OBJECTIVES: To determine seasonal trends in the number of deaths from violence in Bogotá, Colombia from 1997-2006 and to analyze any possible associations with the city's economic cycles. METHODS: An ecologic study based on the official death toll from violence registered between January 1997 and September 2006 in Santa Fe de Bogotá, Colombia. Auto Regressive Integrated Moving Average (ARIMA) time-series trimester models were used to describe the behavior of violent deaths. The explanatory variables for the city's economic cycles were the trimester series of the consumer price index, the unemployment rate, and the total number of people employed; these were related to the death series through transfer function models. RESULTS: Of the 36 575 total deaths from violence in Bogotá, 57.7% were homicides and 23.7% resulted from traffic accidents. The homicide, suicide, and traffic accident deaths showed downward trends, with a noteworthy seasonal pattern in the traffic accident and suicide deaths that peaked during the last two trimesters of each year. Significant associations were found between all the death series from all the different causes and at least one economic series, and in each case a respective transfer function model was identified. CONCLUSIONS: Homicides constituted the largest portion of the violent deaths, and exposed a seasonal pattern to this death type. Significant associations were found between deaths from violence and some economic variables, revealing a cycle tied to the life of the city. The socioeconomic circumstances affecting a society were confirmed to influence the level of violence.
OBJECTIVES: To determine seasonal trends in the number of deaths from violence in Bogotá, Colombia from 1997-2006 and to analyze any possible associations with the city's economic cycles. METHODS: An ecologic study based on the official death toll from violence registered between January 1997 and September 2006 in Santa Fe de Bogotá, Colombia. Auto Regressive Integrated Moving Average (ARIMA) time-series trimester models were used to describe the behavior of violent deaths. The explanatory variables for the city's economic cycles were the trimester series of the consumer price index, the unemployment rate, and the total number of people employed; these were related to the death series through transfer function models. RESULTS: Of the 36 575 total deaths from violence in Bogotá, 57.7% were homicides and 23.7% resulted from traffic accidents. The homicide, suicide, and traffic accident deaths showed downward trends, with a noteworthy seasonal pattern in the traffic accident and suicide deaths that peaked during the last two trimesters of each year. Significant associations were found between all the death series from all the different causes and at least one economic series, and in each case a respective transfer function model was identified. CONCLUSIONS: Homicides constituted the largest portion of the violent deaths, and exposed a seasonal pattern to this death type. Significant associations were found between deaths from violence and some economic variables, revealing a cycle tied to the life of the city. The socioeconomic circumstances affecting a society were confirmed to influence the level of violence.