We investigated six nonobese women ages 23 to 34 who met the Rotterdam criteria17
for diagnosis of PCOS. The women did not have any other medical problems. The patients were relatively lean, with body mass index (BMI) ranging from 18.5 to 26.6 (mean 22 ± 1.4) (
). All the women had a history of oligomenorrhea with less than five spontaneous periods per year since menarche. They gave a history of hirsutism that was improved by using either oral contraceptives and/or spironolactone. Ferriman–Gallwey scores18
ranged from 1 to 7. The low hirsutism scores reflected the treatments they had received. One patient had a history of a large ovarian cyst.
Baseline Characteristics and Hormone Levels
Measurements of follicular stimulating hormone (FSH), luteinizing hormone (LH), prolactin, dehydroepiandrosterone (DHEA) sulfate, total and free testosterone by liquid chromatography tandem mass spectrometry (LC/MS/MS), and 17 hydroxyprogesterone (LC/MS/MS) were performed on morning blood samples at Quest Diagnostics (Madison, NJ). Patients had normal LH, FSH, DHEA sulfate, prolactin, and 17 hydroxyprogesterone levels (). The total testosterone levels ranged from 8 to 44 ng/dl (normal range 2–45 ng/dl) and free testosterone levels ranged from 0.3 to 3.6 pg/ml (normal range 0.1–6.4 pg/ml). Four of the six patients were on oral contraceptives, and their free testosterone levels ranged from 0.3 to 2.6 pg/ml. The two women who were not on oral contraception at study entry had free testosterone levels of 3.6 and 1.6 pg/ml and total testosterone levels of 44 and 22 ng/dl. The ethnic background was five Caucasians and one Asian.
A standardized oral glucose tolerance test was performed to exclude the diagnosis of either diabetes or impaired glucose tolerance. Individuals with cardiovascular diseases, human immunodeficiency virus, other active infections, thyroid disorders, epilepsy, cancer, hepatitis, cystic fibrosis, sickle cell disease, asthma, or renal disease were also excluded. Subjects were not taking medications known to affect insulin sensitivity, carbohydrate metabolism, or lipid metabolism. These medications included glucocorticoids, adrenergic agonists, psychotropic drugs, diuretics, beta blockers, and 3-hydroxy-3-methylglutarylcoenzyme A reductase inhibitors.
Subjects were screened with respect to their exercise and physical activity pattern. Only subjects who performed less than 60 min of strenuous exercise per week were recruited—the rationale being that regular strenuous physical activity may attenuate the effect of CRLA on insulin sensitivity. Subjects underwent a dietary history at enrollment and were placed on a weight maintenance diet in order to avoid the confounding effect of weight loss on insulin sensitivity.
The women underwent baseline tests which included (a) a hyperinsulinemic, euglycemic clamp,19
for which a primed-continuous infusion of regular human insulin was administered at a rate of 40 mU/min/m2
/body surface area for 120 min, blood glucose levels were maintained at approximately basal level with a variable infusion of 20% glucose, average glucose infusion rates were determined during the steady state period between 90 and 120 min, and glucose disposal values (M/I) were calculated as milligrams of glucose infused per minute per kilogram of body mass divided by steady state insulin levels (in μU/ml × 100); (b) a cardiovascular risk panel, including low-density lipoprotein (LDL) pattern size and intermittent density lipoproteins measured using a vertical ultracentrifugation technique (VAP®
panel, Atherotech, Birmingham, AL), and highly sensitive C-reactive protein (hsCRP) and fasting homocysteine levels were also obtained; and (c) measurements of different serum oxidative stress markers, including isoprostane (Kronos laboratories), protein carbonyls, 2-thiobarbituric acid reactive substances (TBARS), and total reactive antioxidant potential (TRAP) (Antioxidants and Mass Spectrophotometry Core Laboratory of the University of California-Davis, Clinical Nutrition Research Unit).
After completion of the baseline tests, the subjects were administered CRLA 600 mg twice daily for 16 weeks, and the tests repeated. Controlled-release alpha lipoic acid was supplied by the Medical Research Institute, San Francisco, CA.
All subjects gave informed consent. The protocols and consent forms were approved by the University of California, San Francisco, institutional review board and by the Clinical and Translation Science Institute, where the study was conducted.
Comparisons between the baseline and final were made using a Student's t-test (two-tailed) and two paired signed rank sum test (Wilcoxon), in which values in each row of the data set were paired. Significance was accepted at p < .05. The p values stated are those obtained with Student's t-test. Nonparametric analyses of the data did not alter the conclusions. All analyses were performed using GraphPad Prism (MS Windows, version 4.03, GraphPad Software, San Diego, CA). Data are presented as mean ± standard error of the mean.