Abstract
Background
Environmental lead exposure continues to be a major societal and public health issue in low- to middle-income countries (LMICs). Lead exposure is associated with numerous adverse health effects, negative individual behaviour, and societal ills. Scientific evidence from high-income-countries (HICs) showed that environmental exposure to lead could contribute to violent and criminal behaviour. This study aimed to determine the prevalence of blood lead levels (BLLs) and its association with violent criminal behaviour among young males in conflict with the law. The objectives of the study were; i) to describe demographic characteristics, behavioural traits (violent and criminal) and blood lead levels; ii) to determine the risk factors linked with elevated blood lead levels in the study participants; iii) to determine causes of high blood lead levels in the study participants; iv) to describe risk factors associated with violent behaviour, v) to examine the association between blood lead levels and criminal offending, and vi) to assess the direct and indirect pathways in the relationship between blood lead and violent behaviour in the study population.
Methodology
To attain the objectives, a cross-sectional analytical study was conducted. Young males in conflict with the law were recruited from two secure facilities in Gauteng Province (GP). Data collection was undertaken in phases and included reviewing participants’ records, blood samples for lead content analysis, and the administration of a questionnaire to determine risk factors. Ten questions extracted from a Youth Self-Reported (YSR) violent behaviour scale were included in the questionnaire to determine violent behaviour, while criminal offending was determined through coding of the crime type captured in the participants records in two categories (violent and non-violent crime). Lastly, follow-up interviews were conducted with participants with a self-reported
vi
history of lead exposure. After entering the collected data onto the RedCap database management software, it was transferred to MS Excel for cleaning prior to exporting to STATA version 15.1 for descriptive and inferential analysis. Binary logistic regression was used to determine and describe risk factors to elevated BLLs. Crude and adjusted odds ratios determined the relationship between criminal offending (binary variable) and risk factors, including violence and blood lead levels as a categorical variable. Continuous blood lead levels and other risk factors were fitted into a linear regression model as an independent variable, and the violent behaviour scale as dependent variable to determine associations. Thereafter, structural equation modeling (SEM) was used to determine the direct and indirect pathways between blood lead levels and violence, taking into consideration the study confounders. The study conformed to ethical standards throughout.
Results
In total, the study sample equaled 192 participants. There sample comprised 111 (58%) non-violent offenders and 81 (42%) violent offenders. The majority of the participants were between the ages of 16 and 18 years (n=146; 76%), and in this age group most were violent offenders (n=86; 59%). The results confirmed that the total study mean for violent behaviour scales was 24.74, and violent offenders were more prone to violent behaviour (scale mean=27.52). The geometric mean blood lead level for the total sample was 3.81 μg/dL, while that for violent offenders and non-violent offenders respectively equalled 4.39 μg/dL and 3.14 μg/dL. Sixty-six participants (34%) had elevated BLLs (mean=10.96 μg/dL), and most of them were violent offenders (n=49; 44%). Binary logistic analyses showed that the risk factors for elevated blood lead levels in the study were previous exposure to lead (p≤0.001) and residing in a house with peeling paint on the exterior (0.024). Linear regression did not show an association between violent behaviour and blood lead levels (p=0.186) after adjusting for confounding factors. However, being older and having been hospitalised due to violence were strong risk factors of violent behaviour, with a history of lead exposure having a negative effect. The final logistic model analysis showed an association between
vii
criminal offending with elevated BLLs (AOR=3.819, 95% CI: 1.486-2.389), drug use (AOR=2.281, 95% CI: 1.088-4.784), and being abused before (AOR=2.813, 95% CI: 1.349-5.866); while a history of lead exposure had a lower impact on criminal offending (AOR=0.004, 95% CI: 0.000-0.047). The SEM pathway analysis showed a significant association between a history of lead exposure and blood lead levels (p<0.001). The direct pathways between violent behaviour with a history of lead exposure (p<0.001), age (p=0.026), having experienced abuse (p=0.004), and hospitalisation due to violence (p<0.001) showed a significant association.
Conclusion
The study showed that blood lead levels were not associated with violent behaviour. However, an decrease in blood lead led to a decrease on the violent behaviour scale. A history of lead exposure had an indirect effect on violent behaviour through blood lead levels. Age, previous experience of abuse and hospitalisation due to violence were risk factors of violent behaviour in the study. The study also found an association between criminal offending and elevated blood lead levels. However, a study on bone lead levels might be necessary to investigate lifetime lead exposure and its association with violent criminal behaviour. The study findings point to a potential role for clean environments in the prevention of lead poisoning and crime, especially violent crimes in South Africa. Lastly, this is the first study to investigate these phenomena; therefore, the results can be used as a baseline in the study population and for developing policies to combat these challenges.
KEY WORDS
Blood lead level; environmental exposure; Youth Self-Reported; violent behaviour; antisocial behaviour, violent crime; non-violent crime; environmental health.