This ecological study, which assessed temporal trends in behavioral risk and protective factors and their association with mortality rates in two of the biggest countries in South America, has revealed interesting and important results from a public health perspective. The most relevant finding was a negative association between regular F&V consumption and mortality rates in Brazilian capitals. The Brazilian cities with higher prevalence of regular F&V consumption between 2006 and 2014 presented lower overall mortality rates, regardless of HDI and educational level. Our study also revealed that this association was only found among women after gender stratification. Results for women also showed that higher prevalence of regular F&V consumption was negatively associated with mortality from cancer.
Previous studies have indicated positive effects of F&V consumption on mortality rates (34–36). A recent meta-analysis of sixteen prospective cohort studies showed an inverse association between higher intake of F&V and overall and CVD mortality (36). Additionally, a study conducted in Australia showed a negative association between F&V consumption and all-cause mortality (35), while a study from Eastern Europe showed that F&V intake was inversely associated with stroke mortality (37).
Data from 18 countries, including Brazil and Argentina, also showed an inverse association between intake of fruits, vegetables and legumes, and total and cardiovascular mortality. Nevertheless, no association was found when only South American countries were analyzed (38). In Brazil, CVD and all-cause mortality attributable to dietary intake increased from 1980s to 2009, but it was related to increases in consumption of ultraprocessed food and decreases in F&V consumption in this period (39).
Importantly, unlike the above-mentioned research studies, our study detected such association at the populational level. The negative association between F&V consumption and mortality rates at the ecological level can be partly explained by the increase in the prevalence of F&V consumption in Brazilian capitals in the last decade. From 2006 to 2014, prevalence of regular F&V consumption in Brazil increased more than 30%, and consumption was always higher among women. This increase may be result of several strategies adopted in Brazil to improve dietary intake patterns of the Brazilian population, e.g., the National Food and Nutrition Policy launched over 20 years ago (40), the Food Purchase Program, the National School Feeding Program (41), in addition to the Global Strategy for Healthy Eating, Physical Activity and Health (42) and the National Health Promotion Policy (43). Moreover, Brazil was one of the first countries to implement a food guide for its population in 2006, and in 2014, the Brazilian Ministry of Health launched an updated food guide focusing on a reduction in the consumption of ultraprocessed foods (44) Therefore, the positive impact of F&V consumption on mortality rates seen in our study may be a consequence of these above-mentioned strategies adopted in the last years, which increased regular F&V consumption by the Brazilian population.
Brazil also succeeded in increasing prevalence of PA in the 8-year-period covered by our study. In 2014, more than 50% of adults living in Brazilian capitals reported at least 150 minutes of PA per week. In contrast, there was a reduction in PA in Argentina from 2005 to 2013. Differences in frequency of PA can be explained by differences in strategies to increase this practice in the two countries.
Implementation of programs to encourage PA as well as mobilization and support for this practice have been some of the priority strategies for health promotion in Brazil since the early 2000s. Along with the implementation of the Global Strategy for Healthy Eating, Physical Activity and Health in 2004 (42) and the National Health Promotion Policy in 2006 (43), two other initiatives were undertaken: the Family Health Support Center was created to support the consolidation of primary health care in Brazil (45), and the Health Academies Program was launched to create public places with infrastructure, equipment and qualified professionals in order to promote self-care and support PA (46). All these strategies and programs have helped Brazil to increase its prevalence of PA, mainly in the state capitals.
In Argentina, on the other hand, policies to include and support PA started to be implemented a little later, more specifically at the beginning 2010s, when the Healthy Argentina National Plan (Plan Nacional Argentina Saludable) was created. This policy was reinforced by the creation of the National Program to Fight Against Sedentarism in 2013 as well as the National Plan for Healthy Eating in childhood and adolescence for prevention of overweight and obesity (ASI Plan) in 2019 (47). Perhaps the effects of these strategies were not captured by our study, as we have observed a decrease in PA from 2005 to 2013 in Argentina, when the policies were just starting to be implemented. Nevertheless, data from the 2018 ENFR showed an increase in PA in Argentina (48), which may be indicative of a positive effect of these policies on prevalence of PA.
The positive impact of different types of PA on mortality rates has been found for a cohort study with 130,000 individuals from high-income, middle-income and low-income countries. In middle-income countries, e.g., Brazil and Argentina, moderate and high PA were associated with lower risk of mortality and major CVD (49). In our study, the increased prevalence of PA observed from 2006 to 2014 resulted in lower overall and cancer mortality rates in Brazilian women. Nevertheless, the effect has been attenuated after adjustment for HDI and educational level. This attenuation might have been due to socioeconomic differences among Brazilian capitals: those capitals with higher PA rates are precisely those with higher HDI and educational level, where mortality rates by NCD are also higher.
Regarding tobacco smoking, prevalence reduced by almost 40% from 2006 to 2014 in Brazil (from 15.2% to 9.4%). In comparison, the reduction of tobacco smoking in Argentina was much slighter: a decrease of only 8% (from 54.6% to 50.1%). However, how has Brazil achieved such a positive result in prevalence of tobacco smoking? Since 2003, when Brazil adopted the Tobacco free initiative MPOWER, the country has taken several actions to reduce this behavior; for example, ban on smoking in public places and means of transportation, smoking cessation treatment, warnings on cigarette packages, ban on cigarette advertisements in the media and at points of sale, and cigarette tax that accounts for more than 80% of the total price of cigarettes, which establishes Brazil as the country with the highest cigarette tax of all members in the Americas region (13).
On the other hand, tobacco smoking rates in Argentina are still some of the highest in South America (50), despite the reduction from 2005 to 2013 observed in our study. Tobacco free initiatives adopted in Argentina were much more modest when compared to Brazil, which may explain differences in tobacco reduction in both countries. In Argentina the WHO Framework Convention on Tobacco Control was also signed in 2003 but it has not been ratified again. Only in 2011 did the country approve Law number 26,687, which regulates advertising, promotion and consumption of tobacco products and provides a free smoke environment. The cost of cigarettes in Argentina was among the cheapest in the world (around 1 USD a pack of 20 cigarettes) until 2016, when the tax on cigarettes and other tobacco by-products was increased (51). Even so, the Argentinean legislation has important gaps; for example, it allows advertising at points of sale and commercial publications for people or institutions that participate in the tobacco production and consumption chain (48).
Both countries presented a similar scenario in trends of excessive alcohol consumption in the last decade: heterogeneity among regions, higher prevalence among men but a sharper increase among women. Although, historically, women have consumed less alcohol than men, this scenario is quickly changing around the world. In Europe, for example, alcohol consumption is increasing gradually among young women (52). This change in women’s behavior can be attributed to the evolution of women’s role in the society of the 21st century, with greater gender equality.
While previous studies have demonstrated a link between excessive alcohol consumption and mortality (53–55), we were not able to find any association in Brazil. In Argentina, however, an intriguing association remained significant after adjustment for confounders: those Argentinean provinces with higher prevalence of excessive alcohol consumption presented lower mortality from CVD. Moderate wine consumption has been suggested to improve cardiovascular health (56,57) and wine has been declared a national drink in Argentina (Law 26.870/13) and exempt of taxes imposed to other alcoholic beverages, which can stimulate its consumption. As the questionnaire used in ENFR measured self-reported alcohol consumption, people might have overestimated their consumption, and this could be a reason for this intriguing result seen in our study.
Unavailability to assess regular F&V consumption in Argentina was the biggest limitation of our study. The questions regarding F&V consumption in the 2005 and 2013 ENFR were different; for this reason, they do not allow comparisons between the two time-points evaluated here. In addition, self-reported information for all behavioral factors analyzed may be considered another limitation of our study. Nevertheless, self-reported information about important risk factors (such as tobacco smoking, alcohol consumption and dietary intake) presents good reliability in epidemiological studies (58). Finally, ecological studies use group-based data, which results in loss of information or concealment of details at the individual level. On the other hand, analyses based on national population surveys from two of the biggest Latin American countries, studying an association hardly explored in both countries, can be considered a strength of our study.