This analysis, carried out on a representative sample of the French population, shows that people who have experienced sexual violence in their lifetime have an increased risk of using psychoactive substances such as tobacco, alcohol and cannabis. Our results align with previous research on the subject.

Several studies have investigated the association between sexual violence and substance use (Fletcher, 2021). This link may be, at least partly, explained by the intense stress and onset of negative emotions caused by sexual violence. In response to a traumatic sexual violence, victims/survivors often seek to self-medicate to avoid and reduce difficult, negative emotions (Ullman et al., 2005). These often-unsuccessful attempts to disassociate further promote substance use (Holahan et al., 2015; Smith et al., 2014), which in turn negatively affect health, and quality of life (Degenhardt et al., 2018; Rooke et al., 2013).

To better understand the association between sexual violence and substance use, we must consider the impact of other factors as well. While it's important to consider individual characteristics such as mental health pre-assault, biological factors, personality, etc., circumstances of sexual violence also play important role. Research shows that the effects of sexual violence are cumulative, meaning that people who have experienced sexual violence more than once in their life are more likely to use substances (Cisler et al., 2011; Hedtke et al., 2008; Walsh et al., 2014). Our findings offer partial support for this hypothesis: both single and repeated sexual violence were associated with increased substance use risk, though the pattern was inconsistent across outcomes and by sex, and no clear dose-response relationship was demonstrated. The heterogeneity of sexual violence types captured in this study (ranging from unwanted touching to rape) may have contributed to the inconsistent pattern observed across outcomes. Future studies should examine whether the nature and severity of sexual violence, in addition to its repetition, modulates the risk of substance use. Similarly, research shows that the long-term consequences of sexual violence are more numerous when sexual violence took place during childhood, like 65% of the people in our sample (WHO, 2007). The type of sexual violence and the degree of violence also play important roles in the incidence of substance dependence (Campbell et al., 2009; Kendler et al., 2000).

Finally, the impact of sexual violence on the victims/survivor's health can depend on a much wider range of factors, as demonstrated by Rebecca Campbell's ecological model (Campbell et al., 2009). Social reactions, for example, play a determining role in a person's mental health, self-esteem and tendency to self-blame. This may partially explain the differences in substance use outcomes between men and women who had sexual violence experiences in our results.

We acknowledge that this study has several limitations. First, the questionnaire does not allow for the confirmation that respondents began using substances after experiencing sexual violence, thus limiting the interpretation of our results. However, our findings on age at sexual violence exposure (mean 13.4 years) indicate that sexual violence occurred at a very young age for respondents in our sample.

Furthermore, the proportions of men and women reporting their sexual violence experiences are most likely underestimated. It is undoubtedly very difficult to talk about such events and it may even lead to dissociative amnesia (Staniloiu and Markowitsch, 2014). However, it should be noted that according to the 2005 Health Barometer, 4.1% of women and 0.6% of men reported experiencing sexual violence in their lifetime (Léon and Lamboy, 2006). Apart from the fact that the survey question was different (the Health Barometer 2005 asked: “In your life, have you been subjected to forced sexual intercourse?”), we can assume that this difference is an outcome of a greater willingness to report sexual violence in a scientific survey, rather than a major increase in the incidence of sexual violence (Bajos et al., 2008). This is a notable development that will perhaps be amplified by the social movements which destigmatized sexual violence subsequent to the collection of our data, but it does not diminish the under-reporting of sexual violence experiences present in our analysis. Moreover, due to the high prevalence of poly-victimization within our sample, disentangling the independent effects of a single incident of sexual violence was not feasible. Furthermore, it should be noted that alcohol and cannabis variables were likely also under-reported in our sample (Le et al., 2022; Stockwell et al., 2016).

The analysis of behavioral differences between men and women in our sample may be underpowered, as suggested by the presented confidence intervals. Only 223 men declared having experienced sexual violence. Finally, this is a cross-sectional study, which does not allow for causality to be determined, and the data is relatively dated since it dates from 2017. To date, it is the last Health Barometer survey that included data on sexual violence in France.

However, this analysis has the advantage of being based on data from a large sample, representative of the general population, making it easier to generalize our results to the entire French adult population. Additionally, this study investigated three distinct psychoactive substances, including different degrees of alcohol and cannabis use. Finally, the multi-thematic nature of the Health Barometer data gives our results a certain robustness. It allowed us to adjust for the impact of a wide range of factors on the association between sexual violence and substance use, and to better take into account the social, demographic and economic diversity of the population.