Second-stage labor duration, epidural analgesia, and early postpartum pelvic floor electromyography: a retrospective cohort study of 1,127 vaginal deliveries

Prolonged second stage of labor is usually analyzed dichotomously, and the association between epidural analgesia and pelvic floor function remains contested, with electromyographic evidence limited and unadjusted for labor duration. We quantified the association of second-stage duration with early postpartum pelvic floor muscle electrophysiology and reassessed the epidural association using standardized surface electromyography (sEMG). Single-center retrospective cohort of 1,127 singleton vaginal deliveries (January 2023 to January 2026) with Glazer-protocol pelvic floor sEMG at postpartum screening (median 49 days). Outcomes were fast-twitch maximum, sustained-contraction mean, and endurance mean amplitudes (µV). Multivariable linear models were mutually adjusted for forceps, perineal injury grade, epidural analgesia, parity, age, BMI, birth weight, and days postpartum, with a 50-dataset multiple-imputation sensitivity analysis and an exploratory bootstrap mediation analysis of the epidural association. Each additional 10 min of second stage was associated with − 0.39 µV of fast-twitch amplitude (95% CI − 0.59 to − 0.19, p < 0.001). No statistical evidence of departure from linearity was identified (P for nonlinearity = 0.55), although power was limited at the tails. The association was stronger in women tested within 42 days postpartum (− 2.20 vs. − 0.88 µV per 30 min; n = 247 vs. n = 880; interaction P = 0.042) and remained stable after multiple imputation. Forceps delivery was independently associated with − 6.08 µV ( p = 0.010); perineal injury grade showed no independent association ( p = 0.438). Epidural analgesia was not directly associated with fast-twitch amplitude (+ 0.93 µV, 95% CI − 1.29 to + 3.15, p = 0.413), sustained contraction ( p = 0.411), or resting tone ( p = 0.996), but was associated with a 20.3-minute longer second stage ( p < 0.001). In exploratory mediation analysis, the indirect association through longer second-stage duration was statistically significant (indirect effect − 0.79 µV, 95% CI − 1.26 to − 0.40), indicating that the lower sEMG amplitudes observed in epidural labors were statistically consistent with an indirect association through longer labor rather than with a detectable direct association of epidural analgesia itself. Second-stage duration showed a quantifiable, graded association with early postpartum pelvic floor muscle electrophysiology, with no statistical evidence of departure from linearity. No direct association between epidural analgesia and postpartum pelvic floor muscle electrophysiology was detected; absence of evidence is not evidence of safety, and the exploratory mediation result should be interpreted as a statistical association, not as a causal pathway. These findings are hypothesis-generating, and whether modifying second-stage management improves pelvic floor outcomes requires prospective evaluation.