Descriptive cross-sectional study conducted from December 2022 to July 2023 in northeastern Benin. The study population consisted of all female students at the University of Parakou. The sample size was calculated using Schwartz’s formula, taking as a reference a study conducted in Mali where the prevalence of sexual assault was 2%. The minimum sample size thus calculated was 188 female students. The sampling technique was initially a non-proportional stratified convenience sampling at the level of the study entities of the University of Parakou (Faculty of Agronomy; Faculty of Law and Political Science; Faculty of Economics and Management; Faculty of Medicine; University Institute of Technology ; Faculty of Letters, Arts and Human Sciences; National training school for senior health technicians and epidemiological surveillance; National School of Statistics, Population and Demography; Nursing and obstetrics training institute) made up of students at various levels of study. In a second stage, simple random sampling was used to obtain the number of female students. All female students regularly enrolled at the University of Parakou and having given their free and informed consent were included in the study.

A digitized survey form was used, containing information on socio-demographic data, clinical data and standardized scales.

The dependent variable was the existence of sexual assault (declarative variable) among female students at the University of Parakou. The psychopathological repercussions linked to sexual assault and explored in these students were: post-traumatic stress disorder assessed with the Post Traumatic Stress Disorder Checklist-5 (PCL-5) scale and current depression screened with the Patient Health Questionnaire-9 (PHQ-9) scale.

Post-Traumatic Stress Disorder Checklist-5 (PCL-5): This is a measurement tool that assesses symptoms of post-traumatic stress disorder. Created in 2013 by Weathers et al. [15] of the National PTSD Symptom Center, it was translated into French by Alain Brunet et al. and validated in 2016. It consists of 20 items that assess nightmares, flashbacks, avoidance, hypervigilance and physical and emotional reactions in relation to the memory of the trauma. Respondents are asked to rate how much they have been bothered by each of the 20 items over the past month on a 05-point scale (Likert scale ranging from 0 to 4). Items are added together to give a total score ranging from 0 to 80. A threshold score of 44 was taken as significant.

Patient Health Questionnaire-9 (PHQ-9)

This is a brief tool used to diagnose and measure the severity of depression. The PHQ-9 is shorter than many other depression screening instruments and can be self-administered. The PHQ-9 includes the two cardinal signs of depression: anhedonia and depressed mood. The PHQ-9 was developed by Robert L. Spitzer, Janet W.B. Williams and Kurt Kroenke in 1999 [16] . Each item is rated on a severity scale ranging from 0 to 3, where the respondent is asked to rate how often each symptom has occurred in the last two weeks (0-not at all; 1-few days; 2-more than half the days or 3-almost every day), producing a total score ranging from 0 to 27. The respondent is also asked to what extent the identified problems have interfered with work, home or social life. However, responses to this item are not scored or included in the total score, which is interpreted as follows:

· 1 - 4: minimal depression

· 5 - 9: mild depression

· 10 - 14: moderate depression

· 15 - 19: moderately severe depression

· 20 - 27: severe depression

The Hospital Anxiety and Depression scale (HAD) [17] is a screening instrument for anxiety and depressive disorders. It was developed by Dr Phillip Snaith and Antony Zigmond in 1983 and translated in 1985 by JP Lépine et al. It comprises 14 items scored from 0 to 3. Seven (07) items relate to anxiety (total A) and seven (07) to depression (total D), giving a maximum score of 21 for each. Completion time was short, between 2 and 6 minutes. Only questions concerning anxiety were taken into account. The final score corresponds to the addition of the items as follows: 1 + 3 + 5 + 7 + 9 + 11 + 13 for anxiety and 2 + 4 + 6 + 8 + 10 + 12 + 14 for depression. To screen for anxiety symptoms, the following interpretation is proposed for each of the scores (A and D): <07: no symptoms; [8 to 10]: doubtful symptoms; ≥11: definite symptoms.

Female Sexual Function Index (FSFI) questionnaire [18]

To assess sexual functioning, we used the Female Sexual Function Index (FSFI) scale, validated in French in 2012 by Dargis L. et al. in Quebec. It comprises 19 questions assessing 06 domains: desire (1, 2); arousal (3, 4, 5, 6); lubrication (7, 8, 9, 10); orgasm (11, 12, 13); satisfaction (14, 15, 16); pain (17, 18, 19). The threshold value used to define impaired sexuality is a score less than or equal to 26.55.

Data were analyzed using SPSS 26 software. Central tendency and dispersion parameters were used for quantitative variables, and proportions for quantitative variables with their confidence intervals (CI). Pearson’s Chi-2 statistical test and Fisher’s exact test were used to compare the dependent variable with the independent variables. Means were compared by Student’s t-test or by Leven and Kolmogorov in the case of a cross between a nominal or ordinal qualitative variable and a quantitative variable. The difference was statistically significant at p ˂ 0.05. Variables associated in bivariate analysis were introduced into the initial multivariate analysis model. The associated variables were identified from the binary logistic regression analysis with determination of the odd ratios (OR) and their 95% confidence intervals (95% CI).

The average age of the female students surveyed was 20.56 ± 2.46 years, with extremes of 16 and 32 years, and a predominance of subjects under the age of 25.

The majority of female students who had been sexually assaulted were single (82.50%) (33) (Table 1).

Most victims had no medical (36; 90%) or surgical (37; 92.50%) history. Only 5% (2) of victims had suffered from mental illness. Obstetrical history showed that 7.50% (3) were primigravida and primiparous with a live child. None of them reported a criminal record (Table 2).

Menarche for the majority of victims had begun before the age of 13 (52.50%). More than half had had their first sexual intercourse after the age of 18 (62.50%), and the majority had consented (21; 52.50%). On average, they had one sexual partner (30; 75.00%) and 56% (14) had frequent sexual intercourse (Table 3).

The most frequent types of sexual assault were rape 45% (18) and sexual touching (16; 40%), frotteurism and voyeurism (3; 7.50%) and finally sexual harassment for 2 of the victims (5.00%). According to the results, some 17 (42.50%) of the victims had been assaulted more than once, and 32.50% (13) had been assaulted by a friend or boyfriend, and 30.00% (12) by a family member. These assaults generally occurred during the day (19; 48.72%) and during home visits (21; 37.5%). In 32.50% (13) of cases, the victims had been subjected to physical coercion and in 20.00% (8) to psychological coercion by the aggressors (Table 4).

To assess the various repercussions of sexual assault, the following scales were used: Post traumatic stress disorder checklist-5 (PCL-5) which assesses symptoms of post-traumatic stress disorder; patient health questionnaire (PHQ-9) which assesses depression; Female Sexual Function Index (FSFI) which assesses female sexual dysfunction and Hospital Anxiety and Depression (HAD) which assesses anxiety.

Psychological repercussions were reflected by depression in 72.5% (29) of cases, including mild depression in 30% (12), moderate depression in 37.5% (15) and severe depression in 5% (2). Anxiety disorders were present in 27.50% (11) of victims. Symptoms of post-traumatic stress disorder were present in 5 victims (12.5%). Sexual dysfunction was noted in 77.5% (31) of cases (Table 5).

In bivariate analysis, the following variables showed a statistically significant association with sexual assault: age (p = 0.018), marital status (p = 0.035), gender (p < 0.001), parity (p = 0.008), virginity (p = 0.002), earliness of sexual intercourse (p = 0.005), consent of first sexual intercourse (p = 0.024) and number of sexual partners (p = 0.009).

Multivariate analysis shows that potential predictors of sexual assault among female students at the University of Parakou in 2023 were gravidity: greater than 1 (OR (IC95%): 15.32 ([1.731 - 135.69]), p = 0.014), significant negative life events (OR (IC95%: 17.24 ([1.99 - 149.24]), p = 0.010), first sexual intercourse before age 18 (OR (IC95%: 4.608 ([1.32 - 16.03]), p = 0.016) and first consensual sexual intercourse (OR (IC95%: 0.21 ([0.07 - 0.62]), p = 0.005). Risk factors for sexual assault were multigestation, significant negative events and early sexual debut (before age 18). Multi-gestational students were 15.33 times more likely to be sexually assaulted than non-gestational students, the presence of a negative event in the student’s life was associated with a high risk of sexual assault, 17.24 times more than those with a positive event, and early sexual intercourse increased the risk of sexual assault by a factor of 4.

Consensual first intercourse, on the other hand, is a protective factor compared to non-consensual first intercourse. The risk of sexual assault is reduced by 79% when the first sexual experience is pleasant (Table 6).

Of the 376 targets surveyed, 40 had been victims of at least one form of sexual assault. This represents a prevalence of 10.63%. This prevalence corroborates that of Hebert et al. in the city of Marseille in France which was 10.3% in 2018. Essiben et al. in the cities of Douala, Yaoundé and Baganté in Cameroon reported a prevalence of 9.5% in 2018 among students from specific universities in these cities [28] .

Nevertheless, this prevalence is much higher than in other authors. In Mali, Théra et al. and Thierno et al. found a prevalence of 2% and 0.22% respectively in 2015 and 2021 [7] [29] ; Niort et al. [27] in France in 2014 a prevalence of 1.5%. Similarly, Enouani et al. [6] found 2.6% in Congo Brazzaville in 2020. The samples were 37 out of a total of 1810 in Théra et al. in Mali; 107 out of 47729 in Thierno et al. in Mali; 592 out of 39469 in Niort et al. in France; and 150 out of 5620 in Enouani et al. in Congo Brazzaville [6] [7] [29] . This low frequency can be explained by the fact that victims do not systematically report the trauma they have suffered. According to the WHO, the reasons for this include inadequate support systems, shame, fear or risk of reprisals, and fear or risk of being blamed or accused. The low proportion should not systematically mean that the phenomenon does not exist, leading to its trivialization.

On the other hand, some authors have reported higher prevalences than in this study: Vaillancourt Morel et al. [24] found a prevalence of 15.32% in 2014 in Quebec; Chérif et al. [22] in 2021 in Donka, Guinea, which was 16%. Puri et al. [3] report a prevalence of 46% in Nepal in 2012. This high figure in Nepal can be explained by the fact that the study was carried out in rural areas among married women who are afraid to denounce their husbands who have abused them within a marriage.

Rape was the most common type of sexual assault, accounting for 45.0% (18). Sexual touching accounted for 40% (16), voyeurism and frotteurism for 7.5% (3) and sexual harassment for 5% (2). The same finding has been made by other authors, with higher proportions than in the present study: Adama et al. [8] in Togo 2013 found a rape rate of 62.2%. Diallo et al. [23] in Dakar in 2017 found rape in a proportion of 79.3%, followed by sexual touching 12.9%. For Dembele et al. [30] in 2021 in Mali, the percentage of rape was 86.27%. This can be explained on the one hand by the fact that vaginal intercourse is the usual and culturally valued form of sexual intercourse in African countries. And on the other hand, the study populations were predominantly minors and young women unable to defend themselves physically against their attackers due to unequal physical strength. Other types of aggression, such as sexual harassment, touching and exhibitionism are often downplayed and are not often reported.

The consequences of sexual assault are such that they can have irreversible effects on the lives of the victims. At the end of the study, the consequences found among victims were mild depression (30%), moderate depression (37.5%) and severe depression (5%); post-traumatic stress symptoms (12.50%); and anxiety disorders (27.50%). In a similar vein, Bourgou et al. [31] in Tunisia observed psychiatric disorders in 58% of cases, with post-traumatic stress disorder in 19.3% and full-blown depressive disorders in 8%. Khemakhem et al. [21] found post-traumatic stress symptoms in 32.7% of subjects and depression in 5.8%. For Hebert et al. [4] , the data also indicate that more victims of sexual assault present post-traumatic stress disorders reaching the clinical threshold of (27.9%; n = 173), compared with non-victims (9.1%; n = 310). Vaillancourt Morel et al. [24] report that these assaults expose victims to a high degree of psychological distress (48.15%) and abandonment anxiety (67.29%).

Several authors discuss the various repercussions experienced by victims, which corroborates our findings. A lack of psychological care for victims and poor follow-up would explain the occurrence and persistence of these psychological repercussions in some victims.

In this study, the non-probability sampling method was used. The sample size was calculated using Schwartz’s formula based on an average prevalence of sexual assault victims estimated at 2% in Mali in 2014 [7] . The minimum size obtained after calculation was 188 students, while the size obtained during data collection was 376 students. The data collected came from subjective statements and the questions asked were sensitive and related to traumatic events in the past. Information bias could therefore not be entirely ruled out.