Background
Gender-based violence persists as a global public health problem. In 2000, there were an estimated 119 000 female homicides worldwide, for an overall age-adjusted rate of 8.8 per 100 000 population [
1]. Of these, the majority of deaths occurred in low- to middle-income countries, with the highest number of female homicides reported for the African Region, at a rate of 11.8 per 100 000 population. South Africa, despite its democratic transformation, the strength of its emerging economy, and widespread structural and social policy changes since 1994, is estimated to have the highest rate of intimate female homicide in the world, with 8.8 per 100 000 women age 14 years and older murdered by an intimate partner in 1999 [
2].
The international body of literature on female homicide victimisation reveals a proliferation of studies during the last decade that seek to explore the patterns and correlates of homicide victimisation among women. This field of research is dominated by studies that have investigated: (1) female homicide victimisation patterns of occurrence within and between countries [
3‐
6]; (2) the associations with social-structural factors, such as gender and economic inequality, deindustrialisation, urbanisation, ethnic heterogeneity and family disruption [
7‐
12]; and (3) the victim-perpetrator relationship [
13‐
16]. Studies to disaggregate female homicide victimisation, to examine the extent, distribution and patterns of occurrence of specific manners of death, such as asphyxia due to strangulation, are more limited.
Strangulation is considered to be a form of mechanical asphyxia [
17]. The mechanisms of death in strangulation include airway occlusion, resulting in hypoxia; occlusion of the neck vessels or compression of the carotid arteries, leading to cerebral ischemia; and carotid sinus reflex, leading to cardiac arrest [
17]. The term strangulation is specifically used to indicate the external pressure applied to the neck either by means of a ligature or the hands [
17]. Descriptions of homicide victimisation by strangulation are largely to be found within the legal and forensic medicine literature. This research has focused on post-mortem accounts [
18,
19]; assessments of survived strangulation and mitigation of risk [
20‐
23]; preliminary formulations of prevention [
20,
23,
24]; and initial descriptions of the epidemiology of homicidal asphyxia in both males and females. The latter descriptive inquiry has illuminated the epidemiological profile of homicidal strangulation in both high-income settings [
25‐
27], as well as low- to middle-income contexts [
24,
28‐
30]. In general, these studies conclude that strangulation is one of the most common forms of violent asphyxia, and accounts for approximately 10–20% of all homicide deaths in a range of countries [
31], thereby representing a notable cause of death in homicide victims.
Almost all attempted or completed homicides by strangulation involve either ligature strangulation (strangulation with a cord-like object) or manual strangulation (done with the hands or forearms, or standing or kneeling on the victim's throat) [
32]. Ligature strangulation is reported as the more frequently recorded method of asphyxial homicide [
18,
28,
30,
33,
34]. In contrast, there appears to be significant differences in homicidal asphyxia patterns between the sexes, with some investigations on strangulation deaths observing a higher female to male ratio [
20,
32,
34‐
36], and others determining a clear male preponderance among victims [
18,
25,
30]. Findings in terms of age distribution presented as varied, with investigations reporting sharp peaks in strangulation rates for 20–30 year olds, 30–40 year olds, as well as 5–12 year olds in the case of female paediatric and adolescent strangulation deaths [
25,
27,
30,
34]. In an overwhelming number of cases studied, the victim's domestic context represented the location of the crime [
24‐
27] which, in instances of female homicidal strangulation, appeared to further support available data signifying the victim-perpetrator relationship to be primarily intimate [
25‐
27,
36].
Many of the findings reported here are based on small numbers of cases that limit the conclusions that can be drawn, and are essentially descriptive in nature and therefore do not take into account socio-political-cultural concepts considered to be central to the understanding of gender-based violence. The results are nonetheless instructive in the preliminary identification of risk profiles for homicidal asphyxia and the development of implied interventions.
Despite the burgeoning scholarship on the subjects of female homicide victimisation and of asphyxial homicide, the research evidence reviewed highlights that systematic investigations directed specifically at strangulation in the context of female homicide victimisation remain meagre. The near absence of such research in South Africa, against the backdrop of findings emerging from the single documented national study of female homicide in South Africa [
2], as well as national-level data exposing the alarmingly high homicide rates in the country [
37], clearly calls for methodical inquiry into homicidal strangulation in women in South Africa. In a context plagued by persistently high rates of both fatal and non-fatal violence against women, accurate descriptions of the extent and occurrence of violent incidents are required for targeting resources, developing relevant interventions, and enabling more reliable comparisons of national and global female homicide victimisation information.
The current study explores the distribution and patterns of homicidal strangulation of women in the four largest South African metropolitan centres, Tshwane/Pretoria, Johannesburg, Cape Town and Ethekwini/Durban, for the period 2001 to 2005. These cities register amongst the highest rates of homicide in the country [
38]. The following three study questions are addressed, namely:
1. What is the incidence of female strangulation in four South African cities?
2. What is the distribution of female strangulation across age and population groups?
3. What are the typical circumstances of occurrence of female strangulation?
Discussion
This study reports that female homicidal strangulation in urban South Africa ranges from 1.71/100 000 in Durban (2003) to 0.70/100 000 in Johannesburg (2003). These rates, especially the former, are high compared to the few reported for other settings, such as the combined overall strangulation rates (for males and females) of 0.17/100 000 for Jordan and the 1.1/100 000 reported for the United States [
28,
42]. The South African female strangulation rates are consistent with the high overall homicide rates reported in this study for the four cities and the available national estimates, with the latter approximated to be more than seven times the global average [
43]. The strangulation rates reported herein are likely to be high relative to those for other African countries, where the overall homicide rate is up to 30% lower than in South Africa [
44].
Homicidal strangulation is more prominent in females than males in the metropolitan centres of the country, unlike the higher male rates observed for the other major external causes of urban homicide death, such as firearm and sharp force injury [
37]. Strangulation presents as a method favoured by assailants that are considerably physically stronger than their victims, as illustrated by this higher proportion of female deaths [
27]. This is also consistent with the reported association of sexual violence with strangulation [
45].
Female homicidal strangulation has since 2001 generally declined in two of the cities, Pretoria and Johannesburg. Durban, despite a more erratic pattern of occurrence, also reported an overall decrease of female strangulation. This trend echoes that for the overall homicide rates across the four cities. This decrease has occurred in the context of supportive national legislation (for example, the Domestic Violence Act), the enhancement of the criminal justice system (such as the establishment of Special Offences Courts), and other social and health initiatives [
46]. Despite the range of policies and interventions that have been promoted, questions have remained about the impact of these on gender-based violence, with various commentators highlighting South Africa's incomplete transition to democracy, particularly as regards the persisting indicators of gender inequity, patriarchal power, and the feminisation of poverty [
47].
Cape Town is the single exception to this declining trend. The Western Cape Province (of which Cape Town is the capital and the largest metropolitan centre) also reported the highest homicide rate and amongst the highest number of reported cases of rape for this period [
48]. These findings support the proposed links between sexual violence and female strangulation, as highlighted by other research [
45], including investigations into sexual murders of elderly females [
31]. Cape Town also reported the highest number of intoxicated strangulation victims, concordant with elevated alcohol consumption and abuse trends in the Western Cape [
49], and previous reports that accentuate the association between alcohol inebriation and vulnerability to strangulation [
25,
27] and, more broadly, urban violence [
50].
Three cities reported the highest rates of strangulation deaths amongst their older populations. Elderly female victims have been reported to be at higher risk for strangulation due to their physical vulnerability and resulting inability to offer much resistance during an attack [
27,
28]. The emphasis of strangulation amongst the older female population thus differs from South African reports on general female homicide (i.e. irrespective of type), which typically occurs amongst 30 to 44 year olds [
44]. The next highest cluster of deaths occurs amongst young and mature adult females. This is largely consistent with other studies on strangulation, which have reported the mean age of victims to be within the 20 to 39 year range [
25,
27,
30], and with general studies on intimate partner and non-intimate homicide, for which the average ages for South African victims are 30.4 and 41.2 years respectively [
2]. This study did not reveal a significant proportion of strangulation victims to be children, unlike the findings reported in other studies [
27,
28], and despite the increasing proportion of child homicidal deaths in South Africa [
44].
This study reported the highest rates of strangulation amongst coloured women, consistent with the indications of elevated intimate femicide rates amongst coloured females [
38]. In this study, the highest rates occurred in Pretoria, Durban and Cape Town. This was followed by high rates amongst both black and white women, again in Cape Town and Durban. In general, the occurrence of high rates across these population groups in Cape Town and Durban suggests that the predisposing contexts and circumstances to strangulation may be present across, and perhaps irrespective, of population group. This study provides only initial indications of these circumstances, and the clarification of these would be the focus of further investigations into the situations that increase vulnerability to strangulation.
Most strangulations occurred from the early morning hours and across the typical working day in Johannesburg and Durban, and to a lesser extent in Cape Town. Pretoria occurrences clustered in the early morning hours until midday. These periods indicate the times of greatest vulnerability for strangulation attacks. These attacks may be directed especially at unemployed, home-based women and occur at a time of day when the majority of the population is at work or school, with consequently less immediate protection and safety. Cape Town, unlike the other cities, reported the highest rates over the weekend, consistent with high alcohol consumption over recreational periods. The seasonal variation in strangulation deaths suggests a pattern of occurrence generally spanning the period from end-winter to summer. Increasing day length has been associated with a rise in aggression in young males, and increasing evening warmth and daylight with increased consumption of alcohol in bars and streets, thereby resulting in a higher probability of inebriated encounters compared with other times of the year [
26]. Across cities, the predominant crime scene is linked to the domestic context, suggesting that perpetration is by an intimate partner or acquaintance, similar to the majority of previous findings reviewed.
The study is the first of its kind in South Africa and contributes to an emerging gendered homicide risk profile for a country with one of the highest homicide rates in the world. The data is based on a well-established mandatory mortality surveillance system, which provides reliable forensic data for the cities. However, complete information on both demographics and the circumstances of cases of death are not always possible due to forensic investigative limitations. There is for example an absence of information on the victim-perpetrator relationship, which would be the focus for further South African studies on the subject. The study provides a description of strangulation victims and the circumstances of occurrence, important for the generation of overall policy and prevention guidelines. This data also lends itself to future multivariate pattern analysis, which will provide further descriptions of the scenarios of occurrence and thus more detailed prevention targets.
Conclusion
In urban South Africa, fatal violence in females is clearly distinct from fatal violence in men. Accordingly, current prevention initiatives that address gender-based violence need to be strengthened, and supplementary evidence-based and gender-specific initiatives need to be developed to especially address the forms of violence that instigate fatalities. Dedicated policing teams; stronger legislation aimed at protecting women; universal screening for strangulation injury in women assessed to be victims of intimate partner violence; community-level substance abuse and domestic violence prevention programmes; and accessible and affordable support services and networks are some of the critical prevention initiatives that require development and support at both the local and national levels. The prevention of female homicidal strangulation is dependent on the generation of frequent, timely, comprehensive and accurate data on victims, perpetrators and the precipitating circumstances. In particular, homicidal strangulation in women needs to be further investigated in relation to intimate partner violence, as well as sex crimes. Importantly, research is required on the experiences of women who survive non-lethal strangulation to further identify risk factors. The prevention of female homicidal strangulation in South Africa will ultimately be strengthened by the creation of a social milieu that promotes equity, safety, health and human rights, and justice.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
This study was devised by SS and AVN. NA analysed the data. All authors made contributions to interpretation of the data. SS and AVN drafted the manuscript. All authors have read and approved the final manuscript.