Artigos

Conteúdo (article):

Shifting trends in obesity growth rates in Brazilian adults between 2006 and 2021

Claudia Cristina Gonçalves PASTORELLO 1, 2 (ORCID: 0000-0002-0646-8777)
Caroline dos Santos COSTA 2, 3 (ORCID: 0000-0002-3522-1546);
Rafael Moreira CLARO 2, 4 (ORCID: 0000-0001-9690-575X);
Maria Laura da Costa LOUZADA 1, 2 (ORCID: 0000-0002-3756-2301)

1 Programa de Pós-Graduação em Nutrição em Saúde Pública. Faculdade de Saúde Pública. Universidade de São Paulo. São Paulo, SP, Brasil. Av. Dr. Arnaldo, 715, Sumaré, CEP - 01246-90 - São Paulo – SP, Brasil. e-mail: claudia.pastorello@usp.br
2 Núcleo de Pesquisas Epidemiológicas em Nutrição e Saúde. Universidade de São Paulo. São Paulo, SP, Brasil
3 Programa de Pós-Graduação em Epidemiologia. Universidade Federal de Pelotas. Pelotas, RS, Brasil.
4 Departamento de Nutrição. Universidade Federal de Minas Gerais. Belo Horizonte, MG, Brasil

 
ABSTRACT
Objective
To analyze and quantify changes in overweight and obesity prevalence growth rates in adults living in Brazil’s capital cities and the Federal District between 2006 and 2021.
Methods
Using data from the annual Surveillance of Risk and Protective Factors for Chronic Diseases Telephone Survey (Vigitel), joinpoint regression analysis was performed to identify potential inflection points in the timeseries data and to estimate annual percentage change (APC) for overweight and obesity prevalence.
Results
Obesity and overweight prevalence increased from 11.9% and 37.3%, respectively, in 2006, to 22.4% and 58.8%, respectively, in 2021. The APC for obesity prevalence fell from 5.8% (95% CI 4.7; 7.0) in the period 2006-2012 to 2.7% (95% CI 2.1; 3.3) in the period 2012 and 2021. A similar trend was observed for overweight prevalence, with the APC falling from 2.9% (95% CI 2.7; 3.2) during the period 2006-2013 to 1.3% (95% CI 1.1; 1.5) during the period 2013-2021. Reductions in the prevalence growth rate were more pronounced among women with up to 8 years of schooling.
Conclusion
The results indicate a slowdown in the growth of overweight and obesity prevalence among Brazilian adults from 2012/2013.
Keywords: Obesity. Overweight. Food and Nutritional Surveillance. Time Series Studies. Health Surveys.
 
RESUMO
Objetivo
Analisar e quantificar mudanças nos padrões de velocidade de crescimento das prevalências de excesso de peso e obesidade em adultos das capitais brasileiras e do Distrito Federal entre 2006 e 2021.
Métodos
Foram utilizados dados do sistema de Vigilância de Fatores de Risco e Proteção para Doenças Crônicas por Inquérito Telefônico de 2006 a 2021. Análises de regressão Joinpoint foram realizadas para a identificar quebras na série temporal e cálculo das diferentes variações percentuais anuais (VPA) das prevalências de excesso de peso e obesidade.
Resultados
A prevalência da obesidade aumentou de 11,9% em 2006 para 22,4% em 2021, e do excesso de peso de 37,3% em 2006 para 58,8% em 2021. A taxa de crescimento anual da prevalência de obesidade, de 5,8% (IC95% 4,7; 7,0) no período de 2006 a 2012, passou a ser de 2,7% (IC95% 2,1; 3,3) entre 2012 e 2021. O excesso de peso apresentou taxa de crescimento anual de 2,9% (IC95% 2,7; 3,2) entre 2006 e 2013, e de 1,3% (IC 95% 1,1; 1,5) entre 2013 e 2021. As reduções de crescimento foram maiores em mulheres com menos de 8 anos de estudo.
Conclusão
Os resultados indicam provável desaceleração no crescimento das prevalências de excesso de peso e obesidade a partir de 2012/2013 nos adultos brasileiros.
Palavras-Chave: Obesidade. Sobrepeso. Vigilância Alimentar e Nutricional. Estudos de Séries Temporais. Inquéritos Epidemiológicos.

INTRODUCTION
Obesity is one of the greatest global public health challenges. Prevalence of overweight has tripled worldwide since 19751, 2, affecting almost one quarter of the global population in 20193. Obesity is associated with a range of noncommunicable diseases (NCDs), including cardiovascular diseases, type 2 diabetes mellitus and some types of cancers1, and is one of the leading factors contributing to the global burden of disease and years of life lost due to disability4.

In view of this problem, the creation of health surveillance policies that enable the monitoring and analysis of trends in obesity and overweight is a priority for national governments. In Brazil, since 2006, the annual Surveillance of Risk and Protective Factors for Chronic Diseases Telephone Survey (Vigitel) has been collecting data on height and weight, socioeconomic and demographic characteristics, behavior and health status among adults living in the country’s capital cities and the Federal District (DF). The annual nature of the survey and the long period covered by data mean it is possible to monitor trends in the obesity epidemic and provide essential information to help shape public policies designed to address this problem and evaluate implementation5.

A study using Vigitel data showed that the prevalence of overweight and obesity increased from 42.6% and 55.4%, respectively, in 2006, to 11.8% and 20.3%, respectively, in 20196. The 2020 and 2021 Vigitel reports indicate that this upward trend was maintained5,7. However, these studies did not explore eventual shifts in trends in time series, which can help identify points of acceleration and deceleration and stabilization or contraction of the epidemic, providing important insights into the potential effects of public policies. This first-of-its-kind study investigates potential shifts in trends in prevalence of overweight and obesity among adults living in the country’s capital cities and the DF between 2006 and 2021.

METHODS
We used Vigitel data for the period 2006-2021. The data are freely available to the public on the national health system’s department of informatics’ (DATASUS) website. Each edition of the Vigitel was approved by the National Research Ethics Committee attached to the Ministry of Health, and verbal informed consent was obtained from all respondents.

The Vigitel sampling process consists of the random selection of 10,000 landline telephone numbers distributed uniformly across the country5. One individual aged 18 years or over is then randomly selected from each household and asked to answer a questionnaire. A mean of approximately 2,000 interviews were conducted annually in each capital, except in 2020 and 2021, when the number of interviews was reduced to 1,000 per capital due to the Covid-19 pandemic. The total population over the period 2006-2021 was 811,753 respondents (an average of 50,735 per edition).

Self-reported weight and height were used to calculate overweight and obesity rates. The respondent is asked to say their approximate weight in kilograms (accepting weights of between 30 and 300 kg) and height in meters (accepting heights of between 1.20 and 2.20 m). Body mass index [weight (kg)/height (m)²] was used to classify obesity (≥ 30 kg/m²) and overweight (≥ 25 kg/m²). Missing data from respondents who were unaware of their weight and height or did not wish to respond were imputed using the Hot Deck method5. Further information about methodology can be found in the annual Vigitel reports.

National overweight and obesity prevalence rates and their respective 95% confidence intervals (95% CI; p<0.05) were described for each year of the study by sex (female and male) and education level (up to 8 years of schooling, 9-11 years; and 12 years and over). Absolute percentage-point (pp) differences in prevalence between the beginning and end of the period (delta) were calculated and the t-test was used to assess the statistical significance of these differences.

Joinpoint v.4.9.1.0 (The National Cancer Institute, MD, EUA) was used to perform the time series analysis and calculate annual percentage change (APC). This software detects linear patterns in the distribution of estimated annual prevalence rates, identifying line segments with similar APC. The null hypothesis of this analysis was that a single linear equation explains year-to-year variability and the alternative hypothesis was that two or more linear equations better explain this variability. In other words, the analysis identifies whether there are inflection points in the time series data, indicating potential shifts in trends. We described APC and respective 95% CI (p<0.05) across all years, comparing the following strata: sex (male and female); and education level (low: up to 8 years of schooling; and high: 12 years and over).

The Vigitel estimated prevalence rates are weighted using preestablished weighting factors to correct for the unequal probability of selecting households with more than one telephone line or adult resident and to match the sex, age and education level strata with the sociodemographic structure of the total adult population in each location in each year (based on census data or official population projections when the former were not available). RStudio v2022.07.2+5768 was used to calculate mean prevalence rates and their respective 95% CI (p<0.05) and to perform the t-test for hypothesis testing.

RESULTS
National prevalence of obesity rose 10.5 pp between 2006 and 2021 (from 11.9% to 22.4%). This trend and the magnitude of change was consistent across all population strata, ranging from 10 pp among men with at least 12 years of schooling and 14.8 pp in women with 9-11 years of schooling. National prevalence of overweight rose 18.2 pp (from 37.3% to 53.8%). The increase of overweight was more pronounced in women with 9-11 years of schooling (23.5 pp) and less pronounced among men with at least 12 years of schooling, with overall growth of 11.3 pp (supplementary table 1).

The results of the time series analysis show that the prevalence growth rate was not even over study period. Between 2006 and 2012, APC for obesity prevalence was 5.8%, compared to 2.7% between 2012 and 2021, showing a significant reduction in the pace of growth (Table 1; Figure 1.a).

The analysis of the data stratified according to sex reveals similar results to those of the general population, with the period 2006-2012 seeing a sharper increase in prevalence of obesity than the period 2012-2021 (APC of 5.9% and 2.7%, respectively, for both sexes). However, trends differed according to education level, with a slowdown in growth of obesity from 2013 among people with at least 8 years of schooling (5.3% of APC up to 2013 compared to 1.7% in the following period). This break was not observed however in the stratum with at least 12 years of schooling, with APC stabilizing at 4.8% throughout the study period (Table 1; Figure 1.b).

Women with a low level of education showed a similar break in trend to that of the general population, with an APC of 5.7% up to 2013 and 1.8% up to 2021. In contrast, among women with a high level of education APC was 6.2% throughout the whole study period. Men showed a similar upward trend in prevalence of obesity across both education groups (APC of 3.6% throughout the whole study period) (Table 1; Figure 1.c).

As with obesity, the rate of increase in prevalence of overweight was not even over the study period among the general population. APC decreased from 2.9% up to 2013 to 1.3% from 2013 up to the end of the time series, revealing a slowdown in growth (Table 2; Figure 1.a).

The findings reveal differences between men and women, with the former showing a significant increase in prevalence of overweight from 2006 to 2015 (APC 2.2%), followed by a stationary trend between 2015 and 2019 (APC -0.1%; not statistically significant), increasing once again up to the end of the time series (APC 2.3%). This stationary trend was not observed among women, who showed an APC of 3.7% up to 2012, falling to 1.9% for the rest of the time series. As with obesity, the results of the time series analysis reveal differences in overweight prevalence growth rates according to education level. APC among the group with up to 8 years of schooling was 5.2% up to 2014, falling to 0.8% throughout the rest of the time series. A point of inflection was not found in the group with at least 12 years of schooling, which showed an APC of 2.3% over the whole period (Table 2; Figure 1.b).

The results of the analysis by sex and level of education showed that trends were not even. Among women with a low level of education, there was a slowdown in the overweight prevalence growth rate, from 2.8% between 2006 and 2014 to 1.0% between 2014 and the end of the time series. A break was not observed in women with a high level of education, with APC remaining stable at 3.9% throughout the study period. Breaks were not observed in men, with APC remaining stable throughout the study period at 1.5% for the group with a low level of education and at 1.1% for those with a high level (Table 2; Figure 1.c).
DISCUSSION
The data on weight and height collected during the period 2006-2021 from a sample of over 810,000 adults living in the country’s capital cities and the DF indicate changing patterns of prevalence of overweight and obesity, revealing an upward trend in rates throughout the period marked by a slowdown in growth from 2012 and 2013. This slowdown was more pronounced in women with a low level of education.

Understanding the causes of this slowdown is a complex task and the results should be interpreted with caution. However, it is worth noting that this slowdown occurred in tandem with a decline in the share of ultra-processed foods in household food purchases9. High rates of consumption of these foods is strongly associated with increased risk of overweight and obesity10, 11, 12, with a study analyzing data from 80 countries showing that increased sales of these products between 2002 and 2016 was positively associated with population‐level BMI trajectories13. Data from household budget surveys shows that there was a 4pp/year rise in purchases of ultra-processed foods between the periods 2002-2003 and 2008-2009; however, between the periods 2008-2009 and 2017-2018 the rate of growth fell to 2pp/year. A similar trend was observed with data limited to urban areas and metropolitan regions, where the population is similar to that of the Vigitel sample9. Previous results from the Vigitel also suggest a significant reduction in the consumption of sodas and artificial fruit juices between 2011 and 2017 (APC of -10.96%; 95% CI -12.4%;-9.5%). This reduction was also more pronounced in women (APC of -11.69%; 95% CI -13.7%;-9.6%) and individuals with a low level of education (APC of -12.14%; 95% CI 14.8%;-9.4%)14.

Possible explanations for these shifts in trends include an increased focus on public policies addressing obesity from the beginning of the 2000s. Measures included the provision of healthy school meals, strengthening of actions to promote healthy eating and exercise in primary health care services, the launch of the strategic action plan to tackle NCDs in Brazil and the integration of actions to promote food and nutrition security between different government sectors15.

Despite the above measures and decline in obesity and overweight growth rates, it is not possible to say that the obesity epidemic in Brazil is under control. Brazil has experienced a major political and economic crisis over the last decade as a consequence of the agendas of recent federal governments. This includes, among other aspects, the imposition of a health spending ceiling in 2016, dismantling of the National Council for Food Security and Nutrition (CONSEA) in 2018 (which was reestablished only in 2022), and increase in non-evidence-based interventions, culminating in the dismantling of a large part of science, education and health policies16. The Covid-19 pandemic, which started to unfold in 2020, complicated this situation even more, with the Latin American Federation of Obesity Societies calling it a “tragedy of two pandemics”17. On the other hand, Brazil’s return to the hunger map is a result of reduced access to food, which may have also partially contributed to the slowdown in growth of obesity among some population strata16.

Another important point is the recent growth in prevalence of overweight among Brazilian children (defined by the World Health Organization as ≥ +2 standard deviations from weight-for-height z-scores). There was an apparently stationary trend in prevalence of overweight in children under 5 up to 2007, with the rate standing at around 7%. This trend was not confirmed however in the last representative survey conducted in 2019, which showed that prevalence of overweight among this age group was significantly higher, at 10.1%18.

Data from the national health survey, which collects household data in both rural and urban areas across all Brazil’s regions, also reveal an increase in prevalence of obesity among adults, from 20.8%, in 2013, to 25.9%, in 2019, which is equivalent to 5.1 pp over the period. The same study shows that the obesity growth rate began to rise once again in the latest period19.

In Mexico, which recently began to implement a comprehensive set of policies to curb consumption of ultra-processed foods, data from the National Health and Nutrition Survey (ENSANUT) suggest improvements in the control of the obesity epidemic. The findings reveal a fall in prevalence of overweight across the adult population between 2020 and 2021, from 35.2% to 32.8% among women and 40.6% to 37.8% among men. Younger age groups showed a decline in prevalence of overweight: schoolchildren aged 5-11 years (from 19.6% to 18.8%); and adolescents (26.8% to 24.7%). However, prevalence of obesity rose slightly among men in the same period, from 31.5% to 31.8%. This increase was more pronounced among women (from 37.6% to 39. 0%)20.

Temporal trends in nutritional status are complex measures. In a study investigating literature from other countries suggesting breaks and decreases in obesity rates, Visscher et al. (2015) identified bias related to measurement errors, lack of representativeness and few timeseries data points, as well as studies whose results were not confirmed by subsequent studies21. One of the limitations of the Vigitel is the fact that the sample is limited to individuals with a landline telephone living in capital cities. However, the estimates of prevalence of obesity are very similar to those of the national health survey, which is countrywide22.

On the other hand, the present study analyzes relatively long breaks in trends and the general sample is representative of the target population and strata analyzed. While the measure of BMI adopted by this study (based on self-reported weight and height) is prone to inaccuracies, this method has been previously validated for use with the adult population in Brazil23. Finally, it is important to highlight that the Covid-19 pandemic affected data collection in the 2020 and 2021 editions of the Vigitel, when the sample was half the usual size7.

CONCLUSION
The results of this first-of-its-kind study show shifting trends in overweight and obesity in Brazil, with a slowdown in rates of growth among the adult population living in capital cities and the DF. This break in trends was more pronounced in women with a low level of education. Future research using Vigitel data should be carried out to verify whether this trend continues.

REFERENCES
1. World Health Organization. Obesity and overweight. [Internet]. 2021 [acessado em 5 maio 2023]. Available at: https://www.who.int/news-room/fact-sheets/detail/obesity-and-overweight
2. Abarca-Gómez L, Abdeen ZA, Hamid ZA, Abu-Rmeileh NM, Acosta-Cazares B, Acuin C, Adams RJ, et al. Worldwide trends in body-mass index, underweight, overweight, and obesity from 1975 to 2016: a pooled analysis of 2416 population-based measurement studies in 128· 9 million children, adolescents, and adults. The lancet. 2017 Dec 16;390(10113):2627-42. https://doi.org/10.1016/S0140-6736(17)32129-3
3. Murray CJ, Aravkin AY, Zheng P, Abbafati C, Abbas KM, Abbasi-Kangevari M, et al. Global burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. The lancet. 2020 Oct 17;396(10258):1223-49. https://doi.org/10.1016/S0140-6736(20)30752-2
4. Dai H, Alsalhe TA, Chalghaf N, Riccò M, Bragazzi NL, Wu J. The global burden of disease attributable to high body mass index in 195 countries and territories, 1990–2017: An analysis of the Global Burden of Disease Study. PLoS medicine. 2020 Jul 28;17(7):e1003198. https://doi.org/10.1371/journal.pmed.1003198
5. Brasil. Vigilância de fatores de risco e proteção para doenças crônicas por inquérito telefônico: estimativas sobre frequência e distribuição sociodemográfica de fatores de risco e proteção para doenças crônicas nas capitais dos. 2021
6. Silva LE, Oliveira MM, Stopa SR, Gouvea ED, Ferreira KR, Santos RD, et al. Tendência temporal da prevalência do excesso de peso e obesidade na população adulta brasileira, segundo características sociodemográficas, 2006-2019. Epidemiologia e Serviços de Saúde. 2021;30. https://doi.org/10.1590/S1679-49742021000100008
7. Brasil. Vigilância de fatores de risco e proteção para doenças crônicas por inquérito telefônico: estimativas sobre frequência e distribuição sociodemográfica de fatores de risco e proteção para doenças crônicas nas capitais dos. 2020
8. RStudio. [acessado em 12 fev. 2023]. Available at: https://www.rstudio.com/products/rstudio/
9. Levy RB, Andrade GC, Cruz GL, Rauber F, Louzada ML, Claro RM, et al. Três décadas da disponibilidade domiciliar de alimentos segundo a NOVA–Brasil, 1987–2018. Revista de Saúde Pública. 2022 Aug 8;56. https://doi.org/10.11606/s1518-8787.2022056004570
10. Askari M, Heshmati J, Shahinfar H, Tripathi N, Daneshzad E. Ultra-processed food and the risk of overweight and obesity: a systematic review and meta-analysis of observational studies. International Journal of Obesity. 2020 Oct;44(10):2080-91. https://doi.org/10.1038/s41366-020-00650-z
11. Moradi S, Entezari MH, Mohammadi H, Jayedi A, Lazaridi AV, Kermani MA, et al. Ultra-processed food consumption and adult obesity risk: a systematic review and dose-response meta-analysis. Critical reviews in food science and nutrition. 2022 Dec 8;63(2):249-60. https://doi.org/10.1080/10408398.2021.1946005
12. Lane MM, Davis JA, Beattie S, Gómez‐Donoso C, Loughman A, O\'Neil A, et al. Ultraprocessed food and chronic noncommunicable diseases: a systematic review and meta‐analysis of 43 observational studies. Obesity Reviews. 2021 Mar;22(3):e13146. https://doi.org/10.1111/obr.13146
13. Vandevijvere S, Jaacks LM, Monteiro CA, Moubarac JC, Girling‐Butcher M, Lee AC, et al. Global trends in ultraprocessed food and drink product sales and their association with adult body mass index trajectories. Obesity Reviews. 2019 Nov;20:10-9. https://doi.org/10.1111/obr.12860
14. dos Santos Lopes M, Edwards B, Lima YD, Ramalho AA. Tendência Temporal do Consumo Excessivo de Refrigerantes nas Capitais Brasileiras e Distrito Federal (2007 a 2019). Desafios - Revista Interdisciplinar da Universidade Federal do Tocantins. 2022 May 17;9(2):129-43. https://doi.org/10.20873/uftv9-10869
15. Jaime PC, da Silva AC, Gentil PC, Claro RM, Monteiro CA. Brazilian obesity prevention and control initiatives. Obesity reviews. 2013 Nov;14:88-95. https://doi.org/10.1111/obr.12101
16. Cátedra Josué de Castro, Linha do tempo | Combate à fome: 75 anos de políticas públicas In: T. Campelo and A. P. Bortoletto, ed. Elefante, Da fome à fome: diálogos com Josué de Castro. São Paulo: 2022. p. 44-61.
17. Halpern B, Louzada ML, Aschner P, Gerchman F, Brajkovich I, Faria‐Neto JR, et al. Obesity and COVID‐19 in Latin America: A tragedy of two pandemics. Official document of the Latin American Federation of Obesity Societies. Obesity Reviews. 2021 Mar;22(3):e13165. https://doi.org/10.1111/obr.13165
18. Casto, Inês Rugani Ribeiro, Tendências temporais de indicadores do estado nutricional no Brasil In: T. Campelo and A. P. Bortoletto, ed. Elefante, Da fome à fome: diálogos com Josué de Castro. São Paulo: 2022. p. 90-95.
19. Ferreira AP, Szwarcwald CL, Damacena GN, Souza Júnior PR. Aumento nas prevalências de obesidade entre 2013 e 2019 e fatores associados no Brasil. Revista Brasileira de Epidemiologia. 2021 Dec 10;24. https://doi.org/10.1590/1980-549720210009.supl.2
20. México, Secretaría de Salud. Instituto Nacional de Salud Pública. Encuesta Nacional de Salud y Nutrición 2021 Sobre Covid-19. 2021
21. Visscher TL, Heitmann BL, Rissanen A, Lahti-Koski M, Lissner L. A break in the obesity epidemic? Explained by biases or misinterpretation of the data?. International Journal of obesity. 2015 Feb;39(2):189-98. https://doi.org/10.1038/ijo.2014.98
22. Instituto Brasileiro de Geografia e Estatística. Pesquisa nacional de saúde: 2019: percepção do estado de saúde, estilos de vida, doenças crônicas e saúde bucal: Brasil e grandes regiões. IBGE, Coordenação de Trabalho e Rendimento
23. Peixoto MD, Benício MH, Jardim PC. Validade do peso e da altura auto-referidos: o estudo de Goiânia. Revista de saúde pública. 2006;40:1065-72. https://doi.org/10.1590/S0034-89102006000700015