Introduction

The Food Guide for the Brazilian Population (¹) introduced a new approach to assessing food consumption based on the Nova food classification (²). The central idea is based on the golden rule, which recommends “always prefer natural or minimally processed foods and culinary preparations to ultra-processed foods” (¹).

Fresh and minimally processed foods are those not altered by industrial processes that may have undergone minimal processing such as removing inedible or unwanted parts, drying, crushing, grinding, fractionating, roasting, boiling, pasteurization, refrigeration, freezing or packaging (²). These foods are generally prepared with processed culinary ingredients, which are obtained from the fresh food group, such as oils, fats, sugar and salt (²). Processed foods are those produced from the fresh food group with added culinary ingredients to preserve or improve sensory qualities (²). Ultra-processed foods are formulations of ingredients, mostly for exclusively industrial use, that result from various industrial processes (²). Ultra-processed foods are associated with increased risk of obesity (³) and chronic non-communicable diseases, such as diabetes, hypertension and depression (⁴,⁵). These foods generally have high energy density, added sugar, sodium, saturated or trans fats, and low amounts of dietary fiber (²).

Despite their harmful effects on health, the consumption of ultra-processed foods increased in Brazil by 5.5% over ten years, from 2008 to 2018 (⁶). They were estimated to account for 20% of total energy intake, with a slightly higher percentage among women (20.3%) compared to men (19.1%) (⁷).

In Brazil, women tend to consume less ultra-processed foods (⁸) in addition to having a lower tendency to accumulate health risk factors, such as smoking, regular consumption of sugary drinks and abusive alcohol consumption (⁹). Socioeconomic gender disparities negatively impact their ability to access healthier foods, such as fresh and minimally processed foods, due to lower income levels among women (¹⁰). This inequality worsens the risk of food insecurity in households headed by women, which are more vulnerable to hunger (¹¹). It is noteworthy that, even with the limitations mentioned, more than half of Brazilian households are headed by women (50.8%) (¹²).

Monitoring the food consumption of the female population, as well as its association with health characteristics, becomes extremely relevant for targeting effective public health measures. The objective of this study was to evaluate women’s adherence to the golden rule of the Food Guide for the Brazilian Population, according to sociodemographic characteristics and their association with health characteristics in Brazil.

Methods

Statistical methods

The prevalence rates (%) and 95%CIs for sociodemographic information and women’s health characteristics were described according to the level of adherence to the golden rule for the total population and according to sociodemographic groups. Differences between prevalence rates were investigated using the 95%CIs.

Logistic regression models were used to estimate crude and adjusted odds ratios (OR) for associations between women’s adherence to the golden rule and their health characteristics. The adjusted OR were estimated considering sociodemographic information (age, education and marital status) as confounding variables. For this comparison, the reference standard was low adherence to the golden rule.

All analyses were performed using Stata version 16.1, taking a 5% significance level. The VIGITEL system obtained free and informed consent from research participants orally prior to the interview. The VIGITEL databases were available for public use on the official website of the Ministry of Health (http://svs.aids.gov.br/download/Vigitel/).

Results

We assessed 102,057 adult women. The majority were between 18 and 35 years old (37.09%), had ≥9 years of schooling (37.71%), did not have a partner (56.58%) and were of Black race/skin color (50.31%, of whom 10.80% were Black and 39.51% were of mixed race). The golden rule adherence score ranged from -9 to +12. The highest prevalence of low adherence to the rule was found among women under 35 years old (64.93%), with an intermediate level of education (56.63%), who did not have a partner (57.17%) and were of Black (56.30%) and mixed (53.95%) race/skin color. High adherence to the rule was more prevalent among those over 50 years old (55.19%), with a high level of education (24.20%), with a partner (24.68%) and without significant differences regarding race/ skin color (Table 2).

Among the health characteristics assessed, 21.76% of the women were obese, 27.10% had hypertension, 14.67% had depression and 8.66% had diabetes. Negative self-rated health was reported by 5.71%. No significant differences were found in relation to obesity and depression between the different levels of adherence. The highest prevalence of hypertension (29.52%) and diabetes (10.14%) was found among those with high adherence to the golden rule. Negative self-rated health was more frequent among those with low adherence (6.73%) (Table 3).

Compared to low adherence to the rule, moderate adherence was inverse to the presence of obesity (adjusted OR 0.86; 95%CI 0.78; 0.93) and negative self-rated health (adjusted OR 0.72; 95% CI 0.62; 0.84). High adherence was inverse to the presence of obesity (adjusted OR 0.72; 95%CI 0.65; 0.79), hypertension (adjusted OR 0.85; 95%CI 0.78; 0.93), depression (adjusted OR 0.69; 95%CI 0.59; 0.82) and negative self-rated health (adjusted OR 0.55; 95%CI 0.45; 0.67) (Table 4).

Discussion

Based on the analysis of more than 100,000 adult women from 26 Brazilian state capitals and the Federal District, this study investigated association between women’s adherence to the golden rule of the Food Guide for the Brazilian Population and their health characteristics, according to sociodemographic characteristics. The findings indicated that women with low adherence to the rule were younger, with an intermediate level of education, without a partner and of Black and mixed race race/skin color. Those with high adherence to the golden rule were older, with a higher level of education and had a partner.

Compared to women with low adherence to the rule, those with moderate adherence had 14% lower odds of being classified as being obese and 28% lower odds of having negative self-rated health. Women with high adherence had 28% lower odds of being classified as obese, 15% lower odds of having been diagnosed with hypertension, 31% lower odds of having been diagnosed with depression and 45% lower odds of having negative self-rated health.

Global rates of chronic non-communicable diseases are projected to increase 17% over the next decade, with women facing the greatest increase, particularly for obesity, diabetes and depression (¹⁵). Although prevalence of hypertension among Brazilian women remained stable between 2006 and 2019 (from 25.4% to 27.2%) (¹⁶), prevalence of diabetes increased by 20% in women, from 7.0% to 8.4 %, between 2013 and 2019 (¹⁷). Depression was another more prevalent condition among women (6% prevalence compared to 4% among men) (¹⁸), being especially relevant among women of reproductive age (13.3%) (¹⁹). Obesity is a critical public health issue for women, with projected rates for 2030 of 30.2% (compared to 28.8% for men) in Brazil (²⁰).

The increase in multimorbidity is a worrying trend among Brazilian women. There was a substantial increase in the combination of obesity, diabetes and hypertension, from 5.5% to 9.6% (0.97 percentage points per year), among adult women in the period 2017-2021 (²¹). This increase was more pronounced in adults who reported negative self-rated health (²¹). Self-rated health is associated with morbidity and mortality, in which health rated as “poor” or “very poor” is a predictor of a higher risk of mortality (²²).

Inadequate eating habits stand out, such as low consumption of fresh and minimally processed foods and high consumption of ultra-processed foods, these being among the most common risk factors for chronic non-communicable diseases (²³). Evidence suggests that adhering to dietary guidelines focusing on increasing the consumption of fruit and vegetables, combined with reducing consumption of soft drinks, could prevent 12 million premature deaths annually worldwide. Improving diet quality could reduce premature deaths among women by up to 25% (²³). Regular consumption of ultra-processed foods increases the risk of obesity, diabetes, cardiovascular diseases and several other chronic non-communicable diseases (²⁴). Changes in traditional eating patterns have been identified in recent years, such as the reduction in bean consumption, signaling the weakening of Brazil’s traditional food culture. Beans are considered an important marker of healthy eating, included in the fresh and minimally processed food group, but are projected to no longer dominate eating habits by 2025 (²⁵).

Women who deviate from dietary guidelines, such as the golden rule of the Food Guide for the Brazilian Population, may be more susceptible to developing health problems (⁶,⁸). The dietary patterns of the female population are considered important in health outcomes of their households, given that women play a fundamental role in planning family meals (¹⁰,¹¹). It is imperative to highlight that the guide itself includes guidance on sharing responsibilities for preparing meals among all family members, in order to avoid women being overloaded with domestic activities (¹,¹⁰).

This situation appears to have deteriorated in recent years, including during the COVID-19 pandemic. Recent trends have shown a 34% drop in grocery purchases made by women and a 146% increase in the use of delivery services (²⁶). There was an increase in consumption of ultra-processed foods, such as cookies, cakes and sweets in 2017-2018 (⁷). Vegetable consumption decreased from 40.7% to 36.8% among women during the pandemic (²⁷).

The potential adverse health effects of consuming ultra-processed foods (⁵,⁴,²⁴) have been publicized, leading to official recommendations to discourage their intake (¹). In 2020, food processing was linked to cardiometabolic risk factors and several chronic non-communicable diseases, including obesity, hypertension, diabetes, cardiovascular disease, cancer and depression (⁵). Encouraging consumption of fresh and minimally processed foods can be as beneficial as avoiding consumption of ultra-processed foods (¹,²³), as consumption of ultra-processed foods often replaces consumption of fresh and minimally processed foods. This benefits women who adhere to the golden rule in their efforts to combat chronic non-communicable diseases.

The results of this study emphasized the importance of existing public policies and regulations aimed at promoting healthy eating practices, such as the Promotion of Adequate and Healthy Eating (²⁸). Brazil has implemented programs and policies related to nutrition, food security and socioeconomic factors that influence food consumption. Among them, the Bolsa Família Program (²⁹), which offers financial support to low-income families, and family farming programs (³⁰), which strengthen local food production. It is important to recognize the weaknesses of current regulations, especially with regard to attempts to tax ultra-processed foods, which is under extensive discussion during the ongoing tax reform in Brazil, but little progress has actually been seen (³¹), as well as regulation of advertising for these products, which still requires significant progress to be made (³²).

Some limitations of this study must be considered. The VIGITEL survey was based on self-reported information via telephone interviews, being subject to memory bias or reporting of socially desirable information. Its reach was restricted to individuals with a landline telephone, which could limit the representation of certain population groups. This bias was controlled by applying weighting factors (¹³), allowing the results obtained to be extrapolated to the total population of the Brazilian state capital cities (¹³). The validity of the information collected by telephone survey has been investigated in previous studies, which compared VIGITEL data with household surveys (³³,³⁴) and found very similar results, demonstrating the reliability of this method when applying sample weights. At a population level, the use of self-reported information has been widely accepted in epidemiological studies, (³⁴-³⁶) being considered a valid parameter for monitoring risk factors for chronic noncommunicable diseases.

International studies (³⁵,³⁶) corroborate the effectiveness of telephone interviews in large-scale research (²⁸). The absence of validated cutoff points to adequately classify levels of adherence to the Food Guide limited the generalizability of the results of this research. Previous studies sought to analyze the validity and consistency of scales of adherence to the Guide in different subgroups of the population (³⁷) and to understand the influence of sociodemographic factors on adherence to the Guide’s recommendations according to characteristics such as schooling, income and age group (⁸,³⁷,³⁸), but without focus on clearly presenting cutoff points for analyzing these scores in the population. The need for future research that adequately validates the cutoff points used to categorize adherence to recommended dietary practices stands out.

This study is relevant when evaluating the golden rule as an independent parameter, while covering both ultra-processed foods and fresh and minimally processed foods together. The use of continuous scales and scores has made it possible to monitor and compare the evolution of health outcomes in the population (⁹,³⁷,³⁸). In clinical practice or in research to assess food consumption, these tools could be beneficial when used to assess adherence or changes in behavior (³⁷,³⁸). It is believed that the applicability of the score can be of great importance for studies monitoring adherence to the Guide’s recommendations, given that it has a simple metric and is easily compared with other studies.

Moderate and high adherence to the golden rule of the Food Guide for the Brazilian Population was inversely associated with obesity, hypertension, depression and negative self-rated health among adult Brazilian women. These findings highlight the relevance of taking the Food Guide recommendations into consideration when promoting women’s health and the importance of public policies aimed at healthy eating habits among this population. We highlight the importance of continuous monitoring of indicators related to the Guide’s impacts on the population’s health. We suggest that future studies be carried out to analyze the temporal trend of adherence to the golden rule and other guidelines contained in the Guide.