Sexual-Function Outcomes in Women With SCI Using Epidural SCS: Evidence for Population-Specific Assessment Tools
DOI: Pending DOI
Synopsis
Sexual function remains a major unmet need for women with spinal cord injury (SCI), yet most neuromodulation trials rely on population-agnostic surveys that fail to reflect post-injury sexual experiences. In a post-hoc exploratory analysis of women with chronic SCI undergoing epidural spinal cord stimulation (eSCS), sexual distress decreased over time, with mixed-effects modeling identifying satisfaction and desire as the main contributors to measured change. These findings illustrate how conventional instruments may mask stimulation-related effects by weighting domains less relevant, or differently expressed, in SCI. This supports the need for population-specific measures and neuromodulation-sensitive biomarkers in future eSCS studies of autonomic and sexual function.
Background
Sexual function is consistently cited as a top priority for restoration among women with spinal cord injury (SCI), yet clinical trials rarely include robust measures of female sexual outcomes. Epidural spinal cord stimulation (eSCS), originally developed for pain, has shown emerging benefits across autonomic domains, including micturition, bowel function, and sexual responses. Early reports suggest that eSCS may modulate sensory, vascular, and segmental circuits relevant to arousal and orgasm, but evidence in women remains sparse and is constrained by reliance on patient-reported outcome measures (PROMs) validated solely in able-bodied populations.
Prominent instruments such as the Female Sexual Function Index (FSFI) may not account for injury-level–specific sensory changes, altered autonomic responses, or divergent definitions of satisfaction within the SCI community. As neuromodulation trials increasingly target autonomic and pelvic-organ function, the field requires outcome measures capable of distinguishing SCI-specific mechanisms from general well-being. This exploratory analysis examines trajectories of sexual-distress scores in women with SCI using long-term eSCS and evaluates which PROM domains most contribute to distress, highlighting gaps in current measurement practices and opportunities for more targeted, physiologically anchored endpoints.
Methods
We performed a post-hoc exploratory analysis of female participants enrolled in the E-STAND trial, a single-arm prospective study evaluating chronic eSCS following motor complete SCI. Participants completed sexual-function questionnaires across ~13 monthly follow-ups, including the Female Sexual Distress Scale–Revised (FSDS-R), Female Sexual Function Index (FSFI), Orgasm Rating Scale (ORS), Neurogenic Bladder Symptom Score (NBSS), and Neurogenic Bowel Dysfunction Score (NBDS). Longitudinal trends in sexual distress were examined using Pearson correlations. To identify which domains contributed most to distress, we constructed generalized linear mixed-effects (GLME) models using participant-level random intercepts. Model reduction proceeded from composite scores to domain-level predictors to identify the most parsimonious model. Missing data were handled via listwise deletion due to intermittent questionnaire deployment and variable completion across early visits.
Results
Seven female participants contributed longitudinal PROM data (93 complete observations). At baseline, six of seven had FSDS-R scores ≥11, indicating clinically significant sexual distress. Across the 13-month follow-up period, FSDS-R scores decreased significantly (p < 0.01), although individual trajectories varied. Three participants demonstrated distress levels that fell below the clinical cutoff by study end.
Model 1 (Composite Scores): In the initial GLME including composite scores from sexual-function (FSFI, ORS), bladder (NBSS), bowel (NBDS) surveys, and visit number, only FSFI (p = 0.001) and Visit (p = 0.046) significantly predicted FSDS-R. Bladder, bowel, and orgasm-quality composites showed no relationship to distress, suggesting that autonomic changes outside sexual domains did not drive the observed improvements.
Model 2 (FSFI Domain Decomposition): When the FSFI was separated into its six validated domains, only desire (p = 0.007) and satisfaction (p < 0.001) remained significant predictors of distress. Other domains traditionally emphasized in able-bodied cohorts—arousal, lubrication, pain—did not contribute meaningfully. Orgasm approached but did not reach significance (p = 0.074).
Final Model: The most parsimonious model retained Visit, Desire, and Satisfaction. Satisfaction showed the strongest negative association with distress, whereas Desire unexpectedly exhibited a positive association, suggesting that rising desire in the context of persistent orgasmic or sensory limitations may heighten distress in neurogenic populations. Model fit was strong (R² = 0.84), with predicted trajectories closely approximating observed scores.
Together, these findings indicate that only a subset of FSFI domains capture variance in distress among women with SCI undergoing eSCS, and that domain hierarchy differs substantially from patterns observed in able-bodied validation cohorts.
Discussion
Neuromodulation trials increasingly target autonomic circuits, yet outcome measurement for female sexual function remains anchored in tools developed for non-neurogenic populations. Our exploratory findings show that although sexual distress decreased during long-term eSCS use, only select FSFI domains, desire and satisfaction, accounted for variance in distress. Domains traditionally linked to autonomic or genital-vascular mechanisms, such as arousal or lubrication, did not meaningfully contribute. This divergence underscores a mismatch between validated PROM frameworks and the sensory, autonomic, and behavioral realities of women with SCI.
For a field moving toward circuit-specific neuromodulation of pelvic and autonomic functions, reliance on PROMs that fail to reflect injury-dependent physiology risks mischaracterizing therapeutic effects. These results highlight the need for (1) SCI-specific PRO development, (2) integration of objective physiologic endpoints (e.g., genital blood-flow measures, autonomic markers, evoked-response signatures), and (3) multimodal designs capable of distinguishing stimulation-driven improvements from psychosocial contributors to sexual well-being. Establishing measurement tools that align with neurogenic mechanisms is essential for accurately evaluating emerging neuromodulation therapies in women’s sexual health after neurological damage.
Disclosures
Supported by the Minnesota Office of Higher Education SCI/TBI Research Grant Program; E.C.B. received partial salary support from a T32 training grant. No relevant financial relationships or equity interests.