Disparities in self-reported exposure to tobacco marketing among youth and young adults from low-socioeconomic status neighbourhoods in Mexico City

ABSTRACT

Exposure to tobacco marketing is positively associated with smoking initiation and behaviours. There is limited literature examining disparities among reported exposure in low- and middle-income countries. This study sought to fill this gap with a survey among 1427 adolescents and 889 adult smokers in Mexico City in 2020. Data were analysed using chi-square and hierarchical regression models. Two-thirds of adolescents noticed cigarette pack displays in stores. Participants from low- and mid-socioeconomic status (SES) neighbourhoods were more exposed to tobacco marketing than their counterparts through several channels. After addressing the shared variance among participants from the same household nested in neighbourhood SES level and controlling for gender, adolescent non-smokers and adult smokers who noticed pack displays were more likely to be susceptible to smoking (OR = 1.8, 95% CI 1.4–2.4) and to smoke more (OR = 1.7, 95% CI 1.21–2.47); adult smokers who noticed tobacco marketing at more places were less likely to be certain about smoking risks (OR = 0.9, 95% CI 0.8–0.9). The results suggest that the tobacco industry targets youth and individuals from low-SES neighbourhoods through several channels. Greater exposure to tobacco marketing was associated with increased susceptibility to smoking and decreased risk perception. These findings support a comprehensive ban on tobacco marketing in Mexico.

KEYWORDS:

• Tobacco industry
• cigarettes
• policy
• global health
• middle-income country

Introduction

Mexico ratified the World Health Organization's (WHO's) Framework Convention on Tobacco Control (FCTC) in 2004. Since ratification, the Mexican General Law for Tobacco Control (GLTC) has included several key FCTC-recommended policies including prohibiting advertising and promotion at the point-of-sale (POS), promotional distribution of tobacco products, and all forms of sponsorship from the tobacco industry (SEGOB, 2010). However, tobacco promotion and publicity aimed at adults, such as in magazines and via mail, remains allowable (SEGOB, 2010).

Exposure to tobacco marketing has been shown to be positively associated with smoking initiation and continuation among adolescents and adults (Madkour et al., 2014; Mantey et al., 2019; Nicksic et al., 2017). Globally, the tobacco industry spends tens of billions of dollars each year to market its products using both direct and indirect tobacco advertising, promotion, and sponsorship (TAPS) strategies (WHO, 2013a). Several studies show that the tobacco industry has implemented strategies targeting youth and people from low socioeconomic status (SES) neighbourhoods (Green et al., 2007; Isip & Calvert, 2020; Lee et al., 2015; van der Eijk et al., 2019; WHO, 2013a). Research in Canada, the United Kingdom, and the U.S. suggests that adolescents are more likely to be susceptible to smoking when exposed to tobacco products display at the POS (Bogdanovica et al., 2015; Johns et al., 2013; MacKintosh et al., 2012; Spanopoulos et al., 2014) and tobacco promotion campaigns (Choi & Forster, 2014; Duke et al., 2019; Soneji et al., 2014). Johns et al. (2013) found that exposure to tobacco marketing increased the probability of smoking initiation among the New York City adolescents. Most studies examining the correlation between tobacco marketing and adolescents’ smoking behaviours have been conducted in high-income countries, with a limited number in lower income countries in Africa (Cadmus & Ayo-Yusuf, 2018; Chido-Amajuoyi et al., 2017; English et al., 2016) and Asia (Arora et al., 2012; Saito et al., 2013). There is also evidence that the tobacco industry has targeted socioeconomically disadvantaged populations (Green et al., 2007). Lee et al. (2015) systematically reviewed 43 studies on POS tobacco marketing (41 conducted in high-income countries and two in lower-income counties) and found that people living in socioeconomically disadvantaged neighbourhoods were disproportionately exposed to tobacco marketing. When exploring disparities in tobacco marketing, many studies have focused on the association between the neighbourhood characteristics (i.e. neighbourhood income or any other indicator of socioeconomic disadvantage) and marketing for tobacco products at the POS (Lee et al., 2015). Studies in the U.S. suggest that there tend to be more exterior tobacco advertisements in lower-income neighbourhoods (Frick et al., 2012) and residents are more likely to experience violations to tobacco advertising and labelling regulations (Rose et al., 2013).

However, very limited research has been conducted to examine disparities in reported exposure to tobacco marketing among youth and individuals from low-SES neighbourhoods in middle-income countries, such as Mexico. Existing literature on tobacco marketing in Mexico is mainly focused on exposure to a few selected marketing channels, rather than a comprehensive consideration of advertising and promotion channels. One study in 2012 found that exposure to tobacco promotion increased overtime for free sample, branded clothes or other items, and tobacco industry sponsored events among adult smokers in Mexico (Pérez-Hernández et al., 2012). RuízHernández et al. (2018) found that women cigarette smokers in Mexico were more exposed to tobacco advertising and promotion through tobacco products displays at the POS and posters. Another study also found elevated exposure to tobacco marketing at the POS among Mexican adolescents (Valdés-Salgado et al., 2006). More recent research in Mexico City found that cigarette pack displays at the POS were the most prominent form of tobacco marketing (Institute for Global Tobacco Control, 2019).

This study seeks to fill the dearth of research examining disparities in reported exposure to tobacco marketing in middle-income countries by exploring tobacco marketing exposure among Mexico City residents through various channels. Two research questions were posed: (1) Does exposure to tobacco marketing and promotion vary by sociodemographic characteristics [i.e. age, gender, SES]? and (2) Is exposure to tobacco marketing and promotion associated with smoking behaviours, susceptibility, and health risk perception?

Methods

Sample and recruitment

This manuscript reports results from a household survey in Mexico City that is the major element of a mixed-methods study in which exploratory focus groups with adolescents and young adults were conducted first in the U.S., and then in Mexico City in the development of the survey instrument (Grilo et al., 2021). The study design, results, and dissemination were also discussed extensively with tobacco control advocates in Mexico to ensure that our research was aligned with tobacco control policy priorities for the country and to obtain their perspectives and feedback to effectively present findings, including by translating key messages into Spanish. These approaches to public involvement helped maximise the relevance and potential impacts of this research (de Freitas, 2017).

Adolescents (13–17 years) and young adult smokers (18–34 years) completed a computer-based, self-administered survey in Mexico City in early 2020. While tobacco use usually initiates during adolescence, smoking behaviours become established during early adulthood (Brook et al., 2008). The inclusion of young adult smokers allowed us to assess their marketing exposure in comparison especially to adolescent smokers, which is particularly relevant in Mexico because part of the tobacco advertising and promotion ban allows certain promotions and advertising when directed at adults (ages 18 years and older) (Valdés-Salgado et al., 2006). Berumen and Associates, a Mexican-based market research firm, conducted household recruitment using a quota-based sampling procedure to assure that gender (men, women), age (13–17, 18–24, and 25–34 years), and SES (low, mid, high) were proportionally represented in the sample. Using population-level demographic and socioeconomic data, Mexico City was divided into low-, mid-, and high-SES regions, which were selected to maximise geographic distribution and achieve SES-quotas. City blocks were then randomly selected within each region.

Recruiters visited all households within each selected block and screened all interested individuals within the same household for eligibility. Inclusion criteria for adolescents were current residents of residency in Mexico City, aged between 13 and 17 years, and able to speak and read Spanish. Adolescents’ parent/guardian provided oral consent before eligibility screening and before starting the survey for eligible adolescents. Inclusion criteria for adults were residency in Mexico City, aged between 18 and 34 years, able to speak and read Spanish, and currently a cigarette smoker (defined as having smoked in the past 30 days). Adolescents and adults provided electronic assent/consent before starting the survey and, upon its completion, received a voucher equivalent to $100 MXN (∼ $5 USD).

Survey design

The survey included demographic items from the Global Youth Tobacco Survey (GYTS)/Global Adult Tobacco Survey (GATS) asking about participants’ age, gender, educational attainment, and SES. Participants’ smoking behaviours were measured by 13 items drawn from GYTS/GATS and ENCODAT, the National Survey on Drugs, Alcohol, and Tobacco Consumption developed by the Mexican Ministry of Health. The first item asked whether the respondent had ever smoked a cigarette. Adolescents who answered ‘no’ were then asked about their smoking curiosity and susceptibility. Those who reported ever having smoked a cigarette were then asked about their lifetime smoking history. Participants who reported having smoked in the past 30 days were asked to indicate how many days they smoked and how many cigarettes smoked per day. Participants who had smoked the equivalent of at least one pack of cigarettes in their lifetime were asked to indicate their brand preference and to answer questions about perceived addictiveness and the harms of smoking. Questions drawn from ENCODAT assessed participants’ exposure to tobacco marketing at the POS and in multiple venues as well as exposure to different tobacco promotions.

Variables

Smoking behaviour and perception of smoking

Number of cigarettes smoked per day. Participants reported the daily number of cigarettes smoked over the past 30 days. Response options were: (1) I did not smoke a cigarette in the past 30 days, (2) Less than 1 cigarette per day, (3) 1 cigarette per day, (4) 2–5 cigarettes per day, (5) 6–10 cigarettes per day, (6) 11–20 cigarettes per day, and (7) More than 20 cigarettes per day. In the sample, 55.4% smoked between 2 and 10 cigarettes/day and 34% smoked one cigarette/day or less. Based on this, categorical responses were recoded as dichotomous: those who smoked less than or equal to 1 cigarette per day versus smoking more than 1 cigarette per day.

Susceptibility to tobacco use. A susceptibility index was used to identify adolescent non-smokers who were at risk of future tobacco use (Nodora et al., 2014; Pierce et al., 1996; Strong et al., 2015). Adolescents were classified as not susceptible to tobacco use (susceptibility = 0) if they responded: ‘not at all curious’ to the question ‘Have you ever been curious about smoking a cigarette?’ and ‘definitely not’ to questions ‘Do you think that you will try a cigarette soon?’ and ‘If one of your best friends were to offer you a cigarette, would you smoke it?’. Adolescents who did not provide these responses to all three questions were considered susceptible (susceptibility = 1) as they did not clearly commit to not smoking. Cronbach's ɑ coefficient among the three questions was 0.83 based on the study sample, showing high index reliability (Cortina, 1993). Spearman's correlation coefficients between the first question and the second, and third question, and the second and the third within this sample were 0.63 (p < 0.001), 0.55 (p < 0.001), and 0.70 (p  <  0.001), respectively. These findings support the reliability and validity of the index in this sample.

Perceived smoking danger. Participants reported their perception of whether smoking was a danger to their health with response options of (0) definitely no, (1) probably no, (2) probably yes, and (3) definitely yes. Categorical responses were recoded as dichotomous: those who responded ‘definitely yes’ as ‘1 – yes’ and others as ‘0 – not sure’.

Tobacco marketing exposure

Noticed cigarette packs in stores. Participants reported whether they recalled noticing (1 – yes) cigarette packs or not (0 – no) on cabinets or counters inside stores during the past 30 days. If they said ‘yes’, they were asked about their perceptions regarding the pack display, i.e. if they thought the pack display was appealing/not appealing, striking/not striking, colourful/not colourful, organised/disorganised. Both measures are part of the receptivity spectrum.

Number of advertising channels noticed. Participants responded whether they noticed any tobacco marketing during the past 30 days, in stores that sell cigarettes, billboards, newspapers/magazines, movie theatres, the Internet, public transportation or stations, public streets, and other places or media. A composite variable was created with a range of 0–9 that counted the number of places or media where tobacco advertising was noticed.

Number of promotion types noticed. Participants responded whether they noticed any tobacco marketing type during the past 30 days including free cigarette samples, discounted cigarettes, coupons, gifts or special discounts with cigarette purchase, clothes or other items with a cigarette brand or trademark, and mail marketing. A composite variable that counted the number of tobacco promotion types noticed was created with a range of 0–6.

Covariates

Sociodemographic characteristics. Participants’ sociodemographic data included gender, age, and SES. Gender was indicated as either (1) male, (2) female, or (3) prefer not to respond. Age was categorised as (1) adolescents (13–17 years old) or (2) young adults (18–34 years old). Neighbourhood SES was categorised as (1) low-SES, (2) mid-SES, or (3) high-SES.

Smoking status. Adolescents and young adults were considered current smokers if they had ever smoked and smoked one cigarette in the past 30 days.

Statistical analysis

Statistical analyses were performed using SPSS® for Windows®, version 27.0 (SPSS Inc., Chicago, IL, USA) and SAS, version 9.4 (SAS Institute). Descriptive statistics were used to describe the studied population and variables. Differences in tobacco marketing exposure, smoking behaviours and sociodemographic characteristics between adolescents and young adults and between smokers and non-smokers were examined by Chi-square test, independent t-test, and analysis of variance (ANOVA). The association between tobacco marketing exposure and adolescents’ and young adults’ smoking behaviours and adolescent non-smokers’ perceived smoking health risks and susceptibility was determined using hierarchical mixed-effect logistic regression models with random intercepts for household effect and two levels: the household and the neighbourhood SES level. Gender was included in the model as a covariate. We used the listwise deletion method to deal with the missing data (2% of the total participants). For each analysis, a two-sided p < 0.05 was used to determine statistical significance.

Ethics approval

The study was approved by Institutional Review Boards at the Johns Hopkins School of Public Health in the U.S. and at the Instituto Nacional de Salud Pública in Mexico.

Results

Sample characteristics

Table 1 presents the characteristics of the study sample (N = 2316) for 1427 adolescents and 889 young adult smokers, by smoking status. Adolescent smokers and young adult daily smokers were more likely to be male (64.4% and 62.4%, respectively) (p < 0.001). Nearly one-third of adolescent smokers (30.7%) smoked more than one cigarette per day. Among adolescent non-smokers, approximately two-thirds (63.1%) were susceptible to smoking. Adolescent smokers were less likely to be certain about the danger of smoking to health than adolescent non-smokers, young adult non-daily smokers, and daily smokers (p < 0.01).

Table 1. Sociodemographic characteristics, smoking behaviours, perceived smoking danger, and tobacco marketing exposure by age group and smoking status (N = 2316).

		Adolescents		Young adults
		Non-smokers	Smokers	Non-daily Smokers	Daily Smokers
		(n = 1163)	(n = 264)	(n = 687)	(n = 202)
Sociodemographic characteristics
Gender	Boy/Man	46.3%	64.4%	46.6%	62.4%
	Girl/Woman	53.4%	35.6%	53.1%	37.1%
	Choose not to identify	0.3%	0.0%	0.3%	0.5%
Neighbourhood	Low SES	29.8%	30.3%	28.7%	33.7%
	Middle SES	38.3%	37.1%	38.6%	34.7%
	High SES	31.9%	32.6%	32.8%	31.7%
Smoking related issues
Number of cigarettes per day	>1 cigarette	–	30.7%	58.4%	85.6%
Danger of smoking to health	Definitely yes	71.8%	61.4%	70.0%	79.7%
Susceptibility to smoking	Susceptible	63.1%	–	–	–
Tobacco marketing exposure
Noticed cigarette packs in stores	% ‘Yes’	60.5%	72.7%	73.9%	76.7%
	% no response	1.3%	1.5%	0.7%	0.0%
Number of advertising channels noticed	Mean	1.78	2.27	2.36	2.25
	(Std. error)	(0.07)	(0.16)	(0.10)	(0.20)
	% no response	1.3%	0.4%	0.4%	0.5%
Number of promotion types noticed	Mean	0.65	0.76	0.84	0.99
	(Std. error)	(0.04)	(0.09)	(0.06)	(0.11)
	% no response	1.8%	0.8%	0.6%	2.0%

Notes: % no response, % of participants who did not respond and % of participants who reported that they did not know; number of advertising channels noticed, ranged 0–9; number of promotion types noticed, ranged 0–6.

Exposure to tobacco marketing

The top three channels through which participants reported being exposed to tobacco marketing were in stores (45.9%), the Internet (32.3%), and newspapers/magazines (27.4%) (Table 2); the top three tobacco promotions that participants reported noticing were clothes or other items with a cigarette brand or trademark (20.0%), discount cigarettes (17.1%), and gifts or special discounts with cigarette purchase (12.8%) (Table 3). Approximately two-thirds of all participants (68.0%) reported that they noticed cigarette packs inside stores (Table 4). Over 70% of those who noticed packs inside stores agreed that their display was appealing or striking and over 80% agreed that the display was colourful or organised. Reported exposure to tobacco marketing did not differ by gender. Percentages of young adult daily and non-daily smokers reporting exposure to tobacco marketing were similar, except for noticing tobacco marketing on posters (higher among non-daily smokers; p < 0.05) and noticing gifts or special discounts with cigarette purchase (higher among daily smokers; p < 0.01).

Table 2. Exposure to tobacco marketing in the past 30 days, by channel (N = 2198).

		Total n	Mean (SD)	Stores	Billboards	Posters	Newspapers/magazines	Movie theatres	Internet	Public trans./stations	Public streets/places	Other sites
			(0–9)	‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)
All Participants		2198	2.05	45.9%	24.3%	26.2%	27.4%	13.0%	32.3%	19.9%	16.3%	12.8%
			(2.54)	(43.8−47.9)	(22.5−26.0)	(24.4−28.1)	(25.6−29.3)	(11.6−14.4)	(30.4−34.3)	(18.3−21.6)	(14.8−17.8)	(11.4−14.1)
Age/Smoking Status	Adolescents	1342	1.87	42.2%	21.0%	23.4%	25.7%	11.1%	31.2%	18.0%	15.5%	12.0%
			(2.42)	(39.5−44.8)	(18.8−23.2)	(21.2−25.7)	(23.4−28.1)	(9.5−12.8)	(28.7−33.7)	(16.0−20.1)	(13.5−17.4)	(10.3−13.8)
	Non-smokers	1090	1.78	40.1%	20.8%	22.1%	24.2%	10.3%	29.9%	17.2%	14.3%	11.7%
			(2.36)	(37.2−43.0)	(18.3−23.2)	(19.6−24.5)	(21.7−26.8)	(8.5−12.2)	(27.2−32.6)	(14.9−19.4)	(12.2−16.4)	(9.8−13.6)
	Smokers	252	2.27	51.2%	22.1%	29.5%	32.4%	14.6%	36.9%	21.8%	20.7%	13.5%
			(2.62)	(45.0−57.4)	(17.0−27.3)	(23.8−35.1)	(26.6−38.1)	(10.3−19.0)	(31.0−42.9)	(16.7−26.9)	(15.7−25.7)	(9.3−17.7)
	Young adult smokers	856	2.34	51.6%	29.4%	30.6%	30.0%	15.9%	34.1%	22.9%	17.6%	13.9%
			(2.71)	(48.3−55.0)	(26.3−32.4)	(27.5−33.6)	(27.0−33.1)	(13.4−18.3)	(30.9−37.3)	(20.1−25.7)	(15.1−20.2)	(11.6−16.3)
	Non-daily	656	2.36	50.9%	30.4%	32.3%	31.2%	15.8%	34.1%	22.7%	17.8%	14.4%
			(2.68)	(47.1−54.7)	(26.8−33.9)	(28.7−35.9)	(27.7−34.8)	(13.0−18.6)	(30.5−37.8)	(19.5−25.9)	(14.8−20.7)	(11.7−17.0)
	Daily	200	2.25	54.0%	26.0%	24.7%	26.0%	16.1%	34.0%	23.4%	17.1%	12.6%
			(2.82)	(47.1−60.9)	(19.9−32.1)	(18.8−30.7)	(20.0−32.1)	(11.0−21.2)	(27.5−40.6)	(17.6−29.3)	(11.9−22.3)	(8.0−17.2)
Gender	Boy/Man	1095	2.05	46.2%	23.7%	25.3%	26.7%	13.4%	31.4%	20.5%	16.6%	12.6%
			(2.58)	(43.3−49.2)	(21.2−26.2)	(22.7−27.8)	(24.1−29.3)	(11.4−15.4)	(28.7−34.2)	(18.2−22.9)	(14.4−18.8)	(10.7−14.6)
	Girl/Woman	1097	2.05	45.5%	24.8%	27.1%	28.1%	12.4%	33.1%	19.4%	15.8%	12.7%
			(2.51)	(42.5−48.4)	(22.3−27.4)	(24.5−29.7)	(25.5−30.8)	(10.4−14.3)	(30.4−35.9)	(17.0−21.7)	(13.6−18.0)	(10.8−14.7)
Neighbourhood SES	Low	653	2.24	45.2%	28.0%	28.4%	31.9%	14.5%	36.3%	22.6%	17.5%	15.9%
			(2.67)	(41.4−49.0)	(24.6−31.5)	(24.9−31.8)	(28.3−35.5)	(11.8−17.2)	(32.6−40.0)	(19.3−25.8)	(14.5−20.4)	(13.1−18.7)
	Mid	835	2.15	46.0%	24.4%	27.5%	29.1%	13.3%	35.3%	21.6%	18.5%	12.3%
			(2.61)	(42.6−49.4)	(21.5−27.3)	(24.5−30.5)	(26.0−32.2)	(11.0−15.6)	(32.1−38.6)	(18.8−24.4)	(15.8−21.1)	(10.1−14.6)
	High	710	1.76	46.3%	20.6%	22.8%	21.3%	11.2%	25.1%	15.5%	12.7%	10.4%
			(2.31)	(42.7−50.0)	(17.6−23.6)	(19.7−25.9)	(18.3−24.3)	(8.9−13.5)	(21.9−28.3)	(12.9−18.2)	(10.2−15.1)	(8.2−12.6)

Table 3. Exposure to tobacco marketing in the past 30 days, by type of marketing (N = 2155).

		Total n	Mean (SD)	Cigarette free sample	Discount cigarette	Coupons	Gifts or discounts with cig purchase	Clothes or other items with cigarette branding	Mail marketing
			(0−6)	‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)
All Participants		2155	0.75	11.2%	17.1%	6.1%	12.8%	20.0%	12.4%
			(1.42)	(9.9−12.6)	(15.5−18.7)	(5.1−7.2)	(11.4−14.2)	(18.4−21.7)	(11.0−13.8)
Age/Smoking Status	Adolescents	1302	0.67	10.6%	16.1%	5.2%	10.5%	18.8%	11.5%
			(1.30)	(8.9−12.3)	(14.1−18.1)	(4.0−6.4)	(8.8−12.1)	(16.7−20.9)	(9.8−13.3)
	Non-smokers	1053	0.65	10.2%	15.4%	4.6%	10.4%	18.4%	11.6%
			(1.27)	(8.3−12.0)	(13.2−17.5)	(3.4−5.9)	(8.6−12.3)	(16.0−20.7)	(9.6−13.5)
	Smokers	249	0.76	12.4%	19.0%	7.6%	10.8%	20.6%	11.4%
			(1.42)	(8.3−16.6)	(14.2−23.9)	(4.3−10.8)	(6.9−14.6)	(15.6−25.6)	(7.5−15.4)
	Young adult smokers	853	0.87	12.2%	18.7%	7.6%	16.2%	22.0%	13.7%
			(1.59)	(10.0−14.4)	(16.0−21.3)	(5.8−9.3)	(13.7−18.6)	(19.2−24.8)	(11.4−16.1)
	Non-daily	660	0.84	13.2%	17.9%	7.8%	14.5%	21.0%	12.6%
			(1.58)	(10.6−15.8)	(14.9−20.8)	(5.8−9.8)	(11.8−17.2)	(17.9−24.1)	(10.1−15.2)
	Daily	193	0.99	8.8%	21.5%	6.7%	21.8%	25.4%	17.5%
			(1.62)	(4.8−12.8)	(15.7−27.3)	(3.2−10.3)	(15.9−27.6)	(19.2−31.5)	(12.2−22.9)
Gender	Boy/Man	1079	0.73	11.0%	16.4%	6.3%	12.3%	19.8%	11.5%
			(1.42)	(9.2−12.9)	(14.1−18.6)	(4.8−7.7)	(10.4−14.3)	(17.4−22.1)	(9.6−13.4)
	Girl/Woman	1070	0.77	11.4%	17.6%	6.0%	13.1%	20.3%	13.2%
			(1.43)	(9.5−13.3)	(15.3−19.9)	(4.5−7.4)	(11.1−15.1)	(17.9−22.8)	(11.2−15.3)
Neighbourhood SES	Low	640	0.81	12.7%	17.3%	8.0%	13.5%	19.3%	15.8%
			(1.51)	(10.1−15.2)	(14.3−20.2)	(5.9−10.1)	(10.8−16.1)	(16.2−22.3)	(13.0−18.6)
	Middle	813	0.75	11.7%	17.3%	6.2%	12.6%	20.8%	10.5%
			(1.43)	(9.5−13.9)	(14.7−19.9)	(4.6−7.9)	(10.3−14.9)	(18.0−23.6)	(8.4−12.6)
	High	702	0.70	9.4%	16.7%	4.4%	12.4%	19.8%	11.5%
			(1.33)	(7.2−11.6)	(14.0−19.5)	(2.9−5.9)	(9.9−14.8)	(16.9−22.8)	(9.2−13.9)

Table 4. Number and percentage of participants who noticed cigarette packs on cabinets or counters inside stores in the past 30 days and ratings of cigarette pack display inside stores (N = 2292).

		Noticed cigarette packs inside stores		Ratings of cigarette pack displays inside stores
		Total N		Total n that noticed cigarette packs inside stores	Appealing	Striking	Colourful	Organised
			‘Yes’ (%) (95% CI)		‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)	‘Yes’ (%) (95% CI)
All Participants		2292	68.0%	1559	71.6%	73.8%	83.6%	84.5%
			(66.1−69.9)		(69.4−73.9)	(71.6−75.9)	(81.8−85.5)	(82.7−86.3)
Age/Smoking Status	Adolescents	1408	63.6%	896	63.4%	64.5%	77.6%	79.4%
			(61.1−66.1)		(60.2−66.5)	(61.4−67.6)	(74.8−80.3)	(76.7−82.0)
	Non-smokers	1148	61.3%	704	56.8%	58.9%	74.1%	75.6%
			(58.5−64.1)		(53.2−60.5)	(55.3−62.6)	(70.9−77.4)	(72.4−78.7)
	Smokers	260	73.8%	192	87.5%	84.9%	90.1%	93.2%
			(68.5−79.2)		(82.8−92.2)	(79.8−90.0)	(85.9−94.3)	(89.7−96.8)
	Young adult smokers	884	75.0%	663	82.8%	86.3%	91.9%	91.4%
			(72.1−77.9)		(79.9−85.7)	(83.7−88.9)	(89.8−93.9)	(89.3−93.5)
	Non-daily	682	74.5%	508	82.3%	86.2%	92.1%	91.7%
			(71.2−77.8)		(79.0−85.6)	(83.2−89.2)	(89.8−94.5)	(89.3−94.1)
	Daily	202	76.7%	155	84.5%	86.5%	91.0%	90.3%
			(70.9−82.6)		(78.8−90.2)	(81.1−91.8)	(86.5−95.5)	(85.7−95.0)
Gender	Boy/Man	1139	68.6%	781	72.7%	75.9%	84.5%	86.0%
			(65.9−71.3)		(69.6−75.9)	(72.9−78.9)	(82.0−87.0)	(83.6−88.5)
	Girl/Women	1148	67.4%	774	70.7%	71.6%	82.8%	82.9%
			(64.7−70.1)		(67.5−73.9)	(68.4−74.8)	(80.2−85.5)	(80.3−85.6)
Neighbourhood SES	Low	687	66.4%	456	72.6%	75.0%	82.9%	84.4%
			(62.8−69.9)		(68.5−76.7)	(71.0−79.0)	(79.4−86.4)	(81.1−87.8)
	Mid	866	68.4%	592	68.8%	71.3%	81.6%	82.6%
			(65.3−71.5)		(65.0−72.5)	(67.6−74.9)	(78.5−84.7)	(79.5−85.7)
	High	739	69.1%	511	74.2%	75.5%	86.7%	86.7%
			(65.8−72.5)		(70.4−78.0)	(71.8−79.3)	(83.7−89.6)	(83.7−89.6)

Tobacco marketing exposure

Adolescent smokers and young adult daily and non-daily smokers reported similar exposure to tobacco marketing overall. Over 80% of adolescent and young adult smokers noticed tobacco displays in stores and found them to be attractive. On average, adolescent smokers and young adult daily and non-daily smokers noticed tobacco marketing in more than two places (range 0–9) (2.27, 2.25, and 2.36, respectively) (Table 1). Adolescent smokers were more likely than non-smokers to report noticing cigarette packs inside stores and finding the display of packs attractive (p < 0.001). Adolescent non-smokers were less likely to report noticing tobacco marketing in stores, on posters, newspapers/magazines, the Internet, and public streets and places than adolescent smokers (p < 0.05) (Table 2). However, adolescent smokers and non-smokers reported similar exposure to all the various types of tobacco promotion (Table 3). Overall, a high percentage of adolescents reported noticing cigarette pack displays inside stores and found them appealing, striking, colourful, or organised (Table 4).

Adolescent smokers were less likely to notice tobacco marketing on billboards (21.0%) and gifts or discounts with cigarette purchase (10.5%) than young adult smokers (29.4% and 16.2%, respectively) (p < 0.05).

Tobacco marketing exposure among smokers from low-SES neighbourhoods

Participants from low-SES neighbourhoods were more exposed to tobacco marketing than their counterparts from mid- or high-SES neighbourhoods through several channels. As presented in Tables 2 and 3, participants from low- and mid-SES neighbourhoods were more likely to notice tobacco marketing on the Internet (36.3% and 35.3%), newspapers/magazines (31.9% and 29.1%), posters (28.4% and 27.5%), billboards (28.0% and 24.4%) and at public transportation or stations (22.6% and 21.6%) compared to their high-SES counterparts (25.1%, 21.3%, 22.8%, 20.6%, 15.5%, respectively) (p < 0.05). Furthermore, a significantly higher percentage of participants from low-SES neighbourhoods reported that they noticed tobacco mail marketing (15.8%) than those from mid- (10.5%) or high-SES neighbourhoods (11.5%) (p < 0.01). About 8% of participants from low-SES neighbourhoods reported exposure to tobacco coupons, which was significantly higher than participants from high-SES neighbourhoods (4.4%) (p < 0.01).

Association between tobacco marketing exposure and smoking behaviour and perceived danger of smoking

The hierarchical mixed effects logistic regression models with a 2-level random intercept were applied, with levels 1 units (households in neighbourhoods) nested in level 2 units (neighbourhoods defined by SES level) to examine the association between tobacco marketing exposure and smoking behaviour and perceived danger of smoking. In the model for susceptibility to smoking, adolescents who noticed cigarette packs inside stores were more likely to be susceptible to smoking after controlling for gender (OR = 1.80, 95% CI 1.37–2.36) (Table 5). Among young adults, the hierarchical logistic regression models demonstrated that exposure to cigarette pack displays inside stores was positively associated with smoking more than one cigarette per day and exposure to tobacco marketing at more places was negatively associated with participants’ perception of smoking danger after controlling for gender (OR = 1.73, 95% CI 1.21–2.47; OR = 0.90, 95% CI 0.83–0.97, respectively) (Table 5).

Table 5. Hierarchical logistic regression models for the association of tobacco marketing exposure and susceptibility to smoking, perception of danger of smoking to health, and amount smoked (N = 2316).

	Adolescent non-smokers (n = 1163)		Adolescent smokers (n = 264)		Young adult smokers (n = 889)
	Susceptibility to smoking	Perceived danger of smoking	Number of cigarettes smoked per day	Perceived danger of smoking	Number of cigarettes smoked per day	Perceived danger of smoking
Fixed effects	Odds ratio (95% CI)	Odds ratio (95% CI)	Odds ratio (95% CI)	Odds ratio (95% CI)	Odds ratio (95% CI)	Odds ratio (95% CI)
Noticed cigarette packs in stores	1.80* (1.37−2.36)	2.20 (0.70−6.99)	4.26 (0.03−708.56)	1.82 (0.02−216.86)	1.73*** (1.21−2.47)	0.77 (0.48−1.21)
Number of places noticed marketing	1.06 (0.99−1.14)	1.08 (0.79−1.47)	1.05 (0.47−2.37)	0.89 (0.38−2.10)	0.99 (0.93−1.06)	0.90***** (0.83−0.97)
Number of promotion types noticed	0.95 (0.84−1.08)	0.73 (0.44−1.20)	0.69 (0.13−3.67)	0.81 (0.17−3.76)	1.10 (0.98−1.23)	0.97 (0.86−1.11)
Sex (Girl/Woman)	0.68** (0.53−0.88)	2.84 (0.97−8.34)	1.03 (0.02−49.54)	0.76 (0.02−35.02)	0.57**** (0.42−0.77)	1.01 (0.71−1.43)
Model fit (−2 log-likelihood)	1444.49	958.60	289.56	312.59	1082.15	988.34

*Significant at p < 0.0001 level.

**Significant at p < 0.004 level.

***Significant at p < 0.003 level.

****Significant at p < 0.0003 level.

*****Significant at p < 0.01 level.

Discussion

To the best of our knowledge, this is the first study to assess broad tobacco marketing exposure and examine its association with smoking behaviours, susceptibility, and perceived health risks in Mexico. Overall, our results showed that adolescents were exposed less to tobacco marketing than young adults; however, adolescent smokers were more exposed to tobacco marketing than non-smokers and, in certain cases, at rates similar to young adult smokers. Nevertheless, levels of exposure among adolescents were high. Results also showed that individuals from low-SES neighbourhoods were more exposed than their higher-SES counterparts. Higher exposure to cigarette displays in stores was positively associated with daily smoking and susceptibility and higher exposure to multiple types of tobacco marketing was negatively associated with smoking risk perception.

In this study, the top channel where participants reported noticing tobacco advertising was stores (45.9%) and the top type of tobacco marketing participants reported was clothes or other items with a cigarette brand or trademark (20.0%). These findings correspond with results from the 2015 Mexico Global Adult Tobacco Survey (GATS) (Pan American Health Organization & Instituto Nacional de Salud Pública, 2015). However, the percentages of participants who noticed tobacco advertising through several channels and types of promotion were greater in our study than GATS because we only recruited young adults who smoked. This study also found that nearly one-third of young adult non-daily smokers noticed tobacco advertising on posters, which was significantly higher than young adult daily smokers (32.3% vs. 24.7%). It may be that this finding is driven by gender differences, with young women being more likely to notice the information on posters than young men; a significantly greater percentage of the non-daily smokers are women compared to daily smokers.

The tobacco industry purposely targets adolescents through aggressive marketing tactics in order to maintain their customer base (Campaign for Tobacco Free Kids, 2014; West et al., 2012). This study demonstrates that levels of tobacco marketing exposure among adolescents were high. Further, through several channels studied, adolescent smokers were exposed to tobacco marketing at similar levels to young adult smokers. Nearly one-third of adolescent smokers (32.4%) reported being exposed to tobacco marketing in newspapers/magazines which is of particular concern considering that the GLTC prohibits this type of marketing for minors (<18 years old). These data suggest that this has not prevented exposure. In addition, among adolescents who noticed cigarette displays inside stores, smokers were more likely to consume more than one cigarette daily and non-smokers were more likely to be susceptible to smoking even after accounting for the covariance among those from the same household nested in the same neighbourhood and controlling for gender. A comprehensive TAPS policy would help to reduce tobacco marketing exposure among adolescents and might subsequently decrease smoking behaviour (English et al., 2016). Banning cigarette pack displays in stores as part of the policy is another effective mechanism to prevent initiation and increase cessation, particularly given the exposure levels reported here (Robertson et al., 2016).

This study also suggested that young adult smokers who noticed tobacco marketing in more places were less likely to be certain about the health dangers of smoking. Cho et al. (2018) found that perceived health risks of smoking were associated with knowledge of toxic constituents in Mexico (Cho et al., 2018). Sustained education and communication about health risks of smoking can enhance people's understanding of smoking-related risks (World Health Organization [WHO], 2013a, 2013b, 2013c), counter the negative effects of tobacco promotion (Moran et al., 2019), and potentially reduce youth smoking prevalence (Duke et al., 2019).

Existing literature indicates that individuals of lower SES or living in lower income neighbourhoods are disproportionately exposed to storefront and outdoor tobacco marketing at POS (Lee et al., 2015; Moran et al., 2019; Seidenberg et al., 2010). This study adds to the literature by examining self-reported tobacco marketing exposure through channels in addition to POS marketing. Participants from low- and mid-SES neighbourhoods in this study were more likely to be exposed to tobacco marketing through the Internet, newspapers/magazines, posters, billboard and public transportation or stations than high-SES participants. The tobacco industry targets low-income individuals with discounting strategies, such as coupons, and other price promotions or incentives (Brown-Johnson et al., 2014; Seidenberg et al., 2010). This study also found that participants from low-SES neighbourhoods were more likely to report exposure to tobacco mail marketing and coupons than participants from high-SES neighbourhood as previously reported (Choi & Forster, 2014). Regulating tobacco price discounting or incentives strategies may help reduce the disparities and decrease smoking prevalence in low-SES neighbourhoods.

Findings from this study support the need for a comprehensive TAPS ban in Mexico and similar settings, although our data also suggest that the partial ban may offer some protection against tobacco marketing for adolescents. Previous studies have shown that partial TAPS bans are less effective in reducing tobacco use, as the tobacco industry can use the loopholes in the regulations to market its products via available channels (Blecher, 2008; Saffer, 2000). Blecher (2008) found that a comprehensive ban resulted in a 7% decrease in per capital tobacco consumption while partial bans had no significant impact on consumption. A comprehensive TAPS ban also has the potential to reduce socioeconomic disparities, protect the youth and low-SES populations from smoking, and decrease tobacco consumption at the population level (WHO, 2013a).

Findings from this study are the results of a large, household-based survey using both validated measures of tobacco use and marketing exposure, and also building on findings from robust, exploratory work. We assessed exposure to tobacco marketing through different channels at the individual level in a middle-income country. The sample was constructed to provide broad socioeconomic diversity. We sought to reduce social desirability bias by allowing participants the opportunity to answer smoking related questions privately using a tablet. As with all research, this work is subject to several limitations. First, we employed quota-based recruitment to enrol adolescents and smoking young adults. Our results may not, therefore, be representative of the general population aged 13–34 years in Mexico City nor other regions of Mexico. In addition, smoking behaviours and marketing exposure were self-reported, and thus subject to response bias. The smoking behaviours section of the survey focused on key individual behaviour questions in relation to smoking and it did not include measures related to parents/friends’ smoking status and living with tobacco users, all potential cofounders that could influence the results. Moreover, because some participants were smokers or were susceptible to smoking, they might have recalled exposure to tobacco marketing at higher rates given their already existing interest compared to non-smokers. The cross-sectional design of this study prevents us from making any claims about the causal nature of the association between tobacco marketing exposure and tobacco use outcomes. Future studies could collect individual-level SES data and utilise a random sampling method or survey weights to establish representativeness and generalisability. It would also be valuable to examine disparities in tobacco marketing exposure in other low- and middle-income countries.

Conclusion

The purpose of this study was to observe tobacco marketing exposure among adolescents and young adults through various channels in Mexico City and explore the association between exposure and smoking behaviours and risk perceptions. Findings from this study are in line with existing research demonstrating that elevated exposure to tobacco marketing is associated with increased daily smoking and susceptibility and decreased perception of smoking health risks. This study also found that adolescents were exposed to high levels of tobacco marketing; and that individuals from low-SES neighbourhoods were more likely to report being exposed to tobacco marketing via several channels and types of promotion than individuals from high-SES neighbourhoods. The harmful effects of tobacco marketing on smoking perceptions and behaviours among vulnerable populations reinforce the need for stronger tobacco advertising, promotion, and sponsorship restrictions in Mexico, including banning the display of tobacco products at the POS and prohibiting advertising and promotion that are currently allowed for adults.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported with funding from Bloomberg Philanthropies’ Bloomberg Initiative to Reduce Tobacco Use (www.bloomberg.org).