Suicide mortality in the city of São Paulo: epidemiological characteristics and their social factors in a temporal trend between 2000 and 2017. Retrospective study

ABSTRACT BACKGROUND: Suicide is one of the leading causes of death worldwide, accounting for one million deaths annually. Greater understanding of the causal risk factors is needed, especially in large urban centers. OBJECTIVE: To ascertain the epidemiological profile and temporal trend of suicides over two decades and correlate prevalence with social indicators. DESIGN AND SETTING: Descriptive population-based longitudinal retrospective study conducted in the city of São Paulo, Brazil. METHODS: A temporal trend series for suicide mortality in this city was constructed based on data from the Ministry of Health’s mortality notification system, covering 2000-2017. It was analyzed using classic demographic variables relating to social factors. RESULTS: Suicide rates were high throughout this period, increasing from 4.6/100,000 inhabitants in the 2000s to 4.9/100,000 in 2017 (mean: 4.7/100,000). The increase in mortality was mainly due to increased male suicide, which went from 6.0/100,000 to the current 8.0/100,000. Other higher coefficients corresponded to social risk factors, such as being a young adult (25-44 years old), being more educated (eight years of schooling) and having white ethnicity (67.2%). Suicide was also twice as likely to occur at home (47.8%). CONCLUSION: High suicide rates were seen over the period 2000-2017, especially among young adults and males. High schooling levels and white ethnicity were risk factors. The home environment is the crucial arena for preventive action. One special aspect of primary prevention is the internet and especially social media, which provides a multitude of information for suicide prevention.


INTRODUCTION
Suicide is one of the three most common causes of death worldwide. The World Health Organization (WHO) has defined it as aggression or a violent act committed against one's own life, with the intention of death. 1 It is currently one of the most important public health problems and is often attributed to how its victims are affected by society and the collective environment in which they live. 2 It has several supporting risk factors, such as psychological, biological and social factors. 3 Suicide is among the 20 leading causes of death worldwide in different age groups, including adolescents, adults and the elderly. One death by suicide occurs every 40 seconds somewhere in the world. 4 Findings from 2012 revealed that there were around 800,000 suicide deaths worldwide, representing an age-standardized overall annual rate of 11.4 per 100,000 inhabitants, within which males and the age group of young adults formed more significant components. 5 In epidemiological studies, it has been estimated that by 2020, there will be a 50% increase in the annual incidence of suicide deaths worldwide, which could exceed the number of deaths from homicide and war. 6,7 Suicide rates have undergone exponential increases in several countries such as France, China, Switzerland, Belgium, Austria, the United States, Japan and Brazil, as well as in countries in Eastern Europe, which report suicide rates above 16 deaths per 100,000 inhabitants. 6,7 However, there is still great difficulty in assessing the dimensions of this problem and accurately recording suicidal acts. 8 Globally, suicides are the second leading cause of premature mortality among individuals aged this is most markedly so when young and middle-aged men who are about to start or have just started their professional and family lives commit suicide. 9 In the United States, in the early 2010s, the costs per each single suicide were estimated to be over $1 million, while estimates from Ireland, Scotland and New Zealand lay between $2.1 and $2.5 million. In the literature, only 6% of the studies come from low-income countries low-income countries such as sub-Saharan Africa. Most studies have focused on measurements of poverty like unemployment and economic status, while neglecting dimensions such as debt, relative and absolute poverty, and support from welfare systems. 9,10 This opens up a huge gap between the overall numbers of suicides in LMIC (78% of suicides worldwide) and knowledge of costs within the respective societies. 9,10 Considering the significance that suicide has and the public health crisis that it has precipitated, a detailed understanding of the different age groups, genders, ethnicities and places with higher prevalence of suicide in large urban centers (where the characteristics of the environment increase the risks of mental disorders, such as depression, anxiety and stress) is essential. Through this, monitoring can be promoted and possible strategies for reducing individuals' risk of suicide can be implemented. [11][12][13] Although studies have shown that there has been a threefold increase in the suicide rate among males, 12 there are still no reports of this rate over the course of two decades, or 18 consecutive years.
Nor are there any reports on its prevalence in specific age groups, among specific ethnicities or in places in emerging countries with large urban populations. This phenomenon results from a complex network of biological, genetic, psychological, sociocultural and economic interactions. 13 Studies have shown that suicide rates are increasing, but most studies have investigated only specific periods of time, or regions and states. [14][15][16] However, greater specificity for actions and local preventive strategies is required.

OBJECTIVES
The aims of this study were to analyze the epidemiological profile and temporal trends of suicide cases in the city of São Paulo, Brazil, over two decades, and to analyze the prevalence of these cases in relation to social indicators.

METHODS
This was a mortality study (time series) that had the aim of characterizing aspects of suicide mortality in the city of São Paulo (SP), Brazil, from 2000 to 2017.
All the data used were obtained from official secondary sources.
The number of suicides was obtained from the mortality informa- Census data were taken from years in which censuses were performed, and were interpolated for the other years. These datasets are publicly available online.
Suicide was defined as death resulting from intentional selfharm, in accordance with the International Classification of Diseases, tenth edition (ICD-10), which uses codes X60 to X84 and Y87 to identify this outcome.
The 2000-2017 time series was composed of annual suicide rates. These were calculated as suicide mortality coefficients (using the numbers of occurrences divided by the general population per 100,000 inhabitants) and as standardized mortality (using the numbers of occurrences divided by the standardized population per 100,000 inhabitants) for the municipality of São Paulo. Analyses stratified according to sex, age group, education level, ethnicity and place where suicide was committed were also performed.
Values corresponding to unknown age were excluded.

Statistical analysis
We used the SPSS software, version 10.0, to identify and estimate suicide rates and standardized mortality coefficients. To calculate mortality coefficients, population data and data on mortality due to external causes were used. The overall mortality coefficient is The median of the participation percentages, rather than the arithmetic mean, was used with the aim of removing the influence that, in this case, would be exerted by possible occurrences, in certain populations, of strongly disagreeing percentages of participation. 18 The denominator that was used to calculate the standardized coefficient was that of the standard population, which was calculated using the median population of the period studied. Historical series were built for the period from 2000 to 2017.

RESULTS
Over the 18 years that made up this analysis, there were 8,726 deaths due to suicide in the city of São Paulo, corresponding to 4.7 deaths per 100,000 inhabitants, while the world averages ranged from 3.5 to 4.0 deaths per 100,000 inhabitants.
The coefficients were standardized using the standard population provided by the World Health Organization (WHO). The gross coefficient of the population increased from 4.08/100,000 in 2000 to 4.69/100,000 in 2017. It also needs to be taken into account that, because of the taboo surrounding suicide, deaths due to this event may be reported as deaths due to an external cause of unknown type. This may have induced underreporting of the problem. The male-to-female ratio increased from four in 2000 to six at the end of the study period, in 2017. The temporal trend showed that higher male rates were maintained throughout the study period, without any abrupt increase.

DISCUSSION
The main objective of this study was to analyze the epidemiological profile and temporal trends of suicide cases in the city of São Paulo over two decades and to analyze the prevalence of suicide in relation to social indicators. The main finding was that there was a great number of deaths (8,726) due to suicide in the municipality studied, in comparison with other countries, corresponding to 4.7 deaths per 100,000 inhabitants, while the world average ranged from 3.5 to 4.0 deaths per 100,000 inhabitants. Another important finding was that there was an increase in the male stan- and Central European countries were in an intermediate range. 14,21 The differential of the present study was that a suicide rate of 4.7 was ascertained in a single municipality. This urban center can be considered to be a reference point, given its large population.
The results from this city showed that there is a need for greater support for preventive and management actions within public policies. 183 countries shows that the male-to-female ratio varies. 9 Several studies have documented the epidemiology of higher suicide rates among males, which in some countries have reached a ratio of 3:1. 14,19 Higher rates have also been observed in small and medium-sized cities and municipalities. 20,22 One exception is India and China, where the suicide rate among females exceeds that of males. 23 The findings of the present study showed that there was higher prevalence of suicide among males than among females.
The explanation for this, according to some authors, [24][25][26] relates to manifestations of masculinity, which involves behaviors that predispose towards suicide, such as competitiveness, impulsivity and greater access to lethal weapons, including firearms.
In addition, failure to fulfill traditional gender roles, which for men means being the economic provider for the family, generates greater stress and anxiety. Men who live within a patriarchal culture are more sensitive to economic setbacks such as unemployment and impoverishment and more prone to suicide. 25  The place of the suicidal act has been well discussed in the literature. The risk of suicide is around three to five times higher in hospital environments than in the population as a whole. 31 Most hospital-related cases are associated with chronic or terminal illnesses that have become painful and debilitating. 31,32 However, in the present study, the opposite was observed, i.e. the prevalence of suicide was twice as high, over the 18-year period, in the home environment. Across the world, there is little data on the places where suicide attempts are made. If present, the quality of such data is low due to a lack of reliable statistics, which relates to underdiagnosis, misdiagnosis or non-diagnosis and reporting. The WHO does not receive information from any country in the world on this topic, although at least data from emergency rooms and somatic hospitals might be obtained, along with some self-reports. The results from the present study corroborate the data of Lovisi et al., 14 which showed that the home was the most frequent scenario for suicides, accounting for 51%, followed by hospitals, with 26%. The predominant means used were hanging (47%), firearms (19%) and poisoning (14%). 14 The limitation of the present study was that the real degrees of underestimation and underreporting of data were not ascertained.
Thus, the true prevalence of suicide may have been greater, considering the extreme difficulty in accurately assessing the scale of suicidal acts and recording them.