The cigarette pack is an important marketing tool for the tobacco industry and has grown in relevance as companies face increasing advertising restrictions in other mediums [1–4]. Pack shape, color, and design all communicate a brand’s “personality” to attract consumer attention [1, 4]. The use of specific colors, imagery, and text descriptors (e.g. light, mild, smooth) are especially important as they can influence perceptions of product taste and strength [5, 6], as well as the health risks associated with use [6–10]. Tobacco companies intentionally tailor packaging features to appeal to specific consumer subgroups based on demographic and lifestyle factors [11–14]. Importantly, the pack functions as a major advertising platform to draw in and recruit new smokers, including women and girls who rated feminine branded packs, such as the bright pink and black Camel No. 9 pack, as more appealing than non-feminine or non-branded, plain packs [15, 16].

The tobacco industry has long marketed smoking as a socially acceptable activity for women in high-income, Western countries like the United States, Great Britain, and Spain [17–22]. Starting in the 1920s, advertising campaigns in women’s lifestyle magazines promoted cigarettes using themes of independence, fashion, and thinness, (e.g. “Reach for a Lucky instead of a sweet”) [21–23], and the use of the cigarette as a symbol of glamour, femininity, and women’s emancipation persisted well throughout the late 20th century [22–24]. With respect to packaging, the print advertising campaigns worked in concert with the design of the cigarette and the cigarette pack itself. Tobacco companies extensively researched women’s personal and social preferences related to smoking [4, 20]. The cigarette pack often carried over the Western, stereotypical feminine appeals in the form of imagery, (e.g. flowers, butterflies [22, 25, 26]) and colors, (e.g. pastels, pink [22, 24]) of the advertising campaign, and slim or ultra-thin cigarettes were designed to reinforce the perception of smoking as a feminine, graceful, stylish activity that aligned with the idealized thin and glamourous images of women promoted in magazine ads [20, 22, 25]. Additionally, both packaging and advertisements offered different technologies that met and fortified the odor and taste preferences of women: reduced odor side-stream smoke, a light or mild cigarette that was smoother to smoke and appeared as a “healthier option,” and improved flavor through the use of menthol, mint and other flavor constituents [4, 20].

Collectively, these marketing strategies were effective at increasing cigarette sales and smoking rates among women [27]. Although women still smoke at a lower rate compared to men in nearly all countries [28], women and girls remain a potential market for long-term growth among the major transnational tobacco companies [26, 29–31]. This is particularly relevant in low-and middle-income countries (LMICs) where shifting norms around the acceptability of women smoking and increased spending power among women may facilitate smoking uptake [22, 29, 32, 33]. In 2018, less than 5% of women living in LMICs used combustible tobacco compared to 17.9% of women living in high-income countries [28]. However, in those LMICs with the highest global burden of tobacco use, rates of smoking among women is more variable. For example, while less than 3% of women in Bangladesh, China, Egypt, Pakistan, India, Indonesia, Pakistan, and Thailand smoked in 2018, the smoking rate among women in Mexico (6.5%), the Philippines (7.0%), Brazil (9.5%), Ukraine (9.9%), Russia (15.7%), and Turkey (17.0%) were much higher and some approached smoking rates among women in higher-income countries like the United States (16.7%) [28].

Further, evidence suggests that major transnational tobacco companies are employing similar marketing strategies previously used in high-income countries to target women in LMICs [22, 30, 33]. For example, cigarette advertising campaigns in countries like India and the Philippines frequently featured fashionably dressed women who conformed to Western style and beauty standards which may associate smoking with aspirations of glamour, sophistication, and independence from traditional economic and social roles held by women in LMICs [22, 26, 30, 33–35]. In addition, slim cigarettes, lipstick pack shaped cigarettes, and “light” or “mild” cigarettes have also been introduced into the marketplace of several LMCIs [22, 26, 35, 36]. Slim cigarettes that appear to be designed for women were associated with increased odds of experimental smoking among adolescents in China [35], and experimental studies conducted among young women in Mexico (16–18 years old) and Brazil (16–26 years old) found that participants rated flavored cigarette packs branded with feminine colors, (e.g. light pastels, pink) as more appealing than flavored, non-branded plain packs [36, 37]. However, little is known about the extent to which cigarette packs sold in LMICs feature feminine imagery or appeals outside of pack shape and reduced harm claims. Only a small number of studies have examined marketing features present on cigarette packs sold in LMICs [38–42], and none to date has systematically explored the presence of feminine appeals.

The current study addresses this gap and utilizes a large dataset of cigarette packs collected across 14 LMICs to examine the presence of feminine marketing appeals on packs sold in LMICs. Feminine appeals were initially informed by stereotypical, Western features, (e.g. color, fashion appeals, flowers, butterflies, and imagery of women) in advertising and pack design discussed in the extant literature and then refined based on input from country-specific experts. We document trends in the use of these overt feminine marketing appeals on packs over time and by multinational tobacco company, as well as the co-occurrence of feminine appeals with other marketing features historically used to appeal to women, (e.g. slim pack shape, flavor, “light” cigarettes, reduce odor claims). Given the role of the cigarette pack as a marketing platform to attract women and girls and increase interest in product use [15, 16, 24, 35], results from this study can elucidate the ways in which women in LMICs may be targeted by tobacco companies through cigarette packaging and inform policy strategies to reduce the impact of this marketing tactic.

This study is part of the larger Tobacco Pack Surveillance System (TPackSS) project to monitor compliance with health warning label requirements on tobacco packages sold in 14 LMICs with the highest number of smokers at the time of initial data collection [43]. The primary aim of the TPackSS project is to purchase a comprehensive collection of unique tobacco packs sold in each TPackSS country and assess whether tobacco health warning requirements are implemented as intended. Data have been published on overall compliance with warning label requirements [44], as well as pack features that may detract from the effectiveness of warning label placement on packages, (e.g. tax stamps covering warnings) [45]. Data have also been used to explore the different marketing appeals, (e.g. English language, sports imagery, organic/natural descriptors, reduced harm terminology) present on tobacco packs across the study countries [38, 39, 41, 42].

TPackSS data collection occurred in two waves. In Wave 1 (2013), data collectors purchased cigarette packs (n = 3,240) from 14 LMICs: Bangladesh, Brazil, China, Egypt, India, Indonesia, Mexico, Pakistan, Philippines, Russia, Thailand, Turkey, Ukraine and Vietnam. In Wave 2 (2015–2017), data collectors purchased cigarette packs (n = 2,336) sold in nine of the original 14 LMICs where health warning label requirements were updated since Wave 1: Indonesia, Russia, Thailand, and Vietnam (packs purchased in 2015); Bangladesh, Brazil, India, and the Philippines (packs purchased in 2016); and China (packs purchased in 2017). Wave 2 data were not collected from Egypt, Mexico, Pakistan, Turkey, and Ukraine. Unlike the other nine countries where health warning label requirements were updated and required a new wave of compliance assessment, the warning label requirements in these five countries remained the same between 2013 and 2016 [46]. Therefore, a new compliance assessment was not required and Wave 2 data were not collected.

Table 2 presents the proportion of cigarette packs purchased in Wave 1 (2013) and Wave 2 (2015–2017) that contained overt feminine appeals by country, tobacco company, and pack-specific attributes. Overall, the proportion of feminine cigarette packs in the study samples significantly decreased from 8.6% (n = 278 packs) in Wave 1 to 5.9% (n = 137 packs) in Wave 2 (p < 0.001).

By country, Ukraine had the highest proportion of unique feminine cigarette packs (21.6%, n = 70 packs) in Wave 1 followed by Russia (20.7%, n = 104 packs), Brazil (8.2%, n = 10 packs), China (7.3%, n = 33 packs), and Vietnam (5.4%, n = 8 packs). In Wave 2, a significantly smaller proportion of packs from Russia (13.5%, n = 68 packs), Brazil (2.0%, n = 3 packs), and Vietnam (0%, n = 0 packs) contained overt feminine appeals (all p’s < 0.05). The proportion of feminine packs purchased in China remained stable over time (7.9%, n = 58 packs in Wave 2). There was no Wave 2 pack collection in Ukraine.

Most cigarette packs included in this study were brands owned by major multinational tobacco companies. Across waves, over 10% of packs from Japan Tobacco International contained overt feminine appeals (14.5%, n = 59 packs Wave 1; 10.2%, n = 24 packs Wave 2, p = 0.143). In contrast, a smaller proportion of packs from British American Tobacco (5.7%, n = 31 packs) and Phillip Morris International (6.6%, n = 34 packs) contained feminine appeals in Wave 1, and the proportion of overtly feminine packs from both tobacco companies significantly declined in Wave 2 (see Table 2). The most common brands across waves included Vogue from British American Tobacco (n = 24 packs Wave 1, n = 8 packs Wave 2), Glamour from Japan Tobacco International (n = 24 packs Wave 1; n = 14 packs Wave 2), Kiss from Richmond Tobacco Trading Ltd. (n = 27 packs Wave 1, n = 14 packs Wave 2), Style from Imperial Tobacco Company (n = 15 packs Wave 1, n = 9 packs Wave 2), and Nanjing from China National Tobacco Corporation (n = 10 packs Wave 1, n = 11 packs Wave 2).

In Wave 1, a larger proportion of alcoholic/energy drink flavored cigarettes (32.3%, n = 11 packs) followed by fruit flavored cigarettes (31.1%, n = 33 packs) and menthol/mint flavored cigarettes (11.8%, n = 32 packs) contained feminine appeals. In Wave 2, the proportion of fruit (10.7%, n = 9 packs) and menthol/mint flavored packs (5.8%, n = 12 packs) with overt feminine appeals significantly decreased (p = 0.001 and p = 0.037, respectively). With respect to pack shape, nearly 36% of slim packs (n = 184 packs) and all lipstick-shaped packs (100%, n = 36 packs) in Wave 1 contained feminine appeals. In Wave 2, a significantly smaller proportion of slim packs (25.3%, n = 69 packs, p = 0.002) and lipstick-shaped packs (28.6%, n = 2 packs, p < 0.001) contained feminine appeals.

Across waves, very few cigarette packs with flavor capsules or the use of “light/lights” or“mild/low” descriptors contained overt feminine appeals. In Wave 1, around half (51.5%, n = 34 packs) of cigarette packs with “reduced odor” claims contained feminine appeals; however, in Wave 2 the proportion of “reduced odor” packs with feminine appeals significantly declined (26.7%, n = 27 packs, p = 0.002).

The proportion of illicit packs sold in a country that contained overt feminine appeals was small overall and decreased from 5.6% (n = 43 packs) in Wave 1 to 3.3% (n = 15 packs, p = 0.094) in Wave 2.

Figure 1 displays the proportion of overtly feminine packs by type of appeal category. Flower-related imagery or descriptors were commonly featured on feminine packs, and at least 50% of feminine packs in Wave 1 and Wave 2 contained flower-based appeals. Fashion-related features were also common and present on 46.0% of feminine packs in Wave 1; however, the presence of fashion appeals significantly decreased in Wave 2 (32.8%, p = 0.011). While imagery and descriptors related to pink were present on approximately one-third of feminine packs in both waves, there was a significant increase in imagery or text associated with women or girls from Wave 1 (11.5%) to Wave 2 (22.6%, p = 0.005). Figure 2 presents examples of stereotypical feminine packs by the four main appeal categories.

In this study, we found that a notable number of cigarette packs sold in 14 LMICs contained overt feminine imagery or textual cues. This finding aligns with previous reports that women in LMICs are an important growth market for the tobacco industry [22, 29, 32, 33], and reinforces the role of the cigarette pack as a marketing vector to communicate a brand’s intended user [4]. Overall, transnational tobacco companies produced a substantial number of the feminine packs purchased in our study. Although we saw some differences in how femininity was characterized across countries, the feminine packs in this study largely utilized many of the same stereotypical Western marketing features, (e.g. fashion, flowers, pink) and pack features, (e.g. flavor, reduced odor, slim pack shape) that were previously employed to target to women in higher income countries [20, 22, 24]. We also found that the use of terms and imagery associated with women and girls in our sample of feminine packs increased over time, reflecting the potential importance of these marketing features in the context of LMICs. Our findings suggest that the major tobacco companies continue to draw on a consistent playbook to rely on Western standards of femininity to attract the attention and reinforce the sensory and pack shape preferences of women in LMICs [22, 26]. Smaller, domestic tobacco companies like the Richmond Tobacco Trading Ltd. also appeared to utilize the same or similar tactics as the transnational companies to market feminine packs, (e.g. Kiss) in their local market [22]. Collectively, these findings demonstrate efforts across tobacco companies to design packs that signal the suitability of smoking among women in LMICs and potentially increase receptivity to smoking among this demographic group through appealing branded packs [15, 16, 35–37].

One major finding from this study is that the proportion of overtly feminine packs in our systematic sample declined in almost all LMICs over time. Across both study waves, we purchased very few or no feminine packs in countries with low smoking rates among women such as Thailand, the Philippines, Indonesia, and India [28]. At the same time, the proportion of feminine packs in countries with higher rates of smoking among women, like Russia, remained relatively high but also decreased over time [28]. We are limited in our ability to fully monitor this trend given the lack of Wave 2 data from Ukraine—the country with the highest proportion of feminine packs in Wave 1. However, when Ukraine is excluded from the analysis, we still observed a similar, though not statistically significant, decrease in the proportion of feminine packs from Wave 1 (7.1%) to Wave 2 (5.9%, p = 0.073).

Overall, there were no changes in tobacco control policies regarding product packaging in the study countries between Wave 1 and Wave 2 that could potentially account for this decline [46]. However, one potential explanation for the decline is that fewer overtly feminine brands and brand variants were on the market from Wave 1 to Wave 2. Tobacco companies create brand variants that differ by color, pack design, and brand or product descriptors, (e.g. “Marlboro Red”; “ultra-light”) to target different groups of consumers [48, 49], and they invest heavily in market research to understand the appeal of these different brands and brand variants [14, 20, 50, 51]. It is possible that the industry only focused on their most successful brands and brand variants and stopped producing those with lower market sales or less appeal to women.

Another possible explanation for the observed decline in overtly feminine packs is that tobacco companies have either slowed or stopped efforts to target cigarettes to women using the explicit feminine cues. In some LMICs, it is possible that overtly targeting women with branded packaging is not culturally acceptable and other approaches are needed [22, 30]. For example, in India brands like Platinum, a cigarette that explicitly targeted women and their preference for silver jewelry, was less successful than other cigarette brands that used more subtle, indirect cues to link smoking with a woman’s sophistication and sexual allure in billboard and magazine advertising [34]. Additionally, there has been a global shift toward marketing gender-neutral (“dual-sex” [20]) cigarettes to women in Western, high-income countries [18, 20], and a general reactance against the use of traditional femininity and female independence to market products directly to women [52–55]. It may be that a similar trend toward less explicitly feminine and more gender non-specific branding and technologies (e.g., flavor capsules) that contain features that appeal to men and women equally is also taking place in the LMICs included in this study. Emerging evidence suggests that women in several LMICs are more likely than men to prefer cigarette packs with flavor capsules [56, 57], although both groups find capsule packs appealing [56, 58]. This corresponds to cross-cultural research in 10 countries, which included India, China, Russia, and Brazil, that suggests brands that have equally high levels of masculine and feminine appeals (referred to as gender androgenous) have greater reported brand value vs. brands that are exclusively feminine or exclusively masculine [59]. Given the increase in the number of capsule packs in our sample from Wave 1 (n = 85) to Wave 2 (n = 137), it is possible that tobacco companies are targeting women through novel and more gender androgenous or gender-non-specific capsule packs.

A notable exception to the overall decline in overtly feminine packs observed in this study may be China, where the proportion of feminine packs purchased in this study remained stable and high despite the low rates of smoking among women in China [28]. It may be more culturally appropriate in China to market cigarettes with overt feminine appeals. Further, China was the only study country with a significant change in marketing restrictions on tobacco advertising between Wave 1 and Wave 2, where tobacco advertising in public places, including retail stores, and mass media was banned, effective September 2015 [60]. It is possible that tobacco companies may continue to focus attention on marketing cigarettes to consumers through appealing pack designs given the more recent limitations on promoting cigarettes through other advertising platforms [4]. Continued surveillance to examine changes in how cigarettes are marketed to women in China is warranted, including whether there is a general shift toward more gender non-specific packaging.

Despite some slightly different conceptualizations of femininity across countries (see Table 1), feminine packs were identified in every country included in this study. These findings can be useful for tobacco control efforts, primarily policies to reduce branding/targeting of cigarette packs such as single presentation requirements and plain packaging. A single presentation policy, like the one adopted in Uruguay in 2009, could reduce the number of brand variants on the market to only one, single variant per brand family and limit any proliferation of packs aimed at women [49]. Plain packaging policies like those recently implemented in study countries Thailand (2019) and Turkey (2020) could remove branded packaging altogether and require all brands be sold in packs with the same standard color, font, and layout [61]. Both policy options—alone or in combination [48, 49]—have the potential to reduce the influence of the pack itself as an avenue to increase product appeal among consumers, including women in LMICs [36, 37].

This study is subject to several limitations. First, data were derived from a select number of LMICs where the global burden of tobacco use is the highest and our findings may not generalize to other LMICs, particularly in under-represented regions like Africa. Given the design of the TPackSS study to assess compliance with new health warning label requirements over time, we are also limited in our ability to examine trends in feminine marketing in some countries, like Ukraine, where no change in warning label rules occurred between waves and Wave 2 data were not collected. Second, we only purchased cigarette packs from the most populous cities in each country. It is possible that cigarette packs not collected in this study were available in smaller cities and rural areas of a country, which would affect our estimate of the proportion of unique overtly feminine packs in a sold in a country. Additionally, our data collection is intended to capture the breadth of packs available, therefore, the proportions presented are not weighted to reflect the relative market share or popularity of each brand or brand variant. Future analyses of brand-level sales data could complement our study and offer insight into the patterns of sales of feminine packs over time. Finally, our coding structure for overt feminine appeals may not have captured all images or descriptors stereotypically associated with femininity. Even though we worked with an expert from each country to ensure the cultural relevance of our predetermined coding categories, it is possible that we underreport the prevalence of feminine packs across countries if specific feminine imagery or terms were not accounted for in our codebook or in the expert review.