ABSTRACT

OBJECTIVE:

to describe the trend of mortality from general cancer and more frequent types among men and women living in the Capitals and other municipalities of the five macro-regions of Brazil between 1978 and 2017.

METHODS:

Time series study with mortality data corrected by redistribution of ill-defined causes. Proportional cancer mortality was calculated for Brazil and regions. The annual percentage change in rates for total cancer and specific types in each segment and in the selected unit of analysis was calculated by generalized linear regression with Gaussian binding.

RESULTS:

the proportion of cancer increased progressively for both sexes from 1978 to 2017. Important differences between the Capitals and the interior of the macro-regions were seen with disaggregated data. The greatest declines occurred for stomach cancer, except in the northern and interior regions of the Northeast, and for the cervix cancer, with a generalized fall, with the exception of the interior of the northern region. Lung cancer decreased among men in the Southeast and South regions and had a generalized increase among women. Breast and prostate cancers tended to decrease in the Southeast and South regions and among residents of the Capitals but showing an increase in the interior of the North and Northeast regions. Colorectal cancer had a general tendency to increase; with stability among men in the Capitals of the South region and among women of the Southeast and Midwest regions and, since 2007, a decrease among women in the South region.

CONCLUSIONS:

Cancer mortality showed great variation among residents of capitals and the interior of the country's major regions. Clear decrease in mortality was seen for the main types in the Southeast and South regions. The North and Northeast regions have patterns compatible with cancers associated with poverty, while the large increase of the cancers related to sedentary lifestyle stand out.

DESCRIPTORS:
Neoplasms, epidemiology; Mortality, trends; Time Series Studies; Health Status Disparities

INTRODUCTION

Cancer is currently the second leading cause of illness and death in the world, with wide variation in incidence and mortality among countries due to socioeconomic status and exposure to risk factors related to social conditions and lifestyle.¹1. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global Cancer Statistics 2018: GLOBOCAN estimates of Incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):394-424. https://doi.org/10.3322/caac.21492
https://doi.org/10.3322/caac.21492... The occurrence of cancer in the world has increased regardless of demographic growth; in 2012, an estimated 14.1 million new cases and 8.2 million deaths were estimated, and a more significant growth in the worldwide incidence of the disease is expected over the next 20 years ²2. Ferlay J, Soerjomaratam I, Dikshit R, Eser S, Mathers C, Rebelo M, et al. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015;136(5):E359-86. https://doi.org/10.1002/ijc.29210
https://doi.org/10.1002/ijc.29210... . In developed countries, although there is still an increase in the incidence of some types of cancer, mortality has shown a significant decrease in the last two decades³3. Hashim D, Boffetta P, La Vecchia C, Rota M, Bertuccio P, Malvezzi M, et al. The global decrease in cancer mortality: trends and disparities. Ann Oncol. 2016;27(5):926-33. https://doi.org/10.1093/annonc/mdw027
https://doi.org/10.1093/annonc/mdw027... .

The analysis of current trends in cancer indicates that the ongoing economic and social transitions in several middle- and low-income countries are related to increased exposure to reproductive and hormonal risk factors and changes in food standards⁴4. Bray F, Jemal A, Grey N, Ferlay J, Forman D. Global cancer transitions according to the Human Development Index (2008–2030): a population-based study. Lancet Oncol. 2012;13(8):790-801. https://doi.org/10.1016/S1470-2045(12)70211-5
https://doi.org/10.1016/S1470-2045(12)70... . Monitoring the magnitude of the incidence and mortality from cancer in a population is a fundamental element for defining priorities and actions directed to its control, as well as for evaluating the effectiveness of the interventions instituted.

In Brazil, cancer mortality in the early 1980s still showed a typical pattern of middle- and low-income countries, with significant rates of cervical and stomach cancers. A transition, however, was already outlined, with an increase in the types associated with the best socioeconomic conditions (breast, prostate and colorectal)⁵5. Azevedo e Silva G, Gamarra CJ, Girianelli VR, Valente JG. A tendência da mortalidade por câncer nas capitais e interior do Brasil entre 1980 e 2006. Rev Saude Publica. 2011;45(6):1009-18. https://doi.org/10.1590/S0034-89102011005000076
https://doi.org/10.1590/S0034-8910201100... marked by differentiated patterns among the macro-regions of the country and among residents of the Capitals and other municipalities.

Although cancer trends have been reported in Brazil and Federative units (FU) between 1990 and 2015⁶6. Guerra MR, Bustamante-Teixeira MT, Corrêa CSL, Abreu DMX, Curado MP, Mooney M, et al. Magnitude and variation of the burden of cancer mortality in Brazil and Federation Units, 1990 and 2015. Rev Bras Epidemiol. 2017;20 Supl 1:102-19. https://doi.org/10.1590/1980-5497201700050009
https://doi.org/10.1590/1980-54972017000... , the existing differences in relation to the place of residence, comparing the data of inhabitants of the Capitals with those of the interior in the five major regions of the country, have not yet been explored in longer series. Knowing the evolution of cancer as a function of the changes that have been occurring in the specific risk patterns and the possibilities of access to treatment is fundamental for the improvement of regionalized control policies.

In addition, it should be considered that the Sistema de Informações sobre Mortalidade (SIM – Mortality Information System) was implemented in Brazil by the Ministry of Health (MH) in the late 1970s, with the death declaration as a source document. Since then, MH has been working to improve its completeness and coverage⁷7. Ministério da Saúde (BR), Secretaria de Vigilância em Saúde, Departamento de Vigilância de Doenças e Agravos Não Transmissíveis e Promoção da Saúde. Saúde Brasil 2014: uma análise da situação de saúde e das causas externas. Brasília, DF; 2015. A vigilância do óbito no Brasil: trajetória de 2008 a 2015; p.45-67.. The series of consolidated data available by SIM until 2017 allows the construction of extended historical series, which is advisable for monitoring the various types of cancer.

This study aims to describe the four-decade trend (1978-2017) of mortality of all cancers and the most frequent types among men and women living in the Capitals and other municipalities of the macro-regions of Brazil.

METHODS

This is a time series study that had as source the SIM data referring to the years between 1978⁸8. Ministério da Saúde (BR), Secretaria Nacional de Ações Básicas de Saúde, Divisão Nacional de Epidemiologia. Estatísticas de mortalidade: Brasil 1978. Brasília, DF: Centro de Documentação do Ministério da Saúde; 1984. and 2017⁹9. Ministério da Saúde (BR), Departamento de Informática do Sistema Único de Saúde. Estatísticas vitais, demográficas e socioeconômicas. Brasília, DF; 2017 [cited 2019 May 15]. Available from: http://datasus.saude.gov.br/informacoes-de-saude/tabnet
http://datasus.saude.gov.br/informacoes-... . For the year 1978, the eighth codification of the Manual of the International Statistical Classification of Diseases Injuries and causes of death (ICD-8) was used, between 1979 and 1995, the ninth (ICD-9), and between 1996 and 2017, the tenth (ICD-10).

The population denominators were those of the Brazilian Institute of Geography and Statistics (IBGE) censuses for the years 1980, 1991, 2000 and 2010⁹9. Ministério da Saúde (BR), Departamento de Informática do Sistema Único de Saúde. Estatísticas vitais, demográficas e socioeconômicas. Brasília, DF; 2017 [cited 2019 May 15]. Available from: http://datasus.saude.gov.br/informacoes-de-saude/tabnet
http://datasus.saude.gov.br/informacoes-... . For the intercensitary years, linear interpolation was performed. For the later period (2011 to 2017), the population projections were extrapolated also linearly, with disaggregation by sex, age group and place of residence (FU and Capitals) based on the previous period.

In men, the main types were: lung (ICD-9: 162; ICD-10: C33-34), prostate (ICD9: 185; ICD10: C61) stomach (ICD9: 151; ICD10: C16), colorectal (ICD-9: 153-154.1; ICD-10: C18-20), and in women were: breast (ICD-9: 170; ICD-10: C50), lung (ICD-9: 162; ICD-10: C33-34), colorectal (ICD-9: 153-154.1; ICD-10: C18-20), and the cervix (ICD-9: 180; ICD-10: L54). Deaths that had ignored information for sex, age and municipality of residence were excluded. As for the year 1978, only proportional mortality from cancer was calculated, the ICD codes referring to the set of cancers (ICD-8: 140-209) and the chapter referring to the ill-defined causes (ICD-8: Ch. 16).

The correction of the information on the underlying cause of death was done according to the methodology proposed by Mathers et al.¹⁰10. Mathers CD, Bernard C, Iburg KM, Inoue M, Fat DM, Shibuya K, et al. Global burden of disease in 2002: data sources, methods and results. Geneva: World Health Organization; 2003 [cited 2019 22]. (GPE Discussion Paper, 54). Available from: https://www.who.int/healthinfo/paper54.pdf
https://www.who.int/healthinfo/paper54.p... and the adaptation made by Girianelli et al.¹¹11. Girianelli VR, Gamarra CJ, Azevedo e Silva G. Os grandes contrastes da mortalidade por câncer do colo do útero e de mama no Brasil. Rev Saude Publica. 2014,48(3):459-67. https://doi.org/10.1590/S0034-8910.2014048005214
https://doi.org/10.1590/S0034-8910.20140... . This adaptation, which consists in proportionally redistributing 50% of deaths with a basic cause classified as undefined (ICD-8: Ch. 16, ICD-9: 780-799; ICD – 10: R00-R99), is necessary because cancer is less found among ill-defined deaths than among well-defined ones¹²12. França E, Teixeira R, Ishitani L, Duncan BB, Cortez-Escalante JJ, Morais Neto OL, et al. Ill-defined causes of death in Brazil: a redistribution method based on the investigation of such causes. Rev Saude Publica. 2014;48(4):671-81. https://doi.org/10.1590/S0034-8910.2014048005146
https://doi.org/10.1590/S0034-8910.20140... . Thus, the correction factors were calculated for each age group (0-4, 5-14,15-19, 20-29, 30-49, 50-59, 60-69, 70-79 and 80 years or more), calendar year, sex and place of residence (capitals and other municipalities) of the five macro-regions of the country. For deaths recorded as cervical cancer, an additional correction was made with redistribution of all deaths classified as malignant neoplasm of the uterus without other specification (ICD-9: 179; ICD-10: C55), maintaining the proportion recorded as deaths from cervical and uterine body cancer¹³13. Gamarra CJ, Valente JG, Silva GA. Correção da magnitude da mortalidade por câncer do colo do útero no Brasil, 1996-2005. Rev Saude Publica. 2010;44(4):629-38. https://doi.org/10.1590/S0034-89102010000400006
https://doi.org/10.1590/S0034-8910201000... .

With corrected data, the proportional mortality from cancer was calculated for Brazil. Since the 1978 data did not cover all regions, the proportional mortality series for macro-regions was started in 1979.

Then, crude and age-standardized cancer mortality rates were calculated(0-4, 5-14,15-19, 20-29, 30-49, 50-59, 60-69, 70-79 and 80 years or more) and for each sex by the direct method, taking as standard population the world population proposed by Segi¹⁴14. Segi M. Cancer mortality for selected sites in 24 countries (1950-57). Senda (JPN): Tohoku University School of Public Health; 1960. for the 1980- 2017 period. These rates were calculated in a disaggregated manner by macro-regions, capitals and other municipalities (interior).

To estimate the trend, a linear regression model was used. For series with structural breaks, the time variable was introduced into the model by means of linear spline by parts, with break points in the years in which the trend changed its trajectory. The identification of the break points was visual and cubic splines with 10 degrees of freedom were used to highlight the points of inflection of the curve, allowing to define break points at specific moments of time and the respective trend in each interval, adjusting by a continuous line segment between the points. This approach was chosen because it concerns time series with trends that vary in time in a non-regular way, which makes inappropriate a linear model for global trend for the entire period.

The models were evaluated via residues to verify if the usual assumptions were met, in particular the existence of residual autocorrelation through the functions of autocorrelation (FAC) and partial autocorrelation (FPAC). Models that showed significant autocorrelation in residues or with an absolute value greater than 0.5 were re-estimated using generalized least squares with a first-order autoregressive model AR(1) to model the autocorrelation and correct the variance of the coefficient estimators. We used the function “gls” with estimation for restricted maximum likelihood (REML) of the package “nlme”¹⁵15. Pinheiro J, Bates D, DebRoy S, Sarkar D; R Core Team. “nlme”: linear and nonlinear mixed effects models. R package version 3.1-140. Vienna (AUS); 2019 [cited 2019 Nov 22]. Available from: https://CRAN.R-project.org/package=nlme
https://CRAN.R-project.org/package=nlme... .

The coefficient of the term for each segment expresses the logarithm of the trend in that interval. Thus, the annual percentage change (APC) of mortality rates was calculated by the formula and respective p values.

For the interpretation of trends, statistical non-significance (p-value above 0.05) was used as a criterion to characterize an APC as stable. The statistically significant APC, when positive, indicated an increasing trend and, when negative, a decreasing trend. The analyses were performed in the R program, version 3.4.1.

RESULTS

Proportional mortality from cancer increased from 1978 to 2017 in men, from 8.3% to 16.2%, and in women, from 9.1% to 18.3%. From 1979, it was possible to obtain data for the regions and it was observed that the lowest proportion was among men from the Northeast region in 1979 (5.4%) and the highest among women from the South region in 2017 (21.6%) (Figure 1).

Mortality rates for the age-adjusted set of cancers showed different magnitude patterns between regions of the country for both sexes or between men and women separately (Figure 1) and the trend was not constant. Oscillations were detected in this period of 38 years. Among men (1999–2017) there was a slight increase (APC = 0.14; p < 0.001); among women (2004–2017) the increase was greater (APC = 0.44; p < 0.001) (Table 1).

Disaggregating by place of residence, four trends were seen for both sexes in the Capitals (declining until 1990, increasing from 1990 to 1997, sharp fall from 1997 to 2005 and lower intensity fall from 2005 to 2017) and only two in the interior (stability until 1990 and increase until 2017). Among men, it was in the Midwest region that the greatest fall was observed, which occurred in the Capitals (1980-1991: APC = −1.65; p < 0.001), and the greatest increase, which occurred in the other municipalities (1980-1985: APC = 6.36; p < 0.001). Among women, the greatest decline occurred in the southern region, both in the Capitals and in the interior, between 1980 and 1984 (APC = −1.96; p < 0.014 and APC = −1.89; p < 0.001, respectively), while the greatest growth occurred among residents of the interior of the northern region (1999-2017: APC = 3.44; p < 0.001).

Among men, lung, prostate, colorectal and stomach cancers accounted for 44.9% of all cancer deaths in 1979 and 43.4% in 2017. Among women, breast, lung, colorectal and cervical cancers accounted for 39.0% of cancer deaths in 1979 and 44.4% in 2017. Figure 3 shows the specific trends for these types of cancer and Table 2 shows the respective APC in the period studied.

The cancers that showed the greatest decline in mortality were stomach and cervix. Markedly, stomach cancer in men declined in all regions throughout the period, with the exception of the northern region, where a tendency to increase was seen in the Capitals after 2010 and in the interior from 2001, and in the interior of the Northeast region between 1996 and 2017, as shown by the APC in the period.

Among women, stomach cancer represented the seventh most frequent type of cancer in 2017 and, although the rates were much lower than those of men, the same decline occurred throughout the period, with the exception of the interior of the northern region, where there was an increase in the trend from 1980 to 1987, a fall from 1987 to 2001 and a resumption of growth from 2001 to 2017. Among residents of the interior of the Northeast region, there was a fall from 1980 to 1991 and an increase from 1991 to 2007, followed by stability thereafter (data not presented).

Cervical cancer, which was the second most frequent among female deaths in 1980, rose to fourth place at the end of the period. Although rates in the Capitals were higher, they were descending in all regions. It is noteworthy that, in general, there was no difference in the magnitude and trend of rates of this type of cancer among residents of the Capitals or the interior of the Southeast and South regions. The APC analysis indicated a fall both in the Capitals and in the interior of all regions, except in the interior of the North region between 2002 and 2017 and in the interior of the Northeast region between 1994 and 2006. In the interior of the Northeast region, however, as of since 2006 rates became stable, which did not happen for residents outside the Capitals in the North region.

Lung cancer had a different behavior according to sex and between the macro-regions and within them. Among men in the Capitals, rates began to fall from 1984 in the Southeast region, 1992 in the South Region, and 1995 in the Midwest region. Within the South-East and South regions, rates declined from 1994 and 1993, respectively. In the capitals of the North and Northeast regions, there was an unsustainable fall and a marked increase in the interior, especially from 2011 to 2017 in the North, and from 1994 to 2017 in the Northeast. In the Midwest region, the rates decreased in the Capitals between 1995 and 2017 and increased in the interior, but with intensity tending to decrease. Among women, there was a generalized tendency to increased mortality from lung cancer throughout the period, with the highest rates observed in the Southern region and the greatest increase in the temporal trend within the northern region throughout the series and within the Northeast region between 1993 and 2017.

Breast cancer in the southern region showed an increase in the Capitals until 1990, and after that a fall until 2017, which did not occur in the interior. In the Capitals of the Southeast region, there was a decline from 1996, while in the interior the fall that occurred between 1996 and 2007 did not continue afterwards. In the North and Northeast regions there was a tendency to increase, especially high in the interior.

Prostate cancer rates have always been higher in the Capitals than in the interior, especially in the North, Northeast and Midwest regions. The largest increases were observed in the interior of the northern regions between 2004 and 2017 and Northeast between 1993 and 2017. In the interior of the Midwest region, there was also a growing trend throughout the period. In the Capitals, between 1997 and 2017, rates declined in the Southeast, South and Midwest regions. Later, declining rates were observed in the interior of the Southeast region between 2004 and 2017, and the South region between 2007 and 2017.

Colorectal cancer was ranked sixth among men and fourth among women throughout the country at the beginning of the period and began to occupy the fourth position among men and third among women at the end. The highest rates were observed in the Southeast and South regions, both for men and women. Among men, from 1999, it was possible to notice an increasing trend in the North region, and from 2000 in the Northeast region, both in capitals and in the interior. In the Midwest region, there was an increase in the Capitals between 1984 and 2017 and in the interior between 1992 and 2017. An increasing trend was also observed in the Southeast and South regions, with the exception of residents of the capitals of the southern region, which from 2002 to 2017 showed rate stability.

Among women, there was also a tendency to increased mortality from colorectal cancer, both in the Capitals and in the interior of all regions, with the exception of the Capitals of the Southeast region, where rates remained stable between 1997 and 2017, and of the Midwest region, with stability between 2009 and 2017. Only in the capitals of the southern region there was a fall from 2007 to 2017.

It is noteworthy that the largest increases in the last two decades have been seen in the interior of the North and Northeast regions, both in men and women. In the interior of the northern region, it occurred for men between 1999 and 2017 (APC = 7.83; p < 0.001) and in the Northeast region, between 2000 and 2017 (APC = 6.96; p < 0.001). For women, the increase in the interior of the northern region occurred between 2000 and 2017 (APC = 6.42; p < 0.001) and in the interior of the Northeast region, between 1999 and 2017 (APC = 5.49; p < 0.001).

DISCUSSION

Cancer mortality in Brazil for both sexes at the end of the study period (90.2/100,000 inhabitants) was similar to that of high-income countries (USA: 91.0/100,000, Canada: 92.8/100,000, United Kingdom: 102.6/100,000, Japan: 85.2/100,000)¹⁶16. Ferlay J, Ervik M, Lam F, Colombet M, Mery L, Piñeros M, et al, editors. Global Cancer Observatory: Cancer Today. Lyon (FR): International Agency for Research on Cancer; 2018. [cited 2019 Nov 22]. Available from: https://gco.iarc.fr/today
https://gco.iarc.fr/today... . The most frequent types of cancer among deaths in the country, with the exception of cervical cancer, are also the most frequent in these countries (lung, breast, prostate, colorectal)⁴4. Bray F, Jemal A, Grey N, Ferlay J, Forman D. Global cancer transitions according to the Human Development Index (2008–2030): a population-based study. Lancet Oncol. 2012;13(8):790-801. https://doi.org/10.1016/S1470-2045(12)70211-5
https://doi.org/10.1016/S1470-2045(12)70... . The temporal trend of almost 40 years observed in the country as a whole for all cancers was not constant, with a fall in men between 1980 and 1990, followed by an increase between 1990 and 1999 that continued less pronounced until 2017. Among women there was a slight fall between 1980 and 1989, followed by an increase until 1996 and a fall until 2004, when rates rose again.

Data from other countries, mostly developed, point to the marked decrease in cancer mortality achieved in the last two decades³3. Hashim D, Boffetta P, La Vecchia C, Rota M, Bertuccio P, Malvezzi M, et al. The global decrease in cancer mortality: trends and disparities. Ann Oncol. 2016;27(5):926-33. https://doi.org/10.1093/annonc/mdw027
https://doi.org/10.1093/annonc/mdw027... ^,1717. Torre LA, Siegel RL, Ward EM, Jemal A. Global Cancer Incidence and Mortality Rates and Trends: an update. Cancer Epidemiol Biomarkers Prev. 2016;25(1):16-27. https://doi.org/10.1158/1055-9965.EPI-15-0578
https://doi.org/10.1158/1055-9965.EPI-15... , attributed to preventive interventions, introduction of screening and, especially, to advances in diagnostic and therapeutic procedures¹⁸18. Welch HG, Kramer BS, Black WC. Epidemiologic signatures in cancer. N Engl J Med. 2019;381(14):1378-86. https://doi.org/10.1056/NEJMsr1905447
https://doi.org/10.1056/NEJMsr1905447... . However, advances in medicine are not always accessible to all individuals, especially in contexts of socioeconomic inequalities¹⁹19. Karanikolos M, Ellis L, Coleman MP, McKee M. Health systems performance and cancer outcomes. J Natl Cancer Inst Monogr. 2013;2013(46):7-12. https://doi.org/10.1093/jncimonographs/lgt003
https://doi.org/10.1093/jncimonographs/l... .

The greatest fall detected in the study period occurred for stomach cancer in all regions, with the exception of the orthern region as a whole and the interior of the Northeast region. The decline, already reported in previous studies⁵5. Azevedo e Silva G, Gamarra CJ, Girianelli VR, Valente JG. A tendência da mortalidade por câncer nas capitais e interior do Brasil entre 1980 e 2006. Rev Saude Publica. 2011;45(6):1009-18. https://doi.org/10.1590/S0034-89102011005000076
https://doi.org/10.1590/S0034-8910201100... , can be explained by a lower exposure to known carrcinogenic factors such as salty food intake and better food preservation²⁰20. Rawla P, Barsouk A. Epidemiology of gastric cancer: global trends, risk factors and prevention. Prz Gastroenterol. 2019;14(1):26-38. https://doi:10.5114/pg.2018.80001
https://doi:10.5114/pg.2018.80001... . However, there is a need to better understand the increase observed among residents of the northern Region and municipalities outside the Capitals in the Northeast regions.

Cervical cancer declined in all regions, but not among women in the interior of the northern region, where rates in 2017 are three times more frequent than in the Southeast region. Some effect of screening based on Pap smear, introduced throughout the country from 1992²¹21. Instituto Nacional de Câncer (BR), Coordenação Geral de Ações Estratégicas, Divisão de Apoio à Rede de Atenção Oncológica. Diretrizes brasileiras para o rastreamento do câncer do colo do útero. Rio de Janeiro: INCA; 2011 [cited 2019 15]. Available from: http://bvsms.saude.gov.br/bvs/publicacoes/inca/rastreamento_cancer_colo_utero.pdf
http://bvsms.saude.gov.br/bvs/publicacoe... , was able to reverse, in part, the trend of rate growth, but there are still women without access to screening in the interior of the North and Northeast regions, a fact reported previously with data until 2011¹¹11. Girianelli VR, Gamarra CJ, Azevedo e Silva G. Os grandes contrastes da mortalidade por câncer do colo do útero e de mama no Brasil. Rev Saude Publica. 2014,48(3):459-67. https://doi.org/10.1590/S0034-8910.2014048005214
https://doi.org/10.1590/S0034-8910.20140... .

The inverse correlation between the Human Development Index (HDI) and national estimates of cervical cancer is well reported in the literature⁴4. Bray F, Jemal A, Grey N, Ferlay J, Forman D. Global cancer transitions according to the Human Development Index (2008–2030): a population-based study. Lancet Oncol. 2012;13(8):790-801. https://doi.org/10.1016/S1470-2045(12)70211-5
https://doi.org/10.1016/S1470-2045(12)70... . The U.S. cervical cancer mortality ratio among poorer and wealthier localities reached 2 between 2012 and 2016²²22. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA Cancer J Clin. 2019;69(1):7-34. https://doi.org/10.3322/caac.21551
https://doi.org/10.3322/caac.21551... . Several indicators can exemplify the inequality in mortality from this type of cancer among Brazilian regions, for example, the number of potential years of healthy life lost at 60 years of age is twice as high in the North region compared to the Southeast region²³23. Szwarcwald CL, Souza Júnior PRB, Marques AP, Almeida WS, Montilla DER. Inequalities in healthy life expectancy by Brazilian geographic regions: findings from the National Health Survey, 2013. Int J Equity Health. 2016;15(1):141. https://doi.org/10.1186/s12939-016-0432-7
https://doi.org/10.1186/s12939-016-0432-... .

The South region has the highest municipal Human Development Index (HDI) in the country (0.766), while the North and Northeast regions, despite having had more accelerated growth between 2000 and 2010, have the lowest²⁴24. Programa das Nações Unidas para o Desenvolvimento (PNUD). Desenvolvimento humano nas macrorregiões brasileiras: 2016. Brasília, DF: PNUD, IPEA, FJP; 2016.. Although there was an improvement in health indicators in Brazil between 1990 and 2016, the burden of disease remains higher in the states of the North and Northeast regions compared to the Southeast and South²⁵25. GBD 2016 Brazil Collaborators. Burden of disease in Brazil, 1990-2016: a systematic subnational analysis for the Global Burden of Disease Study 2016. Lancet. 2018;392(10149):760-75. https://doi.org/10.1016/S0140-6736(18)31221-2
https://doi.org/10.1016/S0140-6736(18)31... .

The experience of high-income countries, such as those in North America, part of Europe and Japan, is an example that investment in effective screening programs has made cervical cancer a rare disease²⁶26. Vaccarella S, Laversanne M, Ferlay J, Bray F. Cervical cancer in Africa, Latin America and the Caribbean and Asia: regional inequalities and changing trends. Int J Cancer. 2017;141(10):1997-2001. https://doi.org/10.1002/ijc.30901
https://doi.org/10.1002/ijc.30901... . Secondary prevention with the use of oncotic cytology, however, presupposes the structured operation of a care network with laboratory quality. Added to this, the effect of the human papillomavirus (HPV) vaccine will have long-term effects and does not replace screening for the disease. In this sense, the introduction of the HPV detection test can accelerate the effectiveness of screening, ²⁷27. Vaccarella S, Franceschi S, Bray F. The incremental benefits of implementing effective cervical cancer screening. Int J Cancer. 2016;138(1):254-5. https://doi.org/10.1002/ijc.29700
https://doi.org/10.1002/ijc.29700... and should be an option to consider in countries such as Brazil, which still has high rates of the disease.

Demographic changes in Brazil, with progressive population aging, increased life expectancy and reduced fertility, added to the increase in obesity²⁸28. Victora CG, Barreto ML, Leal MC, Monteiro CA, Schmidt MI, Paim JS, et al. Health conditions and health-policy innovations in Brazil: the way forward. Lancet. 2011;377(9782):2042-53. https://doi.org/10.1016/S0140-6736(11)60055-X
https://doi.org/10.1016/S0140-6736(11)60... , have as a consequence an increase in the incidence of breast cancer. Although the incidence of breast cancer in Brazil is lower than in the USA and other European countries, mortality from this cancer is higher in all age groups, as a consequence of the high prevalence of cases diagnosed in advanced stage²⁹29. Santos-Silva I, De Stavola BL, Renna Junior NL, Nogueira MC, Aquino EML, Bustamante-Teixeira MT, Azevedo e Silva G. Ethnoracial and social trends in breast cancer staging at diagnosis in Brazil, 2001-14: a case only analysis. Lancet Glob Health. 2019;7(6):e784-97. https://doi.org/10.1016/S2214-109X(19)30151-2
https://doi.org/10.1016/S2214-109X(19)30... . However, even with a tendency to increase the incidence, mortality rates declined in the capitals of the southern region from 1990 to 2017, and Southeast from 1996 to 2017, which may be an effect of better access to diagnosis and treatment of women living in these capitals¹¹11. Girianelli VR, Gamarra CJ, Azevedo e Silva G. Os grandes contrastes da mortalidade por câncer do colo do útero e de mama no Brasil. Rev Saude Publica. 2014,48(3):459-67. https://doi.org/10.1590/S0034-8910.2014048005214
https://doi.org/10.1590/S0034-8910.20140... .

This same trend of decrease in the more developed regions of Brazil happens in relation to prostate cancer, for which there is a decline among residents in the capitals since 1997 and later among those of the interior. This trend may be related to the possibilities of specialized services offer for the treatment of early diagnosed cases³⁰30. Stopa SR, Malta DC, Monteiro CN, Szwarcwald CL, Goldbaum M, Cesar CLG. Acesso e uso de serviços de saúde pela população brasileira, Pesquisa Nacional de Saúde 2013. Rev Saude Publica. 2017;51 Supl 1:3s. https://doi.org/10.1590/s1518-8787.2017051000074
https://doi.org/10.1590/s1518-8787.20170... .

The increase in the incidence of prostate cancer in developed countries seems to be related to the increase in access to health services, as well as to the increase in the registration of cases³¹31. Rawla P. Epidemiology of prostate cancer. World J Oncol. 2019;10(2):63-89. https://doi.org/10.14740/wjon1191
https://doi.org/10.14740/wjon1191... . An independent increase in the adoption of the routine prostate-specific antigen (PSA) detection test was observed, suggesting the interference of the western lifestyle, which leads to increased obesity and physical inactivity³²32. Baade PD, Youlden DR, Krnjacki LJ. International epidemiology of prostate cancer: geographical distribution and secular trends. Mol Nutr Food Res. 2009;53(2):171-84. https://doi.org/10.1002/mnfr.200700511
https://doi.org/10.1002/mnfr.200700511... Although it is unclear why mortality rates have been falling in most Western countries, this fall has been attributed to early detection and improvement in treatment. The introduction of PSA test may have influenced early detection and thus reduced mortality. However, overscreening, which can lead to undesirable effects of treatment, contradicts the use of this test as a screening program³¹31. Rawla P. Epidemiology of prostate cancer. World J Oncol. 2019;10(2):63-89. https://doi.org/10.14740/wjon1191
https://doi.org/10.14740/wjon1191... . The fact that the highest APC of prostate cancer mortality rates have been observed in the North and Northeast regions may suggest that there is a rapid absorption of the sedentary lifestyle coexisting with a low capacity to offer specialized services for diagnosis and treatment³⁰30. Stopa SR, Malta DC, Monteiro CN, Szwarcwald CL, Goldbaum M, Cesar CLG. Acesso e uso de serviços de saúde pela população brasileira, Pesquisa Nacional de Saúde 2013. Rev Saude Publica. 2017;51 Supl 1:3s. https://doi.org/10.1590/s1518-8787.2017051000074
https://doi.org/10.1590/s1518-8787.20170... .

Lung cancer was declining among men from the Southeast and South regions and in residents of the Capitals of the Midwest region. This, however, was not observed among women, for whom, in a generalized way, the rates were ascending in the period studied. In this case, because it is a cancer with still high lethality, the decrease in mortality occurs as a function of the decrease in incidence. The increased mortality from lung cancer among women has already been identified in studies in previous periods³³33. Malta DC, Moura L, Souza MFM, Curado MP, Alencar AP, Alencar GP. Tendência da mortalidade por câncer de pulmão, traquéia e brônquios no Brasil, 1980-2003. J Bras Pneumol. 2007;33(5):536-43. https://doi.org/10.1590/S1806-37132007000500008
https://doi.org/10.1590/S1806-3713200700... ^,3434. Azevedo e Silva G, Noronha CP, Santos MO, Oliveira JFP. Diferenças de gênero na tendência de mortalidade por câncer de pulmão nas macrorregiões brasileiras. Rev Bras Epidemiol. 2008;11(3):411-9. https://doi.org/10.1590/S1415-790X2008000300008
https://doi.org/10.1590/S1415-790X200800... .

The prevalence of smokers in Brazil decreased by 19% between 2008 and 2013, with a decline in all regions, in urban and rural areas, and in most states. There was a drop in all educational levels, but the highest prevalence was seen among blacks and mixed-race³⁵35. Arnold M, Sierra MS, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global patterns and trends in colorectal cancer incidence and mortality. Gut. 2017;66(4):683-91. https://doi.org/10.1136/gutjnl-2015-310912
https://doi.org/10.1136/gutjnl-2015-3109... . Smoking in Brazil is one of the lowest in the world and this may have a positive impact in the future, provided that control, regulation and prevention policies are maintained and strengthened³⁶36. Malta DC, Vieira ML, Szwarcwald CL, Caixeta R, Brito SMF, Reis AAC. Tendência de fumantes na população brasileira segundo a Pesquisa Nacional de Amostra de Domicílios 2008 e a Pesquisa Nacional de Saúde 2013. Rev Bras Epidemiol. 2015;18 Supl 2:45-56. https://doi.org/10.1590/1980-5497201500060005
https://doi.org/10.1590/1980-54972015000... . Although the reduction in the prevalence of smokers occurred similarly in both sexes, it should be considered that the introduction of smoking among women occurred later; therefore, the increase in the occurrence of lung cancer still reflects this condition³⁶36. Malta DC, Vieira ML, Szwarcwald CL, Caixeta R, Brito SMF, Reis AAC. Tendência de fumantes na população brasileira segundo a Pesquisa Nacional de Amostra de Domicílios 2008 e a Pesquisa Nacional de Saúde 2013. Rev Bras Epidemiol. 2015;18 Supl 2:45-56. https://doi.org/10.1590/1980-5497201500060005
https://doi.org/10.1590/1980-54972015000... .

Colorectal cancer, the third and fourth most frequent among women and men, respectively, shows an evolutionary pattern that deserves attention. Although rates in the Southeast and South regions are higher among men, they are three times higher than in the North.

The occurrence of colorectal cancer shows different patterns between countries. In several countries in Eastern Europe, Latin America and Asia, there was an upward trend in incidence and mortality. In Canada, European countries and Singapore, the incidence continues to increase, while there is a decline in mortality. In the group restricted to high HDI countries, there is a downward trend in both incidence and mortality³⁵35. Arnold M, Sierra MS, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global patterns and trends in colorectal cancer incidence and mortality. Gut. 2017;66(4):683-91. https://doi.org/10.1136/gutjnl-2015-310912
https://doi.org/10.1136/gutjnl-2015-3109... .

It is noteworthy that, in the North and Northeast regions, the trends in mortality from colorectal cancer increased more significantly in the interior, but there was no difference in magnitude between sexes. This may be an indication that the increase in obesity and sedentary lifestyle, classic risk factors for this type of tumors³⁷37. Lega IC, Lipscombe LL. Review: diabetes, obesity and cancer-pathophysiology and clinical implications. Endocr Rev. 2020;41(1):33-52. https://doi.org/10.1210/endrev/bnz014
https://doi.org/10.1210/endrev/bnz014... , is occurring widely in the country. This increase may be a consequence of a real increase in the incidence, which, together with the lack of access to specialized diagnosis and treatment, may have an effect on mortality. Large differences in access to health services are seen between regions in Brazil, with a higher proportion of medical appointments in the South and Southeast, among people with better living conditions and in regions with higher HDI³⁰30. Stopa SR, Malta DC, Monteiro CN, Szwarcwald CL, Goldbaum M, Cesar CLG. Acesso e uso de serviços de saúde pela população brasileira, Pesquisa Nacional de Saúde 2013. Rev Saude Publica. 2017;51 Supl 1:3s. https://doi.org/10.1590/s1518-8787.2017051000074
https://doi.org/10.1590/s1518-8787.20170... .

The main limitation of this study stems from the quality of information on deaths in the country. Due to the great discrepancy in terms of completeness of information on the underlying cause of death, especially in the North and Northeast regions in the first two decades of the study period¹²12. França E, Teixeira R, Ishitani L, Duncan BB, Cortez-Escalante JJ, Morais Neto OL, et al. Ill-defined causes of death in Brazil: a redistribution method based on the investigation of such causes. Rev Saude Publica. 2014;48(4):671-81. https://doi.org/10.1590/S0034-8910.2014048005146
https://doi.org/10.1590/S0034-8910.20140... ^,1313. Gamarra CJ, Valente JG, Silva GA. Correção da magnitude da mortalidade por câncer do colo do útero no Brasil, 1996-2005. Rev Saude Publica. 2010;44(4):629-38. https://doi.org/10.1590/S0034-89102010000400006
https://doi.org/10.1590/S0034-8910201000... , correction of ill-defined deaths and those registered as uterine cancer without other specification. Thus, all deaths registered as undefined underlying cause were distributed, which are classified in Chapter 16 of ICD – 9 from 1979 to 1995 and in the corresponding chapter of ICD-10 (18) from 1996.

Some previous studies⁶6. Guerra MR, Bustamante-Teixeira MT, Corrêa CSL, Abreu DMX, Curado MP, Mooney M, et al. Magnitude and variation of the burden of cancer mortality in Brazil and Federation Units, 1990 and 2015. Rev Bras Epidemiol. 2017;20 Supl 1:102-19. https://doi.org/10.1590/1980-5497201700050009
https://doi.org/10.1590/1980-54972017000... ^,3838. Global Burden of Disease Cancer Collaboration. Global, regional, and national cancer incidence, mortality, years of life lost, years lived with disability, and disability-adjusted life-years for 29 cancer groups, 1990 to 2017: a systematic analysis for the Global Burden of Disease Study. JAMA Oncol. 2019;5(12):1749-68. https://doi.org/10.1001/jamaoncol.2019.2996
https://doi.org/10.1001/jamaoncol.2019.2... ^,3939. Bigoni A, Antunes JLF, Weiderpass E, Kjærheim K. Describing mortality trends for major cancer sites in 133 intermediate regions of Brazil and an ecological study of its causes. BMC Cancer. 2019;19:940. https://doi.org/10.1186/s12885-019-6184-1
https://doi.org/10.1186/s12885-019-6184-... follow the methodology proposed by the study Global Burden of Disease (GBD)⁴⁰40. GBD 2015 Mortality and Causes of Death Collaborators. Global, regional, and national life expectancy, all-cause mortality, and cause-specific mortality for 249 causes of death, 1980-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet. 2016;388(10053):1459-544. https://doi.org/10.1016/S0140-6736(16)31012-1
https://doi.org/10.1016/S0140-6736(16)31... , which redistributes recorded deaths among causes considered as garbage codes. In this set of codes, in addition to the deaths recorded in chapters 16 of ICD-9 and 18 of ICD-10, other ill-defined causes and incomplete diagnoses from other chapters are included. In this study, we chose not to include the other nonspecific codes outside chapters 16 of ICD-9 and 18 of ICD – 10 due to the absence of studies that validate the contribution of redistribution of deaths recorded in the other nonspecific codes among specific types of cancer⁴¹41. Ishitani LH, Teixeira RA, Abreu DMX, Paixão LMMM, França EB. Qualidade da informação das estatísticas de mortalidade: códigos garbage declarados como causas de morte em Belo Horizonte, 2011-2013. Rev Bras Epidemiol. 2017;20 Supl 1:34-45. https://doi.org/10.1590/1980-5497201700050004
https://doi.org/10.1590/1980-54972017000... . This option may have slightly underestimated the magnitude of the rates but did not detract from the comparison between regions in the period.

Previous studies that showed the trend of cancer in Brazil, with methodology applying correction of death data, included shorter periods than presented here: 1980-2006⁵5. Azevedo e Silva G, Gamarra CJ, Girianelli VR, Valente JG. A tendência da mortalidade por câncer nas capitais e interior do Brasil entre 1980 e 2006. Rev Saude Publica. 2011;45(6):1009-18. https://doi.org/10.1590/S0034-89102011005000076
https://doi.org/10.1590/S0034-8910201100... , 1990-2015⁶6. Guerra MR, Bustamante-Teixeira MT, Corrêa CSL, Abreu DMX, Curado MP, Mooney M, et al. Magnitude and variation of the burden of cancer mortality in Brazil and Federation Units, 1990 and 2015. Rev Bras Epidemiol. 2017;20 Supl 1:102-19. https://doi.org/10.1590/1980-5497201700050009
https://doi.org/10.1590/1980-54972017000... and 1996-2016³⁹39. Bigoni A, Antunes JLF, Weiderpass E, Kjærheim K. Describing mortality trends for major cancer sites in 133 intermediate regions of Brazil and an ecological study of its causes. BMC Cancer. 2019;19:940. https://doi.org/10.1186/s12885-019-6184-1
https://doi.org/10.1186/s12885-019-6184-... . In the case of cancer, it is important to understand the evolution of incidence and mortality for a longer period, as it was possible to show with data from SIM since 1978. To optimize the correction of death data, it was disaggregated according to place of residence, age group, sex and calendar year. For this reason, the trends presented here differ from those that do not correct the SIM data but are closer to detecting the directions and trend changes in cancer mortality. It should be emphasized that, in the two ICD classifications, the codes for the total number of malignant neoplasms and for the specific types selected, as well as the codes referring to the ill-defined root cause, are fully corresponding, which allowed to analyze the trend throughout the period with certainty that the use of the two versions did not introduce classification bias⁴²42. Pan American Health Organization, Division of Health and Human Development, Program on Health Situation Analysis. Revisions of the International Classification of Diseases (ICD-9 and ICD10): impact on health statistics. Epidemiol Bull. 1996;17(2):1-5..

Another aspect to be considered is the coverage of the death record. Although it has improved significantly in the country over these four decades, under-registration still exists in certain areas, especially in the states of the North and Northeast regions⁴³43. Mello-Jorge MHP, Laurenti R, Gotlieb SLD. Análise da qualidade das estatísticas vitais brasileiras: a experiência de implantação do SIM e do SINASC. Cienc Saude Coletiva. 2007;12(3):643-54. https://doi.org/10.1590/S1413-81232007000300014
https://doi.org/10.1590/S1413-8123200700... ^,4444. Lima EEC, Queiroz BL. Evolution of the deaths registry system in Brazil: associations with changes in the mortality profile, under-registration of death counts, and ill-defined causes of death. Cad Saude Publica. 2014;30(8):1721-30. https://doi.org/10.1590/0102-311X00131113
https://doi.org/10.1590/0102-311X0013111... . Since this study analyzed data from 1978 to 2017, it was decided not to perform the correction of the under-registration because no studies were found that evaluate the coverage of SIM from 2013 onwards. This option may have partially underestimated the rates calculated for the North and Northeast regions. However, since the main types of cancer occur from adulthood, and considering that most of the under-registration occurs in the age group up to one year of age⁴⁵45. Almeida WS, Szwarcwald CL. Adequação das informações de mortalidade e correção dos óbitos informados a partir da Pesquisa de Busca Ativa. Cienc Saude Coletiva. 2017;22(10):3193-203. https://doi.org/10.1590/1413-812320172210.12002016
https://doi.org/10.1590/1413-81232017221... , the underestimation due to under-registration should not have been so high.

As for the application of linear regression by parts to evaluate the trend in time, it should be considered that, if the time interval is too small, trajectory changes may not present statistical significance due to the reduced number of points. However, this did not occur because, in the case of a long study period, the APC that were calculated covered periods of at least four years. In addition, the results of this study considered only those intervals that reached statistical significance. It should also be noticed that the number of deaths in each unit of analysis was not small, considering that the data were disaggregated to the level of the Capitals and the interior of each large region, which decreased the possibility of instability in the estimation of rates and of high heterogeneity in their distribution, that is, lower variance.

Another limitation of linear models for trend analysis is the possibility of residual autocorrelation. In this work, the models diagnosed with autocorrelation in residues were re-estimated via generalized least squares with a first-order autoregressive correlation structure-AR(1). No higher order correlation structures were identified. Therefore, when existing, autocorrelation was adequately treated in the model.

It is worth emphasizing that, at the same time as the reduction of some cancers related to poverty is observed, there is a real increase in new cases of cancer associated with factors arising from reproductive and hormonal history and diet⁴4. Bray F, Jemal A, Grey N, Ferlay J, Forman D. Global cancer transitions according to the Human Development Index (2008–2030): a population-based study. Lancet Oncol. 2012;13(8):790-801. https://doi.org/10.1016/S1470-2045(12)70211-5
https://doi.org/10.1016/S1470-2045(12)70... . Thus, for example, the time crossing point between cervical and breast cancer incidence curves is a typical indicator of cancer transition patterns in middle- and low-income countries⁴⁶46. Bray F, Jemal A, Torre LA, Forman D, Vineis P. Long-term realism and cost-effectiveness: primary prevention in combatting cancer and associated inequalities worldwide. J Natl Cancer Inst. 2015;107(12):djv273. https://doi.org/10.1093/jnci/djv273
https://doi.org/10.1093/jnci/djv273... .

It was seen that, in the Southeast and South regions, this intersection between mortality from breast cancer and cervical cancer occurred before the 1980s among women living in the Capitals. Already in these same regions, for residents in the interior, this transition occurred later. Intermediate situations occurred in the Northeast and Midwest regions, where the crossing point of mortality between these two types of cancer in the Capitals occurred in 1992 and 1993, respectively. In the interior, however, it occurred only in 2008 in the Northeast region and in 2007 in the Midwest region. The most extreme situation is seen in the northern region, where this transition has not yet occurred either in the Capitals or in the interior, indicating a picture found in poor countries⁴4. Bray F, Jemal A, Grey N, Ferlay J, Forman D. Global cancer transitions according to the Human Development Index (2008–2030): a population-based study. Lancet Oncol. 2012;13(8):790-801. https://doi.org/10.1016/S1470-2045(12)70211-5
https://doi.org/10.1016/S1470-2045(12)70... .

In the same logic, among men, the crossing points of mortality curves for colorectal cancer, which is positively associated with HDI³⁵35. Arnold M, Sierra MS, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global patterns and trends in colorectal cancer incidence and mortality. Gut. 2017;66(4):683-91. https://doi.org/10.1136/gutjnl-2015-310912
https://doi.org/10.1136/gutjnl-2015-3109... , and stomach cancer, which is known to be related to poverty⁴4. Bray F, Jemal A, Grey N, Ferlay J, Forman D. Global cancer transitions according to the Human Development Index (2008–2030): a population-based study. Lancet Oncol. 2012;13(8):790-801. https://doi.org/10.1016/S1470-2045(12)70211-5
https://doi.org/10.1016/S1470-2045(12)70... , can be identified. It was in the capitals of the southern region that this crossing took place first; in 1998, the death rates from colorectal cancer exceeded those of stomach. In the same region, for those who live in the interior, this only happened in 2016. In the Southeast region this transition occurred in 2007 in the Capitals and only in 2016 in the interior. In the Midwest and Northeast regions, only in the Capitals the mortality rates from colorectal cancer in men exceeded those of stomach (2014 and 2017, respectively). The pattern of greater need was seen in the northern region, where stomach cancer rates are higher than those of colorectal cancer, both in the Capitals and in the interior.

Some results require a careful examination to understand the mechanisms involved both in relation to the accuracy of death information and in the quality of health care. The marked growth of prostate, colorectal and breast cancers within the North and Northeast regions deserves detailed investigation. One hypothesis would be an improvement in the diagnostic capacity, but with a high proportion of cases detected in late stages²⁹29. Santos-Silva I, De Stavola BL, Renna Junior NL, Nogueira MC, Aquino EML, Bustamante-Teixeira MT, Azevedo e Silva G. Ethnoracial and social trends in breast cancer staging at diagnosis in Brazil, 2001-14: a case only analysis. Lancet Glob Health. 2019;7(6):e784-97. https://doi.org/10.1016/S2214-109X(19)30151-2
https://doi.org/10.1016/S2214-109X(19)30... . The hypothesis of improvement of the information on the underlying cause cannot also be ruled out, even considering that the correction was performed with redistribution of ill-classified deaths. It is possible that in these regions the correction was not enough to approach the actual weight of the occurrence of these tumors. Emphasis should also be given to the widespread growth of lung cancer in women, which was highly expressive in the interior of the North and Northeast regions.

Finally, the increase in rates for all cancers in recent years, particularly in the northern and northeastern regions, is worrying. This may mean a real increase in the incidence due to greater exposure to risk factors related to a sedentary lifestyle, but, in the case of mortality, the difficulties of access, especially to early detection and treatment of tumors of good prognosis such as breast and prostate, can lead to an increase in mortality.

Brazil is a country in economic transition and, during these four decades, has undergone several demographic, social and political modifications⁴⁷47. Vasconcelos AMN, Gomes MMF. Transição demográfica: a experiência brasileira. Epidemiol Serv Saude. 2012;21(4):539-48. https://doi.org/10.5123/S1679-49742012000400003
https://doi.org/10.5123/S1679-4974201200... , which interfere in the trend of specific types of cancer, configuring a specific transition pattern that assumes its own characteristics with important regional contrasts. The declines in mortality, birth and fertility levels observed since 1950 indicate that several changes have not occurred uniformly and simultaneously, but markedly shaped by socioeconomic inequality between and within regions²³23. Szwarcwald CL, Souza Júnior PRB, Marques AP, Almeida WS, Montilla DER. Inequalities in healthy life expectancy by Brazilian geographic regions: findings from the National Health Survey, 2013. Int J Equity Health. 2016;15(1):141. https://doi.org/10.1186/s12939-016-0432-7
https://doi.org/10.1186/s12939-016-0432-... . In the Southeast, South and Central-West regions, this process was more accelerated, while in the North and Northeast, mortality and fertility levels were higher, with younger age structures ⁴⁷47. Vasconcelos AMN, Gomes MMF. Transição demográfica: a experiência brasileira. Epidemiol Serv Saude. 2012;21(4):539-48. https://doi.org/10.5123/S1679-49742012000400003
https://doi.org/10.5123/S1679-4974201200... .

Since the creation of the Unified Health System, incorporated into the Federal Constitution in 1988, there has been a great expansion of public services, with a significant improvement in health equity, but not enough to reduce extreme inequalities throughout the country⁴⁸48. Machado CV, Azevedo e Silva G. Political struggles for a universal health system in Brazil: successes and limits in the reduction of inequalities. Global Health. 2019;15 Suppl 1:77. https://doi.org/10.1186/s12992-019-0523-5
https://doi.org/10.1186/s12992-019-0523-... . In recent years, due to successive economic and political crises, some social indicators are already beginning to point to setbacks, and a scenario of uncertainty is configured for the coming years. A recent study, for example, concluded that the increase in unemployment in the country between 2012 and 2017 led to an excess of 30,000 deaths, mainly from cancer and cardiovascular diseases⁴⁹49. Hone T, Mirelman AJ, Rasella D, Paes-Sousa R, Barreto ML, Rocha R, et al. Effect of economic recession and impact of health and social protection expenditures on adult mortality: a longitudinal analysis of 5565 Brazilian municipalities. Lancet Glob Health. 2019;7(11):e1575-83. https://doi.org/10.1016/S2214-109X(19)30409-7
https://doi.org/10.1016/S2214-109X(19)30... . Demographic changes, socioeconomic inequalities and political crises that have followed in the country over four decades may partly explain the decrease in mortality in the Southeast and South regions and the increase among residents of municipalities outside the Capitals, especially in the North and Northeast regions.

The results found reflect the existence of distinct patterns of magnitude and trend of specific types of cancer that are sometimes contrasting and may reflect processes of illness and death with particular dynamics between the large regions of the country and between residents of capitals and other municipalities. While for residents of the capitals of the more developed regions there is a tendency to decreased mortality, in the poorest regions the significant increases, especially in the interior, make clear a profile of extreme inequity in health.

REFERENCES

Publication Dates

History