Open access Protocol Adaptability to acute stress among women survivors of intimate partner violence: protocol for a mixed-­methods cross-­sectional study in a laboratory setting (BRAW study) Ximena Goldberg ‍ ‍,1 Carme Espelt,1 Diego Palao,1 Roser Nadal,2 Antonio Armario3 To cite: Goldberg X, Espelt C, ABSTRACT Palao D, et al. Adaptability to Strengths and limitations of this study Introduction Intimate partner violence (IPV) is the most acute stress among women common and alarming form of violence against women, survivors of intimate partner ►► The design combines biological and psychosocial affecting around 30% of all women around the world. violence: protocol for a mixed-­ data in an integrative model of the health conse- Using an integrative methodology, we approach IPV as a methods cross-­sectional quences of intimate partner violence. study in a laboratory setting form of chronic exposure to severe stress that alters the ►► A mixed-­methods approach allows to incorporate (BRAW study). BMJ Open stress-­response system of exposed women. The aim of the voices of the survivors in the research process. 2020;10:e036561. doi:10.1136/ this study is to test the hypothesis that sustained exposure ►► Stress reactivity is measured in the laboratory using bmjopen-2019-036561 to IPV in women confers a vulnerability-­to-­stress profile the gold standard Trier Social Stress Task. characterised by higher neuroendocrine and behavioural ►► Prepublication history and ►► The limited sample size will prevent exploring possi- responsiveness associated with a selective attentional additional material for this ble modulatory effects of relevant variables (ie, other paper are available online. To processing bias towards threat. stressful experiences). view these files, please visit Methods and analysis Women between 21 and 50 years the journal online (http://​dx.​doi.​ old from the area of Barcelona (Spain) will be invited to org/​10.​1136/​bmjopen-​2019-​ participate. A sample of 82 women exposed to IPV and 41 036561). women not exposed to IPV will be included and assessed this issue as a global public health problem.1 for attentional bias and response to acute stress in a While women are exposed to several types of Received 23 December 2019 laboratory condition (the Trier Social Stress Task). The gender-­based violence, almost one in three Revised 07 July 2020 ever-­partnered women worldwide (30%) study will include quantitative and qualitative measures of Accepted 13 August 2020 cognitive performance, neuroendocrine activity and face-­ have experienced intimate partner violence to-­face interviews to obtain an integrative description of (IPV).2 This alarming reality applies to the the stress-­response profile of these women. Results are territory of the European Union (EU), where expected to help build resilience strategies with a long-­ a recent EU-­wide survey has estimated the lasting impression on women’s healthy functioning. prevalence of IPV to be 22%.3 It also reflects Ethics and dissemination The study has obtained the on the IPV estimates in Spain: 13% of women approval of the local Ethics Committee (‘Comité de Ética report suffering physical or sexual violence, de Investigación Parc Taulí de Sabadell’; 2 018 551 V.1.2 June 2018). Besides the communication of results in peer-­ and over 26.4% have suffered psychological reviewed papers and scientific congresses, the project and economic IPV.4 will inform guidelines and recommendations through Consistent evidence confirms that survivors policy-­dialogues and workshops with relevant regional and of IPV, compared with women who have not national representatives for future work and prevention suffered IPV, more frequently present chronic strategies. Participants will be invited to be an active diseases such as diabetes, chronic pain, © Author(s) (or their part in the dissemination strategy focussed on raising asthma or cardiovascular disease.5 6 Further- employer(s)) 2020. Re-­use awareness of coping limitations and abilities that women more, mental health-­ related consequences permitted under CC BY-­NC. No themselves will be able to identify throughout the study. commercial re-­use. See rights are largely common among IPV survivors.7 Trial registration details The study has been registered and permissions. Published by It has been estimated that up to 28% of the BMJ. at the ​ClinicalTrails.​gov database (Identifier number: cases of depression can be attributed to life- NCT03623555; Pre-results). For numbered affiliations see time exposure to IPV.8 Only in the USA this end of article. translates into 1 million cases of depression Correspondence to INTRODUCTION per year that could be completely averted if Dr Ximena Goldberg; The burden and prevalence of violence against no women had been exposed to IPV. In this ​xlgoldberg@​tauli.​cat women and girls worldwide has helped raise regard, the possibility to suffer traumatic Goldberg X, et al. BMJ Open 2020;10:e036561. doi:10.1136/bmjopen-2019-036561 1 Open access brain injury has to be considered as this is frequently asso- emotional arousal) and whether this is reflected in an ciated with mental health problems, mostly depression.9 altered HPA responsiveness. The activation of the stress Also relevant to the present study, women survivors of IPV system in response to novel stressful situations is a central are more likely to report a history of childhood trauma, matter as it reflects the person’s capacity to respond to which per se has long-­ term consequences in mental the changing demands that commonly occur at work and health.10 11 In Europe, almost one-­third of the women at home. For example, a job interview could be a stressful exposed to sexual IPV report having experienced sexual circumstance that affected women may have to face after victimisation during childhood.3 A recent multi-­country recovering from IPV. The performance during the inter- study has reported ORs ranging from 1.41 to 3.8 in six out view (ie, getting or losing the job opportunity) will largely of seven study sites.12 depend on the current person’s neuroendocrine and This background clearly supports a relationship behavioural vulnerability to emotionally stressful situa- between IPV and disease. But the biological correlates tions or, on the contrary, on the successful, resilient strat- of such association are still unclear. The most widely egies women may present to cope with acute stress. acknowledged theoretical approach proposes a model of A history of chronic stress can alter responsiveness to chronic stress. Indeed, the victims of IPV are repetitively further acute stressors in a way not predicted by basal exposed to physical, emotional and sexual violence in HPA activity. These alterations can be either sensitisation the context of an intimate relationship that may last for or blunting.19 In this regard, experimental findings indi- years.13 Each single exposure to IPV is expected to trigger cate that both acute and chronic exposure of adult rats the typical neurobiological response to stressful stimuli, to severe stress induced HPA cross-­sensitisation; that is, which includes a wide range of neural and peripheral enhanced response to new acute stressors.20 21 Human stress biological responses, including the activation of studies focussed on the consequences of exposure to the sympathetic-­ adrenomedullary and hypothalamic-­ chronic stress also adopt this perspective. Evidence pituitary-­ adrenal (HPA) axes,14 and the subsequent suggests that the chronic stress exposure implicated in release of catecholamines and cortisol, respectively. caregiving is associated with dysfunctional psychosocial When this stress response presents frequently or persists behaviour related to maladaptive coping strategies when continuously for an extended period of time, the mecha- facing novel stressful circumstances. A key phenomenon nisms that were initially activated to cope with acute stress in this respect is selective attentional processing,22 which extend over time and can eventually lead to pathophysi- refers to an attentional bias early on the process of the ology and psychiatric diseases.15 However, the impact of cognitive approach to any given situation that predis- chronic stress on basal cortisol secretion appears to be poses a person to an enhanced vigilance for threat. This dependent on several critical factors, including the time bias towards threat has been linked to being a victim of since stress onset, the type of challenge and the possibility interpersonal violence,23 and is associated with higher of control.16 HPA axis activity.24 Quite surprisingly, no study has yet Stress-­related biological dysregulations specific to IPV evaluated selective attentional bias in association to acute survivors have been addressed for the first time in a stress responsiveness among women exposed to IPV. recent systematic review.17 Authors reveal that this popu- In the present proposal we aim at identifying potential lation present flattened diurnal cortisol rhythm and an alterations in the adaptive response to acute stress among overall higher diurnal secretion—a pattern expected IPV-­exposed women using valid measures of neuroendo- after exposure to chronic stress. The extent to which crine and behavioural response to acute stress. We propose these changes persist once exposure to the situation has that these alterations persist in the long term even when finished is unclear. Interestingly, in IPV survivors that the exposure to IPV has ceased. Given the innovative developed post-­traumatic stress disorder, but not in those nature of this project, we will register the perceptions of without overt psychopathology, reduced baseline cortisol levels and enhanced suppression by dexamethasone has the participants regarding their experience of acute stress. been reported.18 In sum, previous literature consistently This qualitative information will allow the identification supports that the neurobiological mechanisms involved of variables that may be overlooked during the design of in the stress-­response system are altered following IPV. the study. Also, we will assess the long-­termed marks of Importantly, these alterations can be placed in the core of chronic stress on the global basal levels of cortisol in IPV the neurobiological pathogenesis of the associated health survivors using hair-­cortisol analysis, following previous disorders.16 studies on other stress-­related situations.25 26 Why our study? Objectives of the study Despite the critical importance of the stress-­ response The main (general) objective of our study is to compare system, most research on IPV has only evaluated basal the response to acute stress in a group of women with a activity of the HPA and other biological systems. This has history of IPV as opposed to women without such history left unanswered the question of whether women previ- from the same community. The specific objectives are: ously exposed to IPV present behavioural difficulties 1. To assess the neuroendocrine (HPA axis) and be- when coping with new emotional situations (ie, anxiety, havioural response to acute stress. 2 Goldberg X, et al. BMJ Open 2020;10:e036561. doi:10.1136/bmjopen-2019-036561 Open access 2. To study whether IPV women show selective attention- Because social media will be an important tool for al processing bias towards threat. recruitment, participants in the study are expected to 3. To identify the resilience strategies used by women in belong to areas in Catalonia that will exceed the catch terms of psychosocial schemes (ie, quantitative and zone of the reference Centre. In particular, we expect qualitative information) and neuroendocrine regula- participants based in the area of Barcelona, which covers tion to cope with acute stress and their relationship a population of approximately 1 620 000 inhabitants. with health status. 4. To examine global basal cortisol alterations in the group of IPV-­ exposed women using hair-­ cortisol METHODS AND ANALYSIS analyses. Study design and participants The main hypothesis of our study is that women We will use a mixed-­methods, cross-­sectional design in exposed to IPV, in contrast to a group of women without a compliance with the Good Reporting of a mixed methods history of IPV, will present a vulnerability-­to-­stress profile. study (GRAMMS) framework.27 The key element in The specific sub-­hypotheses of the study are: the study is the inclusion of the Trier Social Stress Task 1. The group of women exposed to IPV will present high- (TSST), which provides a laboratory setting to measure er neuroendocrine and behavioural responsiveness the neuroendocrine and behavioural responses to acute to stress than the non-­exposed group, as measured by stress. Quantitative data will be generated from various self-­reported behavioural scales and salivary cortisol, sources, including biological samples (hair, saliva) and respectively. clinical and behavioural scales that are described in detail 2. A selective attentional processing bias towards threat in the procedures section below. Qualitative data will be will be associated with a higher response to acute stress generated from semi-­structured interviews that will be in the group of women exposed to IPV, as opposed to used to register the participants’ impressions regarding the group of non-­exposed women. their experience during the TSST, with a focus on identi- 3. The performance and resilience strategies during fying emerging themes. The integration of these sources an acute stress task will be self-­perceived as poorer/ of data is the main justification for proposing a mixed-­ weaker among the group of women exposed to IPV methods approach. While the sample size for the quan- in contrast with the self-­perceptions of the group of titative analysis has been calculated during the design non-­exposed women, as assessed using semi-­structured phase of the study, the total number of semi-­structured interviews and quantitative scales. Weaker resilience interviews may vary. Interviews will continue until data strategies will be related to poorer health status. saturation has been reached in the analyses (ie, no new 4. Women exposed to IPV will present higher levels of emerging themes are identified from new interviews, hair cortisol indicating global basal cortisol alterations. expected 30 interviews28). The methods for each stage of the study are presented in more detail in the following sections. STUDY CONTEXT Criteria for the inclusion of women in the study will be Participants will be women from the general popula- mainly guided by their previous exposure to IPV, which tion who will be recruited through advertisements in will define two groups of participants: an IPV-­exposed the community and in social media. The interviews will group with an estimated sample size of 82 women, and take place in the facilities of Parc Taulí Foundation, the a non-­exposed IPV group with an estimated sample size research branch of the Parc Taulí Healthcare Corpora- of 41 women (see sample size calculation for further tion (Corporació Sanitària Parc Taulí, CSPT). Partici- details). The definition of exposure to IPV will follow the pants will not receive monetary compensation for their WHO guidelines: ‘IPV refers to any behaviour within an participation in the research. intimate relationship that causes physical, psychological CSPT is a public healthcare legal entity that manages and sexual harm to those in the relationship’29 and will the third-­level Parc Taulí Hospital along with primary include physical violence, sexual violence, emotional/ healthcare centres, diagnostic and emergency units, and psychological abuse and controlling behaviours. These several transversal services including sexual and repro- forms of violence will also follow WHO’s definitions30 ductive health programmes. It is the single healthcare including but not limited to: (1) Acts of physical violence: provider for the area of the Catalan Eastern Occidental the woman has been pushed, beaten up, choked or burnt Vallès, which counts a total population of close to 500 000 by an intimate partner; (2) sexual violence: any form of people. Parc Taulí Foundation (Fundació Parc Taulí, FPT) sexual coercion by the intimate partner; (3) emotional/ provides support to CSPT in areas of research, teaching psychological abuse: the woman has been humiliated, and innovation, and has been recognised as a University intimidated, threatened by her intimate partner; and (4) Institute affiliated to the Universitat Autònoma de Barce- controlling behaviours: the woman has been isolated, lona. FPT has a strong background in the promotion of monitored for her actions by her intimate partner, and healthy habits and social awareness in the community, restricted her access to financial resources, employment, and for the last decade has dedicated special efforts to education or medical care. In order to warrant chronic enhance this line of work in the area of mental health. exposure to stress as proposed in the rationale of the Goldberg X, et al. BMJ Open 2020;10:e036561. doi:10.1136/bmjopen-2019-036561 3 Open access Figure 1 Procedures of the study. Session 2 (S2) will be scheduled 2 weeks after session 1 (S1) and will take place in the afternoon. CD-­RISC,Connor-­Davidson Resilience Scale; COPE, COPE Inventory; CTQ, Childhood Trauma Questionnaire; GHQ- 12, General Health Questionnaire; h, hours; MAPSAIS, Miller AbusePhysical Symptom and Injury Scale; MINI, Mini-­International NeuropsychiatricInterview; NEO-­FFI, NEO Five-­Factor Inventory; PVS, Partner Violence Screen;SAM, Self-­Assessment Manikin; SF-36, Short Form 36 Health Survey; STAI,State-­Trait Anxiety Inventory; TSST, Trier Social Stress Task; WHO Interview, WHO Violence Against Women Instrument. study, the minimum time of duration of the violent rela- measures along with the reference for the Spanish trans- tionship will be set at 1 year.16 Also, to study the long-­term lations of the scales is available as online supplemental effects of IPV once the exposure has ceased, only women materials. who have already ended the violent relationship for at least 1 year will be included. Session 1 Exclusion criteria will be as following: age below 21 (to All participants will obtain a full description of the study’s allow a margin of accumulated relationship experience aims and procedures and all questions will be answered. during adulthood) and over 50 (excluding menopause), Only those women willing to participate who sign an having any pituitary and/or adrenal gland disorder, informed consent will be included in the study. Socio- currently using steroid-­ based medications, being demographic data will be collected through the use of a currently pregnant, lactating or menopausal and having a standardised self-­report questionnaire. History of IPV will severe illness that may affect cognitive performance and/ be extensively described combining standard measures of or consciousness. No participant will be excluded on the screening (Partner Violence Screen31) and an in-­depth basis of disability, ethnicity, religion or sexual orientation. structured interview among women identified as survivors that include onset and frequency of IPV among other Procedures details (WHO Violence Against Women Instrument29). Women interested in participating will actively contact us Exposure to other forms of interpersonal violence will also by the email address that will be detailed in the advertise- be collected at this point using the Childhood Trauma ment. At first contact, a researcher will assess eligibility Questionnaire (CTQ,32) and the Life-­events Scale.33 A according to inclusion/exclusion criteria. If all criteria description of coping styles and resilient behaviour will are met, a first face-­to-­face session will be scheduled for an be assessed using the Coping Orientation to Problems in-­depth interview, followed by a second session 2 weeks Experienced Inventory34 and the Connor-­Davidson Resil- later to complete the assessment of response to acute ience Scale.35 stress exposure. A visual depiction of the procedures is A comprehensive health profile will also be assessed presented in figure 1, and a detailed description of the during the initial interview that will provide information 4 Goldberg X, et al. BMJ Open 2020;10:e036561. doi:10.1136/bmjopen-2019-036561 Open access on physical and mental health status and history of clin- stress research, and can be briefly described as a mock job ical treatment. The Miller Abuse Physical Symptom and interview. The participants are instructed to imagine that Injury Scale36 and Short Form 36 Scale for self-­perceived having applied for their ‘dream job’, they are now invited health status37 scales will be used for the assessment of to a job interview. Participants are aware that no real job physical symptoms and perception of general health. As is at issue. The TSST consists of three successive phases: part of the comprehensive health profile assessment, a (1) a preparation period (3 min), (2) a free speech task full laboratory test will be included to obtain a descrip- in which the participants have to argue why they are tion of biochemistry markers. For this objective, periph- the best candidate for the job (5 min) and (3) a mental eral blood samples will be collected by a nurse in 15 mL arithmetic task in which participants have to sequentially of capacity tubes. The General Health Questionnaire (12 subtract an odd two-­digit number from an odd four-­digit items version, GHQ-12)38 will be used for screening of number (eg, 17 from 2023; 5 min). The two tasks are mental health status and the Mini International Neuro- performed while standing in an upright position in front psychiatric Interview39 for in-­ depth assessment when of a selection committee consisting of two members, one GHQ-12 suggests mental health disorders. Personality male and one female, dressed in white lab coats, acting traits will be assessed using the Neuroticism-­Extraversion in a reserved manner and providing no facial or verbal Openness Inventory (Five Factor version, NEO-­FFI40). feedback.52 The interview is recorded in a video camera, Finally, a hair sample of more than 3 cm length with a procedure that has been demonstrated effective in trig- an approximate diameter of 30 single hairs will be taken gering further threat.53 The researchers in charge of the from the upper part of the scalp. This technique provides TSST are blinded to the condition (IPV-­exposed or not a measure of the integrated global basal cortisol over exposed) of the participants. the growth period of a specific hair segment, typically The primary measure of the stress response will be 1 cm/month.41 As glucocorticoid levels cannot be exper- the activation of the HPA axis by assessing the release of imentally manipulated in humans, we have biologically cortisol immediately before the TSST (T0), immediately validated this variable measuring hair corticosterone in after (20 min after the start of TSST, T1), and at recovery rats.42 Moreover, we have also technically validated in our (40 min after the start of TSST, T2). Saliva samples will be laboratory the measurement of hair cortisol in humans obtained by means of Salivette Cortisol collection devices using the well-­ characterised Salivary Cortisol Elisa Kit (Sarstedt AG & Co, Germany) and will be stored at −20°C (Salimetrics, LCC, Pennsylvania, USA). Reliability of the on the same day of collection. Salivary cortisol levels will be measure taken at different times is good, with correla- determined by means of a competitive radio-­immunoassay tions of 0.68 to 0.79.41 (RIA) technique developed in our laboratory that uses At this point of the assessment, we will examine selec- anti-­ cortisol antibody (121116) and Iodinated cortisol tive attentional processing by means of the Visual Probe (121126) from MP-­Biomedicals (Valiant Co, Ltd, USA). Task (VPT).43 44 Briefly, participants will be presented We have validated this RIA against the Salivary Cortisol simultaneously with a pair of stimuli, one emotionally Elisa Kit (Salimetrics, LCC, Pennsylvania, USA), showing salient and one neutral for 500 ms time, followed by a a high correlation of r=0.95. The reliability of the salivary probe that replaces one of the two stimuli. Participants cortisol TSST response is moderated, with Spearman will be required to respond as accurately and as quickly correlations over the days between 0.38 to 0.60.54 as possible to the probe. Reaction times will be recorded The behavioural response to acute stress will be the and contrasted. A decreased reaction time to a probe level of anxiety and perceived stress. To fulfil this aim, replacing emotional stimuli compared with the neutral the individuals’ responses to two different scales will be stimuli will provide a measure of bias to be vigilant for registered. The first of these scales will be the state exam- threatening information. Assessment of the VPT will be ination of the State-­ Trait Anxiety Inventory (STAI,55), accompanied by a brief complementary cognitive exam- which will allow to examine the self-­perceived level of ination that will include semantic memory,45 IQ46 and state anxiety at baseline (T0) and after the task (T1). executive functioning.47 48 Cronbach’s alpha reliability of the Spanish adaptation of this measure is high: 0.90 for the Trait Inventory and Session 2 0.94 for State Inventory. The second scale will be the Self-­ After 2 weeks, a second appointment will be scheduled Assessment Manikin (SAM,56), which is a picture-­oriented focussed on the assessment of the response to acute stress. scale to register an emotional response in its three key In order to control for potential differences in circadian features: valence/pleasure of the response, arousal and rhythm associated with endocrine activity, all participants dominance/control. This is a non-­verbal scale specifically will have a light lunch in the cafeteria of the centre 1 hour designed to be applied in transcultural settings. before the start of the laboratory phase. To control for The final half hour of the second appointment will be inter-­individual differences in cortisol levels, heavy phys- dedicated to a face-­to-­face interview aiming to systemat- ical activity 2 hours before lunch, and coffee consumption ically register subjective experience-­based issues related the same day are not allowed.49 50 to stress and coping strategies that may have not been Acute psychosocial stress response will be assessed using included in the quantitative assessment. The interviewer the TSST.51 This is the gold standard in biopsychological will have no previous contact with the women up to this Goldberg X, et al. BMJ Open 2020;10:e036561. doi:10.1136/bmjopen-2019-036561 5 Open access moment. Participants will be invited to a semi-­structured of the experience and will be invited to contribute to interview regarding their feelings, perceptions and atti- possible solutions. This session will have two parts with tudes during exposure to the acute stress (TSST), gener- different purposes. The first will be an open session ating a personal narrative of the experience by women aiming at disseminating the results of the study, including themselves.57 Conversations will be recorded and the specific interventions to raise awareness about the conse- transcripts will be used for analysis. quences of IPV. This first part will target all the partici- pants as well as other stakeholders and social agents that Sample size calculation may benefit from this information. The second session Sample size has been calculated using a priori power anal- will be a closed participatory session targeting the partic- ysis conducted in G*Power V.3.1.9.2 for Windows,58 with a ipants with history of IPV aimed at identifying strategies focus on the neuroendocrine response to acute stress that to build resilience and discussing prevention strategies. will be assessed by means of the changes in the levels of The session will end with the co-­creation of an inventory salivary cortisol during TSST at T0, T1 and T2. It has been of prevention strategies (from survivors and for survivors) estimated that the TSST reliably activates the HPA axis and ideas to communicate them. and triggers a twofold to threefold increase in cortisol in about 70% to 80% of participants.59 Assuming a conser- Data management plan and measures vative effect size of 0.2 based on this reference, and the All participants will be assigned a code at the moment correlation between measures of 0.38 mentioned above,54 of inclusion in the study, and the identification informa- if we want to detect differences among the two groups tion will be saved separately to warrant confidentiality. with a power of 0.95 and an alpha of 0.05, we will need Original data will be transferred to databases that will be a sample size of 41/group to detect a significant effect archived in FPT applying the standard processes of the (repeated measures analysis of variances (ANOVAs) for centre. These same standard processes will be used to the predictor analysis: Test Family: F tests, Statistical test: secure data quality throughout the study. ANOVA, Repeated measures, within-­ between interac- First, a detailed descriptive analysis will be run that tion, Type of power analysis: A priori, two groups, three will include the information from: (1) the initial inter- measurements). Because violence during childhood can view, (2) hair-­cortisol analysis, (3) the results of the VPT also affect the stress system of women,60 the sample size task, (4) the neuroendocrine response trajectories during of the group exposed to IPV will be doubled to include the TSST, and (5) the self-­reported levels of behavioural an even number of women with and without a history responses during the TSST. This information will be of childhood violence, as assessed by the CTQ. There- presented in the form of contingency tables. Group-­ fore, the complete sample will include: (1) the group comparison analysis will be run between the groups of of women exposed to IPV composed of 82 women, 41 IPV-­exposed women (with and without history of child- of them with history of childhood abuse and 41 of them hood maltreatment) and non-­exposed women. Tests will without such history, and (2) the group of women not be selected according to the nature of the variables and exposed to IPV, composed of 41 women. The total sample their sample distributions. size will equal 123 women. Data collection will continue The primary outcome variable in the study is the neuro- until the minimum sample is reached. endocrine response to acute stress, measured on the basis of salivary cortisol at T0, T1 and T2. The secondary Participants and public involvement outcome variables are the behavioural responses to acute The study has been designed with a focus on exposure to stress as measured by self-­ reported behavioural scales: IPV as a main risk factor for a number of health-­related perceived levels of anxiety at T0 and T1 (STAI), perceived issues, particularly mental health disorders. During the levels of valence/pleasure, arousal and control at T0 and design stage the research team was particularly concerned T1 (SAM). Hence, these will be the dependent variables that the exposure to acute stress was clearly justified, and in all analyses (online supplemental table 1). The hypoth- the information obtained through this laboratory condi- esis being tested at this stage will be that the three groups tion could not be collected in other forms. It was also differ in the measures of response to stress both at the highly important that the experimental condition would neuroendocrine (salivary cortisol) and behavioural (self-­ resemble a situation that any person could be presented reported behavioural scales) levels. To test this hypoth- with in real life. The team consulted local experts in the esis, we will use a generalised lineal model, specifically a field of violence against women before deciding on the repeated measures design,61 using group and time as the use of the TSST. explanatory variables (online supplemental table 2). During the recruitment period, different local govern- The following stage will test the hypothesis that bias mental and non-­ governmental organisations will be towards threat as measured by the VPT task is associated involved as consultants and sources of identification of with the response to acute stress in the group of women potential participants. They will also be contacted to with IPV. To test this hypothesis, an independent analysis discuss the final results and potential recommendations. will be run with VPT, group and time as explanatory vari- The final stage of the study will include a workshop ables. Two independent analysis will be run for salivary session where the participants will be placed in the centre cortisol and for behavioural outcomes. 6 Goldberg X, et al. BMJ Open 2020;10:e036561. doi:10.1136/bmjopen-2019-036561 Open access Other variables of potential interest will be included as on violence against women.65 Participants will be compen- covariates in the models: age, level of education, expo- sated for their transportation expenses, but no other sure to other relevant life events, copying styles, resilient compensation will be included in the study. They will be behaviour, general health status (physical symptoms, asked about their experience with the study immediately global cortisol hair concentration, self-­ perception), after completion of the interviews, and this information mental health status, personality traits, and cognition will be collected and used to provide feedback on evalua- (semantic memory, IQ, executive function) (online tion of the study. Finally, all participants will be provided supplemental table 3). Data will be analysed using with information regarding counselling services and SPSS software (IBM SPSS Statistics for Windows, V.23.0. other resources as appropriate. Armonk; New York: IBM Corp). The dissemination of the results of our project will start Regarding the semi-­ structured interviews that follow at the level of the participants during the final workshop, the TSST, the research team will develop the general which is expected to act as a first step for developing guidelines before the start of the study. All interviews will prevention tools and information resources that are be recorded, and the transcriptions of these recordings essentially built by women themselves. We will carry out will be conducted by trained professionals. Content anal- policy-­ dialogues and workshops with relevant regional ysis will be used to extract the fundamental aspects of the and national representatives aimed at enhancing the discourse on the experience of women and their resilience current policies and roadmaps regarding the training strategies.62 The aim will be to search data saturation, this and management in the educational and healthcare is, when new interviews do not provide new findings on areas. Spain counts with relevant programmes in the field the results. Transcriptions will be analysed using software of violence against women,66 and we expect to provide for qualitative analysis, ​Atlas.​ti (​Atlas.​ti 8 Windows, Scien- these initiatives with evidence-­based data that can help tific Software Development GmbH). Following on these build innovative solutions. At the level of the dissemina- data, a second tier for analysis will focus on profile differ- tion of results in the scientific community, the strategy entiation in order to capture the different narratives and includes the publication of the results in international profiles that may arise across participants, and related to peer-­reviewed scientific journals and the presentation in other variables (childhood abuse, socio-­economic posi- national and international congresses. Also, the project tion, education). This will lead to highlight the common is expected to strengthen the way healthcare providers traits within each profile as well as strengths and weak- respond to women who have experienced violence. A nesses. Finally, the integration of quantitative and qualita- series of courses will be developed based on the results of tive data will take place through a mixed-­methods analysis this project and others with similar objectives to inform that will be run to explore relevant aspects identified in the work of psychologists, psychiatrists, social workers, the post-­TSST interviews and that may have not been nurses and any professional association that might be fully considered in the original design.63 The selection of willing to receive the training. The complete set of results the specific variables at this point will be determined by from the study will be used to develop guidelines and results of the previous stages. Integration of the different recommendations for actions that will be distributed sources of data will also be performed during the inter- among professionals. pretation of the results using data presented in theme-­by-­ statistics joint display.64 DISCUSSION Ethics and dissemination The prevention of violence against women in general, This study has been approved by the Ethics Committee of and of IPV in particular, has risen as a priority on the reference (‘Comité de Ética de Investigación Parc Taulí international public health agenda, and research is de Sabadell’). Written informed consent as approved by playing a key role in the detection of protective factors the Ethics Committee will be obtained after a full descrip- and the development of effective interventions.67 Are tion of the study’s aims and design. Participants will be there cognitive implications of exposure to gender-­based informed of the confidentiality of their comments and violence? Which systems underlie these effects, what of the contact information of the principal investigators causes them? Can they be reversed? These are questions to exercise their rights of access, modification, opposi- that exceed the laboratory settings and impact ‘the real tion and cancellation of data, and withdrawal from the world’. Our proposal aims at targeting these traits, which study without any repercussions, following the European not only impair women’s daily functioning but also feed Data Protection legislation (2016/679; Ley Orgánica de a cycle of attitudes, norms and beliefs that justify domi- Protección de Datos de Carácter Personal, 15/1999 del nant notions of masculinity and stigmatise survivors.68 69 13 de Diciembre, LOPD). Registry and use of informa- Hence, the inclusion of a mixed-­methods approach that tion resulting from this study will follow the Declara- integrates subjective reports with neurobiological data is tion of Helsinki agreements. All biological samples will a key aspect of this protocol. be collected and stored according to the corresponding Regarding methodological issues, the TSST is a powerful legislation (Llei 14/2007 de Recerca Biomèdica). The tool to identify dysfunctional patterns of coping that may study will follow WHO’s recommendations for research help explain some critical aspects of the behavioural Goldberg X, et al. BMJ Open 2020;10:e036561. doi:10.1136/bmjopen-2019-036561 7 Open access responses of IPV-­ exposed women. However, we have SGR2017-457). RN is a recipient of an ICREA Academia Award (Generalitat de included the TSST only after extensively discussing the Catalunya 2015-19). possible distress that may be caused to women, and the Competing interests None declared. benefits of including the measure in the study. Our deci- Patient and public involvement Patients and/or the public were involved in the sions throughout the project have been guided by the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details. WHO Practical Guide for Researchers and Activists.70 This Patient consent for publication Not required. document summarises all aspects of research in this field and provides with useful recommendations to assure the Provenance and peer review Not commissioned; externally peer reviewed. project achieves the objective of serving the target women. Open access This is an open access article distributed in accordance with the The safety of respondents and the research team is our Creative Commons Attribution Non Commercial (CC BY-­NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-­commercially, priority and it is our advice that it be that of any other and license their derivative works on different terms, provided the original work is study working with this population. The identification of properly cited, appropriate credit is given, any changes made indicated, and the use any problem in this respect must result in the immediate is non-­commercial. See: http://​creativecommons.​org/​licenses/​by-​nc/​4.​0/. interruption of the assessment. The research team must ORCID iD be trained, and the assessment must be conducted in a Ximena Goldberg http://​orcid.​org/​0000-​0001-​9681-​5826 location different to that where women receive health and social assistance. The main limitation in our study is that the limited sample size will prevent exploring putative modulatory REFERENCES 1 Miller E, McCaw B. Intimate partner violence. N Engl J Med effects of relevant variables such as other lifetime stressful 2019;380:850–7. experiences. Also, we will not be able to test our hypoth- 2 García-­Moreno C, Pallito C, Devries K, et al. Global and regional estimates of violence against women: prevalence and health effects esis among women over 50 years of age due to restric- of intimate partner violence and non-­partner sexual violence. Geneva: tions in the inclusion/exclusion criteria. We would like World Health Organization, 2013. 3 European agency for fundamental rights. Violence against women: an to acknowledge the need for further information and EU-­wide survey main results. Vienna, Austria, 2014. research regarding IPV in the population of women aged 4 Ministerio de Sanidad Servicios Sociales E Igualdad. Macroencuesta 50 years and older, as has been highlighted by others de violencia Contra La mujer 2015. Madrid, 2015. 5 Breiding MJ, Black MC, Ryan GW. Chronic disease and health risk before us.2 The mixed-­methods approach proposed in behaviors associated with intimate partner violence-18 U.S. states/ our project is expected to be challenging because the territories, 2005. Ann Epidemiol 2008;18:538–44. 6 Coker AL, Davis KE, Arias I, et al. Physical and mental health effects qualitative perspective tends to emphasise an inductive of intimate partner violence for men and women. Am J Prev Med method that highlights subjective information while the 2002;23:260–8. 7 Pico-­Alfonso MA, Garcia-­Linares MI, Celda-­Navarro N, et al. quantitative perspective is based on a deductive method The impact of physical, psychological, and sexual intimate male largely based on objectivity and generalisation. In turn, partner violence on women's mental health: depressive symptoms, our study has the potential to provide evidence to serve posttraumatic stress disorder, state anxiety, and suicide. J Womens Health 2006;15:599–611. a deeper understanding of IPV and the vulnerability and 8 Beydoun HA, Beydoun MA, Kaufman JS, et al. Intimate partner resilience processes that IPV-­ exposed women present. violence against adult women and its association with major depressive disorder, depressive symptoms and postpartum This information will allow professionals and institutions depression: a systematic review and meta-­analysis. Soc Sci Med to better understand and address this reality. Ultimately, 2012;75:959–75. 9 Zieman G, Bridwell A, Cárdenas JF. Traumatic brain injury in it is expected that the results of this research will serve domestic violence victims: a retrospective study at the Barrow as the foundation to build evidence-­based tools for the neurological Institute. J Neurotrauma 2017;34:876–80. prevention of re-­victimisation among women exposed to 10 Nelson J, Klumparendt A, Doebler P, et al. Childhood maltreatment and characteristics of adult depression: meta-­analysis. Br J IPV and of IPV in at-­risk groups. Psychiatry 2017;210:96–104. 11 Goldberg X, Serra-­Blasco M, Vicent-­Gil M, et al. Childhood Author affiliations maltreatment and risk for suicide attempts in major depression: a 1 sex-­specific approach. Eur J Psychotraumatol 2019;10:1603557. Mental Health Department, Neuroscience and Mental Health Research Area, Institut 12 Abramsky T, Watts CH, Garcia-­Moreno C, et al. What factors are d’Investigació I Innovació Parc Taulí I3PT, Universitat Autònoma de Barcelona, associated with recent intimate partner violence? findings from the CIBERSAM, Sabadell, Spain who multi-­country study on women’s health and domestic violence. 2 Psychobiology Unit (School of Psychology), Institut de Neurociències, Universitat BMC Public Health 2011;11:109. Autònoma de Barcelona, CIBERSAM, Cerdanyola del Vallès, Spain 13 Garcia-­Moreno C, Jansen HAFM, Ellsberg M, et al. Prevalence of 3 intimate partner violence: findings from the WHO multi-­country study Animal Physiology Unit (School of Biosciences), Institut de Neurociències, on women’s health and domestic violence. Lancet 2006;368:1260–9. Universitat Autònoma de Barcelona, CIBERSAM, Cerdanyola del Vallès, Spain 14 Chrousos GP, Gold PW. The concepts of stress and stress system disorders. overview of physical and behavioral homeostasis. JAMA Contributors XG, RN and AA were responsible for determining the content and 1992;267:1244–52. 15 Chrousos GP. Stress and disorders of the stress system. Nat Rev scope of the study, and for the design. CE and DP were involved in the definition of Endocrinol 2009;5:374–81. the protocol. All authors were involved in study methods and tools. XG drafted the 16 Miller GE, Chen E, Zhou ES. If it goes up, must it come down? manuscript with critical input from the rest of authors, who read and approved the chronic stress and the hypothalamic-­pituitary-­adrenocortical axis in final manuscript. humans. Psychol Bull 2007;133:25–45. 17 Yim IS, Kofman YB. The psychobiology of stress and intimate partner Funding This work has received funding from ‘la Caixa’ Foundation (ID violence. Psychoneuroendocrinology 2019;105:9–24. 100010434), under agreement 2017ACUP00277. XG was supported by the 18 Griffin MG, Resick PA, Yehuda R. Enhanced cortisol suppression Health Department of the Generalitat de Catalunya grant SLT002/16/00237. AA following dexamethasone administration in domestic violence is the principal investigator of a SGR Research Group (Generalitat de Catalunya, survivors. Am J Psychiatry 2005;162:1192–9. 8 Goldberg X, et al. BMJ Open 2020;10:e036561. doi:10.1136/bmjopen-2019-036561 Open access 19 Carroll D, Ginty AT, Whittaker AC, et al. The behavioural, cognitive, 44 Sipos ML, Bar-­Haim Y, Abend R, et al. Postdeployment threat-­related and neural corollaries of blunted cardiovascular and cortisol attention bias interacts with combat exposure to account for PTSD reactions to acute psychological stress. Neurosci Biobehav Rev and anxiety symptoms in soldiers. Depress Anxiety 2014;31:124–9. 2017;77:74–86. 45 Rey A. L’examen clinique en psychologie. [The clinical examination in 20 Belda X, Fuentes S, Daviu N, et al. Stress-­induced sensitization: psychology]. Paris, FR: Presse Universitaires de France, 1964. the hypothalamic-­pituitary-­adrenal axis and beyond. Stress 46 Wechsler D, Coalson D, Raiford S. Wechsler adult intelligence 2015;18:269–79. test: fourth edition technical and interpretive manual. San Antonio 21 Belda X, Nadal R, Armario A. Critical features of acute stress-­induced Pearson, 2008. cross-­sensitization identified through the hypothalamic-­pituitary-­ 47 Golden CJ. Stroop color and word test: a manual for clinical and adrenal axis output. Sci Rep 2016;6:31244. experimental uses. Chicago: Stoelting, 1978. 22 Robinson MD. Running from William James’ bear: a review of 48 Reitan RM. Validity of the TRAIL making test as an indicator of Preattentive mechanisms and their contributions to emotional organic brain damage. Percept Mot Skills 1958;8:271–6. experience. Cogn Emot 1998;12:667–96. 49 Lovallo WR, Al’Absi M, Blick K, et al. Stress-­like adrenocorticotropin 23 Depierro J, D'Andrea W, Pole N. Attention biases in female survivors responses to caffeine in young healthy men. Pharmacol Biochem of chronic interpersonal violence: relationship to trauma-­related Behav 1996;55:365–9. symptoms and physiology. Eur J Psychotraumatol 2013;4:19135. 50 Quigley ME, Yen SS. A mid-­day surge in cortisol levels. J Clin 24 Dandeneau SD, Baldwin MW, Baccus JR, et al. Cutting stress off at Endocrinol Metab 1979;49:945–7. the pass: reducing vigilance and responsiveness to social threat by 51 Kirschbaum C, Pirke KM, Hellhammer DH. The ‘trier social stress manipulating attention. J Pers Soc Psychol 2007;93:651–66. test’ tool for investigating psychobiological stress responses in a 25 Van Uum SHM, Sauvé B, Fraser LA, et al. Elevated content of cortisol laboratory setting. Neuropsychobiology 1993;28:76–81. in hair of patients with severe chronic pain: a novel biomarker for 52 Frisch JU, Häusser JA, Mojzisch A. The trier social stress test stress. Stress 2008;11:483–8. as a paradigm to study how people respond to threat in social 26 Steudte S, Kolassa I-­T, Stalder T, et al. Increased cortisol interactions. Front Psychol 2015;6:14. concentrations in hair of severely traumatized Ugandan individuals 53 Dickerson SS, Kemeny ME, Stressors A. Acute stressors and cortisol with PTSD. Psychoneuroendocrinology 2011;36:1193–200. responses: a theoretical integration and synthesis of laboratory 27 O'Cathain A, Murphy E, Nicholl J. The quality of mixed methods research. Psychol Bull 2004;130:355–91. studies in health services research. J Health Serv Res Policy 54 Kirschbaum C, Prüssner JC, Stone AA, et al. Persistent high cortisol 2008;13:92–8. responses to repeated psychological stress in a subpopulation of 28 Lingard L, Albert M, Levinson W. Grounded theory, mixed methods, healthy men. Psychosom Med 1995;57:468–74. and action research. BMJ 2008;337:a567. 55 Spielberger CD, Sydeman SJ. State-­trait anxiety inventory and 29 Garcia-­Moreno C, Jansen HA, Ellsberg M. WHO multi-­country study state-­trait anger expression inventory. In: Maruish EM, ed. The on women’ s health and domestic violence against women. Initial use of psychological testing for treatment planning and outcome results on prevalence, health outcomes and women’s responses. assessment. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc, Geneva, 2005. https://www.​who.​int/​reproductivehealth/​publications/​ 1994: 292–321. violence/​24159358X/​en/ 56 Bradley MM, Lang PJ. Measuring emotion: the self-­assessment 30 Krug EG, Mercy JA, Dahlberg LL, et al. World report on violence and manikin and the semantic differential. J Behav Ther Exp Psychiatry health. Geneva: World Health Organization, 2002: 360. 1083–8. 1994;25:49–59. 31 Feldhaus KMet al. Accuracy of 3 brief screening questions for 57 Britten N, Pope C, Mays N. Qualitative interviews. In: Qualitative detecting partner violence in the emergency department. JAMA research in health care. Oxford, England: Blackwell Publishing Ltd, 1997;277:1357–5. 2006: 12–20. 32 Bernstein DP, Stein JA, Newcomb MD, et al. Development and 58 Faul F, Erdfelder E, Lang A-­G, et al. G*Power 3: A flexible statistical validation of a brief screening version of the childhood trauma power analysis program for the social, behavioral, and biomedical questionnaire. Child Abuse Negl 2003;27:169–90. sciences. In: Behavior research methods. , 2007: 39, 175–91. 33 Brugha T, Bebbington P, Tennant C, et al. The list of threatening 59 Kudielka BM, Hellhammer DH, Kirschbaum C. Ten years of research experiences: a subset of 12 life event categories with considerable with the trier social stress test (TSST) - revisited. In: Harmon-­Jones long-­term contextual threat. Psychol Med 1985;15:189–94. E, Winkielman P, eds. Social neuroscience: integrating biological 34 Carver CS. You want to measure coping but your protocol’s too long: and psychological explanations of social behavior. 512. New York: consider the brief cope. Int J Behav Med 1997;4:92–100. Guilford Press, 2007. https://​books.​google.​com/​books?​hl=​en&​lr=&​ 35 Connor KM, Davidson JRT. Development of a new resilience scale: id=​mKSqxaHGaicC&​pgis=1 the Connor-­Davidson resilience scale (CD-­RISC). Depress Anxiety 60 Mielock AS, Morris MC, Rao U. Patterns of cortisol and alpha-­ 2003;18:76–82. amylase reactivity to psychosocial stress in maltreated women. J 36 Miller C, Campbell J. Reliability and validity of the Miller abuse Affect Disord 2017;209:46–52. physical symptom and injury scale (MAPSAIS. Chicago: Midwest 61 Daniel W, Cross C. Biostatistics: a foundation for analysis in the Nursing Research Society, 1993. health sciences. 11th edn. Wiley, 2018. 37 McHorney CA, Ware JE, Raczek AE. The MOS 36-­item short-­form 62 Pope C, Ziebland S, Mays N. Qualitative research in health care. health survey (SF-36): II. psychometric and clinical tests of validity analysing qualitative data. BMJ 2000;320:114–6. in measuring physical and mental health constructs. Med Care 63 Lingard L, Albert M, Levinson W. Qualitative research: grounded 1993;31:247–63. theory, mixed methods, and action research. BMJ 2008;337. 38 Goldberg DP, Gater R, Sartorius N, et al. The validity of two versions 64 Guetterman TC, Fetters MD, Creswell JW. Integrating quantitative of the GHQ in the WHO study of mental illness in general health care. and qualitative results in health science mixed methods research Psychol Med 1997;27:191–7. through joint displays. Ann Fam Med 2015;13:554–61. 39 Sheehan DV, Lecrubier Y, Sheehan KH, et al. The Mini-­International 65 WHO Department of Gender Women and Health. Putting women neuropsychiatric interview (M.I.N.I.): the development and validation first: ethical and safety recommendations for research on domestic of a structured diagnostic psychiatric interview for DSM-­IV and ICD- violence against women. Geneva, 2001. 10. J Clin Psychiatry 1998;59:22–33. 66 Butchart A, Mikton C. Global status report on violence prevention 40 Costa PT, McCrae RR. Revised NEO personality inventory (NEO-­PI-­R) 2014. Luxembourg, 2014. and NEO five-­factor inventory (NEO-­FFI). Odessa, FL: Psychological 67 Mikton CR, Tanaka M, Tomlinson M, et al. Global research priorities Assessment Resources, Inc, 1992. for interpersonal violence prevention: a modified Delphi study. Bull 41 Stalder T, Kirschbaum C. Analysis of cortisol in hair-­state of the art World Health Organ 2017;95:36–48. and future directions. Brain Behav Immun 2012;26:1019–29. 68 García-­Moreno C, Zimmerman C, Morris-­Gehring A, et al. Addressing 42 Scorrano F, Carrasco J, Pastor-­Ciurana J, et al. Validation of violence against women: a call to action. Lancet 2015;385:1685–95. the long-­term assessment of hypothalamic-­pituitary-­adrenal 69 García-­Moreno C, Hegarty K, d’Oliveira AFL, et al. The activity in rats using hair corticosterone as a biomarker. Faseb J health-­systems response to violence against women. Lancet 2015;29:859–67. 2015;385:1567–79. 43 MacLeod C, Mathews A, Tata P. Attentional bias in emotional 70 Ellsberg M, Heise L. Researching violence against women. disorders. J Abnorm Psychol 1986;95:15–20. Washington, DC: World Health Organization, 2013. Goldberg X, et al. BMJ Open 2020;10:e036561. doi:10.1136/bmjopen-2019-036561 9