Article| Volume 19, ISSUE 4, P599-603, October 01, 2009

# Assisted reproduction in women over 40years of age: how old is too old?

Published:August 02, 2010

## Abstract

Women’s fertility progressively declines with advanced age due to depletion of the ovarian follicular reserve and poorer oocyte quality. However, many women of advanced age are eager to conceive from their own ova. The aim of the present study was to evaluate the outcome of IVF cycles among older patients. All IVF retrievals performed in the unit in patients ⩾42 years old between 1998 and 2006 were retrospectively analysed. Data were compared with patients <35 years old treated in the unit during the same time period. A total of 843 IVF retrievals in 459 patients ⩾42 years old were assessed. Clinical pregnancy rates per cycle were 7.7%, 5.4% and 1.9% for 42, 43 and 44 years old, respectively. Only one IVF cycle in patients aged 44 years resulted in delivery. None of the 54 cycles performed in women of 45 years or older resulted in a pregnancy. A marked decline in clinical pregnancy and delivery rates, accompanied by an increase in spontaneous abortion rates, was found in patients ⩾42 years old. In view of these results and as the option of egg donation is a promising alternative with delivery rates close to 50%, it appears that IVF treatment should be limited to patients not older than 43 years old with adequate ovarian response.

## Introduction

Advanced maternal age is one of the most challenging fields in reproductive medicine. In most western countries, the average women’s age at marriage and child bearing is ever increasing. Moreover, many single women, mostly around their 40s are now seeking assisted procreation treatments in order to conceive.
Currently, in our centre, over 45% of the IVF cycles are performed in women over 35 years of age. Since natural fecundity also declines dramatically after this age, assisted reproduction in women over 40 years of age raises major medical, moral and social issues.
The progressive decline in female reproductive performance becomes more acute during the fourth decade of life. The hallmark of this decline in fertility is shortening of the menstrual cycle with concomitant increase in basal FSH concentrations. Consequently, assisted reproduction patients with advanced age demonstrate higher degrees of ovarian resistance to trophic hormone administration, decreased pregnancy rates, higher rates of pregnancy loss and higher rates of perinatal genetic anomalies in comparison with younger patient populations (
• Hull M.G.
• Fleming C.F.
• Hughes A.O.
• et al.
The age-related decline in female fecundity: a quantitative controlled study of implanting capacity and survival of individual embryos after in vitro fertilization.
,
• Dor J.
• Soriano D.
• Seidman D.S.
• et al.
High success rate achieved with intracytoplasmic sperm injection for male factor infertility.
,
• Márquez C.
• Sandalinas M.
• Bahce M.
• et al.
Chromosome abnormalities in 1255 cleavage-stage human embryos.
,
• Dal Prato L.
• Borini A.
• Cattoli M.
• et al.
Live birth after IVF in a 46-year-old woman.
).
In Israel, infertility treatments including IVF are covered by the state until the age of 45. On the other hand, other countries prohibit IVF practice even at a lower age limit.
The goal of the present study was to evaluate IVF outcome during a period of almost 8 years in patient populations over 40 years of age, and to help to define new guidelines and policy for infertility treatments in these age groups.

## Materials and methods

The research protocol was approved by the local internal review board. Medical records of all IVF female patients ⩾42 years of age and less than 35, treated between January 1998 and July 2006 (8.5 year period), were analysed retrospectively. Only cycles with fresh embryo transfer using autologous oocytes were included in the study. A control group of women <35 years old was chosen, as fertility rates decrease with increasing age of women and this age is a milestone in the literature.
Three different stimulation protocols were used for the induction of follicular growth, as previously described (
• Elizur S.E.
• Lerner-Geva L.
• Levron J.
• et al.
Factors predicting IVF treatment outcome: a multivariate analysis of 5310 cycles.
,
• Hourvitz A.
• Lerner-Geva L.
• Elizur S.E.
• et al.
Role of embryo quality in predicting early pregnancy loss following assisted reproductive technology.
). In brief, 51.7% (n = 2960) of the patients underwent a long protocol of gonadotrophin-releasing hormone analogue (GnRHa) followed by recombinant FSH (rFSH) or menotropins. In all, 34.6% (n = 1981) of the women underwent a suppression protocol using GnRH antagonists, and 12.3% (n = 704) underwent ovarian stimulation using the flare-up protocol. In 1.3% (n = 81) of cases, a spontaneous cycle was used. Human chorionic gonadotrophin (HCG) was administered when at least two follicles of 17 mm average diameter were observed by transvaginal ultrasound. Oocytes were harvested by transvaginal ultrasound-guided follicular puncture approximately 35–36 h after HCG administration. Conventional insemination or intracytoplasmic sperm injection (ICSI) were performed as indicated. Transfer of two to four of the best cleaved embryos was performed on day 3. In some cases when embryo quality was poor, more than four embryos were transferred after detailed discussion with the couple of the risks of multiple pregnancies. Only fresh embryo transfers were included. Progesterone supplementation was given to all women from 1 day after oocyte retrieval.
Positive pregnancy was defined as a serum HCG concentration of ⩾10 IU/l on day 12 after transfer. A clinical pregnancy was defined as the presence of a gestational sac and visualization of fetal heart beat by ultrasound 2 weeks later. Implantation rate was defined as the ratio between the total number of fetal sacs and the total number of embryos replaced.
Data were collected using the Statistics Package for Social Sciences (SPSS) version 11 for windows program (SPSS, Inc., Chicago, IL, USA). Chi-squared or the Fisher’s exact tests for categorical variables and the unpaired two-way Student’s t-test for continuous factors were used as appropriate. A P-value of <0.05 was considered significant.

## Results

During the period from January 1998 to July 2006, a total of 9081 treatment cycles were performed in the unit. The study group consisted of 843 (9.3%) treatment cycles that were performed in 459 patients aged 42–47 years old. Patients’ basic characteristics and data concerning the relevant oocyte retrieval cycle were similar for all patients aged 42–44 years old (with no difference between women in the different year groups).
Patients’ characteristics are shown in Table 1. The older patients had a significantly longer period of infertility, more previous IVF cycles, and a higher mean basal FSH than the younger group (all P < 0.001). The cancellation rate and gonadotrophin consumption were significantly higher in the older group, with a lower number of oocytes retrieved and fertilized (all P < 0.001; Table 2). Reasons for cycle cancellation were early lutealization, no ovarian response, no ova during retrieval and no fertilization.
Table 1Patient characteristics in the two age groups.
Age group (No. of cycles)
<35years (4883)42years (843)P-value
Mean age29.6 ± 3.143.9 ± 1.0<0.001
No. of previous cycles4.4 ± 3.89.7 ± 5.7<0.001
Duration of infertility (years)5.2 ± 6.68.2 ± 9.9<0.001
Basal FSH (IU/l)7.5 ± 4.99.7 ± 5.6<0.001
Table 2Characteristics of IVF cycles in the two age groups.
Age group (No. of cycles)
<35years (4883)42years (843)P-value
Duration of stimulation (days)10.8 ± 2.9910.2 ± 2.9<0.001
No. of ampoules per day3.1 ± 4.04.9 ± 2.1<0.001
Endometrium >7 mm on day of HCG (n)2705 (55.4)242 (28.7)<0.001
Cancellation rate (n)254 (5.2)122 (14.5)<0.001
No. of oocytes retrieved (mean ± SD)11.5 ± 6.65.5 ± 4.7<0.001
No. of fertilized oocytes (mean ± SD)6.3 ± 4.23.3 ± 3.2<0.001
No. of embryos transferred (mean ± SD)2.7 ± 1.12.8 ± 1.6NS
NS, not statistically significant.
Values in parentheses are percentages.
Table 3 describes the characteristics of the treatment cycles among patients ⩾42 years old according to the woman’s age. No differences were found for any of the parameters between the 42, 43 and 44 years age groups. Therefore, the data concerning these patients could be grouped together. The overall clinical pregnancy rate per retrieval in the 42–47 years old group was 5.3%, with a steep decline from 7.7% at the age of 42 to no clinical pregnancies in women of 45 years and older (Table 4). Out of five women (3.2%) at the age of 44 years who conceived (positive HCG), only one woman (0.6%) delivered. None of the 54 cycles performed in patients aged 45 years or older resulted in a pregnancy. Although the number of embryos transferred was similar for the two age groups (Table 2), clinical pregnancy rates were dramatically reduced in patients ⩾42 years old, with higher rates of biochemical pregnancies and spontaneous abortion (Table 4).
Table 3Patient characteristics stratified according to age in the older age group.
Age group (No. of cycles)
42 (n=336)43 (n=299)44 (n=154)45 (n=54)P-value
No significant differences were found among patients 42–44years old. P-values in the table represent the differences between patients ⩾45years old (column 5) and the other age groups (columns 2–4).
FSH day 3 IU/l9.3 ± 5.89.2 ± 3.610.9 ± 7.216.8 ± 10.20.03
No. of ampoules (mean ± SD)47.9 ± 22.249.4 ± 25.049.9 ± 24.466.9 ± 33.80.02
No. of oocytes retrieved (mean ± SD)6.2 ± 4.85.4 ± 4.74.7 ± 4.33.2 ± 2.7<0.001
Cycles with <4 oocytes retrieved (%) (n)82.1 (276)84.3 (252)89.0 (137)74.1 (40)0.001
No. of fertilized oocytes4.1 ± 3.13.6 ± 3.13.2 ± 3.02.5 ± 2.00.013
Assisted hatching (%) (n)69.3 (233)67.8 (203)71.6 (110)38.4 (21)0.02
No. of embryos transferred (mean ± SD)2.8 ± 1.52.7 ± 1.52.5 ± 1.42.0 ± 1.40.012
Cancellation (%) (n)12.2 (41)14.4 (43)13.0 (20)33.3 (18)0.02
Cycles with no oocytes retrieved (%) (n)5.1 (17)3.9 (12)4.4 (7)10.5 (6)0.03
a No significant differences were found among patients 42–44 years old. P-values in the table represent the differences between patients ⩾45 years old (column 5) and the other age groups (columns 2–4).
Table 4IVF pregnancy outcome according to the age of the patient per cycle.
Maternal age (years)
<35424344454647
No. of cycles48833362991544086
No. of patients1909186158872044
Positive βHCG/cycle (%) (n)29.4 (1435)12.2 (41)6.4 (19)3.2 (5)2.5 (1)00
Clinical pregnancies/cycle (%) (n)24.1 (1179)7.7 (26)5.4 (16)1.9 (3)000
Biochemical pregnancies/cycle (%) (n)4.7 (231)4.5 (15)1.0 (3)1.3 (2)2.5 (1)
Ectopic pregnancies/positive βHCG (%) (n)1.7 (25)
Artificial abortion/clinical pregnancies (%) (n)1.1 (13)7.7 (2)
Spontaneous abortion/clinical pregnancies (%) (n)12.8 (151)34.6 (9)37.5 (6)66.7 (2)
Delivery/cycle (%) (n)20.3 (990)4.2 (14)3.3 (10)0.6 (1)

## Discussion

The data show a significant decline in the ovarian response of the study group (⩾42 years old) and live birth rates. This was evident from the higher daily gonadotrophin consumption (4.9 ± 2.1 versus 3.1 ± 4.0, respectively), and lower number of oocytes retrieved (5.5 ± 4.7 versus 11.5 ± 6.6, respectively) compared with patients <35 years old.
A similar outcome was reported by
• Orvieto R.
• Bar-Hava I.
• Yoeli R.
• et al.
Results of in vitro fertilization cycles in women aged 43–45 years.
, who summarized their IVF results in 276 women (708 cycles) aged 43–45 years old. The overall pregnancy rate was 6.6%, with a delivery rate of 1.98%. Seventy per cent of these pregnancies, however, resulted in miscarriages. However, the authors did not provide details about 44- and 45-year-old subgroups of patients. According to the present data, their performance may well be far lower than the overall results mentioned.
• Klipstein S.
• Regan M.
• Ryley D.A.
One last chance for pregnancy: a review of 2,705 in vitro fertilization cycles initiated in women age 40 years and above.
reviewed 2705 IVF cycles initiated in women aged 40 years and above. At the age range of 41–43 years old, live birth rate per cycle was not different between patients. By age 44, however, the live birth rate per started cycle was significantly lower. Only one live birth occurred among women treated at the age of 45 years, and none occurred beyond this age limit.
• Bopp B.L.
• Alper M.M.
• Thompson I.E.
Success rates with gamete intrafallopian transfer and in vitro fertilization in women of advanced maternal age.
studied 554 IVF cycles and found a 5.1% live birth rate per started cycle in women aged 40–43 years old. There was no live birth beyond the 44th birthday. The authors concluded that women over 44 years of age are unlikely to benefit from IVF using their own eggs. In contrast to other studies cited,
• Spandorfer D.S.
• Bendikson K.
• Dragisic K.
• et al.
Outcome of in vitro fertilization in women aged 45 years and older who use autologous oocytes.
have shown that carefully selected women over the age of 44 years may have some success, especially those with better ovarian reserve with oocyte cohorts larger than five eggs. Obviously this subgroup of patients is very small. The variation between Spandorfer’s data and the present data could be due to a different patient population. In contrast to Spandorfer’s patients, the present ⩾42 years population was defined by repeated failure and low cancellation rates, as IVF is free. Another difference between these two studies is the different sample size, and the fact that the Cornell Medical Center is known for its high success IVF rates (
• Spandorfer S.D.
• Chung P.H.
• Kligman I.
• et al.
An analysis of the effect of age on implantation rates.
).
It appears, in fact, that maternal age is the most significant prognostic factor for IVF success or failure. Numerous studies and observations have consistently shown that oocytes from women over 40 years of age have reduced quality (
• Steer C.V.
• Mills C.L.
• Tan S.L.
• et al.
The cumulative embryo score a predictive embryo scoring technique to select the optimal number of embryos to transfer in an in-vitro fertilization and embryo transfer program.
,
• Orvieto R.
• Bar-Hava I.
• Yoeli R.
• et al.
Results of in vitro fertilization cycles in women aged 43–45 years.
). Embryos derived from such eggs show decreased implantation and survival potentials (
• Márquez C.
• Sandalinas M.
• Bahce M.
• et al.
Chromosome abnormalities in 1255 cleavage-stage human embryos.
). This age-related decline in female reproductive performance has been attributed to a variety of causes, including meiotic irregularities, mitochondrial dysfunction and global reduction in oocyte expression profiles (
• Steuerwald N.M.
• Bermúdez M.G.
• Wells D.
• et al.
Maternal age-related differential global expression profiles observed in human oocytes.
). Fertility decrease is mostly explained by quality loss of the oocyte, as documented by an increase in embryonic aneuploidy, decreased embryo implantation rates and enhanced risk of abortion (
• Navot D.
• Bergh P.A.
• Williams M.A.
• et al.
Poor oocyte quality rather than implantation failure as a cause of age-related decline in female fertility.
). In addition, small contributions to the reproductive ageing process are made by changes in menstrual cycle quality and decrease in endometrial function (
• Broekmans F.J.
Female age in ART: when to stop?.
).
In most developed countries worldwide, there is an ever increasing population of infertile women of advanced age.
The impressive improvements in assisted reproductive technologies over the last two decades have had little impact on the prognosis of women with advanced age. Treatment of such patients still yields low success rates and on top of that, even when a pregnancy is achieved, the toll of pregnancy loss is very high.
Diminished ovarian reserve is a problem that faces patients younger than 40 years too. It has been suggested that a critical number of follicles rather than a critical age determines the time of the menopause (
• Gosden R.G.
• Gougeon A.
• et al.
Accelerated disappearance of ovarian follicles in mid-life: implications for forecasting menopause.
). The time needed to reach the menopause (1000 remaining follicles, average age 51 years) starting from the critical point of 25,000 follicles (average age 37.5 years) is stable at around 13 years (
• Nikolaou D.
• Templeton A.
Early ovarian ageing: a hypothesis. Detection and clinical relevance.
). In addition, the largely genetic factors that determine the age of the menopause will also determine the age of all reproductive milestones that precede the menopause, including onset of subfertility.
A limitation of the current study is that it is retrospective. Although the large sample size may well describe results that can be found from a prospective study, bias may exist due to the nature of the study.
Since the patients are supported by public funds and were determined to try and conceive using their own ova, the cancellation rate was low. In the present study, cycles were not cancelled even if only one follicle was recruited. Therefore, cancellations were only as a result of non-response or, less frequently, as a result of premature ovulation.
When summarizing the mean total number of gonadotrophin ampoules used during the IVF cycle in each age group, and dividing it by the number of oocytes that were retrieved, it seems that three ampoules of gonadotrophins were needed in order to retrieve one ovum in the younger patients (<35), compared with 10 ampoules per ovum in patients >42 years old. In Israel, IVF is free for patients under 45 and paid for by the government for up to two children. This is the reason for low cancellation rates and recurrent trials even in older patients with reduced ovarian reserve. Consistent with the present data, the estimated costs expended on the patients during the study period have been calculated (Table 5).
Table 5IVF cycle cost according to age group.
Maternal age (years)
<3542434445
No. of cycles488333629915454
Cost (US$) of one cycle Cost of one cycle includes gonadotrophin-releasing hormone agonist, clinic assessment, oocyte retrieval and 50% ICSI. 27652765276527652765 Mean no. of total ampoules33.547.949.449.966.9 Mean cost (US$) of ampoules per patient

(@ US$30 per ampoule) 1004.41437148214972007 Deliveries/cycle (%)19.84.23.30.6 Mean cost (US$) for a child18,847100,047128,696710,333
a Cost of one cycle includes gonadotrophin-releasing hormone agonist, clinic assessment, oocyte retrieval and 50% ICSI.
The calculation of each oocyte retrieval cycle was based on the cost of GnRHa (US$200), gonadotrophin ampoules (US$30 per ampoule), clinic assessment (US$950), oocyte retrieval rates (US$1200) and ICSI rates (US$830 for each case). As ICSI was performed for 50% of patients, the ICSI expenditure was calculated as US$415 for each patient. The calculated cost of a live birth for patient ⩽35 years old was US$18,986. The estimated calculated costs per delivery at the ages of 42, 43 and 44 years old for our study group were US$100,728, US$126,835 and US$655,578, respectively.
Chambers et al. summarized the IVF costs of a live birth in Australia according to patient age (
• Chambers G.M.
• Sullivan E.A.
Assisted reproductive technology treatment costs of a live birth: an age stratified cost-outcome study of treatment in Australia.
). The authors claimed that the average cost for an assisted reproduction cycle was US$32,903, with a cost per live delivery of US$182,794 for women aged ⩾42 years.
Cost-effectiveness is hard to estimate, as having a child is mainly a cultural and emotional issue. Jewish tradition, for example, encourages the birth of children. Therefore, previous trials to limit the number of cycles funded by the state were politically unsuccessful and the Israeli parliament refused to pass such laws. The ethical dilemmas of physicians submitting women to a procedure with low success can be dealt with by discussing the subject with the women and arming them with the medical data to allow them to give true informed consent.
As oocyte donation today offers much higher pregnancy and delivery rates in older patients, IVF from autologous ova should be limited as a treatment option only to older patients with adequate ovarian response.

## References

• Bopp B.L.
• Alper M.M.
• Thompson I.E.
Success rates with gamete intrafallopian transfer and in vitro fertilization in women of advanced maternal age.
Fertility and Sterility. 1995; 63: 1278-1283
• Broekmans F.J.
Female age in ART: when to stop?.
Gynecologic and Obstetric Investigation. 2004; 58: 225-234
• Chambers G.M.
• Sullivan E.A.
Assisted reproductive technology treatment costs of a live birth: an age stratified cost-outcome study of treatment in Australia.
Medical Journal of Australia. 2006; 184: 155-158
• Dal Prato L.
• Borini A.
• Cattoli M.
• et al.
Live birth after IVF in a 46-year-old woman.
Reproductive BioMedicine Online. 2005; 11: 452-454
• Dor J.
• Soriano D.
• Seidman D.S.
• et al.
High success rate achieved with intracytoplasmic sperm injection for male factor infertility.
Israeli Journal of Obstetrics and Gynecology. 1997; 8: 169-172
• Elizur S.E.
• Lerner-Geva L.
• Levron J.
• et al.
Factors predicting IVF treatment outcome: a multivariate analysis of 5310 cycles.
Reproductive BioMedicine Online. 2005; 10: 645-649
• Gosden R.G.
• Gougeon A.
• et al.
Accelerated disappearance of ovarian follicles in mid-life: implications for forecasting menopause.
Human Reproduction. 1992; 7: 1342-1346
• Hourvitz A.
• Lerner-Geva L.
• Elizur S.E.
• et al.
Role of embryo quality in predicting early pregnancy loss following assisted reproductive technology.
Reproductive BioMedicine Online. 2006; 13: 504-509
• Hull M.G.
• Fleming C.F.
• Hughes A.O.
• et al.
The age-related decline in female fecundity: a quantitative controlled study of implanting capacity and survival of individual embryos after in vitro fertilization.
Fertility and Sterility. 1996; 65: 783-790
• Klipstein S.
• Regan M.
• Ryley D.A.
One last chance for pregnancy: a review of 2,705 in vitro fertilization cycles initiated in women age 40 years and above.
Fertility and Sterility. 2005; 84: 435-445
• Márquez C.
• Sandalinas M.
• Bahce M.
• et al.
Chromosome abnormalities in 1255 cleavage-stage human embryos.
Reproductive BioMedicine Online. 2000; 1: 17-26
• Navot D.
• Bergh P.A.
• Williams M.A.
• et al.
Poor oocyte quality rather than implantation failure as a cause of age-related decline in female fertility.
Lancet. 1991; 8: 1375-1377
• Nikolaou D.
• Templeton A.
Early ovarian ageing: a hypothesis. Detection and clinical relevance.
Human Reproduction. 2003; 18: 1137-1139
• Orvieto R.
• Bar-Hava I.
• Yoeli R.
• et al.
Results of in vitro fertilization cycles in women aged 43–45 years.
Gynecologic Endocrinology. 2004; 18: 75-78
• Spandorfer D.S.
• Bendikson K.
• Dragisic K.
• et al.
Outcome of in vitro fertilization in women aged 45 years and older who use autologous oocytes.
Fertility and Sterility. 2007; 87: 74-76
• Spandorfer S.D.
• Chung P.H.
• Kligman I.
• et al.
An analysis of the effect of age on implantation rates.
Journal of Assisted Reproduction and Genetics. 2000; 17: 303-306
• Steer C.V.
• Mills C.L.
• Tan S.L.
• et al.
The cumulative embryo score a predictive embryo scoring technique to select the optimal number of embryos to transfer in an in-vitro fertilization and embryo transfer program.
Human Reproduction. 1992; 7: 117-119
• Steuerwald N.M.
• Bermúdez M.G.
• Wells D.
• et al.
Maternal age-related differential global expression profiles observed in human oocytes.
Reproductive BioMedicine Online. 2007; 14: 700-708