Open access Original research Prevalence and associated factors with sexual violence victimisation youth before, during and after the COVID-­19 lockdown: a cross-­sectional study in Spain Belén Sanz-­Barbero ‍ ‍,1,2 Carme Saurina ‍ ‍,1,3 Laura Serra ‍ ‍,1,3 Gemma Renart Vicens ‍ ‍,1,3 Laura Vall-­Llosera Casanovas ‍ ‍,1,3 Laura Otero-­García ‍ ‍,1,4 María José López ‍ ‍,1,5,6 Glòria Perez,1 Carmen Vives-­Cases ‍ ‍1,7 To cite: Sanz-­Barbero B, ABSTRACT Saurina C, Serra L, et al. Strengths and limitations of this study Objectives To analyse the prevalence of sexual violence Prevalence and associated (SV) and associated factors in Spanish young adults in factors with sexual violence ►► This is the first study in Spanish youth that analysed the last year and before, during and after the COVID-­19 victimisation youth before, sexual violence (SV) victimisation in young women lockdown. during and after the COVID-­19 and men during the last 12months, taking in account lockdown: a cross-­sectional Design Cross-­sectional study based on the online ‘SV in sociodemographics, sexual attraction and couple-­ study in Spain. BMJ Open Young People Survey’ (2020). related characteristics. 2021;11:e055227. doi:10.1136/ Setting Non-­institutionalised population residing in Spain. ►► Data collection was carried out at the end of 2020, bmjopen-2021-055227 Participants 2515 men and women aged 18–35 years thus, it was possible to assess the impact of the ►► Prepublication history and old. The participants were obtained from a probability COVID-­19 lockdown on SV victimisation and its as- additional supplemental material based, online closed panel of adults aged 16 or older that sociated factors in young people. for this paper are available is representative of the non-­institutionalised population. ►► The limited sample size and the low frequency of online. To view these files, The sample designed includes quotas by sex, age, region sexual minorities do not allow us to analyse the dif- please visit the journal online and country of origin. ferent sexual minorities independently. (http://​dx.​doi.​org/​10.​1136/​ Outcomes measures and analyses SV victimisation by ►► The sample size of this study, the analysis of sort bmjopen-​2021-​055227). sociodemographics, sexual attraction and couple-­related time periods and the low prevalence of SV during characteristics during the past year and before, during the time periods considered, mean that these results Received 08 July 2021 Accepted 02 November 2021 and after the COVID-­19 lockdown (March–June 2020). should be interpreted with caution. Prevalence ratios were calculated using robust Poisson regression models. Results In Spain, 8.5% of young people experienced SV during the past year. The greatest prevalence was observed INTRODUCTION in women with bisexual attraction (17.5%) and in men Sexual violence (SV) is defined as any sexual with homosexual attraction (14.2%). During the COVID-­19 act or attempt, unwanted sexual comments lockdown, the prevalence of SV victimisation was lower or acts to traffic against a person’s sexuality (1.9%), but unwanted intercourses increased, affecting using coercion by any person, regardless 64.4% of those exposed to SV during the period. People of their relationship with the victim, in any with homosexual or bisexual attraction were more likely to setting, including but not limited to home experience SV in all of the studied periods (PRbefore: 2.01; and work.1 p<0.001; PRduring: 2.63 p=0.002; PRafter: 2.67; p<0.001). Sexual intimate partner violence (Sex-­ © Author(s) (or their Women were more likely than men to experience SV prior IPV) is one of the forms of SV that has been employer(s)) 2021. Re-­use to the lockdown, while no cohabitation increased the studied most. It is primarily perpetrated permitted under CC BY-­NC. No likelihood to experience SV after this period commercial re-­use. See rights by men against women,2 but a high preva- Conclusions SV victimisation in Spanish youth is high. and permissions. Published by During COVID-­19, there were changes in the magnitude lence of Sex-­IPV victimisation has been also BMJ. of factors associated with SV. It seems that SV events reported by men.3 In the USA, an estimated For numbered affiliations see decreased in people who did not live with their partners, 15.8% of women and 9.5% of men have expe- end of article. rienced other forms of Sex-­IPV during their but unwanted intercourses increased. The development of Correspondence to prevention strategies to address SV in youth should take lifetimes.2 In relation to non-­partner SV, it Dr Belén Sanz-­Barbero; into account social inequalities by sex, sexual orientation has been estimated that 7.2% of women over ​bsanz@​isciii.​es and origin. age 15 worldwide have experienced any form Sanz-­Barbero B, et al. BMJ Open 2021;11:e055227. doi:10.1136/bmjopen-2021-055227 1 Open access of non-­partner SV.4 In the European Union-­28 context, a Spanish young adults in the past year and before, during total of 11% of women have experienced Sex-­IPV or non-­ and after the COVID-­19 lockdown period in Spain. partner SV.5 There is an increasing awareness of the magnitude of non-­partner SV/Sex-­IPV, especially in relation to youth.1 METHODS Recent published meta-­analysis shows that 9% of teen- Population and sample agers have experienced sexual dating violence. It has Cross-­sectional study of the SV in Young People Survey been reported that girls have higher rates of victimisation carried out online with young men and women age 18–35 than boys (14% vs 8%).6 In Spain, according to the latest residing in Spain. Calculation of the sample was carried Macro-­survey on Violence against Women, 8.4% of ever-­ out using the most recent data on the prevalence of SV partnered women have suffered Sex-­IPV. In women aged against women in Spain7 corrected for sex. The minimum 16–24, this percentage reaches 12.4%, compared with sample size needed was estimated at 2500 questionnaires the 8.9% of women aged 25 and over. Non-­partner SV with quotas by sex, age, autonomous community (region) is reported to be 1.8 times higher among women aged and country of origin, in order to guarantee a sample error 16–24 as compared with women aged 25 years old and of ±5%, considering a 95% CI and prevalence estimates over (11% vs 6%).7 with a precision level of (±) 0.9. The sample was obtained Non-­ partner SV and Sex-­ IPV have short-­ term and from a voluntary panel that included 138 393 adults over long-­ term health consequences, including a high risk age 16. This panel is designed to be fully representative of depression, suicide and femicide/homicide; sexually of the non-­institutionalised Spanish civilian population. transmitted infections; induced abortion and alcohol The sample design included quotas by sex, age, region consumption.4 5 7 The presence of this negative health and country of origin, in order to represent the popu- effects is alarming.8 The silence and stigma that usually lation between ages 18 and 35 residing in Spain. Panel- surround experiences of non-­ partner SV and Sex-­ IPV lists were invited by email to complete the survey. Those may influence not only the severity of these health conse- who accepted received (via email) an individual link to quences, but also the victim’s help-­seeking behaviour.9 10 complete the online survey. Panellists received only non-­ Addressing SV and Sex-­IPV is a priority. Further research survey-­specific incentives through a point-­based rewards about the associated factors in youth is needed to inform programme, which they could use to purchase products existing primary prevention strategies.11 12 Non-­partner from different stores. After conducting a first pilot study, SV and Sex-­ IPV have been associated with individual, a database of 2525 people was obtained. Once refined, relational and social factors like sexism and traditional it resulted in 2515 registrations. The response rate was gender stereotypes,13 14 exposure to sexually explicit mate- 62.3%. Data collection was carried out from 15 October rial,15 a large age gap between couples or lack of social 2020 to 28 October 2020. support.14 Current knowledge about SV among youth in socially disadvantaged circumstances, due to their sexual Measurements orientation, migrant status or socioeconomic situation, is The main outcome, prevalence of SV victimisation (sex-­ still rather scarce and the studies usually integrate SV with IPV and/or non-­partner SV) during the past 12 months other forms of violence like physical-­IPV.16 (SV12m), was measured using eight items following This study is part of a larger project on SV, and the use national7 and international survey guidelines (see online of pornography in young adults in Spain, which occurs supplemental file 1).21 When one of the participants in the context of the proposal of a new national law to responded affirmatively to having experienced one of cope with SV and guarantee sexual freedom.17 In addi- a number of sexual violence behaviours during last 12 tion, our data collection was carried out at the end of months, we considered the person to have been exposed 2020, thus, it was possible to describe, in young people, to SV12m. Then we asked: ‘Thinking back to how long we the prevalence of SV victimisation and its associated have lived through lockdown (March 14 through May 9, 2020), factors during the last 12 months and in different period did this behavior happen … before lockdown? … during lock- related to the COVID-­ 19 lockdown. Previous studies down? or … after lockdown? (you may check all the applicable have already pointed to a possible increase in the prev- response options). alence of IPV and, above all, its severity, including more Covariables were included that had been associated frequent SV in particular Sex-­IPV.18–20 However, further with SV12m victimisation in prior studies, such as socio- research that distinguishes between the risk of Sex-­IPV demographic variables, sexual attraction and relationship and non-­partner SV in young adults during the period status.6 11 is needed. Socioeconomic variables: sex (man, woman), age, This study was based on the hypothesis that SV risk country of birth (Spain, outside of Spain), highest factors would remain in the different periods analysed. completed studies (no studies, primary studies, secondary However, it is possible that the magnitude of the SV risk studies, higher education); work situation: ‘have you factors associated with SV in cohabiting couples increased worked for pay in the last 12 months?’ (yes/no). during lockdown period. The aim of this paper was to Variables related to relationship status: ‘Do you analyse the prevalence of SV and its associated factors in currently have a partner?’ (yes/no). Those who had a 2 Sanz-­Barbero B, et al. BMJ Open 2021;11:e055227. doi:10.1136/bmjopen-2021-055227 Open access partner were asked about their living situation. Response (10.5%) than in men (6.8%), and in those born outside categories were: yes, we live together/yes, but we live of Spain (12% vs 8.1%). Among men, the prevalence was together only intermittently (phases, weekends)/no, we greater in those who reported homosexual (14.2%) and live in different domiciles. The variable was categorised bisexual attraction (10.6%). In women, the prevalence in terms of continuous cohabitation/non-­cohabitation; increased among those with bisexual attraction (17.5%). In the latter included intermittent cohabitation and non-­ women, SV12m decreased as age increased (p<0.001). It was cohabitation. Information was collected about whether also lower in those who were cohabitating with their partner participants had minor children from either one of the (7.4%). partners. The variable was categorised as: yes, minor chil- During the lockdown (table 2), the prevalence of SV dren of one or both/no minor children. victimisation was 1.8%, and there were no significant Sexual attraction data were collected using the differences by sex. Among the heterosexual population following question: ‘With which of the following affir- the prevalence of SV victimisation was lower in all of the mations do you feel most identified?’: I’m only attracted study periods (5%, 1.3% and 1.6% before, during and to women/I’m normally attracted to women but some- after, respectively) than the bisexual population (11.8%, times I’m attracted to men/I’m attracted to both women 3.8% and 4.5% before, during and after, respectively) and and men/I’m normally attracted to men, but sometimes homosexual men (8.4%, 2.5% and 5.0% before, during I’m attracted to women/I’m only attracted to men/I’m and after, respectively). During the COVID-­19 lockdown, not attracted to men nor women—no response. Those there were no statistically significant differences in the who reported being exclusively attracted to people of prevalence of SV victimisation by place of birth (born the same sex were included in the homosexual category abroad: 2.3% vs born in Spain: 1.8%) but this difference (gay/lesbian, according to sex). Those who reported only became statistically significant after the lockdown (4.0% being attracted to people of the opposite sex were clas- vs 2.1%). Those who cohabitate had a greater prevalence sified as heterosexual. Those who responded not being of SV victimisation before (5% vs 7.6%) and after the attracted to any sex (n=6) were considered missing values. (1.4% vs 2.7%) than those who do not cohabitate, but this difference was not statistically significant during the Statistical analysis lockdown (2.3% vs 1.5%). First, we described the prevalence of SV12m victimisa- The most prevalent behaviours (table 3) among those tion in the total sample and stratified by sex, according to who experienced SV12m victimisation were ‘unwanted the covariables described above. Later, we described the touching of intimate body parts’, reported by 73.6% of prevalence of SV victimisation before, during and after women and 61.6% of men. Unwanted intercourses was the COVID-­19 lockdown period. The differences of the present in 39.4% of people who experienced SV12m. frequency distribution in the covariates were estimated During the lockdown these practices affected 64.4% of using Pearson’s χ2 test. When the expected values in any those exposed to SV during this period. of the cells were below 5 we used Fisher’s exact test. We Tables 4 and 5 show the variables associated with SV prev- also described the frequency of each SV12m behaviours, alence, in different study periods. There is an independent in the whole sample and stratified by sex, in addition to effect of sex on the probability of experiencing SV12m. In the the frequency of each SV behaviour before, during and last 12 months (table 5), the probability of SV victimisation after lockdown. was 54% greater in women than in men, although this differ- Finally, Poisson models with robust variance were esti- ence was not observed neither during nor after the lockdown. mated to analyse the association between the covariables The probability of SV12m was greater among the popula- and SV prevalence victimisation in the different periods. tion born outside of Spain than in the Spanish-­born popu- Unadjusted and later adjusted, prevalence ratios (PR) lation (PR: 1.54 (1.07; 2.19)). This difference lost statistical were calculated. In order to obtain model adjustment, significance during and after the lockdown period (PR: 1.83 variables were included sequentially when they resulted (0.94; 3.57)) approaching statistical significance (p<0.07). in statistical significance in the unadjusted model. Inter- Those with homosexual or bisexual attraction had a greater actions were explored between sex and the covariables probability of experiencing SV than those with heterosexual included in the models. No statistically significant inter- attraction in all of the studied periods, during last 12 months action was identified. (PR:1.98 (1.52;2.59)), prior (PR: 2.01 (1.46;2.76)) during (PR: 2.63 (1.42; 4.88)) and after (PR: 2.67 (1.54; 4.61)) the Patient and public involvement lockdown. Patients or the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research. DISCUSSION In Spain 9 of every 100 young people ages 18–34 have RESULTS experienced SV12m victimisation. The overall preva- Table 1 shows the prevalence of SV12m victimisation in the lence was greater among women (10.5%) and in those whole sample and stratified by sex. In Spain, 8.5% of youth born outside of Spain (12%). By sex, the prevalence was experienced SV12m. The prevalence was greater in women greater among men with non-­ heterosexual attraction Sanz-­Barbero B, et al. BMJ Open 2021;11:e055227. doi:10.1136/bmjopen-2021-055227 3 4 Table 1 Prevalence of sexual violence victimisation in the past 12 months in the total sample and stratified by sex, Sexual Violence in Young People Survey, 2020 Have you experienced sexual violence in the past 12 months? (n.a=41) Total sample Men Women %yes %no Total P value %yes %no Total P value %yes %no Total P value Age (years) 0.002 0.240 <0.001 Open access  18–24 10.8 89.2 859 7.0 93.0 429 14.7 85.3 430  25–29 9.1 90.9 696 8.4 91.6 347 9.7 90.3 349  30–35 6.2 93.8 919 5.4 94.6 465 7.0 93.0 454 Sex 0.001  Man 6.8 93.2 1241 6.8 93.2 1241 __ __ __  Woman 10.5 89.5 1233 __ __ __ 10.5 89.5 1233 Country of birth 0.027 0.107 0.189  Other country 12.0 88.0 301 10.2 89.8 128 13.3 86.7 173  Spain 8.1 91.9 2173 6.4 93.6 1113 10.0 90.0 1060 Level of studies (n.a=22) 0.427 0.761 0.646  Up to secondary 8.1 91.9 760 6.4 93.6 424 10.3 89.7 336  Higher education 8.9 91.1 1692 6.8 93.2 806 10.7 89.3 886 Sexual attraction (n.a=33) <0.001 <0.001 <0.001  Lesbian 0.0 100.0 26 __ __ __ 0.0 100.0 26  Gay 14.2 85.8 120 14.2 85.8 120 __ __ __  Bisexual 15.1 84.9 451 10.6 89.4 160 17.5 82.5 291  Heterosexual 6.7 93.3 1844 5.2 94.8 944 8.3 91.7 900 Paid work in the last 12 months 0.129 0.418 0.155 (n.a=15)  No 7.0 93.0 530 5.6 94.4 251 8.2 91.8 279  Yes 9.1 90.9 1929 7.0 93.0 984 11.2 88.8 945 Currently has a partner (n.a=46) 0.228 0.350 0.576  Yes 8.0 92.0 1628 5.9 94.1 763 9.8 90.2 865  No 10.8 89.2 546 9.0 91.0 289 12.8 87.2 257  I’ve never had a partner 7.9 92.1 254 6.7 93.3 165 10.1 89.9 89 Currently cohabitates (n.a=62) 0.026 0.283 0.01  No 9.7 90.3 679 5.8 94.2 327 13.4 86.6 352  Yes 6.6 93.4 933 5.8 94.2 432 7.4 92.6 501  Never had a partner/I don’t 9.88 90.1 800 8.1 91.9 454 12.1 87.9 346 have a partner Sanz-­Barbero B, et al. BMJ Open 2021;11:e055227. doi:10.1136/bmjopen-2021-055227 Continued Open access (homosexual: 14.2% and bisexual: 10.6%) and in women with bisexual attraction (17.5%). P value 0.231 The prevalence of SV12m identified in this study is greater than that shown in the last Macro Survey on Violence Against Women carried out in Spain.7 This greater prevalence, even though our study period Total 922 201 89 1233 included men and women and the lockdown period, in which there was a reduction in SV, could be related to the use of a different methodology for retrieving infor- mation. While both questionnaires used the same ques- 88.6 93.5 89.9 89.5 %no tions to collect exposure to SV, collection in the Macro Survey cited above took place through computer-­assisted personal interviews. In our study, the people surveyed Women carried out the survey online through a link sent through %yes 11.4 6.5 10.1 10.5 email, thus, they had the opportunity to complete the questionnaire in complete privacy. Also, it is possible that the recognition and declaration of SV among young P value 0.886 people in Spain has increased in recent times given the greater visibility of SV in recent years.22 This work suggests for the first time using empirical data that the prevalence of SV victimisation could have Total 918 135 165 1241 Have you experienced sexual violence in the past 12 months? (n.a=41) decreased in Spain during the lockdown period, though its severity could have increased. This decrease, also observed in other countries through the decrease in the use of emergency services for sexual aggression,23 could %no 93.4 91.9 93.3 93.2 reflect the decrease in SV during this period outside of cohabitation situations. In this sense, the study shows that the behaviours related to ‘unwanted touching of intimate parts’, which frequently occurs in public spaces, %yes Men 6.6 8.1 6.7 6.8 seem to have decreased during the COVID-­19 lockdown period. However, unwanted sexual intercourses probably increased during the lockdown period. Along the same P value 0.644 lines of argument, the probable greater severity of SV identified in this period could be due to the exposure to Sex-­ IPV between cohabiting partners. Even though studies using empirical data during this recent period Total 1840 336 254 2474 are scared, the current evidence suggests that, during this period both in Spain24 as well as in other European coun- tries, there was an increase in calls to helpline centres for victims of gender violence, while there was also a decrease in formal complaints.25 Studies published by Gosangi 91.0 92.9 92.1 91.4 %no Values in bold mean statistically significant associations. show greater severity of injuries in women attended at Total sample gender violence emergency service centres.26 Our results allow us to partially accept the hypothesis %yes raised in this work. Some risk factors, such as belonging 9.02 7.14 7.87 8.5 to sexual minorities, remained associated with SV in all the periods studied. Other factors, such as sex, did not show significant associations with SV during and after the Couple—one/both—has minor confinement period.  I’ve never had a partner The greatest SV risk was observed among women before the lockdown, but not during the lockdown. This Continued could be explained by the decrease in the high level of SV children (n.a=44) experienced by women beyond the SV experienced in the family environment.5 7 n.a, no answer. Related to sexual attraction, our results highlight how Table 1 people with lesbian, gay or bisexual (LGB) attraction  Yes  No Total have a higher prevalence and probability of SV victi- misation in all of the periods analysed; this association Sanz-­Barbero B, et al. BMJ Open 2021;11:e055227. doi:10.1136/bmjopen-2021-055227 5 6 Table 2 Prevalence of sexual violence (SV) victimisation by period related to the COVID-­19 lockdown, SV in Young People Survey, 2020 Have you experienced SV in the last 12 months before the lockdown? Have you experienced SV during the lockdown? Have you experienced SV after the lockdown? Yes % Yes No % No P value Yes %Yes No % No P value Yes % Yes No % No P value Total Open access Age (years) 0.018 0.974 0.114  18–24 68 8.0 784 92.0 16 1.9 833 98.1 22 2.6 829 97.4 851  25–29 45 6.5 648 93.5 12 1.7 680 98.3 21 3.0 671 97.0 692  30–35 43 4.7 872 95.3 17 1.9 898 98.1 14 1.5 901 98.5 915 Sex <0.001 0.472 0.329  Man 55 4.5 1180 95.5 25 2.0 1209 98.0 25 2.0 1210 98.0 1235  Women 101 8.2 1124 91.8 20 1.6 1202 98.4 32 2.6 1191 97.4 1223 Country of birth 0.075 0.478 0.037  Other country 26 8.7 273 91.3 7 2.3 291 97.7 12 4.0 286 96.0 298  Spain 130 6.0 2031 94.0 38 1.8 2120 98.2 45 2.1 2115 97.9 2160 Level of studies 0.224 0.387 0.631 (n.a=22)  Up to 41 5.4 715 94.6 11 1.5 743 98.5 19 2.5 736 97.5 755 complete secondary  Higher 113 6.7 1569 93.3 33 2.0 1647 98.0 37 2.2 1644 97.8 1681 education Sexual attraction <0.001 0.004 <0.001 (n.a=33)  Lesbian 0 0.0 26 100.0 0 0.0 26 100.0 0 0.0 26 100.0 26  Gay 10 8.4 109 91.6 3 2.5 116 97.5 6 5.0 113 91.6 119  Bisexual 53 11.8 395 88.2 17 3.8 430 96.2 20 4.5 427 88.2 447  Hetero 91 5.0 1743 95.0 24 1.3 1807 98.7 30 1.6 1803 95.0 1833 Currently has a 0.426 0.778 0.114 partner (n.a=46)  Yes 98 6.1 1519 93.9 31 1.9 1582 98.1 32 2.0 1583 98.0 1615  No 44 8.1 500 91.9 10 1.8 534 98.2 19 3.5 525 96.5 544  I’ve never had 11 4.3 242 95.7 4 1.6 249 98.4 5 2.0 248 98.0 253 a partner Currently 0.029 0.264 0.021 cohabitates with partner (n.a=62)  No 51 7.6 624 92.4 10 1.5 662 98.5 18 2.7 656 97.3 674  Yes 46 5.0 880 95.0 21 2.3 904 97.7 13 1.4 912 98.6 925 Continued Sanz-­Barbero B, et al. BMJ Open 2021;11:e055227. doi:10.1136/bmjopen-2021-055227 Open access was independent of the rest of the variables. Given the social isolation that took place during the lockdown, it is possible that the prevalence of SV in the LGB popula- Total 1827 2458 797 334 253 tion compared with the heterosexual population in this period, reflects Sex-­IPV. According to the Minority Stress Model,27 people who belong to a minority group are Have you experienced SV after the lockdown? P value 0.702 exposed to excess psychosocial stress due to the fact that they belong to a socially stigmatised and marginalised group. This model distinguishes between external and internal stressors. In the specific case of sexual minori- 97.68 % No 97.0 97.5 98.2 98.0 ties, social homophobia is among the external types of stressors and results in experiences of violence, assault 1782 and discrimination. Internal stressors include internal- 2401 ised homophobia, defined as the level at which people 773 328 248 No who belong to a sexual minority group internalise and integrate into their own identities negative feelings, atti- % Yes tudes, beliefs and behaviours related to their homosex- 3.0 2.5 1.8 2.0 2.3 uality.28 Various studies have suggested that internalised homophobia could be related to greater risk of victimi- Yes sation and perpetration of violence29 towards people of 24 45 57 6 5 the same sexual orientation. Higher levels of internalised Have you experienced SV during the lockdown? P value homophobia have been associated with higher levels of 0.308 IPV in couples of the same sex.28 Bisexual people have been described in the literature as a doubly stigmatised group, exposed to discrimination % No 98.2 98.3 97.0 98.4 98.2 by homosexuals and by heterosexuals.30 In our study, the small sample size and low prevalence of non-­heterosexual people did not allow us to analyse different sexual minori- ties independently. However, at the descriptive level, we 1794 2411 783 324 249 No did observed a higher prevalence of SV among bisexual people, primarily among bisexual women (17.5%). Although there is scarce literature that characterises %Yes perpertrators of sexual violence towards homosexual and 1.8 1.7 3.0 1.6 1.8 bisexual people, the study carried out by Mellins31 shows that 97% of bisexual women and 75% of lesbian women Yes who are sexually assaulted were assaulted by a man. It 14 31 10 45 4 is possible that the greater SV victimisation by bisexual women is due to belonging to two discriminated groups, Have you experienced SV in the last 12 months P value 0.519 that is, to say, due to their being women and being a sexual minority. Some authors suggest that the scarcity of empirical data related to bisexuality has contributed to bisexuality % No 93.1 93.3 93.7 95.7 93.7 Values in bold mean statistically significant associations. being socially falsely stereotyped as a ‘phase’ or ‘fictitious’ sexual identity.30 In this sense, the studies carried out by Friedman et al32 found that 14.4% of undergraduate 1707 2304 students affirmed that ‘bisexuality was not a valid sexual before the lockdown? 742 313 242 No orientation’. Other studies show that bisexual people are perceived as confused about their sexual orientation.33 % Yes 6.9 6.7 6.3 4.3 6.3 Even though research is still limited, there are authors who suggest that bisexual people experience internalised biphobia that could contribute to the risk of IPV in this Continued Yes sexual minority.34 122 156  I’ve never had/ 55 21  I’ve never had 11 The greater probability of women born outside of Spain do not have a of experiencing SV is supported by a both national35 and n.a, no answer. minor children couple have One or both members of international literature.36 Immigrant women in Spain a partner partner have a greater risk of IPV,35 and being killed as a results Table 2 (n.a=44)  Yes of IPV.37 The greater gender inequality in countries of  No Total origin,38 economic dependence, lack of social support, Sanz-­Barbero B, et al. BMJ Open 2021;11:e055227. doi:10.1136/bmjopen-2021-055227 7 8 Table 3 Frequency of violent behaviours between people who experienced sexual violence in the past 12 months, by sex2 Full sample Men Women Open access Lockdown period Lockdown period Lockdown period VS 12m Before During After VS 12m Before During After VS 12m Before During After Sexual violence behaviours (n=213) (n=156) (n=45) (n=57) (n=86) (n=55) (n=25) (n=25) (n=129) (n=101) (n=20) (n=32) 1.-Have you been obligated to engage 6.1 5.8 11.1 5.3 9.3 9.1 20.0 12.0 3.9 4.0 0.0 0.0 in sexual conduct through threats, physical force or harm in some way? 2.-Have you been obligated to engage 10.8 7.1 15.6 14.0 11.6 5.5 20.0 20.0 10.1 7.9 10.0 9.4 in sexual conduct because of your inability to refuse due to being under the influence of alcohol or drugs? 3.-Have you engaged in sex you did not 19.2 19.2 20.0 21.1 16.3 12.7 8.0 28.0 20.9 22.8 35.0 15.6 want due to fear of the consequences if you said no? 4.-Have you been obligated to engage 20.2 17.9 22.2 21.1 18.6 14.5 16.0 28.0 20.9 19.8 30.0 15.6 in sexual conduct when you did not want to? 5.-Has anyone tried, and failed, to 23.5 22.4 17.8 14.0 25.6 21.8 20.0 16.0 21.7 22.8 15.0 15.6 obligate you to have sex against your wishes? 6.-Has anyone touched your intimate 69.5 73.7 44.4 50.9 61.6 69.1 48.0 48.0 73.6 76.2 40.0 53.1 parts—genitals or chest—or has anyone touched you in a sexual way when you did not want it? 7.-Has anyone obligated you to touch 15.5 10.9 20.0 14.0 12.8 7.3 20.0 12.0 17.1 12.9 20.0 15.6 another person in his/her intimate parts—genitals or chest—or have you been forced to engage in sexual touching when you did not want to? 8.-Has anyone obligated you to engage 8.5 7.1 8.9 3.5 15.1 14.5 16.0 8.0 3.9 3.0 0.0 0.0 in a sexual practice other than those mentioned here? 1–4.-Unwanted sexual conduct with 39.4 37.8 64.4 54.4 36.0 32.7 56.0 56.0 41.1 40.6 75.0 53.1 penetration Sexual Violence in Young People Survey, 2020. Sanz-­Barbero B, et al. BMJ Open 2021;11:e055227. doi:10.1136/bmjopen-2021-055227 Table 4 Variables associated with sexual violence in the past 12 months, before, during and after the lockdown, unadjusted analysis, robust Poisson regression In the 12 months prior to the Last 12 months lockdown During the lockdown After the lockdown Variables PR 95% CI P value PR 95% CI P value PR 95% CI P value PR 95% CI P value Sex (ref: man)  Woman 1.55 1.19 to 2.01 0.001 1.85 1.35 to 2.55 0.000 0.81 0.45 to 1.45 0.473 1.29 0.77 to 2.17 0.331 Age 0.96 0.93 to 0.98 0.002 0.96 0.93 to 0.99 0.009 1.01 0.95 to 1.08 0.653 0.98 0.93 to 1.03 0.368 Country of birth (ref: Spain)  Outside of Spain 1.47 1.05 to 2.06 0.026 1.45 0.97 to 2.16 0.073 1.33 0.60 to 2.96 0.479 1.93 1.03 to 3.61 0.039 Sexual attraction (ref: heterosexual)  Lesbian/Gay/ 2.12 1.63 to 2.75 <0.001 2.14 1.57 to 2.91 <0.001 2.58 1.43 to 4.63 0.002 2.68 1.60 to 4.50 <0.001 Bisexual Sanz-­Barbero B, et al. BMJ Open 2021;11:e055227. doi:10.1136/bmjopen-2021-055227 Level of studies (ref: up to secondary study completion)  Higher education 1.12 0.84 to 1.50 0.428 1.24 0.88 to 1.75 0.226 1.35 0.68 to 2.65 0.389 0.87 0.51 to 1.51 0.631 Paid work in the last 12 months (ref:no)  Si 1.30 0.92 to 1.83 0.133 Cohabitation (ref: yes)  No 1.48 1.11 to 1.96 0.008 1.45 1.04 to 2.03 0.030 0.72 0.40 to 1.29 0.266 2.03 1.10 to 3.76 0.024 Minor children, own and/or those of the partner (ref: no)  Yes 0.80 0.53 to 1.21 0.297 0.98 0.63 to 1.54 0.944 1.78 0.89 to 3.56 0.104 0.75 0.32 to 1.73 0.496 Sexual Violence in Young People Survey, 2020. Values in bold mean statistically significant associations. PR, prevalence ratio.; Open access 9 10 Open access Table 5 Variables associates with sexual violence in the last 12 months, before, during and after the lockdown, multivariate analysis, robust Poisson regression Last 12 months In the 12 months prior to the lockdown During the lockdown After the lockdown Variable PR 95% CI P value PR 95% CI P value PR 95% CI P value PR 95% CI P value Sex (ref: man)  Woman 1.54 1.17 to 2.03 0.002 1.77 1.27 to 2.46 0.001 0.70 0.38 to 1.30 0.263 1.32 0.77 to 2.28 0.311 Age 0.96 0.93 to 0.99 0.037 0.97 0.92 to 1.01 0.115 1.00 0.92 to 1.08 0.928 1.01 0.95 to 1.08 0.679 County of birth (ref: Spain)  Outside of 1.54 1.07 to 2.19 0.020 1.51 0.99 to 2.30 0.059 1.40 0.62 to 3.14 0.415 1.83 0.94 to 3.57 0.077 Spain Sexual attraction (ref: heterosexual)  LGB 1.98 1.52 to 2.59 <0.001 2.01 1.46 to 2.76 <0.001 2.63 1.42 to 4.88 0.002 2.67 1.54 to 4.61 <0.001 Level of studies (ref: up to completion of secondary studies)  Higher 1.16 0.84 to 1.60 0.380 1.45 0.99 to 2.13 0.055 1.34 0.65 to 2.77 0.431 0.96 0.53 to 1.74 0.883 education Paid work in the last 12 months (ref: no)  Yes 1.66 1.13 to 2.45 0.010 -- -- -- -- -- -- -- -- -- Cohabitation (ref: yes)  No 1.34 0.93 to 1.94 0.119 1.41 0.99 to 2.20 0.133 0.76 0.34 to 1.71 0.512 2.16 1.00 to 4.64 0.049 Minor children, one's own and/or those of the partner (ref: no)  Yes 1.06 0.65 to 1.74 0.817 1.34 0.76 to 2.36 0.307 1.95 0.83 to 4.59 0.128 1.24 0.48 to 3.21 0.652 Sexual Violence in Young People Survey, 2020. Values in bold mean statistically significant associations. LGB, Lesbian/Gay/Bisexual; PR, prevalence ratio.; Sanz-­Barbero B, et al. BMJ Open 2021;11:e055227. doi:10.1136/bmjopen-2021-055227 Open access presence of minor children in the home and/or barriers Contributors BS-­B: Guarantor of the study. Initiated the collaborative project, to accessing formal resources make it difficult to escape designed data collection tools, monitored data collection, designed statistical analysis plan, cleaned and analysed the data, wrote the manuscript draft and from IPV and aggravate situations in which immigrant edited the manuscript. CS: Participated in the design project, participated in the women are exposed.10 Contrary to expectations, we have statistical analysis design and in the interpretation of data, revised the paper and not identified significant associations between the place approved the final version of the manuscript. LS: Participated in the design project, participated in the discussion of the results, revised the paper and approved the of origin and the victim in the different periods studied. It final version of the manuscript. GR: Participated in the design project, designed data is possible that this lack of association could be explained collection tools, revised and discuss the results, revised the paper and approved the by the lack of precision associated with the sample size. final version of the manuscript. LV-­LC Participated in the design project, participated It is in this sense to point out that although the adjusted in the statistical analysis design, revised and discuss the results, revised the paper and approved the final version of the manuscript. LO-­G Participated in the design prevalence ratios of this variable are not significant, its project, revised and discuss the results, revised the paper based on intellectual value is above the value 1 in all the periods analysed. contribution and approved the final version of the manuscript. MJL: Participated Even though our study shows novel results that advance in the design project, designed data collection tools, participated in the statistical the knowledge of SV, and that our sample includes quotas analysis design, revised and discuss the results, revised the manuscript and approved the final version of the manuscript. GP: Participated in the design project, by sex, age, region and country of origin, in order to participated in the statistical analysis design, revised and discuss the results, represent the population between 18 and 35 years old participated in reviewing, editing and revising the manuscript and approved the residing in Spain the study results should be interpreted final version of the manuscript. CV-­C: Initiated the collaborative project, designed considering its limitations. The sample size of this study, data collection tools, participated in the discussion of the results and revised the paper based on intellectual contribution and approved the final version of the the analysis of concrete time periods, and the low prev- manuscript. alence of SV during the time periods considered mean Funding This research was supported by the CIBER of Epidemiology and Public that these results should be interpreted with caution. In Health of Spain; Grant ESP20PI02. this article, we analyse the prevalence of SV during the Disclaimer This article presents independent results and research. The views last 12 months, and in the 12 months prior, during and expressed are those of the authors and not necessarily those of the Instituto de after the lockdown. These periods are not exclusive, so we Salud Carlos III. cannot compare the periods to each other. Despite this Competing interests None declared. limitation, the results obtained in this study are consistent Patient consent for publication Not applicable. with the results published in previous scientific literature. Ethics approval This study was approved by the Ethics Committee of Alicante The cross-­sectional study design did not allow us to estab- University (ref. UA-­2020-­07-­07). lish a causal relationship, but it did provide the ability to Provenance and peer review Not commissioned; externally peer reviewed. estimate prevalences. In this sense, the sample design, as Data availability statement Data are available on reasonable request. well as the wide range of the panel used, provided the Supplemental material This content has been supplied by the author(s). It has opportunity to approach the exposure of young people not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been in Spain to SV. peer-­reviewed. Any opinions or recommendations discussed are solely those The prevalence of SV observed in Spanish young of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content people is alarming. While COVID-­19 lockdown measures includes any translated material, BMJ does not warrant the accuracy and reliability decreased SV events that occurred outside of cohabita- of the translations (including but not limited to local regulations, clinical guidelines, tion situations, the exposure to unwanted intercourses terminology, drug names and drug dosages), and is not responsible for any error increased during the lockdown. The groups with greater and/or omissions arising from translation and adaptation or otherwise. exposure to SV included women, those born outside of Open access This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-­NC 4.0) license, which Spain and those belonging to the LGB sexual minority. permits others to distribute, remix, adapt, build upon this work non-­commercially, In terms of this last group, more studies are needed that and license their derivative works on different terms, provided the original work is use scales adapted to the different forms of communica- properly cited, appropriate credit is given, any changes made indicated, and the use tion and relation established in sexual encounters among is non-­commercial. See: http://​creativecommons.​org/​licenses/​by-​nc/​4.​0/. different groups.39 The development of more effective ORCID iDs strategies for prevention of SV in young people requires Belén Sanz-­Barbero http://​orcid.​org/​0000-​0001-​6030-​061X a consideration of the social inequalities in sex, in sexual Carme Saurina http://​orcid.​org/​0000-​0003-​4549-​8845 Laura Serra http://​orcid.​org/​0000-​0002-​8835-​6890 orientation and in country of birth that were observed in Gemma Renart Vicens http://​orcid.​org/​0000-​0001-​6037-​9225 this study. Laura Vall-­Llosera Casanovas http://​orcid.​org/​0000-​0001-​7059-​1596 Laura Otero-­García http://​orcid.​org/​0000-​0002-​0171-​2806 Author affiliations María José López http://​orcid.​org/​0000-​0003-​4402-​2239 1 CIBERESP, Madrid, Spain Carmen Vives-­Cases http://​orcid.​org/​0000-​0002-​6797-​5051 2 National School of Public Health, Instituto de Salud Carlos III, Madrid, Spain 3 Research Group on Statistics, Econometrics and Health (GRECS), University of Girona, Girona, Spain 4 Nursing Department. Falculty of Medicine, Universidad Autónoma de Madrid, Madrid, Spain REFERENCES 5 Agència de Salut Pública de Barcelona, Barcelona, Spain 1 World Health Organization/London School of Hygiene and Tropical 6 Medicine. Preventing intimate partner and sexual violence against IIB Sant Pau, Barcelona, Spain women: taking action and generating evidence. Geneva: World 7 Department of Community Nursing, Preventive Medicine and Public Health and Health Organization, 2010. https://www.​who.​int/​violence_​injury_​ History of Science, Alicante University, Alicante, Spain prevention/​publications/​violence/​9789241564007_​eng.​pdf Sanz-­Barbero B, et al. BMJ Open 2021;11:e055227. doi:10.1136/bmjopen-2021-055227 11 Open access 2 Breiding MJ, Smith SG, Basile KC, et al. Prevalence and Available: https://​violenciagenero.​igualdad.​gob.​es/​violenciaEnCifras/​ characteristics of sexual violence, stalking, and intimate partner estudios/​investigaciones/​2021/​pdfs/​Estudio_​Viol​enci​aEnA​dole​ violence victimization--national intimate partner and sexual violence scencia.​pdf [Accessed 17 Jun 2021]. survey, United States, 2011. MMWR Surveill Summ 2014;63:1–18. 23 Muldoon KA, Denize KM, Talarico R, et al. COVID-­19 pandemic and 3 Forsman RL. Prevalence of sexual assault victimization among violence: rising risks and decreasing urgent care-­seeking for sexual College men, aged 18-­24: a review. J Evid Inf Soc Work assault and domestic violence survivors. BMC Med 2021;19:20. 2017;14:421–32. 24 Rodriguez-­Jimenez R, Fares-­Otero NE, García-­Fernández L. 4 World Health Organization. Global and regional estimates of Gender-­Based violence during COVID-­19 outbreak in Spain. Psychol violence against women: prevalence and health effects of intimate Med;7:1–2. partner violence and nonpartner sexual violence. ISBN: 978 92 4 25 Godin M. France to house domestic abuse victims in hotels amid. 156462 5, 2013. Available: https://www.​who.​int/​publications/​i/​item/​ time, 2020. Available: https://​time.​com/​5812990/​france-​domestic-​ 9789241564625 [Accessed 8 Jun 2021]. violence-​hotel-​coronavirus/ [Accessed 17 Jun 2021]. 5 European Union Agency for Fundamental Rights. European Union 26 Gosangi B, Park H, Thomas R, et al. Exacerbation of physical agency for fundamental rights: violence against women survey, 2012. intimate partner violence during COVID-­19 pandemic. Radiology Available: https://​fra.​europa.​eu/​sites/​default/​files/​fra-​2014-​vaw-​ 2021;298:E38–45. survey-​at-​a-​glance-​oct14_​en.​pdf [Accessed 5 Jun 2021]. 27 Meyer IH, Prejudice MIH. Prejudice, social stress, and mental health 6 Wincentak K, Connolly J, Card N. Teen dating violence: a meta-­ in lesbian, gay, and bisexual populations: conceptual issues and analytic review of prevalence rates. Psychol Violence 2017;7:224–41. research evidence. Psychol Bull 2003;129:674–97. 7 Ministry of Equality. Government delegation against gender 28 Badenes-­Ribera L, Sánchez-­Meca J, Longobardi C. The relationship violence. Macro-­survey on violence against women 2019. Available: https://​violenciagenero.​igualdad.​gob.​es/​violenciaEnCifras/​ between internalized homophobia and intimate partner violence in macroencuesta2015/​pdf/​Macroencuesta_​2019_​estudio_​ same-­sex relationships: a meta-­analysis. Trauma Violence Abuse investigacion.​pdf [Accessed 17 Jun 2021]. 2019;20:331–43. 8 Olshen E, McVeigh KH, Wunsch-­Hitzig RA, et al. Dating violence, 29 Carvalho AF, Lewis RJ, Derlega VJ, et al. Internalized sexual minority sexual assault, and suicide attempts among urban teenagers. Arch stressors and same-­sex intimate partner violence. J Fam Violence Pediatr Adolesc Med 2007;161:539–45. 2011;26:501–9. 9 Sanz-­Barbero B, Barón N, Vives-­Cases C. Prevalence, associated 30 Turell SC, Brown M, Herrmann M. Disproportionately high: an factors and health impact of intimate partner violence against women exploration of intimate partner violence prevalence rates for bisexual in different life stages. PLoS One 2019;14:e0221049. people. Sexual and Relationship Therapy 2018;33:113–31. 10 Sanz-­Barbero B, Briones-­Vozmediano E, Otero-­García L, et al. 31 Mellins CA, Walsh K, Sarvet AL, et al. Sexual assault incidents Spanish intimate partner violence survivors help-­seeking strategies among College undergraduates: prevalence and factors associated across the life span. J Interpers Violence 2020;8:886260520976213. with risk. PLoS One 2017;12:e0186471. 11 Lundgren R, Amin A. Addressing IPV and SV among adolescents: 32 Friedman MR, Dodge B, Schick V, et al. From bias to bisexual health emerging evidence of effectiveness. J Adolesc Health disparities: attitudes toward bisexual men and women in the United 2015;56:S42–50. States. LGBT Health 2014;1:309–18. 12 Mujal GN, Taylor ME, Fry JL, et al. A systematic review of bystander 33 Higgins A, Doyle L, Downes C. The LGBT Ireland report: national interventions for the prevention of sexual violence. Trauma Violence study of the mental health and wellbeing of lesbian, gay, bisexual, Abuse 2021;22:381–96. transgender and intersex people in Ireland. GLEN and belong to, 13 Koepke S, Eyssel F, Bohner G. "She deserved it": Effects of sexism 2016. Available: http://​bit.​ly/​2xvY9xi [Accessed 17 Jun 2021]. norms, type of violence, and victim's pre-­assault behavior on blame 34 Rollè L, Giardina G, Caldarera AM, et al. When intimate partner attributions toward female victims and approval of the aggressor's violence meets same sex couples: a review of same sex intimate behavior. Violence Against Women 2014;20:446–64. partner violence. Front Psychol 2018;9:1506. 14 Tarzia L, Tyler M. Recognizing connections between intimate 35 Sanz-­Barbero B, Vives-­Cases C, Otero-­García L, et al. Intimate partner sexual violence and pornography. Violence Against Women partner violence among women in Spain: the impact of regional-­ 2021;27:2687-­2708. level male unemployment and income inequality. Eur J Public Health 15 Tarzia L. Toward an ecological understanding of intimate partner 2015;25:1105–11. sexual violence. J Interpers Violence 2020:088626051990029. 36 Sanz-­Barbero B, López Pereira P, Barrio G, et al. Intimate partner 16 Jennings WG, Okeem C, Piquero AR, et al. Dating and intimate violence against young women: prevalence and associated factors in partner violence among young persons ages 15–30: evidence from a Europe. J Epidemiol Community Health 2018;72:611–6. systematic review. Aggress Violent Behav 2017;33:107–25. 37 Sanz-­Barbero B, Heras-­Mosterio J, Otero-­García L. Perfil 17 Draft organic law for the comprehensive guarantee of sexual sociodemográfico del feminicidio en España y su relación con las freedom. Available: https://www.​igualdad.​gob.​es/​normativa/​ normativa-​en-​tramitacion/​Documents/​APLOGILSV2.​pdf [Accessed denuncias por violencia de pareja [Sociodemographic profile of 17 Jun 2021]. femicide in Spain and its association with domestic abuse reporting]. 18 Rabby MII, Hossain F, Akhi IJ, et al. Sexual violence against COVID-­ Gac Sanit 2016;30:272–8. 19-­Positive girls in India. Asia Pac J Public Health 2021;33:316–7. 38 World Economic Forum. Global gender gap index 2015. Available: 19 Dosdale C, Skarparis K. Supporting survivors of sexual violence and http://​reports.​weforum.​org/​global-​gender-​gap-​report-​2015/​the-​ abuse during the COVID-­19 pandemic. Br J Nurs 2020;29:1159–63. global-​gender-​gap-​index-​2015/ [Accessed 17 Jun 2021]. 20 Johnson K, Green L, Volpellier M, et al. The impact of COVID-­19 on 39 Fernández-­Dávila P. ¿Se puede hablar realmente de actos de services for people affected by sexual and gender-­based violence. “violencia sexual” en los contextos de chemsex? Reflexiones Int J Gynaecol Obstet 2020;150:285–7. desde los entendimientos de los hombres que practican ChemSex 21 United Nations. Guidelines for producing statistics on violence y la cultura sexual gay. [Can we really talk about "sexual violence" against women statistical surveys. United nations, 2014. in the chemsex scene? Reflections from the chemsex users’ 22 Ministry of Equality. Government delegation against gender violence. understandings and the gay sex culture]. Health and Addictions The situation of violence against women in adolescence in Spain. 2021;21:124–37. 12 Sanz-­Barbero B, et al. BMJ Open 2021;11:e055227. doi:10.1136/bmjopen-2021-055227