The Effects of Violence and Poverty on Cardiometabolic Risk Biomarkers in Urban Municipalities in Mexico

Background/Purpose

Poverty and violence within cities frequently concentrate in the same places and evidence suggests these exposures have deleterious and indirect consequences on health by increasing stress. In 2005, the average homicide rate per 100,000 inhabitants (HR) in Mexico was 10, but by 2011, after the so-called War on Drugs, it spiked up to a HR of 24. Contrary to the entanglement of violence and poverty often seen in American and European urban neighborhoods with concentrated disadvantages, in Mexico, poverty remained stable in the last two decades while homicide rates increased in a few poor and affluent urban municipalities. These separate trends make Mexico an ideal site to test the independent effects of violence and poverty on health using an index of cardiometabolic risk biomarkers (CVDR), which has been tested as a sensitive measure of stress.

The study profits from the sharp increase in the HR as a natural experiment in affluent and poor municipalities. Difference-in-difference models allowed the examination of changes within persons on an index of cardiometabolic risk.  

Methods

Data and sample. Secondary data from the Mexican Family Life Survey (MXFLS) was used. The MXFLS is a longitudinal panel study representative of the Mexican population by region and by urban localities that collects well-being indicators and biomarkers. The study uses two waves, 2006 and 2012, with an attrition of 12%. After excluding people who changed municipality (15%), the sample size was 16, 679 people age 15-99 years old.

Measures.  The dependent variable is a reduced version of the Framingham score; an additive ordinal variable (0-5) with high levels on these risk factors: Smoking, Obesity (BMI); Blood Pressure; Total Cholesterol; Diabetes (HBA1c). The independent variables, used as treatments, are dummy variables for HR at the municipality level above 35; change in homicide rates above 10 (HRC) between 2006 and 2012; and the official multidimensional poverty percentage measure above 65. Individual controls were age, gender, victimization, and household income.

Results

There were no significant differences on the CVDR in 2006. On average, respondents in all municipalities increased their mean CVDR scores between 2006 and 2012, especially in affluent ones, but the increase was 8.3% higher for people living in a municipality with HR above 35 and 7% higher in people living in municipalities with HRC above 10. Interactions between treatments showed that people living in municipalities with high violence and low poverty had 10% higher CVDR than people in municipalities with low violence and high poverty. Covariates revealed heterogeneous effects by gender: females living in municipalities with HR above 35 had a 10.6% higher CVDR mean than females living elsewhere; the effects were not significant among males, nor by age or victimization.

Conclusions

Exposure to community violence had deleterious health impacts, beyond victimization, trough stress; even in the young population. Municipal structures of concentrated disadvantages differ in high and medium income countries, so different policies to mitigate violence may ensue. Women were more susceptible to the indirect effects of violence and thus a key population to target social services.