Dual-Energy X-ray Absorptiometry Percent Fat Z-score As A Predictor Of Female Athlete Menstrual Status

In female athletes, menstrual dysfunction is a common consequence of low energy availability in the syndrome of Relative Energy Deficiency in Sport (REDs). Weight gain with subsequent resumption of menses is an important treatment outcome in female athletes recovering from REDs, but an appropriate body mass index (BMI) or % expected body weight (%EBW) target is difficult to determine, particularly as some oligo/amenorrheic athletes have BMI or %EBW in the “normal” range. Body composition, particularly relative fat reserve, may be a useful marker of adequate energy storage to support normal reproductive hormone function. PURPOSE: To determine if dual-energy x-ray absorptiometry (DXA)-derived %fat Z-score can predict menstrual status in female athletes and if a %fat Z-score cut-off of <-1.0 is clinically useful. METHODS: 414 female athletes (ages 14.7-30 yrs) who presented to a tertiary care center for bone health evaluation were classified as amenorrheic (n = 174), oligomenorrheic (n = 78), or eumenorrheic (n = 162) based on menstrual period frequency. We compared %fat Z-score, BMI, and %EBW values across menstrual groups and calculated sensitivity and specificity for amenorrhea and oligo/amenorrhea using traditional BMI/%EBW risk cut-offs (<90% EBW for athletes ≤20.0 yrs and BMI <18.5 for athletes >20.0 yrs) versus a risk cutoff of <-1.0 for %fat Z-score (chosen a priori). RESULTS: There was a decrease across menstrual groups (eumenorrheic to amenorrheic) for mean (SD) %fat Z-score: -0.53 (0.74), -0.80 (0.74), -1.36 (0.74); BMI: 21.5 (2.6), 20.7 (1.8), 19.8 (1.9); and %EBW: 103.4% (12.9%), 98.5% (8.8%), 93.9% (8.2%). Differences were more pronounced for %fat Z-score than BMI or %EBW. Although more specific, sensitivity for predicting amenorrhea using traditional cut-offs of %EBW or BMI was poor (27.6%) versus using a %fat Z-score cut-off <-1 (67.8%), p < 0.001. Sensitivity for predicting oligo/amenorrhea was also better using %fat Z-score (57.1%) than BMI/%EBW (24.6%), p < 0.0001. CONCLUSIONS: Percent fat Z-score < -1 may be a helpful clinical tool to predict menstrual dysfunction and a %fat Z-score > 1, rather than BMI or %EBW, may be a better goal for oligo-amenorrheic patients attempting to resume menses. Further prospective work is needed to determine the clinical utility of %fat Z-score in athletes with REDs.