0021-972X/90/7106-1525$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright© 1990 by The Endocririe Society
Vol. 71, No. 6 Printed in U.S.A.
Effect of Age on Response to Human Menopausal Gonadotropin Stimulation* SAMUEL L. JACOBS, DEBORAH A. METZGER, WILLIAM C. DODSON, AND A. F. HANEY Divisions of Reproductive Endocrinology and Infertility, Departments of Obstetrics and Gynecology, University of Connecticut Health Center (S.L.J., D.A.M.), Farmington, Connecticutt 06032 Pennsylvania State University, Hershey Medical Center (W.C.D.), Hershey, Pennsylvania 17033; and Duke University Medical Center (A.F.H.), Durham, North Carolina 27710
ABSTRACT. Conception rates decline in the latter part of the reproductive years. To examine which ovarian parameters are altered with aging, 486 cycles from 225 ovulatory infertile women undergoing human menopausal gonadotropin (hMG) superovulation and washed irjtrauterine insemination were analyzed. Infertility factors included endometriosis (68%), unexplained infertility (8.4%), malfc factor (12.9%), and ovulatory dysfunction (10.7%). Parameters that demonstrated a linear relationship with increasing age included numbers of ampules of hMG required per cycle (r = 0.79; P < 0.05), days of stimulation (r = 0.73; P < 0.01), estradiol level at the time of hCG (r = -0.92; P < 0.0001), number of follicles larger than 15 mm (r = -0.61; P < 0.05), and rate of rise of estradiol (r = -0.92; P < 0.0001). These same age-depepdent changes were observed in women receiving a standard stimulation protocol (3 ampules hMG be-
T
HE CURRENT trend toward delayed childbearing has focused attention on changes in the natural fecundity of women as they age. Prior studies have demonstrated that a woman's ability to become pregnant declines steadily with age (1). The mechanism of this phenomenon as well as the sequence of events leading to declining reproductive function remain enigmatic. Based on animal models, the decline in reproductive capacity with age has been attributed to a progressive depletion of ovarian follicles (2), production of defective oocytes (3, 4), and reduced competence of the required organs and systems to sustain gestation successfully (5). Response to human menopausal gonadotropins (hMG) in women over 40 yr has been reported to be both similar (6) and inferior (7-9) to that in younger women. In those studies where an age-dependent difference was noted, older women had fewer oocytes retrieved (7-9) and a higher miscarriage rate (10). However, it is unclear from Received March 19,1990. Address requests for reprints to: Deborah A. Metzger, Ph.D., M.D., Department of Obstetrics and Gynecology, L-2090, University of Connecticut Health Center, Farijnington, Connecticutt 06032. * Presented at the 36th Annual Meeting of the Society for Gynecologic Investigation, San Diego, CA, March 15-18, 1989.
ginning on cycle day 2). When standard cycles were limited to the first cycle only, the preovulatory estradiol (r = —0.92; P < 0.005), slope of estradiol rise (r = -0.92; P < 0.005), and number of preovulatory follicles (r = -0.92; P < 0.005) still showed a significant decrease with age. Although the mean estradiol level per preovulatory follicle showed a slight decrease with maternal age, no statistically significant trend was noted. In addition, the cycle day of hCG administration was unaffected by age. With advancing age, there appears to be a decreased ovarian response to an increased amount of stimulation, as measured by steroidogenesis and follicular recruitment; yet the estradiol/follicle remains unaltered, indicating continued health of the follicle. These observations may explain in part the observed decrease in fecundity in older women. (J Clin Endocrinol Metab 7 1 : 1525-1530,1990)
these studies how age is related to alteration in oocyte number and quality. Fecundability in spontaneous cycles has been reported to decrease in a nonlinear function with age, with a significant decrease noted after age 35 yr (11). Whether alterations in ovarian function parallel these changes in fecundity has not been previously explored. hMG induction of ovulation offers the ability to examine several parameters that may contribute to reproductive capacity. First, hMG stimulates the development of a cohort of follicles, which may represent all or a representative fraction of the recruitable follicles in the growing population. This number is an indirect measure of the follicular store available in the ovary (2). Second, other parameters, such as estradiol levels and the amount and duration of hMG stimulation required, give an indication of the health of the follicular apparatus. Recently, controlled hMG superovulation combined with intrauterine insemination has been used to enhance fertility for a variety of infertility conditions (12). These ovulatory patients have made it possible to study specific ovarian parameters in a large number of cycles and, thus, correlate these parameters with advancing age.
1525
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 10:10 For personal use only. No other uses without permission. . All rights reserved.
JACOBS ET AL.
1526
Materials and Methods The subjects of this study were 225 ovulatory infertility patients, aged 23-44 yr, who underwent hMG superovulation and washed intrauterine insemination at Duke University Medical Center between 1986 and 1988. All patients underwent a standard infertility evaluation, and a total of 486 stimulated cycles were studied. The patients' diagnoses included endometriosis (68.0%), male factor (12.9%), ovulatory dysfunction (10.7%), unexplained infertility (8.4%), pelvic adhesions/tubal factor (5.3%), fibroids (3.6%), and other causes (5.8%). The age distribution for each of the diagnostic categories was similar. The overall age distribution of the patients is shown in Fig. 1. Women were initially stimulated with hMG in a standard fashion, with three ampules administered im on a daily basis beginning on cycle day 2. Depending on response, subsequent cycles were started later in the cycle or with a lower or higher hMG dose to modify the ovarian response. Daily monitoring of the ovarian response to stimulation was begun on cycle day 5 or 6, using serial abdominal (until April 1987) or vaginal (after April 1987) ultrasounds and serum estradiol measurements. Approximately half of the patients were scanned abdominally, and half vaginally. No significant differences were noted in the age distribution of the ultrasound modality. Changes in hMG dose were made depending on the ovarian response. When one or more follicles measured 16 mm in maximum diameter, hCG (5000 IU) was administered im. Cycles were cancelled if, despite an increase in hMG dose, estradiol levels failed to rise above baseline levels for 7-10 days or if ultrasound showed no follicular growth beyond 10 mm. This was considered to be a poor response. Additionally, cycles were cancelled if there was considerable risk of hyperstimulation, i.e. an estradiol level greater than 7350 pmol/L and/or more than seven to ten intermediate sized follicles at the time of hCG administration. For some comparisons, patients were divided into 4 age
JCE & M • 1990 Vol 71 • No 6
groups: 23-29, 30-34, 35-39, and 40-44 yr. Otherwise, patient parameters were averaged by 2-yr intervals (odd year rounded up). Cycles were then analyzed for the effects of age on a variety of dependent variables. These parameters included 1) amount of hMG required per cycle, 2) number of days of hMG stimulation, 3) day of hCG administration, 4) preovulatory estradiol level, 5) number of preovulatory follicles, 6) slope of estradiol rise, and 7) estradiol per preovulatory follicle. There were 431 cycles that went to completion. Of these, 272 were standard cycles in which 3 amps of hMG were begun on cycle day 2. First cycles refer to initial standard cycles only and do not include subsequent cycles (n = 146). These age-dependent variables were analyzed using the Student's t test. Each age group was compared to the 23-29 yr age group for all dependent variables. Pearson product-moment correlation coefficients were also determined from the means for each 2-yr age interval in order to analyze the relationship between age and the various parameters. The level of significance was set at P < 0.05. All values are expressed as the mean ± SEM.
Results There were 225 patients who underwent 486 hMGstimulated cycles. Of these, 431 (88.7%) cycles went to completion, and 55 (11.3%) were cancelled. Reasons for cancellation were poor response in 26 (47.3%) cycles and excessive stimulation in 29 (52.7%) cycles. The frequency of cancelled cycles due to poor stimulation increased with age, as shown in Table 1. In contrast, the risk of cancellation from excessive stimulation decreased with increasing age. There was some variability in age-dependent relationships depending on the type of cycles examined. Three categories of completed cycles were examined to evaluate
completed cycles standard cycles first cycles
FIG. 1. The age distribution of all completed cycles. Cancelled cycles were excluded. Standard cycles are those with an initial hMG dose of three ampules on cycle day 2. First cycles refer to the initial standard cycle only. The data represent the number of patients by 2-yr intervals, with odd years rounded,up.
22 24 26 28 30 32 34 36 38 40 42 44 46 Age (years)
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 10:10 For personal use only. No other uses without permission. . All rights reserved.
EFFECT OF AGE ON hMG STIMULATION
1527
TABLE 1. Cancelled cycles
Age groups (yr) Poor stimulation Excess stimulation Total cancelled
Overall
23-29
30-34
35-39
40-44
3/39 (3.4) 8/89 (9.0)
8/213 (3.8) 17/213 (8.0)
13/144 (9.0)° 4/144 (2.8)"
2/40 (5.0) 0/40 (0)
26/486 (5.3) 29/486 (6.0)
11/89 (12.4)
25/213 (11.8)
17/144 (11.8)
2/40 (5.0)
55/486 (11.3)
Percentages are given in parentheses. P < 0.05 vs. age 23-29 yt, by x 2 analysis.
0
the data for confounding variables: 1) first cycles with standard hMG dose, %) all standard cycles, and 3) all completed cycles. First cycles were examined in order to eliminate the bias of altered dose and the attrition of women who achieved pregnancy or those who did not begin subsequent cycles because of poor response. Standard cycles were examined to control for changing dose and day of start of thefapy. The data from all completed cycles were used to observe trends in response where the administration of hMG had been presumably optimized. When cycles were standardized for dose and starting day, there was no trend observed in total hMG dose with increasing age either $mong first cycles (r = 0.19; P = NS) or for all standard cycles (r = 0.52; P = NS; Fig. 2A). In contrast, when all completed cycles were examined, there was a significant correlation between the total dose of hMG needed to achieve adequate stimulation and increasing age (r = 0.79; P < 0.05; Fig. 2A).
40 1
30 25 20 • completed: R=0.79; p15 mm) follicles (Fig. 3C) also decreased linearly with age. In addition to the day of hCG administration, the only
• completed: R=0.73; p
Vol. 71, No. 6 Printed in U.S.A.
Effect of Age on Response to Human Menopausal Gonadotropin Stimulation* SAMUEL L. JACOBS, DEBORAH A. METZGER, WILLIAM C. DODSON, AND A. F. HANEY Divisions of Reproductive Endocrinology and Infertility, Departments of Obstetrics and Gynecology, University of Connecticut Health Center (S.L.J., D.A.M.), Farmington, Connecticutt 06032 Pennsylvania State University, Hershey Medical Center (W.C.D.), Hershey, Pennsylvania 17033; and Duke University Medical Center (A.F.H.), Durham, North Carolina 27710
ABSTRACT. Conception rates decline in the latter part of the reproductive years. To examine which ovarian parameters are altered with aging, 486 cycles from 225 ovulatory infertile women undergoing human menopausal gonadotropin (hMG) superovulation and washed irjtrauterine insemination were analyzed. Infertility factors included endometriosis (68%), unexplained infertility (8.4%), malfc factor (12.9%), and ovulatory dysfunction (10.7%). Parameters that demonstrated a linear relationship with increasing age included numbers of ampules of hMG required per cycle (r = 0.79; P < 0.05), days of stimulation (r = 0.73; P < 0.01), estradiol level at the time of hCG (r = -0.92; P < 0.0001), number of follicles larger than 15 mm (r = -0.61; P < 0.05), and rate of rise of estradiol (r = -0.92; P < 0.0001). These same age-depepdent changes were observed in women receiving a standard stimulation protocol (3 ampules hMG be-
T
HE CURRENT trend toward delayed childbearing has focused attention on changes in the natural fecundity of women as they age. Prior studies have demonstrated that a woman's ability to become pregnant declines steadily with age (1). The mechanism of this phenomenon as well as the sequence of events leading to declining reproductive function remain enigmatic. Based on animal models, the decline in reproductive capacity with age has been attributed to a progressive depletion of ovarian follicles (2), production of defective oocytes (3, 4), and reduced competence of the required organs and systems to sustain gestation successfully (5). Response to human menopausal gonadotropins (hMG) in women over 40 yr has been reported to be both similar (6) and inferior (7-9) to that in younger women. In those studies where an age-dependent difference was noted, older women had fewer oocytes retrieved (7-9) and a higher miscarriage rate (10). However, it is unclear from Received March 19,1990. Address requests for reprints to: Deborah A. Metzger, Ph.D., M.D., Department of Obstetrics and Gynecology, L-2090, University of Connecticut Health Center, Farijnington, Connecticutt 06032. * Presented at the 36th Annual Meeting of the Society for Gynecologic Investigation, San Diego, CA, March 15-18, 1989.
ginning on cycle day 2). When standard cycles were limited to the first cycle only, the preovulatory estradiol (r = —0.92; P < 0.005), slope of estradiol rise (r = -0.92; P < 0.005), and number of preovulatory follicles (r = -0.92; P < 0.005) still showed a significant decrease with age. Although the mean estradiol level per preovulatory follicle showed a slight decrease with maternal age, no statistically significant trend was noted. In addition, the cycle day of hCG administration was unaffected by age. With advancing age, there appears to be a decreased ovarian response to an increased amount of stimulation, as measured by steroidogenesis and follicular recruitment; yet the estradiol/follicle remains unaltered, indicating continued health of the follicle. These observations may explain in part the observed decrease in fecundity in older women. (J Clin Endocrinol Metab 7 1 : 1525-1530,1990)
these studies how age is related to alteration in oocyte number and quality. Fecundability in spontaneous cycles has been reported to decrease in a nonlinear function with age, with a significant decrease noted after age 35 yr (11). Whether alterations in ovarian function parallel these changes in fecundity has not been previously explored. hMG induction of ovulation offers the ability to examine several parameters that may contribute to reproductive capacity. First, hMG stimulates the development of a cohort of follicles, which may represent all or a representative fraction of the recruitable follicles in the growing population. This number is an indirect measure of the follicular store available in the ovary (2). Second, other parameters, such as estradiol levels and the amount and duration of hMG stimulation required, give an indication of the health of the follicular apparatus. Recently, controlled hMG superovulation combined with intrauterine insemination has been used to enhance fertility for a variety of infertility conditions (12). These ovulatory patients have made it possible to study specific ovarian parameters in a large number of cycles and, thus, correlate these parameters with advancing age.
1525
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 10:10 For personal use only. No other uses without permission. . All rights reserved.
JACOBS ET AL.
1526
Materials and Methods The subjects of this study were 225 ovulatory infertility patients, aged 23-44 yr, who underwent hMG superovulation and washed intrauterine insemination at Duke University Medical Center between 1986 and 1988. All patients underwent a standard infertility evaluation, and a total of 486 stimulated cycles were studied. The patients' diagnoses included endometriosis (68.0%), male factor (12.9%), ovulatory dysfunction (10.7%), unexplained infertility (8.4%), pelvic adhesions/tubal factor (5.3%), fibroids (3.6%), and other causes (5.8%). The age distribution for each of the diagnostic categories was similar. The overall age distribution of the patients is shown in Fig. 1. Women were initially stimulated with hMG in a standard fashion, with three ampules administered im on a daily basis beginning on cycle day 2. Depending on response, subsequent cycles were started later in the cycle or with a lower or higher hMG dose to modify the ovarian response. Daily monitoring of the ovarian response to stimulation was begun on cycle day 5 or 6, using serial abdominal (until April 1987) or vaginal (after April 1987) ultrasounds and serum estradiol measurements. Approximately half of the patients were scanned abdominally, and half vaginally. No significant differences were noted in the age distribution of the ultrasound modality. Changes in hMG dose were made depending on the ovarian response. When one or more follicles measured 16 mm in maximum diameter, hCG (5000 IU) was administered im. Cycles were cancelled if, despite an increase in hMG dose, estradiol levels failed to rise above baseline levels for 7-10 days or if ultrasound showed no follicular growth beyond 10 mm. This was considered to be a poor response. Additionally, cycles were cancelled if there was considerable risk of hyperstimulation, i.e. an estradiol level greater than 7350 pmol/L and/or more than seven to ten intermediate sized follicles at the time of hCG administration. For some comparisons, patients were divided into 4 age
JCE & M • 1990 Vol 71 • No 6
groups: 23-29, 30-34, 35-39, and 40-44 yr. Otherwise, patient parameters were averaged by 2-yr intervals (odd year rounded up). Cycles were then analyzed for the effects of age on a variety of dependent variables. These parameters included 1) amount of hMG required per cycle, 2) number of days of hMG stimulation, 3) day of hCG administration, 4) preovulatory estradiol level, 5) number of preovulatory follicles, 6) slope of estradiol rise, and 7) estradiol per preovulatory follicle. There were 431 cycles that went to completion. Of these, 272 were standard cycles in which 3 amps of hMG were begun on cycle day 2. First cycles refer to initial standard cycles only and do not include subsequent cycles (n = 146). These age-dependent variables were analyzed using the Student's t test. Each age group was compared to the 23-29 yr age group for all dependent variables. Pearson product-moment correlation coefficients were also determined from the means for each 2-yr age interval in order to analyze the relationship between age and the various parameters. The level of significance was set at P < 0.05. All values are expressed as the mean ± SEM.
Results There were 225 patients who underwent 486 hMGstimulated cycles. Of these, 431 (88.7%) cycles went to completion, and 55 (11.3%) were cancelled. Reasons for cancellation were poor response in 26 (47.3%) cycles and excessive stimulation in 29 (52.7%) cycles. The frequency of cancelled cycles due to poor stimulation increased with age, as shown in Table 1. In contrast, the risk of cancellation from excessive stimulation decreased with increasing age. There was some variability in age-dependent relationships depending on the type of cycles examined. Three categories of completed cycles were examined to evaluate
completed cycles standard cycles first cycles
FIG. 1. The age distribution of all completed cycles. Cancelled cycles were excluded. Standard cycles are those with an initial hMG dose of three ampules on cycle day 2. First cycles refer to the initial standard cycle only. The data represent the number of patients by 2-yr intervals, with odd years rounded,up.
22 24 26 28 30 32 34 36 38 40 42 44 46 Age (years)
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 10:10 For personal use only. No other uses without permission. . All rights reserved.
EFFECT OF AGE ON hMG STIMULATION
1527
TABLE 1. Cancelled cycles
Age groups (yr) Poor stimulation Excess stimulation Total cancelled
Overall
23-29
30-34
35-39
40-44
3/39 (3.4) 8/89 (9.0)
8/213 (3.8) 17/213 (8.0)
13/144 (9.0)° 4/144 (2.8)"
2/40 (5.0) 0/40 (0)
26/486 (5.3) 29/486 (6.0)
11/89 (12.4)
25/213 (11.8)
17/144 (11.8)
2/40 (5.0)
55/486 (11.3)
Percentages are given in parentheses. P < 0.05 vs. age 23-29 yt, by x 2 analysis.
0
the data for confounding variables: 1) first cycles with standard hMG dose, %) all standard cycles, and 3) all completed cycles. First cycles were examined in order to eliminate the bias of altered dose and the attrition of women who achieved pregnancy or those who did not begin subsequent cycles because of poor response. Standard cycles were examined to control for changing dose and day of start of thefapy. The data from all completed cycles were used to observe trends in response where the administration of hMG had been presumably optimized. When cycles were standardized for dose and starting day, there was no trend observed in total hMG dose with increasing age either $mong first cycles (r = 0.19; P = NS) or for all standard cycles (r = 0.52; P = NS; Fig. 2A). In contrast, when all completed cycles were examined, there was a significant correlation between the total dose of hMG needed to achieve adequate stimulation and increasing age (r = 0.79; P < 0.05; Fig. 2A).
40 1
30 25 20 • completed: R=0.79; p15 mm) follicles (Fig. 3C) also decreased linearly with age. In addition to the day of hCG administration, the only
• completed: R=0.73; p
