The complexity of explaining ethnic differences in suicide and suicidal behaviours.

The paradox of studying suicide or suicidal behaviour is that it appears to be a definable and measurable phenomenon but in practice it is much more difficult to pin down. It is not a very frequent behaviour but this makes studying comparisons in rates between ethnic groups problematic because a small numbers of events, even in relatively large studies, can make differences difficult to interpret. Yet, potentially much can be learnt about risk factors, access to treatment and its effectiveness by studying ethnic differences in rates. Suicidal ideation, suicidal attempts and completed suicide are linked but different phenomena. Kessler (1999) found that 34% of those with suicidal ideation made suicide plans and of those 72% went on to make suicide attempts; certainly suicidal attempts are the most powerful predictor of completed suicide. The link between suicidal ideation and completed suicide is more tenuous, sufficient to say that suicidal ideation does often represent the hopelessness associated with moderate to severe depression. Each is worth studying and findings from one can shed light on the others. A complicating factor is that definitions of what constitutes suicidal behaviours differ by jurisdictions, countries and possibly by ethnic groups. In the respected South African SASH Study ethnicity was classified by Apartheid-era racially classified social group: Black, Coloured, Indian/Other and White. In this classification, the White group is nearly synonymous with more advantaged socioeconomic status. Thus, the association of non-fatal suicidality with higher socioeconomic status and White ethnicity in early childhood and young adulthood from the South African SASH Study (van Pletzen et al. 2012) was unexpected and does not accord with greater levels of adversity and discrimination in non-White ethnic groups in South Africa that might be expected to be associated with increased suicidal risk (Williams et al. 2004). Exclusion of the sickest people from ethnic minorities from the study does not seem to be an explanation as the response rate was high and the proportion of Black respondents in the study was not far from the population norm (Williams et al. 2004). However, elsewhere it is reported from this study that rates of mood disorders and serious or moderate psychiatric disorders tended, although not statistically significantly, to be higher in other ethnic groups than Whites which is out of keeping with the lower levels of suicidality (Williams et al. 2008). Could it be that there is underreporting of suicidal ideas due to it being more stigmatising for the Black and Coloured groups than for the White group to report suicidal ideas or behaviour? This would mean that it is less stigmatising to admit to depression than to report suicidal behaviour is that likely? The prevalence rates of early non-fatal suicidal behaviours of 3.4% (95% CI: 2.6 4.1) in this study are low compared to Wild’s (2004) survey of grade 8 and grade 11 Ethnicity & Health Vol. 17, Nos. 1 2, February April 2012, 3 6