Killing me softly: a meta-analysis examining risk factors associated with suicide among young African American males

Abstract

Increase in suicide rates of African American males since the 1960s has prompted a growing interest among researchers. Although research has increased in this area, suicide remains an issue that is explored far less often among African American males compared to other groups. Moreover, studies exploring risk factors associated with suicide in this group have led to inconclusive results. The current meta-analysis was conducted to synthesize the results of existing literature and to identify risk factors associated with suicide among African American males under the age of 30. Of 25 research articles published between 1970 and 2007 that met inclusion criteria, 48 units of analysis were obtained including 37,927 total subjects. The current analysis identified 57 risk factors that were categorized into 16 constructs: substance use, religion, economic, location, education, family, internalizing, externalizing, ethnic variables, stressor/ conflict, support, medical/somatic, psychological disorder, perception, age, and gender. Risk factors for suicidal behavior were coded and effect sizes between groups were computed. Age yielded the largest magnitude of effect such that suicidal groups of African American males were more likely to be younger than groups of non-suicidal comparisons. Effect sizes of risk factors were also analyzed within four additional domains (attempters, ideators, ideators & attempters, and completrs). Results indicate that age and perception had the largest effects for attempts, while psychological disorders had the least effect. Effects for ideators were largest for substance use and medical problems, while religion was smallest. In a group of ideators and attempters, age was again found to have the largest effect while medical problems had the least. Lastly, factors associated with perception and psychological disorder had the greatest effect for completers while religion had the least effect. Effects of several factors such as religion, location, family, and ethnic variables did not change across suicidal subgroups suggesting that although their effects were not largest, they remained constant across behaviors.