Suicidal thoughts and behaviours in Cape Town: a cross-sectional study of prevalence, social, contextual, and clinical correlates

This cross-sectional study of 613 adults in Cape Town reveals a high prevalence of suicidal thoughts and behaviors strongly linked to mental disorders, community and family violence, and perceived stress, while highlighting that these psychosocial and structural stressors remain significant risk factors even after accounting for psychiatric diagnoses.

The Gist

Imagine a community as a large, bustling garden. In this specific garden (located in Cape Town, South Africa), the soil is very hard, the weather is often stormy, and there isn't enough water for everyone. This paper is like a report card on the health of the plants in this garden, specifically looking at why some plants are thinking about giving up and dying (suicidal thoughts) versus those that are just struggling to grow.

Here is the story of what the researchers found, broken down into simple parts:

1. The Big Picture: A Stormy Garden

The researchers went into two specific neighborhoods in Cape Town. These aren't just any neighborhoods; they are places where people have faced a lot of hardship for a long time—like being forced to move away from their families, high unemployment, and a lot of violence.

They asked 613 adults a very serious question: "Have you ever thought about ending your life?" or "Have you ever tried?"

The Shocking Stat:

• 14% of the people they talked to had thought about suicide in the last month.
• 22% had tried to end their lives at some point in their past.

To put this in perspective, these numbers are much higher than the average for the whole country. It's like finding that in a specific, stormy corner of the garden, nearly 1 in 7 plants are considering giving up, compared to the national average where it's much rarer.

2. The Usual Suspects vs. The Real Culprits

For a long time, doctors and scientists thought that if a plant was dying, it was mostly because of a "sickness inside the plant" (like a mental illness such as depression). They thought if you fixed the sickness, the plant would be fine.

But this study found something different. While mental illnesses (like depression or anxiety) were definitely a big problem, they weren't the only reason the plants were struggling.

The researchers found three "Big Bad Weather Patterns" that were just as dangerous, if not more so:

• The "Heavy Backpack" (Perceived Stress): Imagine carrying a backpack filled with rocks. For some people, this backpack is so heavy they can barely walk. The study found that people with "high stress" (a heavy backpack) were 33 percentage points more likely to think about suicide than those with a light backpack. This was the strongest factor of all.
• The "Broken Fence" (Violence): Imagine living in a garden where the fence is broken, and strangers or even neighbors are throwing rocks at you. People who said, "Yes, I've been hit by violence in my family or my community," were much more likely to have suicidal thoughts. Even after the researchers accounted for mental illness, the violence still mattered.
• The "Gender Gap": The female plants in this garden were more likely to be thinking about giving up than the male plants.

3. The "HIV" Misconception

There was a specific question the researchers wanted to answer: Does having HIV (a virus that weakens the immune system) make you more likely to want to die?

In many parts of the world, people with HIV face huge stigma and fear. However, in this study, having HIV did not make a difference.

• People with HIV and people without HIV had the exact same rates of suicidal thoughts.
• The Analogy: It's like two groups of hikers. One group has a heavy backpack (HIV), and the other doesn't. The study found that both groups were equally likely to want to quit the hike because the trail itself (the poverty, violence, and stress) was too hard, not because of the backpack.

4. The "Perfect Storm" Effect

The researchers used a cool visual tool (like a Venn diagram) to show how these problems overlap.

• Most people who were thinking about suicide had at least two of these problems: a mental illness, a history of violence, or high stress.
• It's rarely just one thing. It's usually the combination. Imagine a plant that is already sick (mental illness), getting hit by rocks (violence), and carrying a heavy backpack (stress). That is a recipe for disaster.

5. What Does This Mean for the Future?

The authors conclude that we can't just treat the "sickness inside the plant" (mental health services) and expect the garden to heal. We have to fix the garden itself.

• Old Way: "Let's give everyone medicine for depression."
• New Way (Based on this paper): "Let's fix the broken fence (stop the violence), help people take off the heavy backpack (reduce stress and poverty), and then offer mental health support."

The Takeaway

This study tells us that in tough, low-income communities, suicide isn't just a medical problem; it's a social problem. You can't fix a broken plant if the soil is toxic and the weather is violent. To save these lives, we need to build better fences, lighten the backpacks, and create a safer environment for everyone, regardless of their HIV status or their mental health diagnosis.

Technical Summary

1. Problem Statement

Suicide is a critical public health challenge, particularly in Low- and Middle-Income Countries (LMICs), which account for nearly 80% of global suicide deaths. While suicidal thoughts and behaviors (STBs) have historically been framed primarily as a consequence of mental disorders, emerging evidence suggests that structural and social determinants (e.g., poverty, violence, unemployment) play a significant, often independent, role.

In South Africa, suicide rates are among the highest globally. However, there is a scarcity of systematic data quantifying the relative contributions of mental disorders versus contextual stressors (such as violence and perceived stress) and HIV status on STBs within high-adversity peri-urban communities. This study aims to fill that gap by examining the prevalence and correlates of STBs in Cape Town, specifically investigating whether HIV status, socioeconomic factors, violence exposure, and perceived stress predict STBs independently of diagnosed mental disorders.

2. Methodology

Study Design and Setting

• Design: Cross-sectional study conducted between March 2023 and September 2024.
• Setting: Three public-sector facilities in the Cape Town metropolitan area (Cape Flats): two primary care clinics in peri-urban communities and one tertiary referral hospital.
• Population: 613 adults (≥18 years), recruited in a 2:1 ratio of People Living with HIV (PLWH) to HIV-negative individuals. Recruitment aimed for equal sex distribution and stratification by age (18–29, 30–44, ≥45).

Data Collection and Instruments

• Outcomes: STBs were assessed using the MINI International Neuropsychiatric Interview (MINI) version 6.0 (based on DSM-IV criteria).
  • Primary Outcome: Past 30-day suicidal ideation (SI).
  • Secondary Outcomes: Past 30-day suicide plan, past 30-day suicide attempt, and lifetime suicide attempt.
  • Hierarchical Outcome: A 4-level ordinal variable (None, Ideation, Plan, Attempt) for secondary analysis.
• Exposures:
  • Clinical: Current mental disorders (Major Depressive Disorder, PTSD, Anxiety, Alcohol/Substance Use Disorders, etc.) via MINI.
  • Psychosocial: Perceived stress measured by the Perceived Stress Scale-4 (PSS-4), categorized as low, moderate, or high.
  • Contextual: Lifetime exposure to violence in the community and family (study-specific questionnaire).
  • Sociodemographic: Age, sex, education, employment, marital status, and HIV status (verified via clinic records or rapid testing).

Statistical Analysis

• Primary Analysis: Logistic regression to estimate associations with past 30-day SI. Seven multivariable models were fitted to isolate effects:
  • Models adjusted for demographics, HIV, violence, stress, and specific mental disorders.
  • Metric: Average Marginal Effects (AMEs) were reported as percentage-point differences in predicted probability to facilitate interpretation of absolute risk.
• Secondary Analysis: Ordinal logistic regression for the hierarchical STB outcome.
• Software: Stata 18 and R 4.4.2.

3. Key Contributions

• Contextual Specificity: Provides high-resolution data from peri-urban South African communities characterized by structural inequities, gang violence, and high HIV prevalence, a setting underrepresented in global suicide literature.
• Disentangling Drivers: Explicitly tests whether social determinants (violence, stress) predict STBs independently of mental disorders, challenging the view that STBs are solely a psychiatric issue.
• HIV Clarification: Offers robust evidence regarding the association between HIV status and STBs in a generalized epidemic setting, contrasting with findings from high-income settings where HIV often correlates with higher suicide risk.
• Methodological Rigor: Utilizes structured diagnostic interviews (MINI) rather than self-report screening tools for mental disorders, reducing misclassification bias.

4. Key Results

Prevalence

• Past 30-day Suicidal Ideation: 14.0% (95% CI: 11.5–17.0).
• Lifetime Suicide Attempt: 22.2% (95% CI: 19.1–25.7).
• Demographics: Prevalence was significantly higher in females (16.8% SI) compared to males (9.0% SI).
• HIV Status: No significant difference in prevalence between PLWH (14.0%) and HIV-negative participants (14.3%).

Associations with Suicidal Ideation (30-day)

• Mental Disorders: Strongest clinical correlates.
  • Major Depressive Disorder (MDD): AME +27.13 percentage points.
  • Post-Traumatic Stress Disorder (PTSD): AME +27.98 percentage points.
  • Alcohol Use Disorder: AME +8.34 percentage points.
  • Dose-Response: Risk increased with the number of comorbid disorders (≥3 disorders: AME +49.43).
• Psychosocial & Contextual Factors:
  • High Perceived Stress: The strongest non-clinical correlate. Participants with high stress had an AME of +33.58 percentage points compared to low stress.
  • Violence Exposure:
    • Family violence: AME +14.09 percentage points.
    • Community violence: AME +7.52 percentage points.
  • Female Sex: AME +8.52 percentage points.
• Non-Associations: No significant evidence of association with HIV status, age, employment, education, or marital status after adjustment.

Persistence of Effects
Crucially, the associations between high perceived stress and community violence with suicidal ideation persisted even after adjusting for all current mental disorders (Model 5). This indicates these factors contribute to suicide risk independently of diagnosed psychiatric conditions.

5. Significance and Implications

Theoretical Implications
The study supports the "Integrated Motivational-Volitional" model and broader structural theories of suicide. It demonstrates that in LMIC settings, structural stressors (violence, poverty-induced stress) are not merely precursors to mental illness but are independent drivers of suicidal ideation. The finding that high stress and violence predict STBs even in the absence of a mental disorder diagnosis suggests that current mental health interventions may be insufficient if they do not address the underlying social environment.

Public Health and Policy

• Beyond Clinical Care: Suicide prevention in peri-urban South Africa requires multilevel interventions. While scaling up mental health services is vital, interventions must also target violence reduction, community safety, and stress management at the structural level.
• HIV Context: The lack of association between HIV status and STBs in this cohort (likely on ART) suggests that for PLWH in generalized epidemics, the risk of suicide may be driven more by the social context of the epidemic (stigma, poverty) than the biological status itself.
• Targeting: Interventions should prioritize females and individuals reporting high perceived stress or exposure to violence, regardless of whether they meet criteria for a specific mental disorder.

Limitations

• Cross-Sectional Design: Cannot establish causality (e.g., reverse causation between stress and ideation).
• Selection Bias: Participants were recruited from health facilities, potentially underestimating prevalence in the general community or missing those who died by suicide (Neyman bias).
• Measurement: Violence exposure was measured via a non-validated study-specific tool.

Conclusion
This study highlights that in high-adversity peri-urban communities, suicidal thoughts and behaviors are driven by a complex interplay of mental disorders, chronic perceived stress, and exposure to violence. Effective prevention strategies must move beyond a purely clinical framework to address the structural and psychosocial determinants of health.