Toward a better measure of midlife sexual function: pooled analyses in nearly 1,000 women participating in MsFLASH randomized trials.

Abstract

Evaluate appropriateness of the current Female Sexual Function Index (FSFI)-19 value of <26.6 to designate female sexual dysfunction (FSD) in postmenopausal women, using the Female Sexual Distress-Revised (FSDS-R) scale to measure distress. Participant-level data containing standardized measures from five completed Menopause Strategies: Finding Lasting Answers for Symptoms and Health trials was pooled. Baseline characteristics and FSFI-19 scores were compared across trials (F-test, homogeneity). FSFI-19 score associations with the FSDS-R were described. Receiver operating characteristic (ROC) curves were plotted to illustrate the choice of optimal FSFI-19 value to predict sexual distress. ROC curves were also estimated adjusting for trial number, clinical center, age, education, race, smoking, and BMI. Nine hundred ninety eight women (79.2% postmenopausal), mean age 55.9 (SD 4.8) had complete FSFI-19, FSDS-R, and covariate data. Baseline mean FSFI-19 score among all participants and sexually active participants was 18.7 (SD 9.5) and 22.0 (SD 7.2), respectively. There was a consistent pattern across the trials of inverse association between poorer sexual function (FSFI-19) and greater sexual distress. Based on the ROC curve showing the likelihood of FSDS-R frequent or greater distress according to cut points of FSFI, the optimal cut point for FSD was FSFI-19 <21 for all participants. This cut point corresponded to sensitivity 87.2% (95% CI, 83.4-91.0), specificity 57.9% (95% CI, 54.3-61.6) and adjusted area under the ROC curve 78.8% (95% CI, 75.8-81.8). A new FSFI-19 cut point of ≥21 should be considered to describe normal sexual function in periand postmenopausal women as opposed to the standard cut point of >26.6. http://links.lww.com/MENO/A915.