Background Although Northern Ireland can, after 30 years of violence, be described as a ‘post-conflict’ society it remains intensely segregated with over 15 km of concrete and metal barriers (‘Peace-lines’) erected to reduce the inter-communal tension at so-called interface areas. To date there has been no exploration of the health of the communities at these interface areas.

Objectives This study aims to use the consumption of antidepressant and anxiolytic medication as proxy indicators of mental health to determine if living close to or very close to a peace-line affects mental health.

Design Data relating to the consumption of antidepressant and anxiolytic medication over the 2 year period October 2008 to September 2010 was provided by an electronic prescribing database covering the entire population. Peace-lines were mapped using geographic information system techniques. Individuals were deemed to be living very close to or close to a peace-line if one passed through their Census Output Area (average population 336) or Super Output Area (average population 1900) of residence. Logistic regression analysis with adjustment for area deprivation was carried out and multi-level modelling to adjust for clustering within General Practitioner (GP) practices.

Results Peace-lines are predominantly located in the most deprived areas with 87% in the most deprived quintile. Overall, 19% of those not close to a peace-line received an antidepressant over the study period compared to 27% of individuals living close or very close to a peace-line. After adjustment for age and sex, the likelihood of receiving an antidepressant increases in areas with proximity to a peace-line, OR=1.67 (95% CI 1.58 to 1.77) and OR=1.76 (95% CI 1.63 to 1.91) for those living close and those living very close a peace-line respectively, compared to the rest of the population. Further adjustment for deprivation attenuated but did not eliminate the relationship; OR=1.14 (95% CI 1.08 to 1.21) and OR= 1.18 (95% CI 1.10 to 1.27). The findings for anxiolytic drug uptake were very similar.

Conclusion The residents of areas near to the peace-lines appear to have worse than expected mental health. However, as these data are cross-sectional it is not possible to determine whether the peace-lines are contributing to or ameliorating health status. This record linkage methodology could be used to examine the effect of selective migration in producing these patterns and any future erection or demolition of these walls.