Demographic, Sociocultural, and Time-Related Patterns of Violent Suicide Attempts

Introduction

Suicide involves directing aggressive and destructive feelings inward, intentionally causing self-harm to end one’s life [1]. Violent suicide attempts (VSAs), often involving highly lethal and externally observable methods, are a disproportionately deadly form of self-harm and pose an escalating challenge to public health, particularly in regions lacking surveillance infrastructure [2].

Suicide remains a leading cause of global mortality, and males exhibit higher rates of both VSAs and suicide-related deaths than females [3,4]. Compared to non-violent methods, VSAs are associated with significantly greater morbidity and mortality [5]. While suicide rates are highest in low- and middle-income countries, a lack of robust national data often hinders epidemiological insight [1]. For instance, in Turkey, the 2018 suicide rate was approximately 4 per 100 000 population, with 3161 deaths, 75.6% of which involved males, but the exact prevalence of VSAs remains unclear [6].

The method of suicide is influenced by multiple factors, including accessibility of means, cultural norms, and socioeconomic context. Determinants such as income level, education, marital status, psychiatric history, and access to mental health services play critical roles and vary considerably across sociocultural and geographic environments [7]. Violent suicide (VS) encompasses methods such as hanging, jumping, firearm use, self-inflicted sharp-object injuries, vehicular accidents, self-immolation, electrocution, drowning, or stepping in front of a train [8]. Individuals using violent methods tend to demonstrate greater suicidal intent compared to those attempting poisoning or overdose [9,10]. Common risk factors include social isolation, unemployment, alcohol and substance abuse, psychiatric disorders, a family history of suicide, and chronic or terminal medical conditions [11].

Prior studies from Turkey and Turkish-origin populations have documented seasonal fluctuations in suicide behavior [12,13], but most focused on general or adolescent populations and did not differentiate violent methods. Research on the timing and environmental correlates of VSAs remains limited. Factors such as ambient temperature, sunlight, and time of day may interact with neurobiological and psychosocial processes influencing suicidal behavior [14–16].

This retrospective observational study aimed to address critical knowledge gaps by systematically examining the demographic, clinical, time-related, and environmental correlates of VSAs in a southern region of Turkey. Specifically, the study investigated factors such as day of the week, time of day, seasonality, environmental variables such as temperature, precipitation, and wind, and how they intersect with gender, suicide method, psychiatric history, and mortality. By providing granular insight into the time- and location-related patterns of VSAs, this research seeks to inform targeted, culturally grounded, and time-sensitive public health interventions and guide future suicide prevention policy in similar regional contexts.

Material and Methods

ETHICAL APPROVAL AND CONSENT TO PARTICIPATE:

This study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki and received approval from the Non-Interventional Ethics Committee of Nigde Omer Halisdemir University Faculty of Medicine (Approval No: 2023/119; Date: 28.12.2023). According to the university’s ethical guidelines, informed consent is not required for retrospective studies. Permission for the use of patient data was granted by the ethics committee. All patient data were anonymized prior to analysis and handled in compliance with institutional data protection policies and GDPR (General Data Protection Regulation)-compatible standards to ensure confidentiality throughout the study.

STUDY DESIGN AND SETTING:

This retrospective analysis was based on data extracted from the hospital information management system (KARMED^®). The study included patients who presented to the Emergency Department of Nigde Omer Halisdemir University Training and Research Hospital between 01/01/2023 and 31/12/2023 with ICD codes X40–49, X60–84, and Z91.5–Z91.6, covering suicide-related and self-harm diagnoses. Groupings were based on ICD-10 codes and confirmed by clinical chart review. In cases involving forensic autopsies, medical examiner documentation confirmed the suicide method and cause.

Additionally, individuals brought to the Forensic Medicine Department for autopsy – those found deceased at the scene of the incident and determined to have commited suicide via violent means – were included. Exclusion criteria encompassed all suicide attempts that did not involve violent methods (such as drug or substance ingestion), cases with inaccessible or incomplete data, and individuals under 18 years of age. Suicide attempts by ingestion were excluded due to their substantially different lethality profiles, and individuals under 18 were excluded due to pediatric ethical restrictions and legal consent boundaries.

The KARMED^® system has been used for electronic medical documentation since 2018 and undergoes quarterly internal audits for data validation by the hospital’s information systems unit. Data extracted from the system were independently verified and cross-checked by 2 researchers to ensure reliability. Outliers and inconsistent entries were reviewed and resolved using original electronic patient records.

DATA COLLECTION:

Collected variables included: age, gender, education level (primary, secondary, high school, undergraduate/graduate), marital status (married, single, widowed, and divorced), suicide method (self-harm with sharp object, firearm injury, jumping from a height ≥5 meters, hanging), presence of alcohol at the time of admission, prior psychiatric history, mortality status after the attempt, time of the attempt (morning: 06: 01–12: 00; noon/afternoon: 12: 01–18: 00; evening: 18: 01–00: 00; night: 00: 01–06: 00), day of the week, and season.

Meteorological data – air temperature, humidity, precipitation, and wind speed on the date of each attempt – were obtained from the Nigde Omer Halisdemir University Meteorology Station. Meteorological data were precisely time-matched with suicide attempt dates. In cases of incomplete hourly records, daily average values were used. Although station-level weather data were used, individual-level exposure could not be measured, and this is acknowledged as a limitation.

Detailed subgroup comparisons were performed to explore associations between demographic characteristics (such as age, gender, marital status, psychiatric history), suicide methods, outcomes such as mortality, time of day, day of the week, and season.

STATISTICAL ANALYSIS:

Statistical analyses were performed using IBM SPSS Statistics for Windows, Version 27.0 (IBM Corp., Armonk, NY, USA). Categorical variables were reported as frequencies and percentages. Age, precipitation, and wind speed, which were not normally distributed, were presented as medians with interquartile ranges (IQR 25–75), while temperature and humidity were reported as means±standard deviation (SD). Normality was assessed using the Shapiro-Wilk test. Parametric tests (Student’s t-test) were used for normally distributed data, and non-parametric alternatives (Mann-Whitney U, Kruskal-Wallis) were used otherwise. Pearson’s chi-square (χ2) and Fisher’s exact tests were used to compare categorical variables, with a p value <0.05 considered statistically significant. These tests were selected based on data type and distributional properties. Subgroup analyses of the variables shown in Tables 1 and 2 were performed using Kruskal-Wallis and chi-square tests. Due to the observational design and sample size limitations, formal multivariable adjustment for confounding factors was not applied. Cases with missing data were excluded from statistical analyses, and no data imputation was performed.

Results

DEMOGRAPHIC CHARACTERISTICS:

The median age was 30 years (IQR: 21–34). Gender distribution revealed 76.7% male (n=66) and 23.3% female (n=20) (male-to-female ratio >3: 1).

Age significantly varied by marital status: widowed/divorced and married individuals were older than single participants (p=0.004).

To explore whether educational attainment varied by gender – potentially influencing the choice of suicide method or risk profile – we examined the association between gender and education level. However, no significant relationship was found between these 2 variables (p>0.05).

MARITAL STATUS AND PSYCHIATRIC HISTORY:

Among female cases, none were married; 80% were single and 20% were widowed or divorced (p=0.032).

When examining the overall sample, a significant association was found between marital status and psychiatric disorder history. Married individuals exhibited a lower prevalence of psychiatric disorders compared to single and widowed/divorced individuals (p=0.003).

SUICIDE METHODS BY GENDER AND MORTALITY:

The most common suicide method was self-harm with sharp objects (37.2%), followed by firearms (27.9%), hanging (18.6%), and jumping from height (16.3%).

Among females, 60% used sharp objects and 40% jumped from heights. None used firearms or hanging (p=0.003). Hanging was significantly associated with higher mortality, while no deaths occurred in sharp-object attempts (p<0.001). Overall, 27.9% of cases were fatal (n=24), of which 91.7% (n=22) involved male patients.

TIME-RELATED AND SEASONAL PATTERNS OF ATTEMPTS:

Most suicide attempts occurred during noon/afternoon (37.2%), followed by evening (27.9%), night (25.6%), and morning (9.3%). All morning attempts occurred on weekdays (p=0.049). All female and 60% of male suicide attempts were also recorded on weekdays (p=0.02).

In spring, most attempts occurred at noon, whereas in summer, attempts were predominantly at night. No suicide attempts were made during nighttime hours in autumn months (p=0.015).

MORTALITY BY TIME OF DAY AND SEASON:

Deaths were most common during the noon and evening periods (each 41.7%), while no deaths occurred during nighttime attempts (p=0.048). Seasonally, mortality was highest in spring and autumn; no deaths were observed in winter (p=0.032).

PSYCHIATRIC DISORDERS AND SEASONAL ASSOCIATION:

Participants with psychiatric disorders were more likely to attempt suicide in autumn, while those without such history predominantly did so in spring (p=0.03).

ENVIRONMENTAL CORRELATES:

Suicide attempts were most frequently observed in spring (30.2%). This season also exhibited significantly higher precipitation than autumn (p=0.02) and summer (p=0.012).

Additionally, average wind speed was significantly higher in spring and summer compared to autumn (p=0.001 and p=0.004, respectively) and winter (p=0.009 and p=0.029, respectively). These meteorological trends occurred concurrently with higher VSA frequency but do not establish causality.

Discussion

LIMITATIONS:

This study has several limitations. It was conducted at a single center in southern Turkey, which may limit the generalizability of results to other regions or cultural contexts. The retrospective design restricts causal inference and limited access to certain variables, such as the severity or chronicity of psychiatric conditions, history of substance misuse, and family psychiatric history. We also lacked systematic data on post-attempt outcomes (such as hospitalization duration or neurological sequelae), which may have affected assessment of attempt severity. Environmental data were based on validated meteorological sources, but individual exposure (such as actual ambient temperature during the attempt) could not be determined. Lastly, the absence of structured psychiatric interviews and follow-up limited insights into long-term mental health trajectories. Future prospective, multicenter studies are needed to validate these findings and explore additional clinical and psychosocial dimensions.

Conclusions

VSAs are a serious public health concern, characterized by high lethality and multifactorial risk dynamics. Our study identified distinct time-related patterns, with VSAs most frequently occurring during spring, on weekdays, and in the early afternoon hours. These patterns, grounded in statistical analysis, suggest opportunities for time-targeted interventions and surveillance efforts that align with high-risk periods. For instance, mental health outreach could be intensified during weekday afternoons in spring months, when VSAs peak.

Men were disproportionately represented among VSA cases, consistent with global trends. Within the female subgroup, all participants were unmarried, and single or divorced status emerged as salient risk factors. Across the full sample, additional contributors to risk included lower educational attainment, psychiatric diagnoses, and acute alcohol use at the time of the attempt. These findings highlight the importance of stratifying suicide risk based on sociodemographic and behavioral profiles.

Recognizing such seasonal and time-specific variations, alongside key vulnerabilities, may help guide future research and support context-sensitive suicide prevention strategies. While our findings derive from a single-center retrospective study and require validation in broader populations, they offer a valuable foundation for future research and potential contributions to public health policy, particularly in regions with similar sociocultural and environmental characteristics.