Discussion

This randomized controlled trial provides robust evidence on the comparative effectiveness of FOPL systems in Bangladesh, demonstrating that all 4 tested formats, including WL, MTL, GDA, and HSR, significantly improved consumers’ ability to identify foods high in nutrients of concern (sodium, saturated fat, and sugar) compared with no label. Among these, WL consistently emerged as the most effective FOPL system, followed by MTL, with findings remaining stable across age groups (adolescents and adults) and sociodemographic characteristics. Similarly, FOPL formats significantly improved psychological responses compared with the control in both adolescents and adults. WL, followed by MTL produced the strongest effects on perceived unhealthiness, health concern, PME, and cognitive elaboration. In addition, WL and MTL showed the greatest effects on perceived understanding, learning, credibility, and support for labeling, with WL most strongly discouraging consumption and increasing health concern. These results underscore the importance of interpretive, simplified labeling approaches in improving consumer understanding within the Bangladeshi context.

Across all product categories, WLs demonstrated the highest effectiveness in enabling accurate identification of unhealthy foods. Their superior performance relative to other FOPL formats was observed among both adolescents and adults, with particularly pronounced effects among adolescents. MTL labels also improved understanding compared with noninterpretive, numerical formats; however, their effectiveness was generally lower than that of WLs. In contrast, GDA labels consistently performed worst, likely reflecting the higher cognitive burden associated with interpreting numerical information and reference intake values. These findings align closely with evidence from randomized and experimental studies conducted in diverse settings, including India [16], South Africa [17,23], and Chile [24], all of which report superior performance of warning-style labels in supporting consumer understanding of unhealthy food products.

The findings of this study demonstrate that FOPL significantly enhances consumers' ability to correctly identify unhealthy products across diverse sociodemographic groups, with WL emerging as the most effective format consistently outperforming other label types. This aligns with growing evidence from LMIC contexts [17,23], where simple, interpretive labels are more effective than nutrient-specific formats such as GDA. Regarding geographic differences, rural adolescents responded slightly more favorably to WL than their urban peers, suggesting that the binary simplicity of WLs may transcend literacy and socioeconomic barriers prevalent in rural settings. However, this pattern reversed among adults, where urban residents outperformed rural counterparts across nearly all label types, including WL and HSR, which presumably reflects greater exposure to packaged food environments and relatively higher nutritional awareness in urban areas. A mass media campaign, especially focused on rural populations, would be valuable to help ensure the WLs work for those citizens. The differences in FOPL effectiveness between adolescents and adults were small. WL was effective in correctly identifying the nutrients of concern for both adolescents and adults.

The stratified analyses of intervention effects by sociodemographic characteristics did not contradict or amplify the main findings, which consistently showed that interpretive labels were more effective. Our subgroup analyses demonstrated that participants showed a similar pattern in the correct identification of products high in nutrients of concern across educational levels. Thus, interpretive FOPL (WLs) remained more effective than GDA labels across both groups. This association likely reflects the ability of individuals with minimal formal education to process and interpret nutrition information. Policy measures should ensure that FOPLs are designed with simple pictorial cues and supported by public education initiatives so that individuals with no schooling can interpret them effectively. The inclusion of this subgroup was important for representativeness, but the relatively smaller sample size may have limited statistical power to detect differences, and these findings should therefore be interpreted with caution. It is also important to note that our classification of education level was based solely on years of schooling, rather than direct measures of functional literacy. Future studies could incorporate cognitive tests of FOPL interpretation to more precisely capture literacy-related differences in comprehension. Moreover, mixed-methods and robust research could be considered to identify which FOPL formats have the strongest potential to improve food choices among individuals with no schooling or who are functionally illiterate.

Interpretive labels, particularly WL followed by MTL, outperformed noninterpretive formats in enhancing attention, comprehension, and behavioral intent. The higher performance of WL among adolescents aligns with previous evidence showing that direct risk-oriented messaging elicits higher understanding and stronger emotional responses among younger consumers [16,17]. These findings are consistent with the WHO’s recommendation that WL serve as a clear, accessible, and effective tool for reducing the consumption of unhealthy foods [13]. The features of WL might give some sort of advantage to the consumers. It reflects simplified and interpretive design, which focuses exclusively on nutrients present in excess and thereby reduces informational overload on food packaging. Unlike GDA labels, which require consumers to interpret numerical values and reference daily intake benchmarks, WL provide direct and explicit signals of unhealthfulness, enabling faster and more accurate judgments [17,25]. Although MTL labels also offer interpretive prompts through color coding, their correct interpretation still requires an understanding of the color–nutrient relationship, which may limit effectiveness in settings with varying levels of nutrition literacy. In addition, rural-urban comparison showed significantly higher performance for WL and MTL than GDA. WL was comparatively higher among rural adolescents and urban adults. Moreover, evidence from international studies similarly suggests that WL are effective across children [26], adolescents [27], and adults [17], and across educational strata, underscoring their potential to promote equitable access to nutrition information.

The improvements observed in secondary psychological outcomes across all FOPL formats indicate that FOPL effectively enhances consumers’ cognitive and affective responses to unhealthy foods. In both adolescents and adults, WL followed by MTL produced stronger perceptions of unhealthiness, greater concern about health consequences, higher PME, and deeper cognitive elaboration compared with no label, consistent with previous evidence [17,28]. The superior performance of interpretive labels, particularly WL, supports behavioral theories suggesting that simple and salient evaluative cues facilitate rapid risk appraisal and message processing more effectively than noninterpretive formats [13]. The similar pattern of effects among adolescents highlights the potential of FOPL to shape early risk perceptions. These findings have important policy implications for LMICs such as Bangladesh, where rising diet-related NCDs and limited nutrition literacy necessitate low-cost, scalable interventions. Adoption of clear, interpretive FOPL, especially WL, could strengthen consumer understanding and support healthier food environments.

Additional outcomes further substantiate the effectiveness of FOPL as a nutrition communication tool. Across both age groups, interpretive labels enhanced perceived understanding, facilitated learning, and enhanced credibility, attention, and concern regarding health consequences. Moreover, products displaying FOPL were more frequently perceived by participants as discouraging consumption, i.e, the presence of interpretive labeling not only conveyed nutritional information but also acted as a deterrent and potentially reducing the appeal of such products. These findings are consistent with prior experimental and real-world evidence showing that interpretive formats enhance risk perception and message processing more effectively than noninterpretive labels [28]. Similar patterns have been reported in India [16], where WLs improved identification of unhealthy products and reduced purchase intentions, supporting the relevance of interpretive FOPL approaches across diverse settings.

Given the rapid growth of packaged food consumption in Bangladesh and the low salience of existing back-of-pack labels, the introduction of mandatory front-of-pack WLs represents a cost-effective, population-level strategy to improve consumer understanding and help curb diet-related NCDs [1,8]. This approach aligns with the Bangladesh Food Safety Authority’s observation that current labeling practices rarely attract consumer attention [11]. The national representativeness of this study, with balanced inclusion of adults and adolescents, enhances the generalizability of the findings and provides insight into intergenerational differences in label interpretation. As dietary habits formed during adolescence often persist into adulthood [29], effective FOPL may play a critical role in promoting healthier food choices among this group, addressing a growing public health concern in Bangladesh [4,5,7].

The findings of this study have important implications for nutrition policy, public health practice, and food system regulation in Bangladesh. The clear superiority of interpretive FOPL systems—particularly WLs—in enabling both adolescents and adults to correctly identify foods high in nutrients of concern provides strong, context-specific evidence to inform national FOPL policy development. These results suggest that adopting WLs could be an effective, low-cost regulatory intervention to improve consumer understanding of unhealthy foods and support healthier purchasing decisions at the population level.

The stronger performance of WL among adolescents is especially relevant for Bangladesh, where young people are increasingly exposed to packaged foods as part of the ongoing nutrition transition [5]. Implementing simple, salient, and easily interpretable labels may help protect this vulnerable group by reducing cognitive demands and supporting rapid, informed choices. For adults, the comparable effectiveness of WLs and MTL labels indicates that interpretive formats more broadly are preferable to numerical systems, such as GDA, which require higher levels of nutritional literacy.

From a policy perspective, these findings provide strong, context-specific evidence to guide the development of an evidence-based FOPL policy in Bangladesh. The clear superiority of WLs in enabling both adolescents and adults to correctly identify foods high in nutrients of concern supports their adoption as a central component of national nutrition policy. Moreover, FOPL implementation should be integrated into broader strategies to address diet-related NCDs, including consumer education, food reformulation, and systematic monitoring of the packaged food supply. Overall, this study offers timely and locally generated evidence to support the introduction of effective front-of-pack WLs as a key public health intervention in Bangladesh.

Strengths and limitations of the study

This study’s coverage of all 8 administrative divisions of Bangladesh improves generalizability. To our knowledge, this is the first study to evaluate FOPL effects among adolescents alongside adults, allowing for direct comparison across age groups. Mock food products reduced brand-related bias and isolated label format effects. Several drawbacks should be considered. The investigation was confined to packaged snacks; therefore, future studies should evaluate more food categories to improve external validity. Additionally, long-term studies are needed to evaluate whether FOPLs enhance diet and health over time. Because of time and resource constraints, the number of rural and urban locations was smaller than expected; however, both adolescents and adults met their predefined sample sizes, providing sufficient statistical power and balance across experimental arms. This slight decrease in geographic coverage is unlikely to have affected internal validity, and the study’s findings are robust because of consistency across sociodemographic subgroups.

In conclusion, the present study demonstrates that interpretive FOPL systems substantially improve consumers’ ability to identify nutrients of concern compared with numerical formats. Across study arms, WL consistently showed the highest effectiveness among both adolescents and adults, with a significantly greater likelihood of correctly identifying products high in sugar, sodium, saturated fat, or calories than GDA labels. MTL labels also outperformed GDAs, although their effects were smaller than those observed for WL. Direct comparisons indicate that WLs offer a modest but meaningful advantage over MTL labels among adolescents, particularly for identifying high-sugar products, highlighting the value of simple, salient, and low–cognitive-burden designs for younger consumers. Among adults, WL and MTLs showed broadly similar effectiveness, with only slight advantages for WL in specific nutrients. WL also most effective at reducing consumers’ intention to purchase these harmful products. Overall, these findings support prioritizing WLs as the most effective FOPL approach for enhancing consumer understanding and informing healthier food choices.