Article Text

Download PDFPDF

Toxicant exposure from smoking a little cigar: further support for product regulation
  1. Wallace B Pickworth,
  2. Zachary R Rosenberry,
  3. Bartosz Koszowski
  1. Battelle Memorial Institute, Public Health Center for Tobacco Research, Baltimore, Maryland, USA
  1. Correspondence to Dr Wallace B Pickworth, Battelle Memorial Institute, Public Health Center for Tobacco Research, 6115 Falls Rd, Baltimore, MD 21209, USA; pickworthw{at}battelle.org

Abstract

Background Although numerous studies have documented the prevalence and increasing use of little cigars and other cigar products, the present study is the first direct, head-to-head laboratory comparison of little cigar and cigarette smoking. The study addressed a fundamental objective to compare exposure and use characteristics of little cigar and cigarette smoking.

Methods Smoking patterns, toxicant exposure and subjective measures were collected and analysed in 21 adults after smoking a little cigar (Winchester) and a cigarette (own brand). Participants were dual users of little cigars and cigarettes.

Results Similar to cigarettes, little cigars delivered substantial nicotine and relatively more carbon monoxide. Puff volume, puff duration and time to smoke were significantly greater after cigarettes, but the temporal pattern of smoking more intensively at the beginning was similar in little cigars and cigarettes. Both little cigars and cigarettes reduced urge to smoke. Participants consistently mentioned that the lower cost of little cigars was a reason for initiation and continuation of their use.

Conclusions The results support the notion that regulation of little cigars is appropriate in light of public health considerations.

  • Smoking topography
  • Nicotine
  • Non-cigarette tobacco products
  • Toxicology
View Full Text

Statistics from Altmetric.com

Introduction

In an attempt to establish a comprehensive federal response that addressed cigarette labelling and advertising, the US Congress passed the Cigarette Labelling and Advertising Regulations Act (1965) that defined a cigarette as a ‘roll of tobacco wrapped in paper’ and a little cigar as a ‘roll of tobacco wrapped in leaf tobacco or any substance containing tobacco’ and ‘to which one thousand units weigh not more than three pounds’.1 ,2 Over the years, popular marketing designations of cigar products have evolved to the current state where three products predominate: little cigars (eg, Winchester), similar to cigarettes in size, filtered with a cellulose acetate filter and packaging (20/package); cigarillos (eg, Black & Mild), intermediate in weight (1–3 g; 120 mm or more), may be filtered, usually flavoured, tipped with a plastic or wooden mouthpiece, sold as singles or in packages of six articles or fewer; large cigars (eg, Phillies Blunt), may be flavoured, typically weigh 3 g or more, may be hand rolled or machine made, sold as singles or in packs of six articles or fewer. Another popular designation of a cigar product is the premium cigar, a hand-rolled imported tobacco that is quite expensive, large and almost always sold individually. The present report uses the term little cigar throughout to refer to products like the Winchester that was tested in the laboratory study.

The Family Smoking Prevention and Tobacco Control Act (FSPTCA, 2009)3 authorised the US Food and Drug Administration (US FDA) to regulate cigarettes, tobacco for roll-your-own cigarettes and smokeless tobacco products (eg, moist snuff, chewing tobacco). This authority has led to important regulations on cigarettes such as the banning of distinguishing flavours (except menthol), prohibition of descriptors such as ‘light’ and ‘ultralight’, and limitation to sales in packages of 20 or more articles. However, these restrictions have not been applied to the sale of cigar products. There are now hundreds of cigar products on the market that are offered in a variety of flavours (eg, fruit, spice, wine); some advertised as ‘light’; some sold individually or in packages of fewer than 20; and their sale can be promoted through the use of coupons, free samples and deeply discounted prices and rebates.4 Furthermore, the price of a pack of little cigars is typically less than a pack of cigarettes making them more appealing to price-sensitive populations including youth.4 The legal definitions of cigars for purposes of continuity, labelling and taxation, which were enacted years ago, served to distinguish cigars from cigarettes; however, this distinction now excludes cigars from the public health-promoting protections of the FSPTCA.

In contrast to declining US cigarette sales, cigar use has steadily increased over the past 10 years.5 ,6 Increased cigar use has been driven by increased use of cigarillos and little cigars among younger, urban, and Hispanic and African-American men.7–9 Typically, little cigar consumers use both cigar product and a conventional cigarette (dual users),9–11 although the epidemiology of US domestic cigar use is obfuscated by multiple definitions of cigar products and the finding that cigar users themselves are uncertain about the classification of the products.10 ,12 In a recent study of African-American cigarillo (Black & Mild) smokers,13 five unique subgroups were identified that were distinguished by use frequency (daily vs non-daily) and cigar manipulation behaviour (blunts and hyping) as well as flavour choice. These results suggest that there may be subgroups of smokers among little cigar users as well. Although there are numerous studies that have documented the prevalent and increasing use of little cigars and other cigar products5 ,7 ,12 ,14–17 and epidemiological studies that indicate cigar use has adverse health effects,18–20 there have been no studies that directly compared little cigar and cigarette smoking in people who ordinarily use both products. In the present laboratory study, smoking behaviour, toxicant exposure and qualitative effects of little cigar smoking were compared with cigarette smoking in dual users of cigarettes and little cigars. Comparative nicotine exposure was emphasised because of the recognition of nicotine as an addictive drug.21 The study was designed to address a fundamental area of importance—a scientific enquiry comparing the exposure and use characteristics of little cigar and cigarette smoking.

Methods

Procedures have been described previously.22 Briefly, participants completed a two-visit study in which they were randomised to smoke either a Winchester little cigar (85 mm, filtered, filter-ventilated) or their own brand of cigarette at each session. The Winchester product was chosen because it has been a market-leading brand in sales (eighth overall—3.2% market share in 2012)23 and was designed to be as close to a cigarette as legally possible.24 As in previous studies,22 ,25 overnight tobacco abstinence was not required prior to the experimental visits. Two postsmoking blood samples were collected to assess peak nicotine levels which may occur slightly later after little cigar compared with cigarette smoking if significant buccal absorption occurs as has been demonstrated with moist snuff.26

Dependent measures

Puff measures

Smoking topography describes how a person puffs (brings smoke into their mouth) a tobacco product.27–29 Measures of topography include the number of puffs, total puff volume (TPV), average puff volume (APV), puff duration (PD), puff velocity (PVel), interpuff interval (IPI) and time to smoke (TTS). It was previously reported in the literature that the pH of the mainstream smoke of cigars changes during smoking which may influence nicotine bioavailability and potentially smoking behaviour.30 ,31 Thus, topography variables were compared for the first three and the last three puffs to assess temporal smoking patterns for cigarettes and little cigars. Furthermore, differences in the first three and last three puffs were observed in cigarettes, make-your-own cigarettes32 and cigarillo smoking.22 Smoking topography was measured using a SPA-D Puff Analyser (Sodim Instruments, MebTEC, Mebane, North Carolina, USA).

Toxicant exposure (tobacco smoke biomarkers)

Plasma nicotine: Venous blood samples were drawn to assess changes in plasma nicotine levels, before and after (5 and 10 min) smoking. The blood samples were centrifuged and the plasma was separated and stored frozen until analysis for nicotine concentration by the Bioanalytical Laboratory at Virginia Commonwealth University (VCU) School of Pharmacy. The lower limit of quantification for the assay was 2.5 ng/mL. Nicotine exposure from cigarettes and little cigars was normalised for the amount of tobacco consumed (ng/mL/g of tobacco) and for mouth-level exposure by adjusting for TPV (ng/mL/1000 mL TPV).

Exhaled carbon monoxide: Exhaled carbon monoxide (COex) is a recognised biomarker of recent tobacco smoke exposure and smoke inhalation.32 ,33 COex was collected using a Breath CO Monitor (Vitalograph Inc, Lenexa, Kansas, USA) at baseline and within 10 min after smoking. COex exposure was normalised for grams of tobacco smoked (ppm/g) and for mouth-level exposure by adjusting for TPV (ppm/1000 mL).

Urinary cotinine: Cotinine, the primary metabolite of nicotine, was analysed from urine samples taken at baseline and used to describe this sample of participants. Cotinine has a longer half-life (16 h) than nicotine (90 min), which provides a more stable assessment of nicotine exposure.34 The urine samples were analysed by Labstat International ULC (Kitchener, Ontario, Canada). The lower limit of quantification for the cotinine assay was 8.78 ng/mL. The results were corrected for creatinine.

Subjective measures

Tobacco use history: Tobacco use history was obtained at baseline. The questionnaire evaluated the participants' demographics, current and smoking history and types of tobacco products used. Information was also collected regarding age of initiation, and brand and flavour preference.

Nicotine dependence: Nicotine dependence was assessed using the Fagerström Test for Nicotine Dependence (FTND)35 at baseline.

Subjective evaluations: The level of nicotine craving was evaluated using the brief version of the Questionnaire on Smoking Urges (QSU)36–38 administered before and 20 min after smoking. The appeal and subjective effects of the products postsmoking were evaluated 20 min after smoking by means of the Duke Sensory Questionnaire (DSQ)39 and Cigarette Evaluation Scale (CES),40 which both employ a seven-point Likert scale with 1=‘not at all’ and 7=‘extremely’. The DSQ contains nine questions related to: puff liking, puff satisfaction, nicotine in puffs, puff strength on the tongue, nose, mouth and throat, windpipe, and chest, and similarity to own brand. DSQ ratings for puff strength on tongue, nose, mouth and throat, windpipe, and chest were collapsed to form an overall strength score (range 7–35). Questions on similarity to own brand were excluded when participants smoked their own cigarette and were therefore not included in statistical analyses. The CES contains 11 questions related to cigarette satisfaction, good taste and effects (dizziness, calmness, concentration, wakefulness, hunger reduction, nausea, irritability, enjoyment of sensations in the throat and chest and reduction of cigarette craving). Several of these items were collapsed to form composite scores of satisfaction (satisfaction and good taste), psychological reward (calmness, concentration, wakefulness, hunger reduction and irritability); DSQ and the CES were administered after smoking. Participants were also asked open-ended questions about their reasons for starting and continuing little cigar use.

Sample size and statistical analysis

Normality was assessed using skewness and kurtosis tests, Shipiro-Wilk tests and visual inspections of the data distribution. It was determined that COex was not normally distributed and was therefore transformed using a natural log transformation for analysis. Repeated measures analysis of variance (rANOVA) was performed to examine differences among dependent variables collected both before and after smoking (ie, CO, plasma nicotine concentration, product weight and QSU). A 2×2 rANOVA with fixed effects for product type (cigarette and little cigar), time (presmoking and postsmoking), and the interaction of product type and time was fitted to the data separately for each outcome of interest. A random effect for participant was also included in the rANOVA model to account for repeated measures within a participant. The model results are presented in table 2 with means and SDs. Additionally, one-way rANOVA models with a random effect for participant were fitted to data collected once during each product use (ie, puff topography, DSQ, CES and measures of exposure controlled for grams of tobacco consumed and 1000 mL of TPV) to determine whether results were significantly different by product. The results of the one-way rANOVA models including the f value and p value for the product type effect are presented in table 3. Paired samples t tests were used for related variables that were not assessed at each smoking condition (ie, age of initiation for each product type). A 2×2 ANOVA with fixed effects for product type, time and the interaction between product type and time was also used to compare topography variables averaged over the first and last three puffs; Scheffe post hoc tests were used to identify significant contrasts. Statistical analyses were conducted using Stata V.13.1 and Statistica V.12.

Results

Participants

The study sample consisted of 21 participants who met eligibility criteria and attended both smoking sessions. Participant characteristics are shown in table 1. The sample consisted of mostly African-American men with an average age of 48 years who were nicotine dependent based on the mean number of cigarettes smoked per day,17 the Fagerström score (6.8) and urinary cotinine corrected for creatinine (1270 ng/mL/g of creatinine).

Table 1

Participant demographics and smoking characteristics

Tobacco use history

Smoking characteristics of the sample are also included in table 1. All participants reported regularly smoking Hats Off little cigars and most regularly smoked Newport cigarettes. Most participants smoked menthol cigarettes and menthol flavoured little cigars. Generally, participants began smoking cigarettes before little cigars; the average age of first cigarette (12.4 years) was significantly less than the average age of first little cigar use (33.9 years, p<0.001).

Dependent measures

Toxicant exposure

Plasma nicotine: The peak plasma nicotine concentration occurred at 5 min postsmoking (compared with 10 min after smoking) and was used in the rANOVA models shown in table 2 below. There was a significant effect of time on plasma nicotine as indicated by p value (p<0.001). There was no effect of product type or interaction with time. The means and SDs of plasma nicotine concentration for both product types before and after smoking are also presented in table 2.

Table 2

2×2 rANOVA models of outcome measures

Exhaled carbon monoxide: As shown in table 2, there was a significant effect of time for COex (p<0.001). There was no effect of product type nor was the interaction significant. The means and SDs of COex both before and after smoking each product are also shown in table 2.

Product weight: Cigarettes and little cigars were weighed before and after smoking. There was a significant effect for product type (p<0.001) and time (p<0.001). There was also a significant interaction between product type and time (p<0.001) showing the effect of product type on product weight over time. The means and SDs of both products before and after smoking are also presented in table 2 below. Based on this finding of a significant interaction term, post hoc analyses were conducted to measure differences in exposure relative to the amount of tobacco that was consumed during a single smoking session. Exposure was also adjusted for TPV. The results of these post hoc analyses are presented in table 3.

Table 3

One-way rANOVA models of outcome measures

Subjective measures

Appeal and effects: The means and SDs for QSU scores before and after smoking are shown in table 2. The results of the 2×2 rANOVA are also shown in table 2. There was a significant effect of time for factor 1 (p<0.001), factor 2 (p<0.001) and total QSU scores (p<0.001). There was also a significant effect of product type on QSU factor 1 scores (p=0.048). The interaction between product type and time was also significant (p=0.044). The effect of product type and interaction was not significant for factor 2 or total QSU scores.

One-way rANOVA models examining the effect of product type on DSQ and CES are presented in table 3 below. There was a significant effect of product type for puff liking (p<0.001), puff satisfaction (p=0.002) and puff strength (p=0.002) on the DSQ. Results from the CES showed a significant effect of product type on sensation (p=0.002), craving reduction (p=0.008), satisfaction (p<0.001) and aversion (p=0.013).

Reasons for initiating and continuing smoking little cigars were assessed by open-ended questions. Results of this qualitative questionnaire are not depicted in tables. Little cigar initiation and continuation were strongly influenced by price considerations; more than any other reason given including use with other substances (drug and/or alcohol), flavouring appeal, peer influence, experimentation and product availability. Specifically, cost was mentioned as a factor in initiation of little cigar use by 38% of the sample and as a reason for continuing use by 52% of the sample.

Puff measures

Summary statistics for smoking topography measurements are presented in table 3 along with the effect of product type in a one-way rANOVA model. There was a significant effect of product type for number of puffs (p<0.001), TPV (p<0.001), TTS (p<0.001), APV (p<0.001), PVel (p<0.001) and PD (p=0.042). The pattern of topography over the course of smoking is illustrated in figure 1 showing the average topography variables for the first three and the last three puffs for cigarette and little cigar smoking. For cigarettes and little cigars, APV (p=0.002 and p<0.001, respectively) and PD (both p<0.001) were larger over the first three puffs compared with the last three puffs. IPI was significantly shorter over the first three puffs compared with the last three puffs for both products (both p<0.001). PVel of the first three puffs was similar to that of the last three puffs for both products.

Figure 1

Puff by puff analysis. Puff parameters for cigarettes and little cigars showing the first three puffs (1, 2 and 3) and the last three puffs (X, Y, and Z) from participants (N=21) smoking. Comparisons of the mean of the first three and the last three puffs indicated significant changes (all F>9.06; p<0.01) changes in puff volume, puff duration and puff velocity and interpuff interval. Post hoc Scheffe tests indicating significant differences for puff velocity, puff duration and interpuff interval (all p<0.01).

Normalised toxicant exposure

Differences in normalised toxicant exposure were assessed using one-way rANOVA with results shown in table 3. There was no effect of product type on the change in plasma nicotine per unit of exposure. However, there was a significant effect of product type on the change in COex per gram of tobacco smoked (p=0.013) and per 1000 mL TPV (p<0.001). The mean changes in normalised toxicant exposure (and SDs) are also presented in table 3.

Discussion

In the past few years, there have been dramatic changes in domestic tobacco use. Cigarette use has continued on a slow decline,5 while the use of some alternative nicotine delivery devices such as electronic cigarettes41 have increased dramatically. In addition, oral nicotine products42 are being introduced and promoted by the tobacco industry as alternatives when smoking is not allowed. There has also been a huge increase in the sales and marketing of cigar products. Cigars are not currently regulated by the US FDA. One type of cigar, little cigars, is about the same size as conventional cigarettes, filtered and sold in packages of 20. Unlike cigarettes, little cigars may be advertised as ‘lights’ and are available in many families of flavours (eg, mint, fruit and liquor). Several epidemiological surveys have demonstrated substantial use of little cigars especially among minority youth. For example, nearly 16% of US young adults reported smoking cigars at least once monthly;9 ,12 cigar use was identified as a significant tobacco product in a national survey by the Centers for Disease Control and Prevention (CDC)7 and further indicated that many choose flavoured cigars.43–45 An interesting advertising strategy has been used to promote the use of little cigars for use in a quick smoke break when smoking a large cigar is impractical.46 In spite of their current popularity, there have been relatively few studies that have directly compared the smoking behaviour and the toxicant exposure of little cigars and cigarettes. In the present study, smoking topography, pharmacological effects and biomarkers of exposure were compared between cigarette and little cigar smoking.

Smoking topography variables indicate significantly different puffing patterns between little cigars and cigarettes. As shown in table 3, compared with cigarette smoking, little cigar smoking was associated with significantly fewer puffs, shorter TTS and PD, and smaller APV, TPV and PVel. IPI was similar in both smoking conditions. Some smoking topography variables are dependent on the size of the product (TTS, TPV, number of puffs) whereas other variables are independent of the size (APV, PVel, IPI). In a previous study, we observed consistency in parameters between factory-made commercial cigarettes and make your own cigarettes.32

In the present study and in studies of cigarillo smoking47 ,48 and cigarette smoking in adults49–51 and adolescence,52 ,53 smoking was not constant over the length of the tobacco rod. Generally, APV, PD and PVel are larger and the IPI is smaller at the beginning of a cigarette than at the end of the cigarette. We observed the same pattern in the cigarette and little cigar smoking in the present study—puffing intensity was greater at the beginning of the smoking. The participants in this study were all experienced cigarette smokers who began smoking cigarettes before little cigars. Tiffany54 suggested that the patterns of smoking are established fairly early in one's tobacco experience and that smoking patterns persist even if the product being smoked changes. The physical properties of the mainstream smoke may also contribute to changes in puff topography. When little cigars are consumed, there is an increase in the pH of the mainstream smoke from 6.5 to 7.830 ,31—this effectively increases the amount of unprotonated readily bioavailable nicotine from approximately 3–38% suggesting that less intensive puffing at the end of the tobacco rod may suffice to deliver the quantities of nicotine derived from more intensive smoking at the beginning of the tobacco rod.

The significant time effect of nicotine exposure in table 2 shows that cigarettes and little cigars increased plasma nicotine from presmoking to postsmoking (20.1 and 17.2 ng/mL increases, respectively). When the plasma nicotine exposure was normalised per gram of tobacco consumed and per 1000 mL TPV, there was no significant effect of product type (table 3). These results indicate that little cigars and cigarettes produced similar increases in nicotine exposure even after accounting for differences in consumption.

There was also a significant effect of time on COex exposure as shown in table 2. COex increased after smoking cigarettes and little cigars (29–35 and 28–36 ppm, respectively). After adjustments for the weight of tobacco consumed and differences in TPV, there was a significant effect of product type on CO exposure. Specifically, compared with cigarettes, little cigars caused nearly double the COex exposure per gram of tobacco consumed or per 1000 mL TPV. Overall, the data emphasise that little cigar smoking is associated with similar delivery of nicotine and CO. However, little cigars deliver much greater amounts of CO compared with cigarettes after adjusting for consumption. Enhanced CO generation from cigar smoke has been observed in studies from our laboratory and others,22 ,33 ,47 ,55 and is usually attributed to less complete combustion of the tobacco filler due to the larger diameter, more dense packing and less porous wrapping.55

Dual users of little cigars and cigarettes were recruited, so that they would be familiar with both products and have smoking histories that are reflective of US little cigar users. Their smoking history, measures of nicotine dependence and urinary cotinine levels were consistent with those of established smokers. Although the participants were older than the little cigar users in national surveys, they were similar in being male, minority and by smoking both cigars and cigarettes. Surveys indicate that about 80% of little cigar and cigarillos users also smoke cigarettes.12 ,56 ,57 The participants in the study consistently mentioned the cost of the little cigars as an important reason for initiating and continuing their use. Little cigars are generally cheaper than cigarettes.58 For example, in a tobacco outlet store in northern Virginia, a package of generic little cigars cost $1.36 (1839) and a premium brand little cigar (Captain Black) was priced at $4.39; whereas a generic pack of cigarettes was priced at $2.52 and a premium cigarette (Newport) was $5.25. Little cigar and cigarette dual users frequently indicated the importance of cost in the initiation and continuation of little cigar use, but, interestingly, cost was not mentioned by dual users of cigarettes and large cigars or cigarillos.59 In the present study, the differences in price were important determinants of use to the low-income participants. Together with their lower price, the ability of little cigars to rapidly deliver equal or greater quantities of nicotine than cigarettes may influence the attractiveness of little cigars to youth and lower income individuals. Further research is needed to identify the relationships among product cost, use behaviour in particular subgroups and risk perceptions and their relationships with toxicant exposure.

There were limitations to the research study that are acknowledged and considered when evaluating the generalisability of the results. We tested a single little cigar product of the many that are commercially available;60 the little cigar that was tested (non-menthol) was not the little cigar ordinarily smoked by the participants. Furthermore, the unfamiliar little cigar smoking was compared with the smoking of a familiar cigarette suggesting that some of the difference could be attributed to novelty of the little cigar, its flavour and the differences in the length of the little cigar and the cigarette (usually 100 mm). We tested exposure to only nicotine and carbon monoxide from the over 4000 components of tobacco smoke identified61 and participants were not tobacco abstinent at the time of the experimental session. More research is warranted on the influence of product design (eg, filter ventilation), flavouring and smoking behaviour on toxicant exposure from cigar smoking.

Nevertheless, we observed that little cigars, like cigarettes, increased plasma nicotine levels, increased CO exposure, and decreased urge to smoke and had similar ratings on satisfaction. Like the smoke from cigarettes, the smoke from little cigars was inhaled and some puffing variables were similar to cigarette smoking. Both cigarette and little cigar smoking delivered equivalent, significant and addictive62 quantities of nicotine. Currently, little cigars are exempt from US FDA regulation because of an outdated, largely arbitrary classification based on manufacturing standards2 and not on public health considerations. Little cigars are sized and appear like cigarettes, packaged and sold like cigarettes, and used like cigarettes. Given these similarities, we propose that little cigars be regulated like cigarettes. The ideal regulation would be through the auspices of the US FDA, so that there would be uniformity across all states, but in lieu of US Federal regulations, there are mechanisms for state and local governments to regulate non-cigarette tobacco products.63

What this paper adds

  • Little cigars are similar to cigarettes in size and packaging and are increasingly popular among youth and minority smokers, but there are no studies that directly compare smoking behaviour and toxicant exposure between cigarettes and little cigars.

  • The results show that, like cigarette smoking, little cigar smoking causes rapid exposure to addictive levels of nicotine and decreases urge to smoke. Little cigar smoking causes more exposure to carbon monoxide and potentially other harmful mainstream smoke components than cigarette smoking. The results support the notion that little cigars should be regulated like cigarettes.

Acknowledgments

The authors gratefully acknowledge the technical assistance of Jennifer Bayne, Alieu Kanu, Jessica Stotler and Carson Smith; Jennifer Potts and Lauren Viray for administrative support and two pharmacy students from the University of Maryland, Daniel Yi and Yevgeniya Kalinina, who provided editorial assistance.

References

View Abstract

Footnotes

  • Contributors All the authors substantially contributed to the conception or design of the work; or the acquisition, analysis or interpretation of data for the work; drafted the work or revised it critically for important intellectual content; approved the final version of the manuscript submitted to the Tobacco Control Journal; agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

  • Funding The research was supported by a grant from National Institutes of Health, National Cancer Institute 1R01CA158045 (Pickworth, PI).

  • Disclaimer The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

  • Competing interests None declared.

  • Ethics approval Battelle Memorial Institute's Institutional Review Board.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.