Abstract:

Objective:

To provide updated information regarding the 12-month prevalence and associated sociodemographic factors for suicide ideation and behavior (plan and attempts) to substantiate preventive programs in Mexico.

Materials and methods:

Cross-sectional nationally representative survey, conducted during 2016 (n=56 877) among those 12-65 years old living in rural, urban and metropolitan dwellings.

Results:

The prevalence of suicide ideation in the last 12-months was 2.3%, 0.8% of the sample reported a plan and 0.7% reported a suicide attempt. All three outcomes were about two times more common among females and suicide plan and attempt were less common among the elderly (50-65 years old). Suicide attempts were more common in urban than in rural areas. The state of Tabasco showed an increased prevalence of ideation, plan and attempts when compared to national average rates.

Conclusions:

Public health measures to diminish and treat suicidal behavior are urgently needed all over the country.

Keywords:
suicide; health surveys; Mexico

Resumen:

Objetivo:

Dar información actualizada sobre la prevalencia de 12 meses y los factores sociodemográficos asociados para la ideación y el comportamiento suicida (plan e intentos) para fundamentar programas preventivos en México.

Material y métodos:

Encuesta transversal representativa a nivel nacional, por sexo y grupos de edad, realizada durante 2016 (n = 56 877) entre personas de entre12 y 65 años de edad en zonas rurales, urbanas y metropolitanas.

Resultados:

La prevalencia de ideación suicida en los últimos 12 meses fue de 2.3%, 0.8% de la muestra informó un plan y 0.7% informó un intento de suicidio. Los tres resultados fueron aproximadamente dos veces más comunes entre las mujeres y el intento de suicidio fue menos común entre los más viejos (50 a 65 años). Los intentos de suicidio fueron más comunes en las zonas urbanas que en las rurales. El estado de Tabasco mostró una mayor prevalencia de ideas, planes e intentos en comparación con las tasas promedio nacionales.

Conclusiones:

Es urgente y necesario que existan medidas de salud pública para disminuir y tratar el comportamiento suicida en todo el país.

Palabras clave:
suicidio; encuestas epidemiológicas; México

Introduction

While the suicide death rate is declining overall around the globe by 26%, suicide has been growing in Mexico during the last decade or so (a 16.6% increase between the years 2000 and 2012).¹1. World Health Organization. Preventing suicide: a global imperative. Geneva: WHO, 2014. Available from: http://apps.who.int/iris/bitstream/handle/10665/131056/9789241564779_eng.pdf;jsessionid=EDAA480817E21E5F8BE7CAE465F1227B?sequence=1
http://apps.who.int/iris/bitstream/handl... This increase is more noticeable among Mexican females (a 55.1% increase) compared to males (a 10.0% increase),¹1. World Health Organization. Preventing suicide: a global imperative. Geneva: WHO, 2014. Available from: http://apps.who.int/iris/bitstream/handle/10665/131056/9789241564779_eng.pdf;jsessionid=EDAA480817E21E5F8BE7CAE465F1227B?sequence=1
http://apps.who.int/iris/bitstream/handl... and among youth (10-19 years old) and those between 30 and 49 years of age.²2. Borges G, García-Pacheco JA, Borsani L. Epidemiology of suicidal behavior in Mexico. In: Pan American Health Organization (ed). Prevention of suicidal behavior. Washington, DC: Pan American Health Organization, 2016:48-56. Available from: http://iris.paho.org/xmlui/bitstream/handle/123456789/31166/9789275119198-eng.pdf?sequence=1&isAllowed=y
http://iris.paho.org/xmlui/bitstream/han... New data available through the Global Burden of Disease study 2015³3. Institute for Health Metrics and Evaluation. Global burden of disease study 2015. Seattle: Institute for Health Metrics and Evaluation, 2015. [accessed August 14,2017]. Available at: Available at: http://ghdx.healthdata.org/gbd-results-tool
http://ghdx.healthdata.org/gbd-results-t... also shows that the age standardized suicide death rate in Mexico increased by 22% between the years 2000 to 2015, more so for women (a 37% increase) than for men (a 18.4% increase). The increase was also more apparent among those aged 15 to 49 years (a 28.8% increase). In the year 2010, suicide ranked in the 11th place for cause of disability adjusted life years in Mexico.⁴4. Lozano R, Gómez-Dantés H, Garrido-Latorre F, Jiménez-Corono A, Campuzano-Rincón J, Franco-Marina F, et al. Burden of disease, injuries, risk factors and challenges for the health system in Mexico. Salud Publica Mex. 2013;55(6):580-94. https://doi.org/10.21149/spm.v55i6.7304
https://doi.org/10.21149/spm.v55i6.7304...

We know much less about suicide ideation and behavior during this period of increased suicide rates in the country. The last national sample of Mexican adults that reported on suicide ideation and behavior (plan and attempts), the Mexican National Comorbidity Survey (M-NCS; Borges and colleagues, 20075. Borges G, Nock MK, Medina-Mora ME, Benjet C, Lara C, Chiu WT, et al. The epidemiology of suicide-related outcomes in Mexico. Suicide Life Threat Behav. 2007;37(6):627-40. https://doi.org/10.1521/suli.2007.37.6.627
https://doi.org/10.1521/suli.2007.37.6.6... )⁵5. Borges G, Nock MK, Medina-Mora ME, Benjet C, Lara C, Chiu WT, et al. The epidemiology of suicide-related outcomes in Mexico. Suicide Life Threat Behav. 2007;37(6):627-40. https://doi.org/10.1521/suli.2007.37.6.627
https://doi.org/10.1521/suli.2007.37.6.6... was conducted in 2001-2002. Although in the year 2005, a large survey for adolescents (12-17 years old) living in the great Mexico City metropolitan area reported on suicide ideation and behavior,⁶6. Benjet C, Borges G, Medina-Mora ME, Zambrano J, Aguilar-Gaxiola S. Youth mental health in a populous city of the developing world: results from the Mexican Adolescent Mental Health Survey. J Child Psychol Psychiatry. 2009;50(4):386-95. https://doi.org/10.1111/j.1469-7610.2008.01962.x
https://doi.org/10.1111/j.1469-7610.2008... ^,77. Borges G, Benjet C, Medina-Mora ME, Orozco R, Nock MK. Suicide ideation, plan and attempt in the Mexican Adolescent Mental Health Survey. J Am Acad Child Adolesc Psychiatry. 2008;47(1): 41-52. https://doi.org/10.1097/chi.0b013e31815896ad
https://doi.org/10.1097/chi.0b013e318158... this survey was geographically restricted. A summary of these surveys dates from year 2010⁸8. Borges G, Orozco R, Benjet C, Medina-Mora ME. Suicide and suicidal behaviors in Mexico: Retrospective and current status. Salud Publica Mex. 2010;52(4):292-304. and it clearly shows the limitations of both surveys to report on 12-month prevalence of suicide ideation and behavior because of the rarity of these phenomenon in Mexico and elsewhere.⁹9. Borges G, Nock MK, Haro JM, Hwang I, Sampson N, Alonso J, et al. Twelve month prevalence of and risk factors for suicide attempts in the WHO World Mental Health Surveys. J Clin Psychiatry. 2010;71(12):1617-28. https://doi.org/10.4088/JCP.08m04967blu
https://doi.org/10.4088/JCP.08m04967blu... Since a prior suicide ideation and attempt are among the most powerful risk factors for suicide,¹1. World Health Organization. Preventing suicide: a global imperative. Geneva: WHO, 2014. Available from: http://apps.who.int/iris/bitstream/handle/10665/131056/9789241564779_eng.pdf;jsessionid=EDAA480817E21E5F8BE7CAE465F1227B?sequence=1
http://apps.who.int/iris/bitstream/handl... ^,1010. Franklin JC, Ribeiro JD, Fox KR, Bentley KH, Kleiman EM, Huang X, et al. Risk factors for suicidal thoughts and behaviors: a meta-analysis of 50 years of research. Psychol Bull. 2017;143(2):187-232. https://doi.org/10.1037/bul0000084
https://doi.org/10.1037/bul0000084... understanding the dynamics of these outcomes is crucial for preventive efforts in Mexico.

Large datasets reporting on 12-month prevalence suicidal ideation and behavior with national representativeness are sorely needed to substantiate public health programs in this area. Here, we report the 12-month prevalence and associated sociodemographic factors of suicide ideation, plan and attempt for a large nationally representative survey among those 12 to 65 years old living in rural, urban and metropolitan dwellings. The survey was conducted during 2016 (n=56 877).

Materials and methods

Datasets

The dataset is the 2016 Encuesta Nacional de Consumo de Drogas, Alcohol y Tabaco (Encodat 2016, for its acronym in Spanish - National Survey on Consumption of Drugs, Alcohol and Tobacco),¹¹11. Secretaría de Salud, Instituto Nacional de Psiquiatría Ramón de la Fuente Muñiz, Instituto Nacional de Salud Pública, Comisión Nacional contra las Adicciones. Resumen metodológico. In: Secretaría de Salud, Instituto Nacional de Psiquiatría Ramón de la Fuente Muñiz, Instituto Nacional de Salud Pública, Comisión Nacional contra las Adicciones. Encuesta nacional de consumo de drogas, alcohol y tabaco 2016-2017. México: Secretaría de Salud, 2017 [cited 2017 Aug 14]. Available from: Available from: https://www.gob.mx/cms/uploads/attachment/file/234804/RESUMEN_METODOL_GICO.pdf
https://www.gob.mx/cms/uploads/attachmen... previously called the National Addiction Survey (ENA). This survey is the latest in a series of nationally representative household surveys that focus on alcohol, tobacco and drug use behavior that have been conducted approximately every five years in the general population, conducted previously in 1988, 1998, 2002, 2008 and 2011.¹²12. Villatoro J, Medina-Mora ME, Fleiz BC, Moreno-López M, Robles NO, Bustos-Gamiño M, et al. El consumo de drogas en México: resultados de la Encuesta Nacional de Adicciones, 2011. Salud Ment. 2012;35(6):447-57. The response rate of the Encodat 2016 was 74%. The 2016 survey was planned as representative of national urban and rural households. The sampling procedure for the Encodat 2016 was probabilistic, multistage and stratified. The selection of sample units was carried out in multiple stages. In the first stage of sampling, Basic Geostatistical Areas (AGEBs) or localities (similarly to the US ZIP codes) were selected within each stratum. In the second stage, localities within the Rural AGEB and contiguous sets of dwellings within the locations were chosen. In the Urban and Metropolitan strata, blocks (“manzanas”) within the AGEB and housing inside the blocks were selected. Lastly, individuals were selected from households with an adolescent aged 12 to 17 or an adult aged between 18 and 65. All selections were probabilistic. Weights were generated to account for the selection process and the survey is representative of the national population within the age and geographical areas. The Encodat interviewed a total of 56 877 respondents. While the goals of the Encodat 2016 are to generate information on the prevalence of tobacco, alcohol and drug use,¹³13. Secretaría de Salud, Instituto Nacional de Psiquiatría Ramón de la Fuente Muñiz, Instituto Nacional de Salud Pública, Comisión Nacional contra las Adicciones. Consumo de drogas: prevalencias globales, tendencias y variaciones estatales, In: Secretaría de Salud, Instituto Nacional de Psiquiatría Ramón de la Fuente Muñiz, Instituto Nacional de Salud Pública, Comisión Nacional contra las Adicciones. Encuesta nacional de consumo de drogas, alcohol y tabaco 2016-2017. Ciudad de México: Secretaría de Salud, 2017 [cited 2017 August 14]. Available from: Available from: https://www.gob.mx/cms/uploads/attachment/file/234856/CONSUMO_DE_DROGAS.pdf
https://www.gob.mx/cms/uploads/attachmen... it also includes other outcomes of interest for the country, such as the scale of 12-month prevalence of suicide ideation, plan and attempt, that is comparable to the scale used before in the M-NCS. The survey questionnaire was applied through a face-to-face interview, where sections of the individual questionnaire which might cause underreporting (the most sensitive ones) were applied via a computer assisted audio program with headphones (ACASI system) self-administered by the respondent, under the direction of the interviewer to ensure an atmosphere of confidentiality while contextual topics were explored through face-to-face interviews.

Ethical considerations

The Encodat 2016 survey protocol was approved by the Research and Ethics Committees of both the National Institute of Psychiatry and the National Institute of Public Health. All the participants were read a letter of informed consent and information was only gathered on those who gave their consent.

Variables

Suicide ideation, plan and attempt

As mentioned above, the Encodat used the same scale for recent suicide ideation and behavior as the MNC-S, and prior reports of prevalence and associated factors for suicidal behaviors in the M-NCS surveys have been extensively published before.⁵5. Borges G, Nock MK, Medina-Mora ME, Benjet C, Lara C, Chiu WT, et al. The epidemiology of suicide-related outcomes in Mexico. Suicide Life Threat Behav. 2007;37(6):627-40. https://doi.org/10.1521/suli.2007.37.6.627
https://doi.org/10.1521/suli.2007.37.6.6... ^,1414. Nock M, Borges G, Ono Y. Suicide: global perspective from the WHO World Mental Health Surveys. New York: Cambridge University Press, 2012. For this report, we are interested in estimating the prevalence of three outcomes, suicide ideation, suicide plans and suicide attempts (suicidal behavior) in the year 2016. Because of limitations in the length of the questionnaire and interview time, we did not include non-suicidal self-injury behavior in the Encodat. Respondents were asked about 12-month experiences of suicidal ideation (‘Have you seriously thought about committing suicide?’), suicide plans (‘Have you made a plan for committing suicide?’), and suicide attempts (‘Have you attempted suicide?’). Those positive for a 12-month suicide attempt were further asked whether: 1. Whether this attempt led to a hospitalization or medical treatment for the injuries, and 2. To characterize if: (a) if the attempt was serious and it was only because of luck that they did not die; (b) if the attempt was serious but they knew the method was not foolproof, and (c) if the attempt was a cry for help and they did not want to die. Because self-administered surveys have been shown to yield higher rates of reporting of embarrassing behaviors than interviewer-administered surveys,¹⁵15. Turner CF, Ku L, Rogers SM, Lindberg LD, Pleck JH, Sonenstein FL. Adolescent sexual behavior, drug use, and violence: increased reporting with computer survey technology. Science. 1998;280(5365): 867-73. https://doi.org/10.1126/science.280.5365.867
https://doi.org/10.1126/science.280.5365... these experiences were listed in a self-administered booklet and referred to by letters (events ‘A’, ‘B’ and ‘C’) for respondents who were able to read. In the Encodat 2016, these experiences were gathered by the use of the Audio Computer-Assisted Self-Interview Software (ACASI) system. The ACASI technology allows respondents to listen to prerecorded survey questions through headphones and record responses using a touch screen or keypad or, alternatively, they may simultaneously read the questions from a tablet, or they may close or hide the screen for complete privacy.

Sociodemographic factors

Sociodemographic factors previously shown to be related to suicidal behaviour,¹⁴14. Nock M, Borges G, Ono Y. Suicide: global perspective from the WHO World Mental Health Surveys. New York: Cambridge University Press, 2012. such as sex, age/cohort: adolescents (12-17 years old), adults (18-49 years old) and mature adults (50-65 years old), years of education (in four categories: 0-6, 7-9, 10-12 and 13+), occupation (paid job, student, homemaker, retired and other), marital status (married/cohabiting, separated/divorced/widowed and never married), country region: North-west (Baja California, Baja California Sur, Nayarit, Sinaloa and Sonora), North (Coahuila, Chihuahua, Durango, Nuevo León, San Luis Potosí, Tamaulipas and Zacatecas), Central-west (Aguascalientes, Colima, Guanajuato, Jalisco and Michoacán), Central-east (Ciudad de México, Guerrero, Hidalgo, Estado de México, Morelos, Puebla, Querétaro and Tlaxcala) and South-east (Campeche, Chiapas, Oaxaca, Quintana Roo, Tabasco, Veracruz and Yucatán) and geographical area (rural, urban and metropolitan), were used to report on common associated factors for these three outcomes.

Analyses

First, we compared the distribution of key individual (age and sex) and demographic (geographical area, country region and state level) variables and the prevalence of the outcomes (suicide ideation, plan and attempt) for the Encodat 2016. Significance tests for these cross-tabulations were conducted using design-based Pearson χ² tests. Next, we estimated model-adjusted prevalence ratios (PRs) based on Generalized Linear Models (GLM) with log link and binomial distribution.¹⁶16. Cummings P. Methods for estimating adjusted risk ratios. Stata J. 2009;9(2):175-96. Thus, PRs (the prevalence rate in the exposed divided by the prevalence rate in the unexposed), were computed for each type of suicidal behavior in adjusted models with all sociodemographic variables, in the total population. The Encodat 2016 is also the first survey with a large enough sample size per national jurisdictions (i.e, states) that will allow us to report the prevalence of suicidal behavior by state level and to test whether differences in prevalence existed when compared to national averages. For this variable (state level), instead of using one single state as the reference category, we used the grand mean, so that each state is compared against the national average prevalence rates for ideation, plan and attempt. Thus, PRs for each state are interpretable as statistically different (an increase or decrease) from the marginal national mean prevalence.^**Stata 13 Base Reference Manual. College Station, TX: Stata Press, 2013. https://www.stata.com/manuals13/r.pdf All analyses incorporated weights developed for these surveys, as described above. For our GLM models, we estimated standard errors and 95% confidence intervals (CIs) using the Taylor series linearization method with STATA version 13 1 to adjust for the design effects, stratification, clustering, and unequal weighting of the observations. ^‡‡ Stata Corp. Stata 13 survey data reference manual. College Station, TX: Stata Press, 2013. https://www.stata.com/manuals13/svy.pdf

Role of the funding source

The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility to submit for publication.

Results

Table I shows the distribution of Encodat by key demographic factors and the overall prevalence of suicide ideation, plan and attempt. The overall prevalence of suicide ideation was 2.3%, 0.8% reported a plan and 0.7% reported a suicide attempt.

Table IIa presents the distribution of outcomes by sex and age groups. All three prevalences among males were about half the size of the corresponding prevalences among females. Among males, the prevalence of ideation, plan and attempts were somehow similar across all age groups, with a non-significant tendency for lower prevalences of plan and attempts among the older males. For females, there was a decreased prevalence of ideation, plan and attempts (all statistically significant) in older females compared to younger females. Among those with a suicide ideation (table IIb) 35.8% made a plan, and 32.4% made an attempt. Among those that reported an ideation and a plan, 68.2% made an attempt. Among those that attempted suicide, 27.3% (112/497) were attempts leading to some form of medical treatment. A large percentage of attempts, 75% (373/497) were planned attempts.

We examined the distribution of medical consequences, self-rated intent and methods used in the attempt (data not showed). In summary: among those with a planned attempt (n=373) 31.2% had some form of medical treatment and among those with an unplanned attempt (n=124) 15.4% had some form of medical treatment (p=0.019). The attempts were self-rated by their intent as: (a) serious (34.9%) (b) serious but not foolproof (26.9%) and (c) a cry for help (38.2%) (data not shown). While we observed a trend, there was not a statistically significant relationship between intent and medical consequences: 35.3% of all serious attempts lead to medical treatment, 26.2% of serious but not foolproof lead to medical treatment and 20.8% of those considered a cry for help lead to medical treatment (p=0.283).

Among those with an attempt, the most common method used was stabbing/puncture object (38%), followed by overdose using either controlled or over the counter drugs (17% each), hanging (15%), and firearms, overdose of illicit drugs, poisoning, drowning, jumping from high places or transit injuries, all of them with prevalences below 10%. Serious attempts were more likely among those who tried to overdose with controlled drugs compared to those not overdosing (45.8 vs 32.6%; p=0.013); a similar tendency was observed among those who tried hanging vs not (54 vs 31.6%; p=0.032).

As per table IIb female ideators were more likely to make a plan (39.1%) than male ideators (29.1%) (p=0.05), but not an attempt (27.9 vs 34.6%, p=0.17). We found no differences by sex or age groups in these two variables of medical consequence and intent. Among those that had an ideation, adolescents (12-17 years) were more likely to make a plan than those in the 50-65 age group (43.2 vs 19.9%, p=0.0007) and were more likely to make an attempt (41.7 vs 23.2%, p=0.0105). When compared to adolescents (12-17 years old), those in the older age group (50-65) were less likely to make a plan or attempt, but among those that attempted suicide, their attempt was as likely to lead to medical treatment (18.4 vs 29.0%, p=0.1995). Some differences by sex and age groups were found among methods for attempt suicide (data not shown). Males were more likely to try to hang themselves or use firearms, while females were more likely to overdose with either controlled or over the counted drugs. The only difference by age group was observed among those using stabbing/puncture objects, who were younger (data not shown).

Table III presents the distribution of prevalences by three levels of geographical aggregation, which is by geographical area, in 5 regions and at the state level. Inspecting the crude prevalences, we found no differences by geographical area. At the regional level, there were no differences of ideation (marginal non-significant at p=0.076), plan (marginal non-significant at p=0.069) and attempt. On the other hand, when we disaggregate these prevalences by state level, we found significant differences for ideation (borderline significant at p=0.051), plan and attempt.

Table IV presents the results of a multivariate model for the prevalence of suicide ideation, plan and attempt using our key individual variables and the information on geographical area information and the state. As apparent from this table, males had a consistently lower prevalence ratio compared to females in all outcomes; those more educated had consistently lower prevalence ratios and those “never married” consistently higher prevalence ratios. Those in the older age group had lower prevalence ratios of plan and attempt (as those between 18-49 years old). While those in rural areas had similar crude prevalences, the prevalent ratio was of statistical significance for ideation and attempt among urban areas. The states of Campeche, Tabasco and Veracruz had higher prevalence ratios of ideation when compared to the national average, while Coahuila and Mexico City had lower prevalence ratios. Tabasco had an increased prevalence ratio and Coahuila had a lower prevalence ratio for a plan. For attempt, Chihuahua and Tabasco had increased prevalence ratios while Chiapas and Hidalgo had had lower prevalence ratios.

Discussion

Summary

The main findings of this study were that: (1) the overall prevalence of suicide ideation was 2.3%, 0.8% reported a plan and 0.7% reported a suicide attempt; (2) the prevalences among males were about half the size of the corresponding prevalences among females, and the prevalences were higher among the younger females, and (3) in general, after statistical adjustments, there remained few differences in prevalence of the three outcomes across the country, but Tabasco still remained as the Mexican state with the highest rates for all outcomes. Our data on the distribution of suicide attempt by medical consequences (27.3% of all attempts lead to medical treatment), intent (34.9% were serious attempts) and methods (stabbing/puncture object was mentioned by 38% of all attempts), have the potential to help to calculate more precisely the burden of suicide for Mexico and to formulate public policies.

The Encodat 2016 is the largest survey ever done in Mexico reporting on the national (metropolitan, urban and rural) and state level data on suicide ideation, plan and attempts. There is simply no prior study in Mexico that fully allows us to compare our results. Nevertheless, a prior report shows that in year 2001-2002 the prevalence of suicide ideation, plan and attempts among urban setting respondents within 18-65 years old were 2.4% for suicide ideation, plan was 0.9% and attempt was 0.5%.5 On another national survey, that used some slightly different questions for suicide ideation,¹⁷17. Borges G, Medina-Mora ME, Orozco R, Oueda C, Villatoro J, Fleiz C. Distribución y determinantes sociodemográficos de la conducta suicida en México. Salud Ment . 2009;32(5):413-25. the prevalence of any suicidal ideation was reported to be 8.8% (ranging from 4.2 to 6.2%, depending on the question used), the prevalence of a suicide plan was 1.22%, and 0.79% of the sample reported to attempt suicide within the past 12 months, with the prevalence of suicide attempts that required medical attention being 0.13%. Because of differences in the scope of the population sampled and age groups included we cannot make a direct comparison between these prevalences. There is, nevertheless, some limited evidence that suicidal behavior, among the young adults in the Metropolitan Mexico City, is increasing.¹⁸18. Borges G, Benjet C, Orozco R, Medina-Mora ME. The growth of suicide ideation, plan and attempt among young adults in the Mexico City metropolitan area. Epidemiol Psychiatr Sci. 2017;26(6):635-43. https://doi.org/10.1017/S2045796016000603
https://doi.org/10.1017/S204579601600060... Monitoring changes in the prevalence of suicidal behavior, together with changes in suicide death, is a next step in the analyses of the Encodat 2016.

The main associated factors explored in this report, female sex, young age groups, those with fewer education and those never married, were similar to previous reports in Mexico⁸8. Borges G, Orozco R, Benjet C, Medina-Mora ME. Suicide and suicidal behaviors in Mexico: Retrospective and current status. Salud Publica Mex. 2010;52(4):292-304. and elsewhere.¹⁹19. Nock MK, Borges G, Bromet E, Cha CB, Kessler RC, Lee S. Suicide and suicidal behavior. Epidemiol Rev. 2008;30(1):133-54. https://doi.org/10.1093/epirev/mxn002
https://doi.org/10.1093/epirev/mxn002... As has been argued recently,¹⁰10. Franklin JC, Ribeiro JD, Fox KR, Bentley KH, Kleiman EM, Huang X, et al. Risk factors for suicidal thoughts and behaviors: a meta-analysis of 50 years of research. Psychol Bull. 2017;143(2):187-232. https://doi.org/10.1037/bul0000084
https://doi.org/10.1037/bul0000084... individual risk factors for suicide have poor predictive power. Efforts to create risk indexes, adaptable for specific populations, seem like a better and more fruitful approach.¹⁷17. Borges G, Medina-Mora ME, Orozco R, Oueda C, Villatoro J, Fleiz C. Distribución y determinantes sociodemográficos de la conducta suicida en México. Salud Ment . 2009;32(5):413-25. While we have focused here on the impact of sociodemographic factors on suicide ideation and suicidal behavior, other factors such as family influences, psychological traits, mental disorders and substance abuse, among others, should be considered. The use of the large array of associated factors for suicidal behavior in the Encodat 2016 will be a matter of a future report from our group.

This is the first time that a household survey in Mexico is capable of reporting state level information for suicide ideation, plan and attempt. After statistical controls, few differences in these prevalences remained. The most consistent of these findings was the excess of prevalence for all outcomes in the State of Tabasco, a state that has had for a long time one of the highest rates of death by suicide in the country.⁸8. Borges G, Orozco R, Benjet C, Medina-Mora ME. Suicide and suicidal behaviors in Mexico: Retrospective and current status. Salud Publica Mex. 2010;52(4):292-304. While a few reports have delved into the increase risk for suicide in Tabasco,²⁰20. Fresán A, González-Castro TB, Peralta-Jiménez Y, Juárez-Rojop I, Pool-García S, Velázquez-Sánchez MP, et al. Gender differences in socio-demographic, clinical characteristics and psychiatric diagnosis in/of suicide attempters in a Mexican population. Acta Neuropsychiatrica. 2015;27(3):182-8. https://doi.org/10.1017/neu.2015.6
https://doi.org/10.1017/neu.2015.6... ^,2121. Aguilar-Velázquez DG, González-Castro TB, Tovilla-Zárate CA, Juárez-Rojop I, López-Narváez ML, Fresán A, et al. Gender differences of suicides in children and adolescents: Analysis of 167 suicides in a Mexican population from 2003 to 2013. Psychiatry Res. 2017;258:83-7. https://doi.org/10.1016/j.psychres.2017.09.083
https://doi.org/10.1016/j.psychres.2017.... to the best of our knowledge the Encodat 2016 is the first report to show that suicidal behavior is also increased in this state. Future work in this area should try to explore the special circumstances and associated factors that may be present in Tabasco, but this is beyond the scope of the current report.

Limitations

While this is the largest study so far in Mexico on the prevalence of suicide ideation, plan and attempt, the small 12-month prevalence of these behaviors prevented us from carrying out additional subgroup analyses.⁹9. Borges G, Nock MK, Haro JM, Hwang I, Sampson N, Alonso J, et al. Twelve month prevalence of and risk factors for suicide attempts in the WHO World Mental Health Surveys. J Clin Psychiatry. 2010;71(12):1617-28. https://doi.org/10.4088/JCP.08m04967blu
https://doi.org/10.4088/JCP.08m04967blu... ^,2222. Kessler RC, Berglund P, Borges G, Nock M, Wang PS. Trends in suicide ideation, plans, gestures, and attempts in the United States, 1990-1992 to 2001-2003. JAMA. 2005;293(20):2487-95. https://doi.org/10.1001/jama.293.20.2487
https://doi.org/10.1001/jama.293.20.2487... Since there is no data in the country that compares suicide in urban and rural places, this matter is speculative at the moment. Future work in the country should pay much more attention to rural areas, where also larger percentages of traditional local ethnicities live. The Encodat 2016 used in some of their respondents the ACASI system module of interview. Given the small prevalence of suicide ideation and behavior reported here, we cannot rule out that these changes played a part in our findings. Future work from our group will delve into possible similarities or differences in the sample by method of interview, but this is beyond the scope of this current work. Suicide ideation and behavior is a complex and multi-causal phenomena and research including a broader array of determinants, besides the sociodemographics considered here, are needed.

Conclusion

Suicidal behavior was widely spread all over the country with very few locations standing out as at higher risk. A national program for suicide prevention, instead of local short-term activities, is needed. Our results may help to substantiate some key points for such preventive program. While focusing in preventive activities on more specific groups as detected here, such as the youth, females, single people and those with lower educational attainment may yield better results, efforts to reach everyone are needed. At least one in every four attempts will lead to medical treatment and one in every three attempts is self rated as serious, suggesting that physicians and other clinical personnel could be trained as in important gatekeeper to prevent that attempts latter develop into a fatal incident.

Funding

The Encodat 2016 was partially sponsored by the National Commission Against the Addictions. The National Institute of Psychiatry and the National Institute of Public Health elaborated the survey study design and were jointly responsible for the data collection.

References

Publication Dates

History