Policy Awareness Outweighs Health Knowledge: A Structural Equation Modeling Analysis of Tobacco Quit Intention Determinants in Vietnam’s Provincial Global Adult Tobacco Survey 2024

Abstract

Tobacco use remains a critical public health challenge in Vietnam, contributing to significant health and economic burdens. Despite comprehensive tobacco control policies, cessation rates remain low. This study investigates how tobacco control awareness, media exposure, and health knowledge influence quit intentions among Vietnamese smokers. A cross-sectional analysis of 17 387 adult current smokers from the 2024 Provincial Global Adult Tobacco Survey (PGATS) was conducted. Structural equation modeling (SEM) tested direct and mediated pathways between tobacco control awareness, anti-/pro-tobacco media exposure, health knowledge, and quit intentions, adjusting for age and gender. Multiple imputation addressed missing data (3.7%), and model fit was assessed using AIC, BIC, and SRMR. Health knowledge and anti-tobacco media exposure were the strongest predictors of quit intentions (β = 0.107, P = 0.002; β = 0.250, P < 0.001, respectively). Pro-tobacco media exposure significantly strengthened quit intentions among smokers already planning to quit (interaction β = 0.445, P = 0.026). Multi-group analyses revealed health knowledge and media effects were significant only among male smokers. Educational stratification showed lower-educated smokers benefited more from health information, while higher-educated smokers responded better to anti-tobacco media campaigns. Individual-level health knowledge and anti-tobacco media campaigns drive quit intentions among Vietnamese smokers. The pro-tobacco media paradox, operating through psychological reactance among motivated quitters, highlights opportunities for targeted counter-advertising. Gender and educational disparities underscore the need for tailored interventions: clear health messaging and sophisticated media campaigns.

Keywords

tobacco control policy awareness structural equation modeling quit intentions global adult tobacco survey

Introduction

Tobacco use remains one of the leading preventable causes of global morbidity and mortality, claiming approximately 8.7 million lives annually and generating economic losses exceeding US$1.4 trillion.^1,2 In Vietnam, the burden is particularly severe, with 35.3% of adult men and 1.1% of adult women identified as current smokers,³ resulting in an estimated 100 300 tobacco-attributable deaths annually.⁴ Recent estimates from the World Health Organization (WHO) indicate that, in 2022, the total economic cost of smoking in Vietnam-including healthcare expenditures and productivity losses due to illness and premature death from both active and passive smoking-amounted to 1.14% of the country’s gross domestic product (GDP), equivalent to approximately 108 trillion Vietnam Dong (VND) (about US$4.5 billion).⁵ Despite Vietnam’s ratification of the WHO Framework Convention on Tobacco Control in 2004 and the enactment of comprehensive tobacco control legislation in 2012, the national annual smoking cessation rate remains below 5%. Understanding factors influencing cessation intentions is therefore critical to informing the development of more effective intervention strategies. Previous research has identified several key factors associated with cessation intentions, including tobacco control policy awareness (β = 0.31, 95% CI: 0.22-0.40), health knowledge (β = 0.20, 95% CI: 0.13-0.27), and exposure to anti-tobacco media campaigns (β = 0.18, 95% CI: 0.11-0.25).^6-8

Tobacco control awareness specifically measures respondents’ knowledge of workplace smoking restrictions (as mandated by Article 11 of the 2012 Law), enforcement activities observed in public venues, and familiarity with regulatory mechanisms established under Decree 176/ND-CP/2013.^9,10 This construct captures both policy presence and perceived implementation effectiveness, distinguishing between formal policy adoption and real-world compliance experiences.^11,12 Health knowledge encompasses awareness of tobacco-related health risks emphasized in the mandated pictorial health warnings covering 50% of cigarette packaging surfaces, as required by the 2012 legislation.¹³ Media exposure measures encounter with both anti-tobacco campaigns, including those funded through Vietnam’s dedicated tobacco control fund established by the 2012 Law, and pro-tobacco marketing activities that persist despite comprehensive advertising bans.^13,14

Theoretical models such as the Health Belief Mode¹⁵ and Theory of Planned Behavior¹⁶ traditionally emphasize health knowledge as the primary driver of cessation intentions. However, emerging evidence from policy implementation research suggests that contextual factors, particularly perceived enforcement of tobacco control measures, may operate through different pathways than health education approaches.^17,18 Vietnam provides a pertinent case study for investigating these dynamics, having adopted comprehensive tobacco control measures under the Law on Prevention and Control of Tobacco Harms (2012). This legislation mandates smoke-free environments in public places, enforces comprehensive bans on tobacco advertising, promotion and sponsorship, requires pictorial health warnings on packaging, and implements tax increases on tobacco products. However, enforcement and compliance vary considerably, with observational studies documenting smoking rates of 23.5% in government buildings, 37.9% in restaurants, and 82.3% in bars despite existing prohibitions.^19,20 The structural equation modeling (SEM) approach allows for empirical testing of whether policy awareness demonstrates stronger direct effects on quit intentions compared to traditional health education pathways, while also examining potential mediation through enhanced health knowledge.¹⁸

Drawing on data from the Provincial Global Adult Tobacco Survey (PGATS) 2024, this study employs SEM approach to examine how Vietnam’s multi-faceted tobacco control approach influences cessation intentions among current smokers. Specifically, we aim to: (1) Quantify the independent effects of tobacco‐related health knowledge, anti‐tobacco media exposure, and policy awareness on smokers’ intentions to quit; and (2) Assess how individual‐level education and gender modify the impact of health education vs policy‐driven interventions on quit intentions, thereby informing the relative value of targeted risk communication and structural policy enforcement.

Methods

Study Design and Participants

This cross-sectional study analyzed data from the PGATS 2024 in Vietnam, a national household survey targeting non-institutionalized adults aged 15 years and older with a total of 88 800 participants. The survey was conducted by the Hanoi University of Public Health in collaboration with the Vietnam Tobacco Control Funds (VNTCF) and Provincial Centers for Disease control and Prevention (CDC) at 37 provinces/cities throughout the country to assess tobacco use behaviors, cessation attempts, exposure to media campaigns, and awareness of tobacco control policies.

Sample Size and Sampling Methods

The formula for calculating the sample size for each province/city to determine 1 proportion in the population is applied:

n = (1 - p) \frac{Z_{1}^{2} - α / 2}{ε^{2} p}

in which:

n is the minimum sample size,

p is the smoking rate for men = 38.9% (according to PGATS 2022-2023),

$Z_{1 - \frac{α}{2}}$ = 1.96 (Statistical significance = 95% and 2-sided test),

ε² is the acceptable relative error = 10%.

The sample size according to the formula is 600 men. Adjusted according to the design factor due to the selection of the 2-stage cluster model (design effect = 2), the required sample size is 1200. Choose a gender-stratified sample, so the sample size needs to be doubled for both genders. The final study sample size was 2400 per province (1200 men, 1200 women aged 15 and over). The smoking rate among women in Vietnam is still low at <2%, so this survey focuses on assessing the situation of tobacco use among men, so the sample size is mainly calculated for men. To get more data, it was done on women with a sample size similar to that of men. The sample size for the overall study in 37 provinces/cities was included in the general data set nationwide: 2400 × 37 provinces/cities = 88 800 people.

A multi-stage, stratified random sampling design was employed to ensure representativeness across Vietnam’s diverse regions. In the first stage, primary sampling units (PSUs) were selected proportional to population size across all provinces/cities, with urban and rural stratification. In the second stage, households within each PSU were randomly selected. One person aged 15 or older in each household was considered eligible to participate in the study. If the household has 2 or more people, the Kish method is applied to select only 1 participant. People with mental illness or poor health are excluded in this study.

Data Collection

Data were collected between August and November 2024 through face-to-face interviews conducted by trained field researchers, utilizing electronic data capture via the REDCap mobile application. The overall response rate was 92.8%. To minimize social desirability bias, interviews were conducted in private settings, with confidentiality assured to all participants.

Measurements

Five key constructs were assessed using validated scales:

Health knowledge was assessed by creating a combination variable measuring awareness of tobacco-related health risks of respondents, including stroke, heart attack, lung cancer, diabetes, chronic obstructive pulmonary disease (COPD), high blood pressure, decreased sexual ability, gastric ulcer, and nasopharyngeal-esophageal cancer (9 items). Responses were binary (yes/no), with higher scores indicating greater health knowledge.

Anti-tobacco media exposure was measured using multiple items assessing exposure to anti-smoking messages in newspapers/magazines, television, radio, large bulletin boards, social networks, website, and email during the past 30 days (7 items). Responses were coded as “Yes” (1), “No” (2), or “Don’t know” (7).

Pro-tobacco media exposure was similarly assessed through multiple items measuring exposure to cigarette advertisements in retail stores, television, radio, large bulletin boards, posters, newspapers/magazines, social networks, website, email, public transportation and restaurants/coffee shops/bar/pub during the past 30 days (11 items).

Tobacco control awareness was operationalized using 3 items measuring: (a) knowledge of workplace smoking restrictions as specified in Article 11 of Vietnam’s 2012 Law on Prevention and Control of Tobacco Harms (e07); (b) know about the Law on Prevention and Control of Tobacco Harms (hvn3); and (c) awareness of regulatory mechanisms established under government implementing decrees (ee1). This measurement approach distinguishes between passive policy knowledge and active enforcement perception, allowing examination of which aspect more strongly predicts cessation intentions.

Quit intention was measured through 2 items: attempts to quit smoking in the past 12 months and whether health warnings on cigarette packages triggered thoughts about quitting. Participants affirming both indicators were included in the SEM analysis.

Sociodemographic characteristics were collected and included as control variables, including gender, age, and educational level (from secondary and lower are defined low educational level; high-school and above are defined high educational level).

Statistical Analysis

Exploratory Factor Analysis

Prior to confirmatory factor analysis, exploratory factor analysis (EFA) was conducted to empirically examine the underlying factor structure of each construct. EFA was performed separately for each theoretical construct to validate the hypothesized dimensionality and identify the optimal number of factors within each domain. For each construct: Health knowledge variables (9 items), Anti-tobacco media exposure variables (7 items), Pro-tobacco media exposure variables (11 items) and Tobacco control awareness variables (4 items), we assessed sampling adequacy using the Kaiser-Meyer-Olkin (KMO) measure (≥0.60 acceptable) and Bartlett’s test of sphericity (P < 0.05). Factor extraction used principal axis factoring, with the number of factors determined through multiple criteria: eigenvalues >0.7 (Kaiser criterion).²¹ Oblique rotation (Promax) was applied to allow for correlated factors, which is theoretically justified given potential relationships between tobacco-related constructs. Factor loadings ≥0.40 were considered acceptable for item retention, with cross-loadings <0.30 preferred to ensure clean factor structure. Results of the EFA were presented in the Supplemental Table S1.

Confirmatory Factor Analysis

Following EFA validation, confirmatory factor analysis (CFA) was conducted to assess the measurement model for each latent construct using the factor structures identified through EFA.²² This sequential EFA-CFA approach ensures empirical validation of theoretical constructs while maintaining methodological rigor. The CFA evaluated factor loadings and their statistical significance, the composite reliability (CR ≥0.70) and average variance extracted (AVE ≥0.50). The CFA also provide some model fit indices which will be used to assess models: Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC) and standardized root mean square residual (SRMR).

Structural Equation Model Testing

The structural model tested hypothesized relationships, including direct effects from anti-tobacco media exposure and tobacco control awareness to health knowledge, and direct effects from health knowledge, media exposure (both anti- and pro-tobacco media exposure), and tobacco control awareness on quit intentions, controlling for age and gender. Mediating pathways through health knowledge were examined using the product of coefficients method. The model can be illustrated through Figure 1.

Figure 1.

Mediation Pathways Diagram

Data Management and Analysis Procedure

Missing data (3.7% overall) were addressed through multiple imputation with 20 datasets under the missing at random assumption. Standardized path coefficients were calculated to facilitate effect size comparisons. All analyses were conducted in Stata/SE 18.0 using mi and sem packages, incorporating survey weights to adjust for the complex sampling design. Statistical significance was determined at P < 0.05.

Ethical Considerations

The study protocol was approved by the Ethics Committee of Hanoi University of Public Health (Decision No. 353/2024/YTCC-HD3 dated 15 August 2024). Written informed consent was obtained from all participants, with parental consent for those aged 15-17 years. Data were anonymized to ensure confidentiality. The study adhered to the Declaration of Helsinki and followed STROBE guidelines for reporting observational studies.

Results

Sample Characteristics

This analysis examined data from 17 387 Vietnamese adults current smokers out of 88 800 people surveyed in the PGATS 2024. The structural equation model incorporated multiple imputation to address missing values (3.7% overall).

Table 1 summarizes the demographic and key baseline characteristics of the 17 387 current smokers included in this study. The sample was overwhelmingly men (96.8%), with only 3.2% women. Participants were predominantly middle‐aged, with 38.3% aged 25-44 years and 45.7% aged 45-64 years; younger smokers (15-24 years) comprised 4.9%, and those 65 years and older 11.1%. Educational attainment was modest: 70.2% of smokers had completed secondary education or less, while 29.8% had a high school diploma or higher. Female smokers were significantly more likely than men to have only secondary or lower education (83.0% vs 69.8%).

Table 1.

Demographic Characteristics of Study Participants (N = 17 387)

	Gender
	Male	Female	Total	Likelihood Ratio test
N	16 827 (96.8%)	560 (3.2%)	17 387 (100.0%)
Age groups
15-24	838 (4.98%)	16 (2.86%)	854 (4.91%)	<0.001
25-44	6480 (38.51%)	178 (31.79%)	6658 (38.29%)
45-64	7727 (45.92%)	222 (39.64%)	7949 (45.72%)
65+	1782 (10.59%)	144 (25.71%)	1926 (11.08%)
Educational level
Secondary and lower	11 731 (69.80%)	464 (83.01%)	12 195 (70.23%)	<0.001
High school and above	5075 (30.20%)	95 (16.99%)	5170 (29.77%)
Have plan to quit smoking
No plan	13 504 (92.44%)	444 (92.50%)	13 948 (92.44%)	0.958
Have plan to quit	1105 (7.56%)	36 (7.50%)	1141 (7.56%)
Health knowledge score	0.77 (0.29)	0.74 (0.32)	0.77 (0.29)	0.018
Anti-tobacco media exposure score	0.34 (0.35)	0.32 (0.38)	0.34 (0.35)	0.449
Pro-tobacco media exposure score	0.02 (0.09)	0.02 (0.10)	0.02 (0.09)	0.958
Tobacco control awareness score	0.58 (0.37)	0.47 (0.40)	0.57 (0.37)	<0.001

Overall, 7.6% of smokers reported having a plan to quit in the next year, with no difference by gender. The mean health knowledge score, reflecting awareness of tobacco‐related risks, was 0.77 (standard deviation (SD) = 0.29) for the full sample; women scored slightly lower than men (0.74 vs 0.77). Exposure to anti‐tobacco media campaigns was modest (mean = 0.34, SD = 0.35) and equivalent across genders, while pro-tobacco media exposure was rare (mean = 0.02, SD = 0.09) with no gender difference. Finally, the tobacco control awareness score, capturing knowledge of workplace smoking restrictions and enforcement activities, averaged 0.57 (SD = 0.37); women demonstrated significantly lower awareness than men (0.47 vs 0.58).

Measurement Model

In Table 2, the measurement model demonstrated generally strong indicator performance for most constructs. For Health Knowledge, 3 items loaded robustly (λ = 0.73-4.71), explaining 8.1%-52.4% of variance, with composite reliability (CR) of 0.67. Anti-Tobacco Media exhibited comparable strength (λ = 0.74-3.29; CR = 0.71), while Pro-Tobacco Media items showed moderate loadings (λ = 1.00-4.95; CR = 0.57). Tobacco Control Awareness, a formative index of policy knowledge and enforcement perception, had lower internal consistency (λ = 0.86-1.97; CR = 0.38), reflecting its multidimensional nature rather than a single latent trait.

Table 2.

Factor Loadings and Reliability Measures for Latent Constructs

Latent construct	Indicator	Factor loading (λ)	% variance explained by factor	Composite reliability (α)
Health knowledge	h02a	4.71	0.5237	0.67
	h02b	1.22	0.1355
	h02c	0.73	0.0812
Anti-tobacco media	g01a	3.29	0.4699	0.71
	g01b	1.01	0.1438
	g01c	0.74	0.1061
Pro-tobacco media	g04a	4.95	0.4504	0.57
	g04b	1.05	0.0950
	g04c	1.00	0.0882
Tobacco control	e07	1.97	0.4914	0.38
	ee1	1.00	0.2499
	hvn3	0.86	0.2586

Structural Model Results

Direct Effects

In the main‐effects SEM (Table 3), both health knowledge and anti‐tobacco media exposure were independently and positively associated with quit intentions. Specifically, a one-point increase in the health knowledge index was associated with a 0.107% increase in the latent quit‐intention score (95% CI: 0.039-0.175; P = 0.002), and a one‐unit increase in anti‐tobacco media exposure corresponded to a 0.25% increase in quit intention (95% CI: 0.185-0.314; P < 0.001). Neither pro‐tobacco media exposure (β = 0.054; P = 0.4) nor tobacco control awareness (β = 294.1; P = 0.56) was related to quit intentions. Among the controls, older age was positively associated with quit intentions (β = 0.002; 95% CI: 0.0006-0.0034; P = 0.004), while gender showed no effect.

Table 3.

Structural Model Results - Main Effects and Interaction Model

	Main effects model			Interaction model
	β (SE)	95% CI	P-value	β (SE)	95% CI	P-value
Direct effects on quit intention
Health knowledge	0.107 (0.035)	0.039–0.175	0.002	0.106 (0.035)	0.038–0.174	0.002
Anti-tobacco media	0.250 (0.033)	0.185–0.314	<0.001	0.251 (0.033)	0.186–0.315	<0.001
Pro-tobacco media	0.054 (0.064)	−0.072–0.180	0.4	−0.004 (0.086)	−0.172–0.163	0.96
Tobacco control awareness	294.1 (504.92)	−695.5–1283.7	0.56	274.1 (254.17)	−224.0–772.3	0.281
Interaction effects
Pro-tobacco media * education	-	-	-	−0.015 (0.156)	−0.320–0.290	0.923
Pro-tobacco media * plan to quit	-	-	-	0.445 (0.200)	0.052–0.837	0.026
Control variables
Age	0.002 (0.001)	0.0006–0.0034	0.004	0.002 (0.001)	0.0006–0.0033	0.005
Gender	−0.055 (0.051)	−0.156–0.046	0.286	−0.056 (0.052)	−0.157–0.045	0.278
Model fit
AIC	34 112.559			18 136.648
BIC	34 384.760			18 501.514
SRMR	0.024			0.024

Interaction Effects

When interaction terms were added (Table 3), the only significant moderator was prior quit plan: the pro‐tobacco media × plan‐to‐quit interaction was positively associated with quit intentions (β = 0.445; 95% CI: 0.052-0.837; P = 0.026), indicating that exposure to pro‐tobacco messages predicted higher quit intention only among those who already intended to quit. The pro‐tobacco media × education interaction was non‐significant (β = −0.015; P = 0.923), and the main‐effect estimates for health knowledge (β = 0.106; P = 0.002) and anti‐tobacco media (β = 0.251; P < 0.001) remained unchanged. Model fit indices were equivalent between models (SRMR = 0.024), and information criteria favored the interaction model modestly (AIC = 18 136.6 vs 34 112.6).

Subgroup Analysis

In subgroup analyses (Table 4), pathways to quit intention varied notably by gender and educational level. Among men, health knowledge exerted a modest but significant positive effect on quit intention (β = 0.106, 95% CI: 0.037-0.174), whereas the association was null among women (β = −0.176, 95% CI: −1.136-0.783). Similarly, anti‐tobacco media exposure predicted higher quit intentions for men (β = 0.247, 95% CI: 0.181-0.313) but not for women (β = 0.045, 95% CI: −0.837-0.926). Pro‐tobacco media exposure and tobacco control awareness showed no significant effects in either gender subgroup.

Table 4.

Multi-Group SEM Results by Gender and Educational Level

	Gender		Educational level
	Male	Female	Low	High
	β (95% CI)	β (95% CI)	β (95% CI)	β (95% CI)
Health knowledge → quit intention	0.106 (0.037–0.174)*	−0.176 (−1.136–0.783)	0.132 (0.038–0.226)*	0.064 (−0.035–0.163)
Anti-tobacco media → quit intention	0.247 (0.181–0.313)*	0.045 (−0.837–0.926)	0.170 (0.060–0.280)*	0.299 (0.212–0.387)*
Pro-tobacco media → quit intention	0.019 (−0.161–0.200)	−0.417 (−1.967–1.133)	0.078 (−0.132–0.288)	0.095 (−0.710–0.899)
Tobacco control awareness → quit intention	0.716 (−0.826–2.258)	0.371 (−0.696–1.439)	0.893 (−0.433–2.220)	−233.758 (−1285.255–817.739)
Model fit by group
AIC	24 131.508	996.669	17 484.808	5675.235
BIC	24 194.957	1097.936	17 551.757	5982.921
SRMR	-	-	-	0.027

Note: * Statistical significance with p<0.05.

Educational stratification revealed that health knowledge was more strongly linked to quit intentions among lower‐educated smokers (β = 0.132, 95% CI: 0.038-0.226) than among higher‐educated smokers (β = 0.064, 95% CI: −0.035-0.163). Anti‐tobacco media exposure also exhibited a larger effect in the highly educated group (β = 0.299, 95% CI: 0.212-0.387) compared with the lower‐educated group (β = 0.170, 95% CI: 0.060-0.280). Neither pro‐tobacco media nor tobacco control awareness significantly predicted quit intention across educational strata, underscoring the consistency of these null effects across subpopulations.

Discussion

In our nationally representative sample of 17 387 Vietnamese smokers, health knowledge and anti-tobacco media exposure were the only robust, independent predictors of quit intentions. A one-unit increase in the health knowledge index was associated with an increase in latent quit intention (95% CI: 0.039-0.175; P = 0.002), consistent with findings from PGATS 2020 showing that greater risk awareness predicts stronger cessation motivation.¹⁴ Likewise, each unit increase in exposure to anti-tobacco messages also corresponded to an increase in quit intentions (95% CI: 0.185-0.314; P < 0.001), aligning with evidence that mass media campaigns significantly boost quit attempts in adult smokers.²³

By contrast, pro-tobacco media exposure and tobacco control awareness had no significant direct effects on quit intentions (P > 0.40 for both paths), suggesting that policy awareness alone is insufficient to motivate quitting without strong risk messaging or individual knowledge. Notably, exploratory interaction analyses revealed that pro-tobacco media exposure strengthened quit intentions among smokers already planning to quit (interaction β = 0.445; 95% CI: 0.052 - 0.837; P = 0.026), a pattern consistent with psychological reactance theory, whereby promotional messages provoke defensive strengthening of opposing behaviors in motivated individuals.²⁴ These results underscore the primacy of targeted health education and anti-tobacco media in Vietnam’s cessation efforts and highlight conditions under which pro-tobacco content may paradoxically reinforce quit motivation.

Gender and Educational Disparities

Multigroup analyses revealed pronounced gender differences in drivers of quit intentions. Among male smokers, greater health knowledge (β = 0.106; P = 0.002) and anti-tobacco media exposure (β = 0.247; P < 0.001) were significant predictors of cessation motivation, whereas these associations were null among female smokers (health knowledge: β = −0.176; P = 0.72; anti-tobacco media: β = 0.045; P = 0.92). Such gender‐specific patterns align with recent evidence from Southeast Asia showing that women’s smoking behavior is more strongly influenced by psychosocial stressors and social norms than by risk communication or media exposure.^4,25 Although our women subsample was limited, these findings underscore the need for gender‐tailored interventions, such as stress‐management programs and peer‐support networks specifically designed for female smokers.²⁶

Educational stratification further highlighted divergent pathways to quit intentions. Smokers with lower formal education exhibited a stronger relationship between health knowledge and quit intention (β = 0.132; P = 0.006) than those with higher education (β = 0.064; P = 0.20), suggesting that foundational risk information may have greatest impact among less‐educated individuals.²⁷ In contrast, anti-tobacco media exposure exerted a larger effect in the highly educated subgroup (β = 0.299; P < 0.001) compared with lower-educated smokers (β = 0.170; P = 0.003), reflecting greater media literacy and campaign receptivity among more educated individuals.²⁸ These stratified effects are consistent with broader literature demonstrating socioeconomic gradients in health communication effectiveness, where tailored messaging strategies improve outcomes across varying levels of health literacy.^29,30

Pro-tobacco Media Paradox and Interaction Effects

Although pro-tobacco media exposure had no significant main effect on quit intentions in the overall sample (β = 0.054; P = 0.40), interaction analyses revealed a significant pro-tobacco media × quit-plan interaction (β = 0.445; 95%: CI: 0.052-0.837; P = 0.026), indicating that promotional messages increased cessation motivation - but only among smokers who had already formed quit intentions. Predicted quit-intention scores rose by approximately 0.32% (latent scale) from the 10^th to the 90^th percentile of pro-tobacco media exposure in this subgroup, whereas no dose-response relationship was observed among smokers without prior quit plans.

This paradoxical effect accords with psychological reactance theory, which posits that people perceive persuasive appeals as threats to their autonomy and may respond by strengthening the threatened behavior.³¹ In tobacco communications, similar “boomerang effects” have been observed: adolescents exposed to tobacco‐industry-sponsored anti‐smoking ads reported increased curiosity and positive attitudes toward smoking, consistent with reactance,³² and controlling health messages have been shown to backfire when perceived as directive.³³ The structural modeling work further demonstrated that reactance comprises intertwined anger and counterarguing processes that can reverse intended message effects,²⁴ while Rains’ meta‐analysis confirmed that reactance often attenuates health‐promotion effectiveness, especially among audiences with high motivation to change.³⁴

Our findings imply that pre-existing quit motivation fundamentally shapes how tobacco-related media are processed: promotional content may serve as a cue for self-affirmation among those already inclined to quit, but risks reinforcing smoking among the unmotivated. Public health practitioners should therefore segment audiences by readiness to quit when designing media interventions, ensuring that unmotivated smokers receive supportive, efficacy-enhancing messages while leveraging reactance dynamics to bolster cessation among those already primed to quit.

Policy Effectiveness in the Country Context

Our finding that tobacco control awareness had no direct effect on quit intentions (β = 0.089; P = 0.56) likely reflects persistent gaps between policy adoption and on-the-ground enforcement. Although Vietnam’s Law No. 13/2012/QH13 bans smoking in all indoor public places - including workplaces, schools, restaurants, and public transport - compliance remains poor. National survey data report secondhand smoke exposure in 81.8% of restaurants and 90.1% of bars/cafés, yet only 13.2% of respondents who witnessed violations intervened to enforce the ban.¹⁰ Observational assessments of hotels and restaurants confirm these shortcomings, with full compliance observed in just 5% of hotels and 0.06% of restaurants before targeted enforcement interventions.²⁰ Stakeholder interviews further identify “insufficient sanctions” and “limited monitoring resources” as primary barriers to effective enforcement.¹⁹ Nonetheless, recent policy advances demonstrate renewed governmental commitment. In November 2024, the National Assembly unanimously approved a resolution banning the production, importation, and use of e-cigarettes and heated tobacco products, effective January 2025.³⁵ Concurrently, VNTCF launched a national “Say No to E-Cigarettes and Heated Tobacco” media campaign in mid-2023, targeting youth via prime-time television, social media platforms, and cinema advertising.³⁶ These actions bolster the MPOWER recommended by WHO - “Monitor,” “Protect,” and “Warn” components but must be paired with strengthened “Enforce” measures – such as expanded inspections, community reporting systems, and meaningful penalties - to translate policy presence into behavior change. Our findings also have broader relevance for other low- and middle-income countries with similar tobacco use patterns and enforcement challenges. Clear health-risk messaging for lower-educated groups, sustained mass media campaigns, and gender-tailored cessation supports are policy approaches that could be adapted in comparable settings across Asia and beyond.

Strength and Limitations of This Study

This study benefits from a large, nationally representative sample and rigorous factor‐analytic and SEM techniques that incorporate both full information maximum likelihood estimation and multiple imputation to minimize bias. However, its cross‐sectional design precludes causal inference and temporal ordering of exposure and intention, and reliance on self‐reported media exposure, policy awareness, and quit intentions may introduce recall and social desirability biases. Among them, self-reported quit attempts may be underestimated, as respondents often forget or disregard shorter cessation efforts lasting only a few days. This potential recall bias could result in conservative estimates of quit intentions in our study. The tobacco control awareness construct, modeled formatively, exhibited low internal consistency (α = 0.38), limiting precision in estimating its effects. The overwhelming predominance of male smokers (96.8%) constrains the generalizability of findings to female smokers and reduces power to detect gender‐specific associations. Finally, unmeasured factors, such as nicotine dependence, local enforcement variation, and exposure to interpersonal smoking cues, may confound observed relationships, underscoring the need for longitudinal and mixed‐methods research to validate and extend these results.

Conclusion and Recommendations

Our analysis of 17 387 Vietnamese smokers demonstrates that individual‐level factors, specifically, tobacco‐related health knowledge and exposure to anti‐tobacco media campaigns, are substantially more influential in shaping quit intentions than awareness of policy measures alone. Although awareness of smoke‐free regulations did not exert a direct effect on cessation motivation, its potential to enhance risk perception warrants further investigation. Besides, exposure to pro‐tobacco media was associated with increased quit intentions among smokers who had already formed plans to quit. Stratified analyses revealed that smokers with lower educational attainment derived greater benefit from straightforward health‐risk messaging, whereas those with higher education responded more robustly to mass media interventions. Moreover, motivational effects of both knowledge and media exposure were markedly stronger in men than in women, indicating an urgent need for gender‐sensitive cessation strategies that incorporate social support and stress‐management components for female smokers.

To advance Vietnam’s tobacco‐control agenda, it is recommended that community and primary‐care education be expanded with clear, actionable messages for low‐education groups; that mass media campaigns be sustained and tailored through adjustments to message complexity based on audience literacy and quit readiness; that “challenge”-style counter‐advertising be employed to leverage reactance among motivated smokers; that gender-tailored supports, such as peer groups and stress‐coping programs for women, be implemented; and that enforcement of smoke‐free laws be strengthened via increased inspections, graduated penalties, and community reporting. By integrating enforcement, education, and targeted messaging, these combined strategies are expected to reduce smoking prevalence and advance Vietnam toward its tobacco‐free goals. Tobacco control strategies in other low- and middle-income countries may also benefit from clear health messaging, sustained mass media campaigns, and gender-tailored supports adapted to similar contexts.

Supplemental Material

Supplemental Material - Policy Awareness Outweighs Health Knowledge: A Structural Equation Modeling Analysis of Tobacco Quit Intention Determinants in Vietnam’s Provincial Global Adult Tobacco Survey 2024

Supplemental Material for Policy Awareness Outweighs Health Knowledge: A Structural Equation Modeling Analysis of Tobacco Quit Intention Determinants in Vietnam’s Provincial Global Adult Tobacco Survey 2024 by Hoang Le Tu, Tran Thi Tuyet Hanh, Phan Van Can, Phan Thi Hai, Duong Tu Anh, Thanh Nguyen Thi Minh, Nguyet Ha Thi Minh, Ha Le Thi Thu, Quan Hoang Long, Tuan Anh Hoang Khac, Phan Thị Thu Trang, Xuan Quy Luu, Thanh Quoc Pham, Hoang Van Minh in Tobacco Use Insights.

Footnotes

Acknowledgements

The authors gratefully acknowledge the support of the Vietnam Tobacco Control Fund (VNTCF) for providing funding and technical assistance for the Provincial Global Adult Tobacco Survey (PGATS) 2024, which formed the basis of this research. We also wish to express our sincere appreciation to Professor Jin-Kyoung Oh of the National Control Center, Graduate School of Cancer Science and Policy, for her invaluable guidance and mentorship throughout this manuscript.

ORCID iDs

Hoang Le Tu

Tran Thi Tuyet Hanh

Author Contributions

HVM, TTTH, HLT, PVC, PTH and DTA conceived and designed the study, agreed with the results, conclusions and came up with arguments for the manuscript. HLT analyzed the data. HLT, TTTH and HVM wrote the first draft of the manuscript. All the authors involved in data collection process, contributed to the first draft of the manuscript, made critical revisions and agreed on the final version of the manuscript. HLT, TTTH and HVM reviewed the final manuscript and approved it for submission, which was done by TTTH.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The PGATS 2024 were supported and funded by the Vietnam Tobacco Control Fund under Decision No. 114/2024/QPCTHTL-DHYTCC.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

Raw data were generated at Hanoi University of Public Health, Vietnam. Derived data supporting the findings of this study are available from the corresponding author on request.*

Supplemental Material

Supplemental material for this article is available online.

References

World Health Organization . WHO Global Report on Trends in Prevalence of Tobacco Use 2000-2025. 4th ed. Geneva: World Health Organization; 2021.

Goodchild

Nargis

Tursan

Espaignet

. Global economic cost of smoking-attributable diseases. Tob Control. 2018;27(1):58-64. doi:10.1136/tobaccocontrol-2016-053305

Hoang

Tuyet Hanh

, et al. Provincial Global Adult Tobacco Survey (GATS) 2024. Hanoi: Vietnam Hanoi University of Public Health, Vietnam Tobacco Control Fund (VNTCF); 2024.

Global Action to End Smoking . State of smoking and health in Vietnam. https://globalactiontoendsmoking.org/research/tobacco-around-the-world/viet-nam/. Accessed April 13, 2025.

World Health Organization . Align tobacco tax with Viet Nam’s goals for health and prosperity, urges WHO. https://www.who.int/vietnam/news/commentaries/detail/align-tobacco-tax-with-viet-nam-s-goals-for-health-and-prosperity--urges-who. Accessed May 3, 2025.

Hyland

Borland

, et al. Individual-level predictors of cessation behaviours among participants in the International Tobacco Control (ITC) four country survey. Tob Control. 2006;15(Suppl 3):iii83-iii94. doi:10.1136/tc.2005.013516

Weaver

Jazwa

Popova

Slovic

Rothenberg

Eriksen

. Worldviews and trust of sources for health information on electronic nicotine delivery systems: effects on risk perceptions and use. SSM Popul Health. 2017;3:787-794. doi:10.1016/j.ssmph.2017.09.003

Trofor

Papadakis

Lotrean

, et al. Knowledge of the health risks of smoking and impact of cigarette warning labels among tobacco users in six European countries: findings from the EUREST-PLUS ITC Europe surveys. Tob Induc Dis. 2018;16(2):A10. doi:10.18332/tid/99542

Nguyen

Capra

Johnstone

, et al. Change in compliance of staff at 4 Vietnam universities after the enactment of smoke-free environment decree. Environ Health Insights. 2020;14:1178630220972957. doi:10.1177/1178630220972957

10.

Nguyen

TTH

, et al.

Smoke-free environment policy in Vietnam: what did people see and how did they react when they visited various public places?

J Prev Med Hyg. 2019;60(1):E36-e42. doi:10.15167/2421-4248/jpmh2019.60.1.942

11.

Nyi

Saw

Myat Cho

Kariya

Yamamoto

Hamajima

. Tobacco control law awareness, enforcement, and compliance among high school students in Myanmar. Nagoya J Med Sci. 2018;80(3):379-389. doi:10.18999/nagjms.80.3.379

12.

Lipperman-Kreda

Paschall

Grube

. Perceived enforcement of school tobacco policy and adolescents' cigarette smoking. Prev Med. 2009;48(6):562-566. doi:10.1016/j.ypmed.2009.03.008

13.

WHO Framework Convention on Tobacco Control (FCTC) . Viet Nam - first comprehensive tobacco control law adopted. https://extranet.who.int/fctcapps/fctcapps/fctc/implementation-database/news/viet-nam-first-comprehensive-tobacco-control-law-adopted. Accessed June 28, 2025.

14.

Hoang

Tuyet Hanh

Khue

, et al. Intention to quit and predictive factors among current smokers in Vietnam: findings from adult tobacco survey 2020. Tob Use Insights. 2022;15:1179173x221098460. doi:10.1177/1179173x221098460

15.

Rosenstock

. Historical origins of the health belief model. Health Educ Monogr. 1974;2(4):328-335.

16.

Ajzen

. The theory of planned behavior. Organ Behav Hum Decis Process. 1991;50(2):179-211. doi:10.1016/0749-5978(91)90020-T

17.

Rennen

Nagelhout

van den Putte

, et al. Associations between tobacco control policy awareness, social acceptability of smoking and smoking cessation. Findings from the International Tobacco Control (ITC) Europe surveys. Health Educ Res. 2014;29(1):72-82. doi:10.1093/her/cyt073

18.

Quadri

MFA

John

Kaur

, et al. Poor implementation of tobacco control measures and lack of education influences the intention to quit tobacco: a structural equation modelling approach. BMC Public Health. 2022;22(1):1199.

19.

Dang

Tran

Nguyen

, et al. Customers’ perceptions of compliance with a tobacco control law in restaurants in Hanoi, Vietnam: a cross-sectional study. Int J Environ Res Publ Health. 2018;15(7):1451. doi:10.3390/ijerph15071451

20.

Giang

Diep

Van Minh

, et al. Improvement in compliance with smoke-free environment regulations at hotels and restaurants in Vietnam after an administrative intervention. Environ Health Insights. 2020;14:1178630220939927. doi:10.1177/1178630220939927

21.

Costello

Osborne

. Best practices in exploratory factor analysis: four recommendations for getting the most from your analysis. Practical Assessment. 2005;10(1):1-9.

22.

JjgjoH-Ssa

Psychology

H-

. The integration of EFA and CFA: one method of evaluating the construct validity. 2015;15(6):15-19.

23.

Durkin

Brennan

Wakefield

. Mass media campaigns to promote smoking cessation among adults: an integrative review. Tob Control. 2012;21(2):127-138. doi:10.1136/tobaccocontrol-2011-050345

24.

Steindl

Jonas

Sittenthaler

Traut-Mattausch

Greenberg

. Understanding psychological reactance: new developments and findings. Z Psychol. 2015;223(4):205-214. doi:10.1027/2151-2604/a000222

25.

Nguyen

, et al. Factors associated with retention in quitline counseling for smoking cessation among HIV-positive smokers receiving care at HIV outpatient clinics in Vietnam. PLoS One. 2024;19(12):e0316250. doi:10.1371/journal.pone.0316250

26.

Tsai

Yang

Kuo

. Gender differences in smoking behaviors in an Asian population. J Womens Health. 2008;17(6):971-978. doi:10.1089/jwh.2007.0621

27.

The Cancer Council Victoria . Tobacco in Australia: examining the effectiveness of public education campaigns. 2019. https://www.tobaccoinaustralia.org.au/chapter-14-social-marketing/14-4-examining-effectiveness-of-public-education-c. Accessed June 29, 2025.

28.

Kim

Seward

Fortuna

McKee

. Korean American women′s experiences with smoking and factors associated with their quit intentions. ISRN Addict. 2013;2013(1):796570. doi:10.1155/2013/796570

29.

Özmen

. Causal effect of education on tobacco use in low-and-middle-income countries. Nicotine Tob Res. 2023;25(8):1474-1480. doi:10.1093/ntr/ntad056

30.

Nguyen

Minh

Tuyet Hanh

, et al. Factors associated with intention to quit among Male smokers in Vietnam: results from the Vietnam population-based provincial global adult tobacco survey in 2022-2023. Tob Use Insights. 2025;18:1179173X251334258. doi:10.1177/1179173x251334258

31.

Grandpre

Alvaro

Burgoon

Miller

Hall

. Adolescent reactance and anti-smoking campaigns: a theoretical approach. Health Commun. 2003;15(3):349-366. doi:10.1207/s15327027hc1503_6

32.

Henriksen

Dauphinee

Wang

Fortmann

. Industry sponsored anti-smoking ads and adolescent reactance: test of a boomerang effect. Tob Control. 2006;15(1):13-18. doi:10.1136/tc.2003.006361

33.

Stautz

Frings

Albery

Moss

Marteau

. Impact of alcohol-promoting and alcohol-warning advertisements on alcohol consumption, affect, and implicit cognition in heavy-drinking young adults: a laboratory-based randomized controlled trial. Br J Health Psychol. 2017;22(1):128-150. doi:10.1111/bjhp.12221

34.

Rains

. The nature of psychological reactance revisited: a meta-analytic. Review. 2013;39(1):47-73. doi:10.1111/j.1468-2958.2012.01443.x

35.

Southeast Asia Tobacco Control Alliance (SEATCA) . National assembly unanimously bans e-cigarettes, starting in 2025. https://seatca.org/national-assembly-unanimously-bans-e-cigarettes-starting-in-2025/. Updated November 2024. Accessed June 29, 2025.

36.

Vietnam Tobacco Control Fund (VNTCF) . Vital strategies. Vietnam – tobacco control – “say no to e-cigarettes and heated tobacco to protect the young generation”. 2023. https://www.vitalstrategies.org/resources/vietnam-tobacco-control-say-no-to-e-cigarettes-and-heated-tobacco-to-protect-the-young-generation/. Accessed June 29, 2025.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.13 MB