Assessing gender disparities in excess mortality of heroin or cocaine users compared to the general population

Abstract Background Previous analyses of excess mortality in drug users compared with the general population have almost always been based on mortality ratios, reporting much higher figures in women than men. This study tests the hypothesis that being a heroin or cocaine user adds more death risk in women than men in Spain. Methods A retrospective cohort of 15,305 heroin users (HUs) and 11,905 cocaine users (CUs) aged 15–49 starting drug treatment in 1997–2007 was recruited in Spain and followed until December 2008 to determine vital status and cause of death. Excess mortality in men and women compared to the general population was assessed with directly age-standardized rate ratios (SRRs) and differences (SRDs). Results SRR was significantly higher in women than men for all causes (14.7 vs. 9.4), natural causes (8.7 vs. 6.2), overdose (331.6 vs. 163.9) and other external causes (46.9 vs. 11.8) among HUs; and for overdose (170.8 vs. 40.5) and other external causes (21.0 vs. 4.7) among CUs. However, the opposite happened with SRD for all causes (1294 vs. 1845 deaths/100,000 person-years), natural causes (675 vs. 1016 deaths/100,000 person-years) and overdose (331 vs. 619 deaths/100,000 person-years) among HUs, while no significant SRD gender disparities were observed among CUs. Conclusion Compared with the general population, being a heroin user adds greater absolute risk in men than women, but this does not happen with cocaine users. Similar results would likely have been found in most published cohort studies if this indicator had been used; the exclusive use of relative indices of disparity as in previous meta-analysis can be extremely misleading.