Suicidal behavior in Colombia (2016–2025): multivariate analysis of risk determinants, vulnerability profiles and territorial inequities

Abstract

From 2016 to 2024, Colombia’s Public Health Surveillance System (SIVIGILA) recorded nearly 259,000 suicide attempts. When the projected estimate for 2025 is included, this figure rises to approximately 297,000 attempts. In 2023, the rate reached 78.3 attempts per 100,000 inhabitants, the highest value in the surveillance series, representing a cumulative increase of 132.3% compared with 2016. This study reconstructs the series using a multivariate design based on multiple correspondence analysis (MCA), binary logistic regression, and k-means clustering, drawing on open-access records from the National Institute of Health, DANE, and SISPRO, without access to identified microdata. The findings show that suicidal behavior in Colombia is not a homogeneous phenomenon, but is closely associated with sex, age, socioeconomic status, and geographic location. Women accounted for between 63% and 65% of reported suicide attempts throughout the series, whereas men accounted for 78% of completed suicides. Individuals aged 15–29 years represented more than 60% of reported cases. The multivariate analysis identified five distinct risk profiles and showed that the co-occurrence of an established suicide plan, untreated depression, and a previous suicide attempt was associated with risk in a non-additive manner. Overall, the findings suggest that public policies addressing suicidal behavior should be tailored to the characteristics of different risk profiles and geographic contexts, rather than relying on universal approaches.