to lower tar and nicotine-yield cigarettes. This behavioural phenomenon, often referred to as compensation, has been extensively studied over the past four decades, and the published data has been extensively reviewed (Stephen et al. 1989; Scherer 1999). A reduction in the nicotine yield of the cigarette has been cited as one of the reasons for compensatory behaviour changes. It has been advocated that reducing tar but not reducing nicotine yield may prevent or reduce the compensatory response to low-yield cigarettes (Russell 1976; Warburton 1985; Bates et al. 1999). However, some researchers have studied the independent effects of tar and nicotine yields and have concluded that the tar yield of a cigarette is an important determiner of smoking behaviour (Stepney et al. 1981; Sutton et al. 1982; Armitage et al. 1988; Fairweather 1989; Hasenfratz et al. 1993; Baldinger et al. 1995). It has been hypothesised that that some components of tar have sensory properties and smokers may smoke more intensely to compensate for a reduction in the sensory properties of lower tar cigarettes (Stepney 1981; Hasenfratz et al. 1993). Nicotine also plays an important role in some of the sensations perceived by the smoker during the inhalation
435
of cigarette smoke (Kochhar and Warburton 1990; Pritchard et al. 1996; Rose et al. 1999). Thus, there is a complex situation involving the potential roles of nicotine pharmacology and the sensory effects of both tar and nicotine in the control of smoking behaviour, particularly the compensatory behavioural response to reduced tar and nicotine-yield products (Creighton and Lewis1978). The aims of the current study were to assess the influence of varying nicotine yields at constant tar yield on human puffing measures, nicotine deliveries under human smoking conditions and the sensory response to mainstream cigarette smoke. These assessments would allow an evaluation of the degree of compensation and the various possible causes of changes, if any. These objectives have been identified as important by the Institute of Medicine in their recent report (Institute of Medicine 2001).
Methods Cigarette samples The tobacco blend used in a commercially available US blend style cigarette was passed through a denicotinisation process at Flavex Germany. This process removed 91% of the nicotine content in the tobacco. Cigarette samples were prepared using different combinations of ‘normal’ and ‘denicotinised’ tobacco blend. All the cigarette samples were made to the same physical specifications. The cigarettes were 84 mm in length and incorporated a 20-mm cellulose acetate filter. The filters were ventilated at a nominal 50% ventilation level. The machine-smoked (ISO) yields, ventilation levels and cigarette pressure drop values for the five samples produced are shown in Table 1. The experimental cigarettes were matched carefully in terms of ventilation levels and total cigarette pressure drop values. There were some differences among the samples for ‘tar’ yields, CO yields and puff numbers, but these were relatively small compared with the differences in nicotine yields. Consequently, the experimental samples exhibited a wide range of nicotine to ‘tar’ ratios at a relatively constant ‘tar’ yield.
A comprehensive, smoking behaviour and sensory evaluation study was conducted. The participants each visited the laboratory on ten occasions and smoked two test samples of the same type at each laboratory visit. Smoking sessions were conducted between 1015 hours and 1500 hours. The participants were instructed to refrain from smoking for a period of at least 1 h prior to attending the study sessions. There was a rest period of 30 min between the two cigarette smoking sessions. Thus, each participant smoked each test sample four times, and the order of sample presentation for the laboratory visits was randomised across participants. Smoking behaviour was monitored by placing the test cigarettes into an orifice-type holder/flowmeter (based on the original design of Creighton et al. 1978) attached to a custom-built smoker behaviour analyser. The following measures were derived from the flow-rate and pressure signals obtained from the cigarette holder/ flowmeter: 1. Puff number (the number of puffs taken per cigarette) 2. Mean puff volume (the volume drawn during the puff, expressed as a mean value per cigarette) 3. Mean puff duration (the time interval from the start to the end of the puff, expressed as a mean value per cigarette) 4. Mean inter-puff interval (the time between the end of a puff and the start of the subsequent puff, expressed as a mean value per cigarette) 5. Mean integrated puff pressure (the area under the puff pressurepuff duration curve, expressed as a mean value per cigarette) 6. Mean maximum puff flow-rate (the peak flow rate achieved during the puff expressed as a mean value per cigarette) 7. Mean maximum puff pressure (the peak pressure achieved during the puff expressed as a mean value per cigarette) 8. Total puff volume (the sum of the individual puff volumes taken from the cigarette). Exhaled carbon monoxide (CO) ‘boost’ measurement Exhaled CO levels were measured prior to smoking and 1 min after extinguishing the cigarettes. The participants took a deep breath, held their breath for 15 s and then exhaled slowly through the mouthpiece of a Bedfont Mini Smokerlyser to measure exhaled CO levels. Exhaled CO ‘boosts’ were determined from the differences between the post- and pre-smoking CO levels. Measurement of nicotine deliveries
Participants Evaluations were conducted with 13 participants who had previously received training in the sensory evaluation of cigarette smoke. The participants were regular smokers of commercial or hand-rolled cigarettes. Seven were smokers of full-flavour products, three smoked lights products and three were smokers of handrolled cigarettes. Permission to recruit participants for the study was given by the University of Reading ethics committee. These participants gave written, informed consent to participate in the study
Table 1 Nicotine-free dry particulate matter (NFDPM or ‘tar’), nicotine and carbon monoxide (CO) yields, filter ventilation levels and total pressure drop values – ISO conditions (1 puff/min, 35 ml puff volume, 2-s duration, butt length: overwrap+3 mm)
Smoking behaviour measurement
NFDPM (‘tar’) (mg/cigarette) Nicotine (mg/cigarette) CO (mg/cigarette) Puff no. Nicotine to ‘tar’ ratio ‘Tar’ to nicotine ratio Filter ventilation (%) Total pressure drop (mmH2O)
The amounts of nicotine delivered to the smokers were determined from butt analysis. After each smoking session, the cigarette butts were stripped of residual tobacco and stored at room temperature in aluminium or amber glass containers to reduce exposure to light. Filters were subsequently split longitudinally and nicotine was extracted by shaking in methanol containing sodium hydroxide (1 mg/ml) and n-heptodecane (0.4 mg/ml) as an internal standard. The nicotine was measured by megabore gas chromatography, quantifying by reference to a set of standards (Jacob and Byrd 1999).
RS33
RS32
RS30
RS31
RS34
9.5 0.77 7.9 9.8 0.081 12.3 51 77.7
8.8 0.48 7.3 9.8 0.055 18.3 51 78.7
8.7 0.81 7.3 8.9 0.093 10.7 52 76.1
8.1 0.10 6.7 8.9 0.012 81.0 48 76.0
7.7 0.22 6.6 9.0 0.029 35.0 51 83.9
436 The cigarettes were machine-smoked under the following smoking regimes: 1. 2. 3. 4.
35 60 90 90
ml ml ml ml
puff puff puff puff
volume, volume, volume, volume,
Results Influence of sessions and replicates
2-s puff duration, one puff per minute 2-s puff duration, one puff per minute 2-s puff duration, one puff per minute 1.5-s puff duration, one puff per minute.
Nicotine filtration efficiencies (FE) were calculated from the data obtained from the above machine smoking regimes according to the following formula: FEð%Þ ¼ 100 Nfilt = Nfilt þ Npad where Nfilt is the amount of nicotine retained in the filter and Npad is the amount of nicotine trapped on the Cambridge pad. The amounts of nicotine delivered to the participants were derived from the butt nicotine values using the following: Nicotine yield ¼ Butt nicotine 100=FEa Butt nicotine where FEa is the average nicotine filtration efficiency obtained over the range of machine smoking regimes. Sensory evaluation The participants assessed the following sensory properties of the test samples after completing each smoking behaviour session: 1. Draw effort (the perceived amount of effort expended during puffing on the cigarette) 2. Mouthful of smoke (the sensation of the amount of smoke entering the mouth during the puff) 3. Impact (the short-lived sensation experienced at the back of the throat during the inhalation of smoke) 4. Irritation (the lingering, tingling, peppery sensation perceived primarily in the throat during and after smoke inhalation) 5. Flavour amplitude—the amount of flavour (gustation and olfaction) perceived irrespective of the character of the flavour. The participants had been previously trained in the recognition and intensity scaling of these five sensory attributes. Each attribute was assessed on a scale of 0–5, with zero representing an absence of the attribute, and five a very high intensity of the attribute. In addition, the participants were asked to rate their opinion on the overall acceptability of each product using a scale of 0–100.
Each smoker attended two smoking behaviour sessions on separate days (session effect). Two cigarettes of the same type were smoked at each monitoring session (replicate effect). The mean €SD values for the behavioural and sensory variables are shown in Table 2. This table also contains the P values for the replicate and session effects obtained from a two-way analysis of variance (ANOVA) test. The behavioural and sensory data were extremely consistent across both replicates and sessions. The twoway ANOVA did not reveal any statistically significant session, replicate or interaction effects. Consequently, the data obtained from the four monitoring sessions with each test sample were combined for each subject and the data in subsequent tables are based on the averages of the four monitoring sessions per subject and per sample. Smoking behaviour The mean and standard deviation values for the smoking behaviour variables were calculated from the 13 participants and 5 test samples and are shown in Table 3 below. A one-way ANOVA test was calculated to determine the significance of the between-sample differences. The ANOVA revealed that the differences in the nicotine to ‘tar’ ratios of the samples did not significantly influence the puffing behaviour patterns, i.e. puff number and interval, total and average puff volume, integrated pressure and puff duration. The individual puff volume responses to the samples are shown in Fig. 1. Additionally, the pre- to post-exhaled CO boosts were not significantly influenced by the experimental samples used in the study. However, the nicotine yields obtained by the smokers were significantly influenced by the machine-smoked nicotine yields or the nicotine to ‘tar’ ratios of the
Table 2 Influence of session and replicate effect on smoking behaviour and sensory variables Session 1 Puff no. 15.4€3.2 Puff interval (s) 20.4€4.7 Puff duration (s) 1.78€0.78 33.3€8.2 Integrated pressure (cmH2O.s) Total puff volume (ml) 833€170 Puff volume (ml) 55.5€12.6 Nicotine delivery (mg/cigarette) 0.68€0.46 CO boost (ppm) 7.6€6.4 Draw effort 2.7€1.1 Mouthful 2.6€0.8 Impact 1.7€1.1 Irritation 1.9€1.1 Flavour amplitude 2.6€0.9 Acceptability 53.7€25.9
Session 2
Replicate 1
Replicate 2
Session P value
Replicate P value
15.1€2.9 21.3€6.8 1.77€0.79 33.2€8.0 819€186 55.3€12.3 0.68€0.46 8.1€8.1 2.6€0.9 2.5€0.8 1.8€1.1 2.0€1.0 2.7€1.0 52.1€27.7
15.2€3.2 20.9€5.5 1.79€0.80 33.0€8.1 819€175 55.2€12.5 0.69€0.46 7.5€5.9 2.7€1.1 2.5€0.8 1.7€1.1 1.9€1.1 2.7€1.0 53.7€27.4
15.3€3.2 20.7€6.2 1.76€0.76 33.5€8.1 832€181 55.6€12.3 0.67€0.45 8.1€8.6 2.6€1.0 2.6€0.8 1.8€1.1 2.0€1.0 2.6€0.9 52.1€26.3
0.365 0.230 0.958 0.894 0.532 0.897 0.917 0.607 0.254 0.605 0.677 0.628 0.247 0.473
0.825 0.745 0.754 0.633 0.520 0.808 0.832 0.516 0.764 0.172 0.380 0.299 0.786 0.616
437 Table 3 Mean€standard deviation values for the smoking behaviour measures (n=13 for each sample)
ISO -nicotine yield (mg/cigarette) Puff no. Puff interval (s) Puff duration (s) Int. press (cmH2O.s) Puff volume (ml) Total puff volume (ml) Nicotine delivery (mg/cigarette) CO boost (ppm)
RS33
RS32
RS30
RS31
RS34
P value
0.77 16.0€2.6 20.9€4.4 1.75€0.78 32.5€7.7 55.4€12.4 877€197 1.11€0.21 8.2€6.2
0.48 16.1€3.0 20.8€3.6 1.80€0.82 34.5€7.9 56.7€11.6 890€152 0.69€0.10 7.4€4.1
0.81 14.4€2.3 21.4€5.3 1.69€0.78 30.6€7.9 53.6€13.3 755€166 1.17€0.23 7.4€4.2
0.10 14.1€2.3 21.1€4.5 1.77€0.74 32.9€7.4 55.6€11.5 766€141 0.09€0.02 8.9€10.1
0.22 15.6€3.3 20.0€3.7 1.84€0.83 35.6€8.6 55.8€13.0 839€136 0.34€0.06 7.2€4.0
NA 0.184 0.952 0.991 0.545 0.981 0.109 0.0001 0.949
Fig. 1 Individual puff volumes obtained from the five experimental cigarettes (y-axis). The values on the x-axis refer to the machinesmoked nicotine yields of the experimental cigarettes. The top panel contains results for participants 1 to 7, and participants 8 to 13 are shown in the bottom panel
Fig. 2 Amounts of nicotine obtained by the individual participants from the five experimental cigarettes (y-axis). The values on the xaxis refer to the machine-smoked nicotine yields of the experimental cigarettes. The top panel contains results for participants 1 to 7, and participants 8 to 13 are shown in the bottom panel
samples. The nicotine yields obtained by the participants for each sample are plotted in Fig. 2. The machine-smoked nicotine yields were highly correlated with the nicotine yields obtained under human smoking conditions (r=0.998). This relationship is shown in Fig. 3. The machine and human smoking nicotine-yield data has been used to calculate indices of compensation (CI). The following formula, based on that described by Scherer (1999), was used to calculate the compensation indices for nicotine intake: CI ¼ 1 ð% change in human nicotine yield = % change in machine smoke yieldÞ
Fig. 3 Plot of mean €SD machine-derived nicotine yields (x-axis) against nicotine yields obtained by the smokers (y-axis)
438 Table 4 Compensation index (CI) calculations with sample RS30 (8.7/0.81) as the reference. The figures in parentheses refer to machinesmoked tar and nicotine yields Sample
% Change in machine nicotine yield
% Change in human nicotine delivery
CI (nicotine intake based)
CI (nicotine adjusted CO boost)
RS33 RS32 RS34 RS31
–4.9% –40.7% –72.8% –87.7%
–5.1% –41.0% –70.9% –92.3%
–0.04 –0.01 0.03 –0.05
–3.07 –0.26 –0.04 0.01
(9.5/0.77) (8.8/0.48) (7.7/0.22) (8.1/0.10)
Table 5 Mean €SD sensory evaluation results (n=13 for each sample)
ISO–nicotine yield (mg) Draw effort Mouthful Impact Irritation Flavour amplitude Acceptability
RS33
RS32
RS30
RS31
RS34
0.77
0.48
0.81
0.10
0.22
2.44€0.90 2.71€0.44 2.16€0.76 2.15€0.88 2.83€0.61 60.1€22.6
CI ¼ 0 represents no compensation CI ¼ 1 represents complete compensation 0 < CI < 1 represents partial compensation Compensation indices were also calculated using the CO boost and nicotine-yield data as an attempt to incorporate an uptake biomarker, namely CO boost. The Scherer (1999) equation was modified as follows: CI ¼ 1 ½ðCO boostnew Nnew =COnew CO boostold Nold =COold Þ=CO boostold Nold =COold = ðNnew Nold Þ=Nold where ‘new’ is one of the lower nicotine-yield cigarettes and ‘old’ is the reference cigarette. N and CO are the machine-derived nicotine and CO yields, respectively. The sample with the highest machine-smoked nicotine yield, RS30, was used as the reference sample for the calculation of compensation indices following the reduction in nicotine machine yields. Table 4 contains the results of the compensation index calculations. The calculated compensation indices gave no evidence for compensation, irrespective of whether the compensation was calculated using nicotine deliveries to the smokers or CO boost data. Sensory evaluation results Table 5 contains the mean€SD sensory data grouped by sample type. There were no statistically significant differences in the draw effort, mouthful, irritation or flavour amplitude characteristics of the test samples. There was a statistically significant difference between the samples in the intensity of the impact sensation, with the very low nicotine-yield samples ( RS31 and RS34)
2.79€0.71 2.49€0.59 1.61€0.74 1.68€0.81 2.65€0.72 54.7€25.1
2.23€0.70 2.79€0.44 2.33€0.56 2.31€0.84 2.80€0.83 59.0€22.9
2.76€0.82 2.58€0.72 1.50€1.06 1.82€0.73 2.50€0.86 45.9€24.3
2.93€0.90 2.22€0.59 1.22€0.82 1.62€0.80 2.41€0.80 44.8€28.1
P value NA 0.180 0.109 0.003 0.147 0.572 0.361
producing lower impact scores than the higher nicotineyield samples (RS30 and RS33). Although not statistically significant, the very low nicotine-yield cigarettes (RS31 and RS 34) were rated as being less acceptable than the higher nicotine-yield cigarettes (RS30 and RS33).
Discussion It has often been stated that the compensatory response, i.e. increase in puffing intensity, which generally occurs following a switch to a lower yield cigarette, occurs as a result in the reduction in the nicotine yield of the cigarette (Royal College of Physicians 2000). This nicotine regulation concept led to a proposal for the development of cigarettes reduced in ‘tar’ yield but with maintained nicotine yields. It was believed that maintaining nicotine yields of low ‘tar’ cigarettes at levels associated with full flavour cigarettes would minimise smoker compensation and thereby present a ‘safer’ alternative to the conventional low ‘tar’, low nicotine cigarette (Russell 1976; USDHHS 1981; ISCSH 1984). Prior to the development of this concept, most studies investigating smoker compensation had not attempted to separate the effects of nicotine and ‘tar’ and had thus ignored the potential role of the reduction in ‘tar’ yield in the compensatory response. However, there were a few studies conducted in the 1970s and early 1980s that attempted to separate the effects of ‘tar’ and nicotine on aspects of smoking behaviour and these produced conflicting results. Goldfarb et al. (1970) assessed the smoking rates and psychological effects of smoking lettuce leaf cigarettes with and without added nicotine. They observed that varying the nicotine yield of the lettuce leaf cigarettes had no effect on cigarette consumption rates but influenced the perceived strength of
439
the cigarettes. The higher nicotine-yield cigarettes were rated as being stronger than the lower or zero nicotineyield cigarettes. The authors concluded that “the habit itself often exhibits functional autonomy from the physiological effects of nicotine”. Dunn and Freiesleben (1978) experimentally enhanced the nicotine yields of cigarettes by adding nicotine citrate to the tobacco blend. Nicotine enhancement did not influence daily cigarette consumption rates, puff numbers, puff durations or relative puff volumes. Stepney (1981) observed reductions in mouth-level exposure to ‘tar’ when smokers switched to either a low ‘tar’–low nicotine cigarette or a low ‘tar’–medium nicotine cigarette. However, there was no evidence that the reduction in ‘tar’ exposure was greater with the low ‘tar’–medium nicotine cigarette than with the low ‘tar’–low nicotine cigarette. Indeed, the smoking behaviour responses in terms of puff numbers and puff volumes were similar for the two low ‘tar’ cigarettes, despite the difference in the nicotine yields of the two cigarettes. Thus, Stepney (1981) suggested that smokers may modify their behaviour in response to sensory cues produced by ‘tar’, and that the amount of smoke taken from a cigarette may be affected by a complex interaction between the sensory effects of ‘tar’ and nicotine. In contrast to these above three studies, Herning et al. (1981) and Gust and Pickens (1982) reported that changing nicotine yields independently of ‘tar’ yields influenced puffing behaviour. Both these studies used experimental, unfiltered cigarettes produced by the University of Kentucky, which had very high ‘tar’ yields (around 25–30 mg) and nicotine yields of 0.4, 1.2 and 2.5 mg. Herning et al. (1981) reported that the puff volume was higher on the lowest nicotine-yield sample, but that puff number, puff duration and inter-puff interval were similar across the three samples in the 24 smokers, who were studied. Gust and Pickens (1982) reported that for the six smokers in their study, increasing nicotine yield was associated with a decrease in average puff volume, puff duration and puff number. More recently there have been a number of studies comparing commercially manufactured cigarettes made from denicotinised tobacco with those made from conventional tobacco in the attempt to differentiate the effects of nicotine and ‘tar’ yields on smoker compensation. Robinson et al. (1992) compared two cigarettes that were matched for ‘tar’ yield (around 9 mg) but differed radically in nicotine yield (0.08 mg vs 0.6 mg). They found no differences in puffing behaviour between the two cigarettes. About the same time, Hasenfratz et al. (1993) conducted a smoking behaviour and smoke uptake study using three cigarettes. One was a commercial ‘lights’ cigarette with a ‘tar’ yield of 10.1 mg and a nicotine yield of 0.81 mg. The second product was similar in ‘tar’ yield to the commercial product, but incorporated denicotinised tobacco and had a nicotine yield of only 0.08 mg. The third cigarette was a commercial ‘ultralights’ product with a ‘tar’ yield of 1.83 mg and a nicotine yield of 0.22 mg. Puffing intensities were similar for the
commercial and denicotinised 10-mg ‘tar’ yield cigarettes. Thus, there was no evidence that the reduction in nicotine yield was associated with an increase in puffing intensity. These researchers also measured nicotine intake/nicotine yield, and CO intake/CO yield ratios as indices of compensation for the three cigarettes. The commercial and denicotinised products produced similar ratios but the ratios were double for the ultra-lights cigarette. Consequently, Hasenfratz et al. (1993) concluded that gustatory and olfactory sensations arising from cigarette smoke ‘tar’ play a greater role in the regulation of smoking behaviour than was previously believed. In the same laboratory, Baldinger et al. (1995) extended the work of Hasenfratz by measuring more extensive indices of puffing topography. They observed that smokers took similar total puff volumes from the commercial and denicotinised light cigarettes but increased total puff volume when smoking the ultra-light cigarettes. Thus, they concluded that puff volumes were influenced by changes in the ‘tar’ yield, but not the nicotine yield of the cigarette. In our study, ‘tar’ yields and physical characteristics of the test cigarettes were held relatively constant. Varying the relative amounts of commercial and denicotinised tobacco in the blends produced a wide range of nicotine yields across the test samples (0.10 mg/cigarette to 0.81 mg/cigarette). However, these changes in nicotine yields at a constant ‘tar’ yield of ca. 8–9 mg/cigarette were not associated with significant changes in puff numbers, volumes, durations, pressures or intervals. Our observations imply that these puffing variables are not controlled by the nicotine yield of the cigarette. Thus, they are consistent with the findings and conclusions of Stepney (1981), Hasenfratz et al. (1993), and Baldinger et al. (1995). As the change in the nicotine yield of our test samples did not influence the puffing measures, the relative differences in the ISO nicotine yields of the samples were maintained in the nicotine yields obtained by the smokers. There was a very high correlation (r=0.998) between ISO and ‘human smoking’ nicotine yields. The calculation of CI, using the sample with the highest nicotine yield as a reference point, revealed CI values close to zero for each lower nicotine-yield sample. Our finding that there was no evidence of compensation for the reduction in nicotine yields in our constant ‘tar’ yield samples is also consistent with the results of Hasenfratz et al. (1993), who observed that compensation occurred when tar yields, but not nicotine yields, were reduced. Although our data clearly show that there was no behavioural compensation for the reductions in the nicotine yields of the samples, it must be stressed that the behavioural measures were focussed on puffing profiles and mouth exposure levels of nicotine. This does not rule out the possibility that the participants compensated for the reduction in nicotine yields by changing inhalation profiles, e.g. inhalation depth and duration. However, as changes in inhalation patterns can influence the magnitudes of CO boosts (Rawbone et al. 1978;
440
Zacny et al. 1987), the observation that the compensation indices based on CO boosts gave no evidence for compensation implies that inhalation profiles were not influenced by the nicotine yields of the test samples. Nicotine plays an important role in some of the sensations perceived by the smoker during the inhalation of cigarette smoke. Kochhar and Warburton (1990) measured the puff-by-puff sensory responses obtained from two cigarettes matched in ‘tar’ yield but differing in nicotine yield. They observed that the higher nicotineyield cigarette produced higher sensory intensities in the throat during cigarette smoke inhalation. Similar findings were reported by Pritchard et al. (1996) who demonstrated that nicotine played an important role in the sensory impact of cigarette smoke and that sensory factors were important for product acceptance and satisfaction. Our results on the sensory effects of changing nicotine but not ‘tar’ yields support these previous findings that nicotine makes a contribution to the sensory properties of cigarette smoke. The sensation of impact was significantly decreased as the nicotine yields were reduced. However, changing the nicotine yield of the cigarette did not influence other sensations such as flavour amplitude and mouthfeel. It has been suggested that nicotine absorbed from cigarette smoke into the systemic circulation stimulates nicotinic receptors in the brain and gives rise to the sensation of ‘impact’ or ‘hit’ (Bates et al. 1999). However, this mode of action is unlikely as the impact sensation occurs very rapidly during cigarette smoke inhalation, and there would be a minimum latency period of 10 s before the nicotine inhaled in a bolus of cigarette smoke could be absorbed and transported to receptor areas of the brain. The timing of the impact sensation suggests that nicotine is acting peripherally rather than centrally. Rose et al. (1998) have hypothesised that nicotinic receptors on peripheral nerve endings in the respiratory tract may be involved in some of the sensory effects of smoking. Recently, Pankow (2001) suggested that the strength of the impact sensation is related to the amount of nicotine evaporating from smoke particles in the throat region. This information, together with data from electrophysiological studies in animals that show tobacco smoke stimulates rapidly adapting afferent receptors in the bronchi (Dixon et al. 1974) and larynx (Boushey et al. 1974), implies that the sensation of impact is initiated by the stimulation of afferent nerve endings in the upper airway by nicotine. The sensation of impact is used by smokers to gauge the inhalation ‘strength’ of a cigarette. A reduction in the ‘tar’ and nicotine yield of the cigarette is associated with a reduction in the perceived strength of the cigarette (Gordin et al. 1987). Indeed, results from surveys of the attitudes of smokers to ‘light’ and ‘ultra-light’ cigarettes support the view that lower yield cigarettes are perceived as being less strong or milder than higher yield ‘regular’ cigarettes (Shiffman et al. 2001). Recently, it has been hypothesised that the reduction in perceived strength is not caused by a reduction in the delivery of ‘tar’ and
nicotine to the smoker. Rather, it has been argued that it occurs as a consequence of an increase in puff volume and subsequent dilution of the smoke by air entering the ventilation holes in the filters of low ‘tar’ cigarettes (Kozlowski et al. 2001). The levels of filter ventilation in the samples used in our current study, and the observed puff volumes taken by the smokers were essentially the same across all of the samples. Thus, it is clear that such a dilution and puff volume effect cannot account for the reduction in the magnitude of the impact sensation observed with the reduced nicotine-yield samples. The most plausible explanation of the reduction in impact magnitude is the reduction in nicotine delivery to the smoker, which we measured. Such a nicotine dose–response effect rather than an ‘air dilution’ effect is likely to be responsible for a reduction in perceived strength of a more highly ventilated lower yield cigarette than a less ventilated higher yield cigarette for two reasons. First, the smoking process involves two distinct phases: the puff process, during which the closure of the soft palate restricts access of the smoke to the oral cavity only (Rodenstein et al. 1985); followed by the inhalation/exhalation process, during which the smoke is mixed with inspired air and is swept from the oral cavity, via the pharynx and larynx, into the lungs (Fairweather 1989). If, as suggested by Kozlowski et al. (2001), filter ventilation results in a smoker taking a larger puff volume and obtaining the same amount of ‘tar’ and nicotine they had obtained from an unventilated cigarette, then the smoke entering the mouth from the ventilated cigarette would be less concentrated than the smoke from the unventilated cigarette. This concentration reduction in the absence of an absolute amount reduction may result in a reduced sensation in the mouth during puff process, e.g. the sensation of mouthful. However, many key ‘strength’ related sensations, e.g. impact and throat irritation, are not initiated during the puff process where dilution may be a factor. These sensations are initiated during the inhalation process. A dilution effect during the puff ceases to be relevant at this point because the smoke components will experience dilution in the residual volume of the mouth and then with a much larger volume—400–1000 ml (USDHHS 1984)— of air during inhalation. Thus, any concentration effect in the puff will be swamped by the much larger dilution processes occurring as inhaled air is mixed with the smoke in the mouth prior to its transport to the sites of the ‘inhalation’ sensations. Second, although many studies report that puff volume is increased following a switch from higher to lower ‘tar’ and nicotine-yield cigarettes, most studies report that smokers receive lower amounts of ‘tar’ and nicotine following such a switch despite the increase in puff volume (Scherer 1999). Thus, a reduction in the concentration of smoke components within the puff will, in general, be associated with a reduction in the amounts of smoke components, such as ‘tar’ and nicotine which is delivered to the smoker during the puff process. Conse-
441
quently, a smoker’s perception of a reduction in the strength of a cigarette following a switch to a lower yield product is likely to be caused by reductions in the yields of smoke components, such as nicotine. Our observations showed that the puffing measures were very similar across the test samples despite the fact that there were large differences between the samples in the nicotine yields and strength of the impact sensation. This finding implies that nicotine-induced sensations initiated during the inhalation phase of smoking do not have a controlling influence on puffing behaviour. This conclusion is consistent with the results of Buzzi et al. (1985) and Dixon et al. (1998) who reported that the nicotine yield of the cigarette did not influence puff-bypuff changes in volume and duration. Additionally, Dixon et al. (1998) demonstrated that puff volume and duration were influenced by sensory responses occurring during the puff process, e.g. mouthful, but not by inhalation sensations. It could be argued that increasing nicotine to ‘tar’ ratios of mainstream cigarette smoke might not have an effect on the compensatory response to reduced-‘tar’ yield cigarettes. This concept of reducing the ‘toxins to nicotine’ ratio was originally proposed by Russell and others (Russell 1976; Warburton 1985) and has more recently been endorsed by Bates et al. (1999). The concept assumes that the compensatory increase in puffing intensity is controlled by the nicotine yield of the cigarette. However, our results suggest that nicotine yield is not a major controlling factor in puffing behaviour. Hence reducing ‘tar’ yields but maintaining nicotine yields may not prevent or attenuate the compensatory response to reduced-yield cigarettes. Finally, although denicotinised/maintained-‘tar’ cigarettes do not result in compensatory changes in smoking behaviour, they appear to be less acceptable to smokers than conventional cigarettes (Baldinger et al. 1995; Rose et al. 1998; Pickworth et al. 1999). Thus, it is clear that nicotine plays an important role in smoking via an effect of nicotine on peripheral sensory mechanisms, a direct pharmacological effect on the central nervous system or a combination of both.
References Armitage AK, Alexander J, Hopkins R, Ward C (1988) Evaluation of a low to middle ‘tar’/medium nicotine cigarette designed to maintain nicotine delivery to the smoker. Psychopharmacology 96:447–453 Baldinger B, Hasenfratz M, Battig K (1995) Switching to ultralow nicotine cigarettes: effects of different tar yields and blocking of olfactory cues. Pharmacol Biochem Behav 50:233–239 Bates C, McNeil A, Jarvis M, Gray N (1999) The future of tobacco product regulation and labelling in Europe: implication for the forthcoming European Union Directive. Tob Control 8:225– 235 Boushey HA, Richardson PS, Widdicombe JG, Wise JCM (1974) The response of laryngeal afferent fibres to mechanical and chemical stimuli. J Physiol 240:153–175 Buzzi R, Nil R, Bttig K (1985) Development of puffing behaviour along burning time of a cigarette—no relation to alveolar
inhalation or nicotine delivery of the cigarettes? Psychopharmacology 86:102–107 Creighton DE, Lewis PH (1978) The effect of different cigarettes on human smoking patterns. In: Thornton RE (ed) Smoking behaviour: physiological and psychological influences. Churchill-Livingstone, Edinburgh, pp 277–288 Creighton DE, Noble MJ, Whewell RT (1978) Instruments to measure, record and duplicate human smoking patterns. In: Thornton RE (ed) Smoking behaviour: physiological and psychological influences. Churchill-Livingstone, New York, pp 289–301 Dixon M, Richardson P, Stransky A, Szereda-Przestaszewska M, Widdicombe JG (1974) Changes in laryngeal calibre due to inhalation of dust. Revue de L’Institut d’Hygine des Mines 29:55–58 Dixon M, Reeves N, Saunders P, Shepperd CJ (1998) Sensoryinduced changes in puffing behaviour. Proceedings of the CORESTA Smoke and Technology Groups joint meeting. Brighton, England, pp 39–51 Dunn PJ, Freiesleben ER (1978) The effects of nicotine-enhanced cigarettes on human smoking parameters and alveolar carbon monoxide levels. In: Thornton RE (ed) Smoking behaviour: physiological and psychological influences. Churchill-Livingstone, Edinburgh, pp 195–202 Fairweather FA (1989) The possible role of factors other than nicotine in compensatory smoking. In: Wald N, Froggatt P (eds) Nicotine and the low tar programme. Oxford University Press, Oxford, pp 212–219 Goldfarb TL, Jarvik ME, Glick SD (1970) Cigarette nicotine content as a determinant of human smoking behaviour. Psychopharmacologia 17:89–93 Gordin HH, Perfetti TA, Mangan PP (1987) A quantification of sensory responses related to dynamic cigarette performance variables. Tob Sci 31:23–27 Gust SW, Pickens RW (1982) Does cigarette nicotine yield affect puff volume? Clin Pharmacol Ther 32:418–422 Hasenfratz M, Baldinger B, Bttig K (1993) Nicotine or tar titration in cigarette smoking behaviour? Psychopharmacology 112:253– 258 Herning RI, Jones RT, Benowitz N, Mines AH (1981) Puff volume increases when low nicotine cigarettes are smoked. BMJ 283:187–189 Institute of Medicine (2001) Clearing the smoke: assessing the science base for tobacco harm. National Academies Press, Washington ISCSH (1988) Fourth report of the Independent Scientific Committee on smoking and health. Her Majesty’s Stationary Office, London Jacob P, Byrd GD (1999) Use of gas chromatographic and mass spectrometric techniques for the determination of nicotine and its metabolites. In: Gorrod JW, Jacob P (eds) Analytical determination of nicotine and related compounds and their metabolites. Elsevier, Amsterdam, pp 191–224 Kochhar N, Warburton DM (1990) Puff-by-puff sensory evaluation of a low to middle tar, medium nicotine cigarette designed to maintain nicotine delivery to the smoker. Psychopharmacology 102:343–349 Kozlowski LT, O’Connor RJ, Sweeney CT (2001) Cigarette design. In: 2001 smoking and tobacco control, monograph 13. Risks associated with smoking cigarettes with low machine-measured yields of tar and nicotine. USDHHS National Institutes of Health, Washington, pp 13–37 Pankow JF (2001) A consideration of the role of gas/particle partitioning in the deposition of nicotine and other tobacco smoke compounds in the respiratory tract. Chem Res Toxicol 14:1465–1481 Pickworth WB, Fant RV, Nelson RA, Rohrer MS, Henningfield JE (1999) Pharmacodynamic effects of new de-nicotinized cigarettes. Nicotine Tob Res 1:357–364 Pritchard WS, Robinson J, Guy TD, Davis RA, Stiles MF (1996) Assessing the sensory role of nicotine in cigarette smoking. Psychopharmacology 127:55–62
442 Rawbone RG, Murphy K, Tate ME, Kane SJ (1978) The analysis of smoking parameters: inhalation and absorption of tobacco smoke in studies of human smoking behaviour. In: Thornton RE (ed) Smoking behaviour: physiological and psychological influences. Churchill-Livingstone, Edinburgh, pp 171–194 Robinson JH, Pritchard WS, Davis RA (1992) Psychopharmacological effects of smoking a cigarette with typical ‘tar’ and carbon monoxide yields but minimal nicotine. Psychopharmacology 108:466–472 Rodenstein DO, Stanescu DC (1985) Pattern of inhalation of tobacco smoke in pipe, cigarette, and never smokers. Am Rev Respir Dis 132:628–632 Rose J, Westman E, Behn F, Johnson M, Goldberg J (1998) Blockade of smoking satisfaction using the peripheral nicotinic antagonist trimethaphan. Pharmacol Biochem Behav 62:165– 172 Rose JE, Behm FM, Westman EC, Coleman RE (1999) Arterial nicotine kinetics during cigarette smoking and intravenous nicotine administration: implications for addiction. Drug Alcohol Depend 56:99–107 Royal College of Physicians (2000) Nicotine addiction in Britain. A report of the Tobacco Advisory Group of the Royal College of Physicians. London Russell MAH (1976) Low tar, medium nicotine cigarettes: a new approach to safer smoking. BMJ 1:1430–1433 Scherer G (1999) Smoking behaviour and compensation: a review of the literature. Psychopharmacology 145:1–20
Shiffman S, Pillitteri JL, Burton SL, Rohay JM, Gitchell JG (2001) Smokers’ beliefs about ‘light’ and ‘ultra light’ cigarettes. Tob Control 10[Suppl 1]: i17–i23 Stephen A, Frost C, Thompson S, Wald N (1989) Estimating the extent of compensatory smoking. In: Wald N, Froggatt P (eds) Nicotine and the low tar programme. Oxford University Press, Oxford, pp 100–115 Stepney R (1981) Would a medium nicotine, low tar cigarette be less hazardous to health ? BMJ 283:1292–1303 Sutton SR, Russell MAH, Iyer R, Feyerabend C, Saloojee Y (1982) Relationship between cigarette yields, puffing patterns and smoke intake: evidence for tar compensation ? BMJ 285:600– 603 US Department of Health and Human Services (1981) The health consequences of smoking. The changing cigarette: a report of the Surgeon General. US Department of Health and Human Services, Washington US Department of Health and Human Services (1984) The health consequences of smoking—chronic obstructive lung disease. Report of the Surgeon General. US Department of Health and Human Services, Washington Warburton DM (1985) Nicotine and the smoker. Rev Environ Health 5:343–390 Zacny JP, Stitzer ML, Brown FJ, Yingling JE, Griffiths RR (1987) Human cigarette smoking: effects of puff and inhalation parameters on smoke exposure. J Pharmacol Exp Ther 240:554–564
