Study design overview.

This study employs a mixed-methods prospective cohort design integrating multimodal neurological assessments and semi-structured interviews (Fig 1). Participants will be enrolled into either a sexual strangulation group or a non-strangulation control group based on screening criteria and will undergo baseline neurological and behavioral assessments. Acute neurological responses will be evaluated following a sexual event involving strangulation (or non-strangulation sex in controls), and longitudinal follow-up assessments will be conducted every six months for up to 30 months to characterize cumulative exposure effects and temporal changes in neurological health.

Screening and enrollment.

Potential participants will be identified primarily through a campus-wide probability survey focused on sexual health and behavior that is administered during the first two to three years of the study, supplemented by community-based recruitment strategies, including university’s classified sites and social media platforms. Individuals who indicate willingness to be contacted for follow-up research will be randomly stratified (via Randomization function within R), with first- and second-year students taking the priority spots to accommodate those who are likely be in school throughout the entirety of 30 months follow-up. Invited participants will complete a prescreening questionnaire to confirm eligibility criteria, including age, sexual strangulation history, and key exclusionary medical or neurological conditions. Eligible individuals will be invited to a virtual consent meeting, during which electronic written consent will be obtained. Following the virtual consent meeting, participants will undergo an additional virtual meeting, and research staff will administer QuickSCID-5 (Quick Structured Clinical Interview for DSM-5 Disorders) to assess current and history of mental health diagnosis. When participants endorse any symptoms related to psychosis and/or schizophrenia, research staff will then conduct a SCID-5 Research Version – Psychotic module B/C to determine the eligibility of participants [32]. During the virtual meeting, research staff will also administer The Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) [33]. When participants meet the inclusion criteria and are free of exclusionary criteria, they will be enrolled in the study and assigned a study-specific ID. Prior to baseline visit, participants are asked to abstain from the following behaviors and caffeine: abstain from sex involving choking and any activities involving pressing or squeezing of the neck (e.g., martial arts chokeholds) 48 hours prior to lab visit, no strenuous exercise (e.g., sports games, cross-fit) 24 hours prior to the visits, no mild to moderate exercise on the day of lab visit, no sexual activities on the day of lab visit, and no caffeine on the day of lab visit.

Baseline assessment.

Between enrollment and the baseline laboratory visit, participants will complete a series of online questionnaires assessing demographics, medical and neurological history, sexual behavior history, and Adverse Childhood Experiences (ACES). At the baseline visit, participants will undergo various assessments, including testing day questionnaires to gauge various contextual information (menstrual cycle, exercise), capillary blood sampling to test blood fatty acid components, urinalysis to verify pregnancy status, blood sampling for biomarker analysis, MRI, cognitive testing, oculomotor assessments, retinal imaging, and mental health questionnaires. Baseline assessments are designed to characterize neurological health prior to acute or longitudinal exposure assessment and to allow examination of baseline group differences between participants with frequent sexual strangulation exposure and non-strangulation controls. The baseline visit will end with setting up a MyCap account within Research Electronic Data Capture (REDCap) for participants for acute visit purposes. Participants in both the strangulation group and control group will participate in the acute aim (aim 2), but control participants who are not sexually active will not be participating in the acute aim.

Acute assessment following sexual activity.

Participants eligible for the acute aim will be monitored following baseline assessment using brief daily electronic surveys to determine the occurrence of sexual activity and the presence or absence of strangulation. For participants in the strangulation group, an acute laboratory visit will be scheduled within approximately 24 hours following a sexual event involving strangulation. For control participants, an acute visit will be scheduled within approximately 24 hours following a sexual event without strangulation. Participants who do not engage in sexual activity during the monitoring period will remain eligible for longitudinal follow-up but may not contribute to acute analyses. The acute visit will replicate key components of the baseline assessment, including blood collection, neuroimaging, cognitive testing, oculomotor and retinal assessments, and mental health questionnaires. These data will be used to characterize short-term neurological responses associated with sexual strangulation relative to baseline and control conditions.

Longitudinal follow-up.

All participants will enter a longitudinal follow-up phase consisting of laboratory visits every six months for up to 30 months following baseline assessment. Each follow-up visit will include repeat multimodal neurological assessments identical to baseline measures, allowing evaluation of changes in neural cellular, physiological, and functional integrity over time. Longitudinal assessments are designed to capture cumulative exposure effects and temporal trajectories of neurological health associated with sexual strangulation. Participants who change sexual behavior during the study (e.g., controls initiating sexual strangulation or strangulation participants discontinuing the behavior) will continue follow-up when feasible. Data collected prior to behavioral change will contribute to primary analyses, while post-change data will be used in exploratory analyses to examine within-individual changes.

Semi-structured interviews.

Semi-structured interviews will be conducted at baseline (within 1 week from the baseline neurological assessments), following acute laboratory visits (within 72 hours), and after each longitudinal follow-up visit (within 1 week from the baseline neurological assessments). Interviews will assess participants’ experiences with sexual strangulation, including frequency, duration, methods, perceived consent, contextual factors, associated symptoms, and changes in behavior over time. These interviews will provide behavioral and contextual data to complement neurobiological findings and to inform interpretation of neurological outcomes.

Blood fatty acid composition.

At baseline and all follow-up timepoints, a capillary blood sample from a fingertip will be collected and sent to OmegaQuant Analytics for blood fatty acids composition analysis.

Psychiatric and mental health questionnaires.

The following questionnaires will be used to gauge symptoms related to various psychiatric and mental health conditions, including Patient Health Questionnaire-9 (PHQ-9) for depression, Generalized Anxiety Disorder-7 (GAD-7) for anxiety, Perceived Stress Scale (PSS) for perceived stress, adult ADHD self-report scale (ASRS) for ADHD-related symptoms, Multidimensional Personality Questionnaire (MPQ) for personality traits, Extended–Hurt, Insulted, Threaten, Scream (E-HITS) for history of domestic/intimate partner violence with an additional question regarding strangulation experience, Penn State Electronic Cigarette Dependence Index (PS-ECDI) for e-cigarette use, Pittsburgh Sleep Quality Index (PSQI) for sleep, Migraine Disability Assessment (MIDAS) for migraine, Life Events Checklist-5 (LEC-5) to screen for lifetime traumatic events, and PTSD Checklist for DSM-5 (PCL-5) to assesses symptoms of PTSD.

Blood biomarker assessment.

Capillary blood sampling using the Tasso blood collection device will be performed each test session. A trained phlebotomist will thoroughly clean the area with an alcohol swab and collect two sets of 400 μL of capillary blood sample using the Tasso device. Plasma levels of NF-L, Tau, UCH-L1, GFAP, p-tau 181, and p-tau 217 will be assessed by Quanterix’s Simoa SR-X assays. Various cytokines and chemokine markers, including S100B, IL-1b,1a,2,6,8,10,17; CCL2,3,5,11; and VCAM, PCAM, angiopoietin, and neuregulin, will be assessed by custom Luminex panels.

Neuroimaging assessment.

Neuroimaging assessments will take place in the Indiana University’s Imaging Research Facility (IRF), which houses a research-dedicated 3T Siemens Prisma MRI scanner equipped with a 64-channel head/neck coil.

High-resolution anatomical images (T1), which will be used for morphometry analyses, will be acquired using 3D MPRAGE pulse sequence: TR/TE = 2400/2.3ms, TI = 1060ms, flip angle = 8, field-of-view = 256 mm, matrix = 320x320, bandwidth = 160 Hz/pixel, iPAT = 2, resulting in 0.8 mm isotropic resolution. Both resting state fMRI (rs-fMRI) and task-based fMRI data will be collected using an SMS single-shot EPI sequence: FOV = 216 mm, TR/TE = 800/30 ms, flip angle = 52°, matrix = 90 × 90, resolution = 2.4 mm isotropic, and multiband acceleration factor = 6, with 100 volumes for rs-fMRI and 240 volumes for task-based fMRI. The rs-fMRI data will be acquired while the participant relaxes with eyes open and passively views a crosshair for 12 minutes. For task-based fMRI, the BART task [34], which has two, 8-min blocks, will be administered. As participants continue to inflate a balloon for greater reward, the probability of an explosion increases to mimic real-world risk-reward decision making. For each trial, the screen will display the image of a purple balloon above a small green rectangle, which indicates that she/he needs to make a decision (inflate or win). Participants will accumulate their win/wager throughout the two, 8-min block, experiments. Potential regions of interest (ROI) include the dorsal anterior cingulate cortex (dACC), nucleus accumbens (NAc), and insula due to their involvement in reward and decision-making processes [35]. Our DTI protocol is adapted from the Adolescent Brain Cognitive Development study [36], which is optimized for NODDI analysis. Two consecutive DTI sessions with opposite phase encoding directions will be performed with a simultaneous multi-slice (SMS) single-shot spin-echo echo-planar imaging (EPI) pulse sequence (TE = 89.4ms; TR = 3590s, 1.5 mm isotropic resolution). Each session has 80 images with different diffusion weightings and gradient directions (7 b = 0 s/mm2, 6 directions with b = 500 s/mm2, 15 directions with b = 1000 s/mm2, 15 directions with b = 2000 s/mm2, and 60 directions b = 2500 s/mm2). The total scan time is 10 minutes. Quantitative Susceptibility Mapping (QSM) will be collected using a three-dimensional flow-compensated multiecho gradient-echo sequence with a repetition time of 45 msec, a total of five echoes (time of the first echo, 13 msec) with echo spacing of 6 msec, flip angle of 20°, field of view of 24 × 24 cm, matrix of 512 × 256, and 88 sections acquired with 1.5-mm thickness.

Cognitive function.

Cognition will be measured using the NIH Toolbox Cognition Battery [37]. Multiple forms will be used to minimize practice effects. This battery consists of 7 instruments testing 5 cognitive domains: 1. Dimensional Change Card Sort Test (cognitive flexibility); 2. Flanker Inhibition Test (inhibitory control); 3. Picture Sequence Memory Test (episodic memory); 4. Picture Vocabulary Test (language); 5. Oral Reading Recognition Test (language); 6. Sorting Working Memory Test (working memory); and 7. Pattern Comparison Processing Speed Test (processing speed). The battery takes 20 minutes to complete on a tablet.

Near point convergence (NPC).

NPC will be measured based on our established protocol [38–42]. Using the accommodative ruler, a target (14-point letter) will be moved toward the eyes at a rate of 1–2 cm/s. NPC will be recorded when participants report diplopia has occurred, or the tester observes eye misalignment. The assessment will be repeated twice, and the mean NPC value will be used for analyses.

Optical coherence tomography/angiography (OCT/A).

Retinal neural structure will be imaged using high-definition spectral domain OCT, which will be acquired using a Zeiss Cirrus 5000 OCT scanner. The OCT variables examined will be macula CSF thickness (an indicator of the gain or loss of neurons or glia in the inner nuclear, ganglion cell and nerve fiber layer), and cup-to-disc ratio (an indicator of neurodegeneration at the optic disc). Retinal vascular structure will be acquired using the OCT angiography (OCT/A) capability of the Cirrus 5000. The primary OCT/A variable examined will be foveal avascular zone (FAZ) area reflecting the size of the central portion of the macula which contains no blood vessels, and which increases in size with loss of capillaries in the surrounding region).

Semi-structured Interviews.

The interview will assess participants’ relationship status, relationship quality, sexual agency, and overall sexual experiences, including partnered strangulation. For participants in the strangulation group, we will assess strangulation-related characteristics (onset, frequency, duration, single vs. repetitive, reasons for continued engagement, perceived prevalence among their peers); strangulation method (e.g., both hands, ligature); and responses from being strangled (e.g., visual changes, dizziness, disorientation, loss of consciousness, neck pain, bruising, euphoric sensations, as well as their levels of enjoyment and pleasure. The interview will also include questions related to less common “rough sex” behaviors that may be associated with neurological sequelae (e.g., face/head slapping, smothering, gagging). Participants will be asked about contextual issues relevant to their sexual experiences, which may include arousal, orgasm, intimacy, and/or trust. We will transcribe and code the interview data. The study will explore various dimensions of consent, including the timing and frequency of consent, who initiates consent discussions, and whether the act has ever become uncomfortable despite initial consent. If discomfort has occurred, participants will be asked whether they have discussed these experiences with their partner. Interviews will be audio-recorded and transcribed, with the audio files then deleted. Interview data will be coded and analyzed following a thematic analysis approach consisting of five steps following Nowell et al. [43]

Baseline group analysis.

To test whether neurological outcomes differ at baseline between the strangulation and control groups, we will use multivariable regression models with each baseline neurological outcome as the dependent variable and group (strangulation vs control) as the primary independent variable. For continuous outcomes (e.g., biomarker concentrations, imaging metrics, cognitive scores, NPC distance, OCT measures), we will use linear regression (or generalized linear models with appropriate functions if distributional assumptions are not met). For binary outcomes (e.g., presence/absence of symptoms or threshold-defined impairment), we will use logistic regression. Models will adjust for prespecified baseline covariates that are clinically relevant and/or imbalanced between groups, including age, BMI, prior concussion history, and baseline mental health symptoms. Because baseline neural health may be influenced by contextual factors, we will additionally consider time since last sexual event, substance use, sleep, and ovarian hormone levels/menstrual cycle phase as covariates when relevant to specific outcomes. For imaging outcomes, scanner-related and acquisition covariates (e.g., motion indices) will be included as standard quality-control adjustments. We will control for multiple comparisons using a false discovery rate (FDR) approach within each modality (biomarkers, imaging, ocular/retinal, cognition, mental health). As a complementary approach that leverages correlation among outcomes, we will conduct multivariate testing within domains (e.g., principal-component-based composite scores) to evaluate whether a domain shows an overall baseline difference by group prior to interpreting individual endpoints. Where appropriate, we also may consider exploratory mediation and explanatory models to help explain baseline group differences. We will use mediation analyses to estimate indirect effects while adjusting for confounders.

Acute effects.

To examine the acute neurological effects of sexual strangulation, we will use a series of generalized linear mixed models (GLMMs) (or linear mixed-effects models for normally distributed outcomes) with each neurological variable as the dependent variable and visit type (baseline vs. post-acute) as the primary within-person predictor. Models will include a participant-specific random intercept to account for repeated measures. The primary interest will be the group (strangulation vs. control) × time (baseline vs. post-acute) interaction. Within-group time-effects will also be examined to assess group-specific time trends. Prespecified covariates will consider age, prior concussion history, and mental health diagnosis and/or mental health symptoms, as well as potentially adjusting for time since the most recent sexual event. We will also explore by including menstrual cycle phase and/or ovarian hormone levels as covariates. For neuroimaging outcomes, standard acquisition/quality covariates (e.g., motion indices) will be included. To address multiple comparisons across numerous endpoints, FDR will be applied within modality (biomarkers, imaging, ocular/retinal, cognition/mental health).

Cumulative effects.

To determine whether cumulative exposure to sexual strangulation is associated with longitudinal changes in neurological health, we will model repeated neurological outcomes collected every 6 months over 30 months using GLMMs (or linear mixed-effects models for continuous outcomes), a similar model structure to the acute aim. Each model will include a participant-specific random intercept to account for within-person correlation. Time will be modeled as a categorical variable (baseline, 6, 12, 18, 24, 30 months) for primary analyses to allow non-linear trajectories, and as a continuous variable in sensitivity analyses.

Primary longitudinal analysis will focus on: (1) overall group differences in trajectories between the strangulation and control groups via group × time interaction, and (2) within-group change over time (especially within the strangulation group) to identify when divergence from baseline emerges. Covariates will include age, prior concussion history, mental health diagnosis, and other clinically relevant time-varying factors assessed at each visit (e.g., mental health symptoms, substance use). For imaging outcomes, motion indices will be included.

To directly test the exposure-response hypotheses, we will incorporate time-varying measures of sexual strangulation exposure (e.g., frequency since last visit and cumulative frequency across follow-up; and, where available, intensity/duration and presence of alteration in consciousness) as predictors of neurological outcomes within the strangulation group. These models will estimate whether higher exposure is associated with worse neurological outcomes and whether this association strengthens over time by including exposure × time terms when appropriate.

Switching behavior.

During follow-up, some participants may initiate or discontinue sexual strangulation. We will conduct exploratory analyses among “switchers” using (a) time to initiation/cessation as an “intervention” and (b) within-person pre/post comparisons using mixed models with an indicator for post-switch period. These analyses will be labeled exploratory and interpreted as hypothesis-generating.

Semi-structured interview.

Interview data will be coded and analyzed following a thematic analysis approach consisting of five steps following Nowell et al. [43] Our analyses will be guided by the following questions: 1. What styles of strangulation do participants report in terms of neck placement, intensity, and being strangled once vs repetitively in a typical encounter? 2. How does one’s experiences with sexual strangulation change over time in the context of frequency, duration, intensity, and methods that they use? 3. What are the signs and symptoms related to sexual strangulation and to what extent do they change over time? 4. Why do people continue to engage in strangulation or decide to stop? As our earlier interview data were collected from only a single time point, it is critical to understand engagement in sexual strangulation prospectively and to examine change over time. Although these data will be novel and extend the literature on their own, they will also support interpretation of the neurological outcomes. For example, motives of engagement whether it is enjoyable/wanted or simply to please their partner may modulate neurological changes over time.