—” title: “Analysis of Suicide Rates in West Africa author:”BA Bocar Alassane : 20230350 , IDANI Mechack Triomphant : 20210293 , Ki Elvire Ravelle : 20240923 , NIKIEMA Rakiswendé Faridatou 20210169 ,”
date: “2025-04-08” output: html_document —
#General Research Objective:
This study explores the causes of suicide in West Africa through a statistical analysis of socio-economic, health, and psychological data from 16 countries. The objective is to better understand the influencing factors using Principal Component Analysis (PCA), linear regression, and hierarchical clustering.
Introduction
Introduction
Suicide is a complex, timeless phenomenon that affects all societies, defined as an intentional act of deliberately ending one’s life. This phenomenon has been growing in recent years, representing a major public health challenge that also affects West Africa, although it is often minimized due to cultural taboos in this region. The factors leading to suicide are multidimensional, including psychological, social, economic, biological, and environmental aspects. In this region, socio-economic difficulties are numerous: unemployment, poverty, and social inequalities lead to despair and hopelessness among many people. Additionally, psychological disorders, stigma, and insufficient mental health care services further complicate the situation. Understanding these determinants is essential to implement suicide prevention strategies and contribute to a more peaceful living environment. It is in this context that an analysis of the causes of suicide can be conducted on the population of West Africa. State of the Art Many people die by suicide each year, and the causes of this phenomenon are a combination of interconnected factors that can be distal or immediate. Faced with this reality, several studies on the causes of suicide have been conducted; however, those exclusively focused on West Africa remain very limited. Among these factors, immediate health-related causes are linked to mental and psychological disorders, which are widespread. For example, Zhuoyang Li et al. show that there is “an increased risk of suicide associated with mental disorders, particularly affective disorders, schizophrenia, anxiety disorders, and personality disorders” (1). Psychiatric illnesses increase this risk. Most West African societies have inherited a social organization based on tradition and religion. This corresponds to a form of social satisfaction, but social exclusion, stigma, and the expectations of adhering to social and religious rules can expose some individuals to significant mental vulnerability, leading to major depression. Indeed, according to Wanyoike and Becky Wanjiku, depression affects a person’s mental, emotional, and physical capacities (2), which can lead to irreversible actions. Social pressure and social integration influence suicidal behaviors, as Johnson and Barclay D. (3) note, particularly the stigma that exposes individuals to an increased risk of suicide (4). The social impact is also very significant, as “Manning argues that suicide stems from increased inequality and decreased intimacy” (5). The authors then discuss theoretical models explaining the importance of social support for maintaining mental health, highlighting the close links between mental health and social support identified in population and clinical surveys. Economic recessions also affect population health, with poor mental health often linked to financial difficulties and poverty. Francis Tchégnonsi Tognon et al., in a descriptive study in Cobly (Northern Benin) on suicide in Sub-Saharan Africa, note that poverty ranks first, accounting for 32.7% of 52 suicide cases in this area (6). Low income, poverty, and unemployment affect living conditions, making them very unfavorable, especially in developing countries with low incomes (7). BAHI Jean Joel adds that being denied a job or being fired can create permanent anxiety in individuals, leading to psychological and mental disorders, which are precursors to suicidal acts (8). Mahudjro AHOYO’s studies on the socio-economic determinants of suicide among individuals highlight that some individuals resort to suicide due to an inability to repay loans after being granted multiple credits, which plunge them into misery (9). Social expenditures, which are obvious priorities, become constraints for leaders, creating stress and anxiety. According to Adam Skinner, “transitions from employment to unemployment tend to produce a significant increase in psychological distress” (10). The rise in unemployment is accompanied by an increase in mental disorders, implying a higher probability of suicidal behavior. Jahoda argues that “distress among the unemployed is the consequence of the absence of five latent functions of employment” (11). The observed increase in suicide may thus precede pronounced economic disruption. The lack of vision, combined with state poverty, would also expose populations to suicidal behaviors. Dr. Yahaya Abou, Coordinator of the National Mental Health Program in Niger, has highlighted the impact of insufficient mental health policies in Niger on suicide, focusing on the issue of access to quality and affordable mental health care (12). In terms of mental health, Niger, like most West African countries, lags significantly behind WHO standards (2015 data) (12). N’Guessan Elkhanan N’GORAN and Yann ZOLDAN studied the cultural representation of depression and suicide among West African Francophones in Quebec and Côte d’Ivoire, emphasizing that the lack of resources, shortages of qualified caregivers, diagnostic errors, social stigma, and taboos related to mental disorders are obstacles to accessing and administering effective care (13). These subjective considerations could be indicators of individual behavior, including their choice to seek or not seek help from specialists or even to commit acts such as suicide (13). Substance use can provide temporary relief from emotional suffering, but it has severe repercussions on health and social integration. Additionally, tobacco and drug use cause behavioral disorders, as noted by Farber, Nuri B. (14), making social relationships difficult. It is also important to note that stopping these substances has even more severe consequences, such as major depression or psychiatric disorders (15). In this sense, depression associated with behavioral disorders or alcohol and drug dependence can lead to suicide. Drug use creates physical or psychological dependence, or both. It is harmful to individuals and society, as it creates addiction, meaning the subject is at the mercy of an intense desire. We are thus discussing the “concept of disease applied to drug use” (Fernandez L, Sztulman H.) (16). Behaviors such as kleptomania, pyromania, and trichotillomania also fall under impulse control disorders but are classified in a distinct category. Despite this, these behaviors share striking similarities with addictive disorders. Indeed, they are characterized by difficulty resisting the impulse to perform a potentially harmful act, whether for oneself or others. This difficulty is often accompanied by a gradual increase in tension before the act is carried out, followed by a sensation of relief or immediate gratification (16). Medication self-intoxication should not be overlooked, as many men and women, due to easy access to medication, resort to suicide (17). Depression affects not only the individual suffering from it but also their surroundings and represents a heavy social and economic burden, according to the WHO (13). To conduct our study, we began by collecting data from the Our World in Data database, then processed and analyzed this data. Finally, the results of this analysis were interpreted spatially across these 16 West African countries using maps # 1.Présentation de la zone d’étude
West Africa, located in
the western region of the African continent, is an area of great
importance due to its geographical, cultural, economic, and
socio-political diversity. This region, composed of 16 member countries
of the Economic Community of West African States (ECOWAS), presents
particular characteristics that make it an essential field of
observation for the study of various issues, notably in terms of
poverty, mental health, and suicide. Our study focuses on the 16 West
African countries, namely: Benin, Burkina Faso, Cape Verde, Côte
d’Ivoire, The Gambia, Ghana, Guinea, Guinea-Bissau, Liberia, Mali,
Mauritania, Niger, Nigeria, Senegal, Sierra Leone, and Togo.
This study explores the causes of suicide in West Africa through a statistical analysis of socio-economic, health, and psychological data from 16 countries. The objective is to better understand the influencing factors using Principal Component Analysis (PCA), linear regression, and hierarchical clustering.
Loading Required Libraries
library(FactoMineR)## Warning: le package 'FactoMineR' a été compilé avec la version R 4.4.3library(factoextra)## Warning: le package 'factoextra' a été compilé avec la version R 4.4.3## Le chargement a nécessité le package : ggplot2## Warning: le package 'ggplot2' a été compilé avec la version R 4.4.3## Welcome! Want to learn more? See two factoextra-related books at https://goo.gl/ve3WBalibrary(ggplot2)
library(corrplot)## Warning: le package 'corrplot' a été compilé avec la version R 4.4.3## corrplot 0.95 loadedlibrary(car)## Warning: le package 'car' a été compilé avec la version R 4.4.3## Le chargement a nécessité le package : carData## Warning: le package 'carData' a été compilé avec la version R 4.4.3library(carData)
library(Hmisc)## Warning: le package 'Hmisc' a été compilé avec la version R 4.4.3##
## Attachement du package : 'Hmisc'## Les objets suivants sont masqués depuis 'package:base':
##
## format.pval, unitslibrary(psych)## Warning: le package 'psych' a été compilé avec la version R 4.4.3##
## Attachement du package : 'psych'## L'objet suivant est masqué depuis 'package:Hmisc':
##
## describe## L'objet suivant est masqué depuis 'package:car':
##
## logit## Les objets suivants sont masqués depuis 'package:ggplot2':
##
## %+%, alphalibrary(clusterSim)## Warning: le package 'clusterSim' a été compilé avec la version R 4.4.3## Le chargement a nécessité le package : cluster## Le chargement a nécessité le package : MASSlibrary(DataExplorer)## Warning: le package 'DataExplorer' a été compilé avec la version R 4.4.3library(FactoInvestigate)## Warning: le package 'FactoInvestigate' a été compilé avec la version R 4.4.3library(Factoshiny)## Warning: le package 'Factoshiny' a été compilé avec la version R 4.4.3## Le chargement a nécessité le package : shiny## Warning: le package 'shiny' a été compilé avec la version R 4.4.3Data Import
data <- read.csv("data_2.csv", header = TRUE, sep = ";", dec = ",", row.names = 1)
head(data)## Suicide.rate mental.disorders poverty unemployement
## Benin 11.8 0.10 0.51 0.02
## Burkina Faso 15.4 0.10 0.31 0.05
## Cote d_Ivoire 12.6 0.10 0.51 0.10
## Gambie 8.3 0.12 0.25 0.06
## Ghana 8.1 0.11 0.14 0.03
## Guinee 8.2 0.11 0.26 0.06
## Gini.coefficient Working.poverty HDI Drug
## Benin 0.53 16.43 0.50 0.06
## Burkina Faso 0.56 28.97 0.45 0.06
## Cote d_Ivoire 0.63 12.08 0.53 0.07
## Gambie 0.57 15.57 0.49 0.06
## Ghana 0.60 18.80 0.60 0.07
## Guinee 0.47 16.52 0.47 0.063. Data Exploration and Correlation Matrix
mat_cor <- cor(data)
col <- colorRampPalette(c("#BB4444", "#EE9988", "#FFFFFF", "#77AADD", "#4477AA"))
corrplot(mat_cor, method="color", col=col(200), type="upper", order="hclust",
addCoef.col = "black", tl.col="black", tl.srt=90, sig.level = 0.1, insig = "blank", diag=FALSE)
rcorr(as.matrix(data))## Suicide.rate mental.disorders poverty unemployement
## Suicide.rate 1.00 0.09 0.36 -0.28
## mental.disorders 0.09 1.00 -0.40 0.06
## poverty 0.36 -0.40 1.00 -0.09
## unemployement -0.28 0.06 -0.09 1.00
## Gini.coefficient 0.38 0.11 0.19 0.07
## Working.poverty 0.12 -0.30 0.45 -0.49
## HDI -0.01 0.33 -0.32 0.33
## Drug 0.13 0.48 -0.07 0.40
## Gini.coefficient Working.poverty HDI Drug
## Suicide.rate 0.38 0.12 -0.01 0.13
## mental.disorders 0.11 -0.30 0.33 0.48
## poverty 0.19 0.45 -0.32 -0.07
## unemployement 0.07 -0.49 0.33 0.40
## Gini.coefficient 1.00 -0.06 0.35 0.47
## Working.poverty -0.06 1.00 -0.52 -0.23
## HDI 0.35 -0.52 1.00 0.69
## Drug 0.47 -0.23 0.69 1.00
##
## n= 15
##
##
## P
## Suicide.rate mental.disorders poverty unemployement
## Suicide.rate 0.7478 0.1849 0.3085
## mental.disorders 0.7478 0.1356 0.8317
## poverty 0.1849 0.1356 0.7508
## unemployement 0.3085 0.8317 0.7508
## Gini.coefficient 0.1593 0.6867 0.4937 0.7988
## Working.poverty 0.6609 0.2695 0.0909 0.0607
## HDI 0.9853 0.2312 0.2472 0.2243
## Drug 0.6499 0.0730 0.8048 0.1407
## Gini.coefficient Working.poverty HDI Drug
## Suicide.rate 0.1593 0.6609 0.9853 0.6499
## mental.disorders 0.6867 0.2695 0.2312 0.0730
## poverty 0.4937 0.0909 0.2472 0.8048
## unemployement 0.7988 0.0607 0.2243 0.1407
## Gini.coefficient 0.8260 0.1948 0.0737
## Working.poverty 0.8260 0.0454 0.4127
## HDI 0.1948 0.0454 0.0041
## Drug 0.0737 0.4127 0.0041##The analysis of the correlation matrix reveals key relationships that shed light on potential factors influencing suicide rates in West Africa:
Suicide Rate:
Positively correlated with poverty and working poverty → suggests that economic hardship increases suicide risk.
Negatively correlated with unemployment and HDI → possibly due to social safety nets or underreporting; high HDI reflects better health, education, and well-being access.
Mental Disorders:
Positively correlated with HDI and drug issues → may reflect better health system performance and diagnostic capacities.
Negatively correlated with poverty → likely due to underreporting in low-income areas.
Poverty & Working Poverty:
Strongly positively correlated with each other.
Negatively correlated with HDI and drug → indicating limited development and health service access in impoverished contexts.
Unemployment:
Negatively correlated with suicide rate and poverty → may reflect specific socio-economic contexts. Positively correlated with HDI and drug → possibly due to better infrastructure and health services. Gini Coefficient:
Moderately positively correlated with suicide rate and drug issues, suggesting that income inequality may increase social and psychological stress. HDI & Drug:
Very strongly correlated → reflecting the impact of development on awareness and reporting of health issues. ## 4. Principal Component Analysis (PCA)
pca_1 <- PCA(data, scale.unit = TRUE, ncp = 11, graph = FALSE)
fviz_eig(pca_1, addlabels = TRUE, ylim = c(0, 65))
fviz_pca_var(pca_1, col.var = "cos2", gradient.col = c("blue", "green", "red"), repel = TRUE)
fviz_pca_ind(pca_1, col.ind = "contrib", gradient.cols = c("blue", "green", "red"), repel = TRUE)
fviz_pca_biplot(pca_1, repel = TRUE, col.var = "blue", col.ind = "red")
variables
where the arrows represent the variables, and the cos2 value shows the quality of representation. The direction of the arrow indicates the direction of the relationship, and the length indicates the strength of the variable’s contribution to the dimensions. A longer arrow means the variable is significantly correlated with the considered dimension and well-represented. A shorter arrow means the variable is weakly correlated and poorly represented. The angle formed by two variables indicates the strength and direction of the correlation between them. Long arrows close to each other indicate a strong positive correlation, while long arrows far apart or opposite indicate strong negative correlations. For clarity, we associated colors with the variables based on length: green for well-correlated and well-represented variables, blue for moderately correlated and represented variables, and red for weakly correlated and poorly represented variables.
individus
PCA with two main dimensions, where the countries represent the studied individuals, and the cos2 value shows the quality of their representation on the factorial plane. We associated colors to facilitate understanding of representation: orange for well-represented individuals and red for poorly represented ones. Thus, countries close to each other on the graph share similar characteristics and could have similar expectations or performances in the considered dimensions.
5. Hierarchical Clustering on Principal Components (HCPC)
res.PCA <- PCA(data, graph=FALSE)
res.HCPC <- HCPC(res.PCA, nb.clust=3, consol=FALSE, graph=FALSE)
plot.HCPC(res.HCPC, choice='tree')
plot.HCPC(res.HCPC, choice='map', draw.tree=FALSE)
plot.HCPC(res.HCPC, choice='3D.map', ind.names=FALSE, centers.plot=FALSE)
### dendogramme First Group: Ghana, Mauritania, and Côte d’Ivoire These
countries share common characteristics marked by a relatively higher
level of human development, as well as greater exposure to mental
disorders and modern social pressures. This could explain suicide rates
more influenced by complex psychological and social factors, typical of
societies transitioning toward greater urbanization and modernity.
Second Group: Niger, Nigeria, Mali This group seems to occupy an
intermediate zone. These countries share unstable socio-economic
conditions where poverty is very present, but modern psychological
factors (such as mental disorders or drugs) remain less visible or
documented. These are countries that could quickly shift into one group
or another, depending on their social or economic evolution. Third
Group: Senegal, Sierra Leone, The Gambia, Burkina Faso, Liberia,
Guinea-Bissau, Guinea, Benin, and Togo This group consists of countries
with similar profiles, marked by persistent economic difficulties, poor
access to mental health care, and precarious living conditions. Suicide
here would be more influenced by structural factors like extreme
poverty, unequal access to education, and limited future prospects.
These countries represent the most socially vulnerable zone, where
suicide may be linked to profound socio-economic despair.
6. Multiple Linear Regression
model1 <- lm(Suicide.rate ~ mental.disorders + poverty + unemployement + Gini.coefficient + Working.poverty + HDI + Drug, data = data)
summary(model1)##
## Call:
## lm(formula = Suicide.rate ~ mental.disorders + poverty + unemployement +
## Gini.coefficient + Working.poverty + HDI + Drug, data = data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.9706 -1.2879 -0.5536 0.6004 5.3674
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -6.33740 17.21700 -0.368 0.724
## mental.disorders 48.42089 127.64153 0.379 0.716
## poverty 8.90902 8.31831 1.071 0.320
## unemployement -39.53680 39.96403 -0.989 0.355
## Gini.coefficient 19.30495 25.70199 0.751 0.477
## Working.poverty -0.05631 0.12083 -0.466 0.655
## HDI -4.37163 27.36070 -0.160 0.878
## Drug 58.84370 280.78634 0.210 0.840
##
## Residual standard error: 3.182 on 7 degrees of freedom
## Multiple R-squared: 0.3712, Adjusted R-squared: -0.2575
## F-statistic: 0.5905 on 7 and 7 DF, p-value: 0.74827. Prediction Plot
data_vis <- data.frame(valeurs_reelles = data$Suicide.rate, predictions = predict(model1))
ggplot(data_vis, aes(x = valeurs_reelles, y = predictions)) +
geom_point() +
geom_smooth(method = "lm", se = FALSE, color = "blue") +
labs(x = "Observed Values", y = "Predicted Values", title = "Linear Regression") +
theme_minimal()## `geom_smooth()` using formula = 'y ~ x'
regression
This graph illustrates a simple linear regression, highlighting the relationship between actual values (independent variable) and predictions (dependent variables). In this specific case, we used: The blue line represents the regression model used to predict values based on actual values. It is determined by the linear regression equation. The black dots represent individual observations used to build the regression model. Each point indicates how an actual value compares to the prediction generated by the model. Points below and above the line represent deviations between actual values and predictions. The in-depth analysis of the data reveals that suicide in this region is the result of a complex interaction between various socio-economic and psychosocial factors. On one hand, economic indicators such as poverty and working precariousness (Working Poverty), as well as income inequalities measured by the Gini coefficient, create an environment conducive to increased psychological distress. On the other hand, variables related to mental health, particularly mental disorders, and drug use, reflect a psychosocial vulnerability that amplifies suicide risk.
8. Automated Report (FactoInvestigate)
Inves_1 <- PCA(data, graph = TRUE)
rapport_1 <- Investigate(Inves_1)9. Interactive Interface
PCAshiny(data)Discussion
At the conclusion of the Principal Component Analysis (PCA), two main dimensions emerge significantly in explaining the disparities related to suicide in West Africa: The first dimension, which we have named Economic Growth and Capacity to Meet Basic Needs, brings together strongly correlated variables such as the Human Development Index (HDI), substance use disorders (Drug), unemployment, and in-work poverty. The second dimension, which we have named Social Vulnerability and Psychological Fragility, is mainly influenced by the suicide rate, the Gini coefficient, and the level of poverty. The results highlight that mental disorders, addictions (alcohol/drugs), poverty, unemployment, and social inequalities are structural factors deeply linked to suicide rates in the region. For example, a high contribution of mental disorders (12.53%) to the first dimension suggests a direct link between mental health and economic conditions. This reality is amplified by a significant correlation with the HDI (24.76%), indicating that countries with lower human development are at greater risk. The analysis of the Gini coefficient, which contributes nearly 29% to the second dimension, shows that economic inequality, measured by wealth distribution, is a deep driver of social distress, potentially leading to extreme cases of despair such as suicide. The suicide rate, which loads heavily on the second dimension (32.99%), is particularly high in countries that combine low economic development, extreme poverty, significant inequalities, and unaddressed psychological fragility. These findings are consistent with many studies linking precariousness, lack of psychological support, loneliness, and marginalization to increased suicidal behavior. In West Africa, the lack of resources, specialized mental health centers, and public policies targeting these issues further weakens individuals at risk.
- Proposed Solutions To reduce suicide risks and improve mental and social well-being, we recommend: Strengthening public mental health systems, including training professionals and raising awareness. Improving economic governance to reduce inequalities (Gini), notably through inclusive fiscal and social policies. Supporting decent employment, particularly for youth and working poor. Implementing addiction prevention programs, especially for alcohol and drug use. Investing in education and human development as levers for social resilience.
CONCLUSION Our study has enabled the consideration of the underlying causes of persistent suicide rates in West Africa by integrating socio-economic, psychological, and environmental factors. It reveals that certain variables play a determining role in the persistence of suicidal behaviors. In particular, human development indicators (HDI), the prevalence of mental disorders, drug use, and in-work poverty are major factors in countries such as Ghana, Mauritania, and Côte d’Ivoire. Conversely, variables such as the suicide rate itself, the Gini coefficient, and overall poverty level largely explain the social vulnerability and psychological insecurity observed in countries like Niger, Nigeria, and Mali. These results underscore the importance of adopting integrated policies that simultaneously address economic and health dimensions to reduce suicide risks. It appears essential to strengthen mental health services, improve access to decent employment, and reduce inequalities. Moreover, looking ahead, it would be relevant to enrich the study by collecting institutional and political data in order to better understand the cultural and structural dynamics that contribute to the development of suicidal behaviors.
###Mapping results
