Serum and follicular fluid leptin during in vitro fertilization:: Relationship among leptin increase, body fat mass, and reduced ovarian response

The satiety factor leptin is expressed in several reproductive tissues, but its role in the control of reproductive physiology is not well understood. We studied leptin concentrations in the sera and follicle fluids of 52 women [body fat mass percentage (BFM%) range, 19.6-38.8%] undergoing pituitary down-regulation and ovarian hyperstimulation for in vitro fertilization (IVF) treatment. Fasting serum samples were collected 1) at maximal suppression before the initiation of gonadotropin treatment, 2) at maximal ovarian hyperstimulation, 3) at the time of oocyte retrieval, and 4) 16 days later when all subjects were under exogenous luteal support using 600 mg progesterone daily. Follicular fluid (FF) was obtained at oocyte retrieval from two representative preovulatory follicles in both ovaries. During ovarian hyperstimulation there was a significant 60% increase in serum leptin concentrations from 10.9 +/- 1.1 (SEM) to 15.7 +/- 1.5 ng/mL (P < 0.01) between suppression and maximal hyperstimulation, demonstrating that the ovarian functional state can affect serum leptin concentrations. A serum leptin increase of 22-198% during ovarian hyperstimulation was evident in 43 subjects, whereas in 9, leptin concentrations remained unchanged. A positive correlation between leptin change and BFM% (r = 0.55; P < 0.0005) was observed in the 43 leptin responders. The follicular fluid leptin level was similar to that in serum. In separate linear regression analysis, BFM% contributed to 59-64%, body mass index to 46-56%, and weight to 46-55% (all P < 0.001) of the variability in leptin concentrations at the 4 time points. The 20-fold increase in serum estradiol concentrations during IVF was not significantly correlated with changes in leptin concentrations. On the contrary, the relative serum leptin increase was negatively associated with the ovarian response to hyperstimulation, as revealed by the numbers of follicles (b = -0.28; r(2) = 8.1%; P < 0.05) and oocytes retrieved (b = -0.39; r(2) = 15.2%; P < 0.01). This relationship was further reflected in a positive correlation between the percent increases in leptin and FSH concentrations (r = 0.39; P < 0.01). The significant relationship of high leptin and reduced ovarian response was also maintained when the cumulative dose of FSH was used as a covariable. Reduced ovarian response was not a function of body mass index, BFM%, basal leptin levels, or insulin concentrations. Fasting serum insulin concentrations remained unchanged in response to IVF, but were positively correlated to serum leptin concentrations at all four time points. Our data suggest that leptin production may be influenced by the ovarian functional state. During IVF a high relative leptin increase is associated with adiposity and a reduced ovarian response. These observations support the possibility that high leptin concentrations might reduce ovarian responsiveness to gonadotropins. Hence, leptin might explain in part why obese individuals require higher amounts of gonadotropins than lean subjects to achieve ovarian hyperstimulation.