Analysis of deaths due to gender-based violence: An autopsy-based cross-sectional study from Mumbai

ABSTRACT

Introduction:

Gender-based violence (GBV), including domestic violence, honor killings, and dowry deaths, represents one of the most extreme forms of violence against women. These deaths are often misclassified as accidents or suicides due to societal biases, leading to underreporting and obscuring the true impact of GBV on women’s mortality.

Aims and Objectives:

This study aimed to evaluate the prevalence of GBV-related deaths, with a focus on cases misreported as accidents or suicides. The objective was to determine the proportion of female deaths attributable to GBV and to highlight the challenges in reporting and investigating these cases.

Materials and Methods:

A retrospective, cross-sectional study was conducted at a medical teaching institution in Mumbai, analyzing 5 years (May 2017–April 2022) of autopsy data involving unnatural female deaths. Autopsy records, police reports, and victim/relative statements were examined to identify potential GBV cases and discrepancies in death classification.

Results:

The study found a significant proportion of female deaths attributed to GBV, primarily within domestic settings. Many cases initially reported as accidents or suicides showed discrepancies in injury patterns and circumstances. The majority of GBV victims were young women aged 15–44 years, with burns being the most common cause of death.

Conclusion:

The study emphasizes the need for improved investigation and classification of unnatural female deaths. Domestic violence emerged as the most common factor in these deaths, with intimate partners being primary perpetrators, with most incidents occurring within the home. It also stresses on enhanced protocols to accurately identify and address GBV-related fatalities.

Introduction

Gender-based violence (GBV), particularly against women, remains a pervasive global public health issue, contributing to significant morbidity and mortality. GBV encompasses a wide range of harmful acts, including physical, sexual, and psychological violence, perpetrated by intimate partners or family members. In India, despite legislative efforts to protect women, GBV-related deaths continue to be underreported or misclassified as accidents or suicides, masking the true extent of the problem.[1,2] Data from the National Family Health Survey-5 (published in December 2021) reveals that 29.3% of ever-married women in India have experienced spousal violence.[1] However, many deaths, particularly those involving burns, poisoning, or hanging, remain misclassified due to sociocultural stigma, inadequate investigation, or pressure from relatives.[3,4] As a result, the true burden of GBV-related deaths is difficult to ascertain, hindering effective policy implementation and interventions.

The United Nations defines violence against women as “any act of gender-based violence that results in, or is likely to result in, physical, sexual, or psychological harm or suffering to women, including threats of such acts, coercion, or arbitrary deprivation of liberty, whether occurring in public or private life.”[5] This definition emphasizes the multifaceted nature of GBV, which often remains hidden within domestic settings, manifesting as intentional homicides or suicides that are the result of prolonged psychological abuse.[1,6] In India, cultural practices such as dowry contribute to a significant number of unnatural deaths among women, often reported as accidental burns or suicides, despite evidence of foul play.[7,8] Dowry-related violence, for instance, accounts for a substantial portion of GBV-related deaths, particularly in rural areas, where sociocultural pressures to conform to traditional gender roles are stronger.[9] These deaths often go unrecognized as GBV due to a lack of investigative rigor, further complicating data collection and policy formulation.

This study aimed to address this critical gap by examining the prevalence and characteristics of GBV-related deaths among women. By analyzing 5 years of autopsy data from a medical teaching institution in Mumbai, we seek to determine the proportion of deaths attributable to GBV, focusing on cases misclassified as accidents or suicides. This study further explores the challenges in reporting and investigating such deaths, advocating for enhanced data collection and policy reform to mitigate the impact of GBV on women’s lives.[10,11]

Materials and Methods

Study design and setting

This retrospective cross-sectional study was conducted at a tertiary hospital in Mumbai. The study period spanned from May 2017 to April 2022.

Study population and sampling

The sampling frame included records of autopsies performed on women, girls, and non-binary individuals who died due to unnatural causes within the specified period. A convenience sampling technique was employed. Inclusion criteria were autopsies on individuals with a history suggestive of GBV, including deaths certified as unnatural, of undetermined intention, accidental exposure, or intoxication, and those with injuries indicative of GBV. Cases with discrepancies between victim statements, post-mortem examination findings, and contextual information were included to identify patterns suggestive of intentional harm, even if initially classified as accidental. Exclusion criteria included deaths from natural causes, chronic or mental illnesses, and other pathological conditions. Accidental deaths such as falls, train accidents, or road crashes were reviewed for indications of GBV and excluded if such indicators were absent.

Data collection

A data collection tool was developed to capture household and individual characteristics, details of the deceased, mechanism and manner of injury, and perpetrator information. Data were sourced from autopsy records, police statements, inquest reports, and chargesheets. Data extraction involved translating records from Marathi to English, and all relevant information was compiled and verified against available chargesheets or FIRs. Missing information was noted. Data were entered into an Excel spreadsheet for analysis. Descriptive statistical measures such as mean, median, and proportions were used for data analysis.

Ethical considerations

The study protocol was reviewed and approved by the institutional ethics committee. Due to the retrospective nature of the study and the use of anonymized data, informed consent from the relatives of the deceased was waived. Ethical guidelines adhered to include ICH-GCP, CDSCO-GCP, the Declaration of Helsinki, and ICMR 2006 guidelines. The study followed guidelines for structured abstracts, data management practices, and adherence to ethical standards to enhance data collection and reporting accuracy.

Results

During the study period, a total of 6190 autopsies were conducted, of which 1467 were female cases. Of these female autopsy cases, 57.3% (840) were determined to have died from unnatural causes. GBV was identified in 12.3% (181 cases) of these unnatural deaths. Of the 181 cases with suspected GBV, 23% (42 cases) had a documented history of violence based on statements from victims or their relatives, while the remaining 77% (139 cases) exhibited signs of potential GBV, such as discrepancies in injury sites, patterns, circumstances, autopsy findings, and statements from relatives, despite lacking a clear history of violence. No cases involving non-binary individuals or persons with disabilities were identified.

Figure 1 illustrates the year-wise distribution of deaths attributed to GBV. Of the 181 GBV cases, 48% (86 cases) were suicides, 47% (85 cases) were accidents, and 6% (10 cases) were homicides. In terms of residence, 77% (139 cases) of the victims lived in urban areas, while 23% (42 cases) resided in rural areas. Deaths occurred in home or private spaces (99%, 179 cases), with only 1% (2 cases) occurring outside the home.[2,9,11] Among urban residents, 50% (69 cases) died by suicide, whereas in rural areas, 57% (24 cases) died due to accidents [Table 1]. Figure 2 shows the age distribution. Out of the 181 cases, 75% (136 cases) were aged 15–44 years, with the highest proportion (45%) being between 15 and 29 years. The mean age of the victims was 34.8 years, with a 95% confidence interval ranging from 32.30 to 37.32. The median age was 30.5 years (age range: 11 months–87 years). When considering with manner of death, the median age of victims who died by suicide was 28 years, accidental deaths was 32 years, and homicides was 24.5 years.[9,10,11,12] Regarding marital and pregnancy status, 67% (121 cases) of victims were married, 29% (53) were unmarried, and 4% (7 cases) were either widowed or divorced.[4,9,13,14] Among the married victims, 4% (5 cases) were pregnant at the time of death.[15] Figure 3 illustrates the nature of injuries in GBV cases. Burns were the leading cause of death, accounting for 58% (105 cases) of the deaths, followed by hanging in 20% (36 cases) and poisoning in 16% (29 cases). In addition, 3% (5 cases) were reported as deaths due to fall from a height, and 3% (6) resulted from aggravated assaults. Figure 4 shows that fire or flames were the most prevalent method of killing in accidental cases (75% (64 cases)). For suicides, drugs and chemical substances were used in 51% (44 cases), while bodily force (suicidal jump and hanging) accounted for 48% (41 cases).[15,16] Half of all homicide deaths 50% (5 cases) involved drugs and chemicals. The underlying reasons for acts of violence in 114 cases were studied. Of these, 58% (66 cases) were related to marital problems, 29% (33 cases) to family issues, and 13% (15 cases) to intimate relationships or affairs. Marital disputes and family issues were identified as primary factors accounting for 87% (99 cases) of all GBV-related deaths, while unsuccessful intimate relationships contributed to 13% (15 cases) of the deaths. In terms of perpetrators, husbands or intimate partners were responsible in 61% (69 cases) of incidents. Other family members were identified in 35% (40 cases), while both the husband and a family member were involved in 4% (5 cases) of cases.[4,17] Figure 5 shows the post-incident survival duration of victims. Of the 181 cases, 21% (46 cases) were brought dead to the hospital. Victims of hanging and jumping typically survived less than a day. Assault victims survived for an average of 3 days (range: 0–16 days), poison victims survived 5 days (range: 0–43 days), and burn victims survived 6 days (range: 0–102 days). In 37% (67 cases), the reason for death remains unclear due to fragmented data.[16] There were discrepancies in post-mortem examination findings (e.g. injury site or pattern) when compared to history provided by the relatives or the information accessed from the police reports. Table 2 summarizes the sources of variables and missing data in the study. Significant gaps in sociodemographic data related to both victims and perpetrators were noted in the autopsy and police reports, highlighting the data gaps.

Figure 1.

Year-wise distribution of cases

Table 1.

Manner of death in relation to the area of residence

Manner of death	Urban Area	Rural Area
Suicidal	69 (50%)	17 (41%)
Accidental	61 (44%)	24 (57%)
Homicidal	9 (6%)	1 (2%)
Grand Total	139 (100%)	42 (100%)

Figure 2.

Distribution of cases by age

Figure 3.

Distribution of cases by nature of injury

Figure 4.

Distribution of cases by method of killing

Figure 5.

Distribution of cases by post-incident survival duration

Table 2.

Summary of the source of variables and missing data in the study

Variables	Variable name	Source record	No. of cases in which the variable was available	No. of cases in which variable was not available
Related to Victim	Education status	Autopsy reports	14 (8%)	167 (92%)
	Employment status	Autopsy reports	43 (24%)	138 (76%)
	Family income of household	Autopsy reports	0	181 (100%)
	No. of children	Autopsy reports	137 (76%)	44 (24%)
	Record of physical, sexual, or psychological violence/harassment	Autopsy reports	14 (8%)	167 (92%)
Related to perpetrator	Age of perpetrator	Relative statement	17 (9%)	164 (91%)
	Education status	Relative statement	0	181 (100%)
	Employment status	Relative statement	0	181 (100%)
	Intoxication with controlled drugs or other psychoactive substances	Relative statement	0	181 (100%)
	Income	Relative statement	0	181 (100%)

Discussion

The study examines the prevalence and characteristics of deaths resulting from GBV in females of all ages, including non-binary individuals, based on autopsy reports. This study underscores the alarming prevalence of 12.3% of GBV as a significant contributor to female mortality. Our findings highlight that many deaths, categorized as accidental or suicidal, are rooted in GBV, pointing to the hidden nature of this violence.

Our analysis also revealed that most victims were young, between the ages of 15–29 years, highlighting the heightened vulnerability of this demographic to intimate partner violence and marital conflict. These deaths were often the result of prolonged psychological or physical abuse, further contributing to the underreporting of violence within domestic settings.[13] This pattern is consistent with other research that identifies younger, married women as disproportionately affected by GBV, particularly in cases involving dowry disputes.[16] In our study, the deaths occurred in home or private spaces (99%, 179 cases), with only 1% (2 cases) occurring outside the home, aligning with other research findings.[2,11,12] When considering the manner of death, the median age of victims who died by suicide was 28 years, accidental deaths was 32 years, and homicides was 24.5 years. This aligns with other studies that have shown that most victims of unnatural deaths are typically 21–40 years old.[9,11,12] Regarding marital and pregnancy status, 67% (121 cases) of victims were married, 29% (53) were unmarried, and 4% (7 cases) were either widowed or divorced, which is consistent with findings from another research.[9,12,13,14]

Burns were the leading cause of death, accounting for 58% (105 cases) of the deaths, consistent with other research findings.[15] The high prevalence of burns, accounting for 58% of deaths, is particularly concerning as burns are frequently reported as accidental, particularly in dowry-related incidents.[15] In the context of India, dowry-related deaths, often referred to as bride-burning, are a significant issue across all social classes. When dowry demands are not met, it can lead to severe marital conflict, and in some instances, women may choose suicide as a means of escaping such brutality. Unfortunately, burn injuries are frequently reported as accidents and are not thoroughly investigated, which obscures the underlying violence and the true nature of these deaths. This underscores the need for improved investigative protocols to differentiate between accidental and intentional burns. The prevalence of hanging (20%) and poisoning (16%) as methods of death also mirrors findings from other studies on unnatural deaths, where these methods are commonly misreported.[18] Half of all homicide deaths (50% (5 cases)) involved drugs and chemicals, consistent with other studies showing the prevalence of accidents and suicides.[15,16] While this finding aligns with previous research, especially in the context of dowry-related violence, our study also brings attention to the systemic underreporting and misclassification of such deaths, which obscures the true scope of GBV.

In terms of perpetrators, husbands or intimate partners were responsible in 61% (69 cases) of incidents. Other family members were identified in 35% (40 cases), while both the husband and a family member were involved in 4% (5 cases) of cases, consistent with findings from other studies.[12,15] A key finding of this study is the critical role played by intimate partners, who were responsible for 61% of the GBV-related deaths. The majority of incidents occurred in private homes, where violence remains hidden from public scrutiny. This reflects the broader issue of domestic violence, where victims often suffer in isolation, exacerbated by cultural norms that discourage reporting.[11] The significant role of other family members (35%) as perpetrators further illustrates the embedded nature of violence within domestic and familial structures. This is consistent with other studies, which also emphasizes that marriage, marital discord, and domestic environments often exacerbate the risk of GBV-related deaths.

One of the novel contributions of this study is its identification of the underlying violence in deaths often classified as accidents or suicides. Most unnatural deaths in women are often classified as accidents or suicides due to families refraining from reporting foul play or fear of social stigma. The fact that many victims survive for days after sustaining burn injuries, compared to more immediate fatalities from hanging, poisoning, fall from height, or assault, highlights a potential window of opportunity for intervention into the circumstances of death as many cases of GBV remain undetected due to superficial categorization. There were discrepancies in post-mortem examination findings (e.g., injury site or pattern) when compared to history provided by the relatives or the information accessed from the police reports, as highlighted by the UNODC report on gender-related killing of women and girls, 2021 where 4 out of 10 cases lack such information.[5]

The absence of contextual information in a substantial portion of cases is a limitation as it prevents a more nuanced understanding of the social and interpersonal dynamics leading to these deaths. This lack of data is not unique to this study; it reflects a broader, global issue in GBV research, where the social stigma associated with reporting violence and the fear of legal or societal repercussions result in underreporting. Compared to other research, this study strengthens the argument for improved investigation protocols, particularly in cases of female deaths labeled as accidental or suicidal.

A key limitation of our study is the reliance on forensic autopsy data, which may not fully capture the scope of GBV-related deaths, particularly in cases where deaths are misclassified due to reporting biases. Furthermore, the absence of detailed contextual information in a significant number of cases limits the ability to fully explore the underlying causes of violence, which remains a challenge in many GBV studies worldwide.[5] In addition, cultural, and societal biases play a role in both the reporting and investigation of these deaths, potentially skewing the data.

Our findings have significant practical implications. There is an urgent need for forensic and law enforcement agencies to develop standardized protocols for investigating deaths suspected to involve GBV. Training in recognizing signs of underlying violence, particularly in cases initially reported as accidental or suicidal, is critical. Moreover, improving the quality and consistency of death reporting through enhanced data collection practices is vital. Institutional reforms aimed at strengthening ethical data collection and capacity building among stakeholders could lead to more reliable estimates of GBV-related mortality, thereby influencing policy and prevention strategies.

In conclusion, the study highlights the pervasive nature of GBV, particularly in intimate partner relationships, and stresses the need for systemic changes in how these cases are identified, reported, and addressed. It further emphasizes the critical need for improved investigative protocols and comprehensive data collection to reveal the true extent of GBV-related deaths. Enhancing the quality of death reporting is essential for generating reliable estimates of the true impact of GBV. An integrated approach, involving both legal and public health frameworks, is essential to reducing GBV-related mortality. While our findings are specific to a particular region, they reflect a broader global trend of underreporting and misclassification of GBV-related deaths. Future research should focus on cross-national comparisons and longitudinal studies that investigate the long-term health impacts of GBV.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

The study’s technical and financial support provided by the Data Impact Program, Vital Strategies, New York 10005, USA under the Data for Health Initiative of Bloomberg Philanthropies