Trends and Spatial Patterns of Suicide Among Adolescent in Ecuador, 1997-2016

1. INTRODUCTION

Suicide is a global public health problem, ranking among the top 20 leading causes of mortality. According to the World Health Organization (WHO), every 40 seconds a life is lost due to self-harm; that accounts for 800,000 deaths per year, 1.4% of total deaths [1].

People younger than 19 are especially susceptible; suicide was the second leading cause of mortality among females and the third among males in the last decade - 6% of all deaths in 2015 (32,499 girls and 34,650 boys) — were caused by suicide, the majority of them via self-harm [2, 3]. As national reports can't account for all suicides (due to underreporting or misreporting); suicide could be the leading of death worldwide among young people [1].

Suicidal behavior - suicide ideation, planning suicide, attempting suicide, and suicide itself — has the highest rates in low and middle-income countries (LMIC); they account for 75% of youth suicides globally [2]. In America, for instance, 65,000 suicides have been noted per year [4]. A comparison of the last two decades (1990-1999 and 2000–2009) showed the average suicide rate per 100,000 young people in South America increased from 1.04 to 2.32 for boys, and from 1.45 to 2.30 for girls [5]. Guyana, Suriname, and Ecuador have reported the highest rates of suicide in America for girls (Suriname reported the highest rate of suicide for boys only) [5, 6]. Among Andean countries, Ecuador has the second highest suicide rate (7.1 per 100,000 population) after Bolivia (18.7/100,000), and ahead of Colombia (6.1/100,000) and Peru (5.8/100,000).

During the last 15 years, 13,024 suicides were officially reported in Ecuador [7]. Data from the National Institute of Statistics and Census (INEC) indicate that between 1997 and 2016, an average of 191 adolescents died per year [8]. Given the high social stigma associated with suicide, it is likely that current data underestimate actual rates, especially in people younger than 19 years old [4]. Previous studies have looked at the epidemiology of suicide in the Ecuadorian general population [7, 9], but none of them explored the specific trends among teenagers.

The objective of this study is to describe the temporal trend of suicide in adolescents between 10-19 years old, from 1997 to 2016, and to analyze the presence of spatiotemporal clusters of high mortality rates and the spatial distribution due to suicide in adolescents in the country, from 2011 to 2016

2. MATERIAL AND METHODS

3. RESULTS

3,824 people between 10-19 years committed suicide in Ecuador during the investigated period (1997–2016), out of which 52.3% of deaths occurred in males (n=1,999) and 47.7% in females (n=1,825). There were 4 registered suicide cases for children under 10 years of age in 2015, two suicides in children 8 years old and two suicides in children 9 years old. Stratified by area of residence, 67.1% (n=2,565) lived in urban areas, 29.9% (n=1,144) in rural areas, and 3% (n=115) in unspecified areas. Intentional self-harm by hanging, strangulation, and suffocation represented 52% of all cases (n=1,986). Self-poisoning and exposure to other and unspecified chemicals and noxious substances represented 21.9% of cases (n=840), while self-poisoning and exposure to pesticides represented 17.3% of cases (n = 662). These three ICD-10 codes (X70, X69, and X68, respectively) accounted for 91.2% of suicides (Table 1).

We did not find significant differences between the monthly distribution of suicides and sex (p=0.14).

3.2. Spatial-temporal Cluster Analysis

Our space-temporal analysis identified one significant cluster for high occurrence of suicide in the years from 2011 to 2013; it included 83 cantons in 13 provinces (Bolivar, Cotopaxi, Chimborazo, Napo, Pastaza, Tungurahua, Orellana, Azuay, Cañar, El Oro, Guayas, Morona Santiago, and Zamora Chinchipe), totaling 295 suicides; with a RR of 2.51 (LLR: 81.07; p <0.001), and an annual mortality rate of 3.1 per 100,000 population (Fig. 1a).

3.4. Spatial Distribution

At the canton level, average annual suicide rates per 100,000 population ranged from 0 to 12.93; smoothed mortality rates ranged from <0.001 to 12.93. The Global Moran I index for the study period shows a positive spatial autocorrelation (0.27; p=0.001). We identified a high-risk cluster (High/High) that included 14 cantons located in 6 provinces (Cotopaxi, Napo, Pastaza, Azuay, Morona Santiago, and Zamora Chinchipe). Clusters with low rates (Low/Low) included 18 cantons located in 6 provinces (Esmeraldas, El Oro, Guayas, Loja, Manabí, and Santa Elena) (Fig. 2).

4. DISCUSSION

We found that suicide among young people increased in Ecuador during the study period; the regression analysis showed a significant growing trend in men living in rural and urban areas, suicide among young women also showed a growing trend although not significant.

Hanging, chemical poisoning, and the ingestion of pesticides caused 90% of all suicides.

Both the spatial-temporal and the spatial analyses identified the same primary high-risk cluster; with most of the provinces located within the cluster in the Amazon region and in the Southern Highlands (RR=2.89); the cluster included the coastal province Guayas. The secondary cluster, located in the north, spans the three main geographical regions of the country. Cotopaxi, Napo, Pastaza, Azuay, Morona Santiago, and Zamora Chinchipe were identified within a high-risk cluster by the Global Moran I index; it should be noted that these six provinces appeared as a high-risk area in three different analyses. Low rates were found mostly in the coastal region.

Our temporal trend findings differ from studies conducted in high-income countries where suicide rates in adolescents have either stabilized or decreased [19]. It is, however, comparable to those of other countries in Latin America [20, 21], as well as those of previous studies in the country. Suicide is highly prevalent in Ecuador, being most common in urban areas [7, 9].

While male suicide rates significantly increased across almost all groups, suicide continues to be the leading cause of death among female adolescents. Female suicide rates increased in the youngest (10-14 years old) group, with most of the effect focused in rural areas. This trend is similar to that found in a study in OECD (Organization for Economic Co-operation and Development) countries where female suicide rates in highly rural countries like China and India exceeded those of males [22]. Male suicides rates also increased significantly in rural areas (10-14 years old APC=5.6% & 15-19 years old APC=4.83%). It is believed that economic, political, and financial problems in the continents during this period contributed to this outcome [6]. Further studies are required to understand the influence of aspects like socio-economic status and gender in these settings.

Suicide in adolescents varies significantly between genders; societal expectations of gender may play a role by emphasizing man’s obligation to be strong, independent, and prone to risks - men, for instance, are less likely to seek help during periods of depression or suicidal behavior [19, 20, 23]. Thus, explaining why men in Ecuador and in other countries kill themselves more often than women. It is unknown whether suicide attempts are more common among women in Ecuador, as suggested by the literature. While a school-based survey by the Ministry of Public Health cited by Svanemyr et al. [24] showed that 17.9% of teenagers between 13 and 15 years considered attempting suicide, the study was not segmented by gender. Further studies are required to understand the prevalence of both suicide ideation and suicide attempt.

Hanging and poisoning remain the most common mechanisms of death [7, 9], with males using more violent means than females, and as such being more likely to be successful [25]. Intake of chemicals and harmful substances could be aggravated by the ease of access, low cost, and lack of regulation (especially using organophosphates and carbamates, which are frequently used pesticides in banana and flower plantations) [26, 27].

As shown in our spatial analyses, two regions were at a higher risk: the Amazon region and the Southern Highlands region. Provinces in the Amazon are populated mainly by indigenous people. Suicide rates of indigenous people are considerably higher than those of non-indigenous [28, 29]; this is also true for children and teenagers [19, 30]. Further studies are required to understand the relationship between indigenous vs. non-indigenous risk factors. The Southern Highlands, on the other hand, had the highest migratory density per canton according to the 2001 Population and Housing Census [31]; a previous study looking at the association between parental migration and health behavior found that adolescents (13-18 years) were at higher risk of suicide ideation [32].

A previous study by Ortiz-Prado et al. [7] suggested a relationship between high altitudes areas and suicide. However, our results contradict this hypothesis. Most provinces (4 of 6) at high risk in all of our spatial analysis are located in the Amazon, with altitudes between 0 and 200 meters. These authors also suggested family dysfunction, school failure, sexual abuse, alcohol, drug use, and immigration patterns could aggravate suicides. While this escapes the scope of our analysis, we can assert that we did not observe significant variations between months, which might have been related to school failure — students are exposed to stress factors including a new environment, new academic expectations, financial hardships, and new schedules [33]. Researchers should look at these and other phenomena, such as malnutrition [34], child abuse [35] or high migratory flows to gain insights on the causes of suicide in the country.

Only 1.2% of the budget of the Ministry of Public Health (MSP) was directed towards mental health strategies; 59% of this amount was allocated to psychiatric hospitals distributed across the country [36]. In 2014, 270 psychiatrists worked in public health facilities - 45 less than in 2002 - and were distributed as follows: 162 in the Highlands, 103 in the Coast, and 5 in the Amazon; there were no specialists in the Galápagos Islands [37] - a significant shortage of access to mental health services. The budget deficit, lack of national preventive strategies, and a limited number of ​​mental health professionals might have worsened suicide in the country. As of 2016, the Ministry proposed to establish an intervention strategy for people with suicidal ideation and self-inflicted injuries in coordination with educational institutions, community, and family members. Such a strategy includes preventive tasks, comprehensive health care, as well as guidance and management to family members [38]. In spite of these actions, it is unlikely that such strategies will achieve a ten percent reduction in suicide rates by 2020, as proposed by WHO's comprehensive mental health action plan.

Our study is strengthened by the use of jointpoint regression model; which avoids the pre-specification of periods by the researcher. The statistical method developed by Kulldorf identifies the distribution of the disease in time and space, evaluating the significance of clusters in high-risk, low-risk, and all-risk categories; as well as an early detection of epidemics in epidemiological surveillance [39].

Our study was limited by the difficulty in establishing an association between the observed trends and potential associated clinical factors. We also could not associate clusters with risk factors in each geographical area. We have already pointed out the necessity of future studies in this regard.

We were also limited by the level of data granularity. Until 2010, for instance, all registries lacked information on ethnicity and location detail at the level of cantons; limiting the coverage of our spatiotemporal analysis. Similarly, it was not until recently that the issuing of medical certificates of death were made mandatory. Since most Ecuadorians are Catholic, families sometimes take steps to prevent the mention of suicide as the cause of death [9]. Finally, suicide reports lack information about mental health history, as well as familial and/or housing conditions.

The systematic review of Bennett et al, recommend coordinated implementation of school-based suicide prevention, prevention of repeat suicide attempts in youth who seek care and who do not seek care; according to sex and gender differences; and ethnic group within a national collaborative youth suicide research-to-practice network [40]. The role of the network would be to identify and facilitate increased implementation of promising programs linked to rigorous evaluation and to eliminate the use of ineffective ones at the regional, provincial, and national level [40]. Wasserman et al. concluded that the school environment is the best system we have to perform primary prevention programs designed to improve mental health and give information about unhealthy lifestyles among youth, whilst at the same time raising the general awareness-level about mental health and mental problems [41].

CONCLUSION

Adolescent suicide rates trend increased during the study period. The spatial analysis indicates the presence of high occurrence clusters in the Amazon and the Southern Highlands regions.

These results alert the Ecuadorian population about the mental health of adolescents and the need to implement programs that counteract major contributors to suicide risk, bullying, and violence in educational centers and familial environments. Our analyses can be used to develop comprehensive strategies to improve the quality of life at a national level by focusing on the most vulnerable: victims of poverty, migration, and ethnic minorities.

REFERENCES