Magnitude of daily energy deficit predicts frequency but not severity of menstrual disturbances associated with exercise and caloric restriction

We assessed the impact of energy deficiency on menstrual function using controlled feeding and supervised exercise over four menstrual cycles (1 baseline and 3 intervention cycles) in untrained, eumenorrheic women aged 18-30 yr. Subjects were randomized to either an exercising control (EXCON) or one of three exercising energy deficit (ED) groups, i.e., mild (ED1; -8 +/- 2%), moderate (ED2; -22 +/- 3%), or severe (ED3; -42 +/- 3%). Menstrual cycle length and changes in urinary concentrations of estrone-1-glucuronide, pregnanediol glucuronide, and midcycle luteinizing hormone were assessed. Thirty-four subjects completed the study. Weight loss occurred in ED1 (-3.8 +/- 0.2 kg), ED2 (-2.8 +/- 0.6 kg), and ED3 (-2.6 +/- 1.1 kg) but was minimal in EXCON (-0.9 +/- 0.7 kg). The overall sum of disturbances (luteal phase defects, anovulation, and oligomenorrhea) was greater in ED2 compared with EXCON and greater in ED3 compared with EXCON AND ED1. The average percent energy deficit was the main predictor of the frequency of menstrual disturbances (f = 10.1, beta = -0.48, r(2) = 0.23, P = 0.003) even when weight loss was included in the model. The estimates of the magnitude of energy deficiency associated with menstrual disturbances ranged from -22 (ED2) to -42% (ED3), reflecting an energy deficit of -470 to -810 kcal/day, respectively. This is the first study to demonstrate a dose-response relationship between the magnitude of energy deficiency and the frequency of exercise-related menstrual disturbances; however, the severity of menstrual disturbances was not dependent on the magnitude of energy deficiency.