Having a baby when aged 45 years or older remains uncommon in Australia, at 0.1–0.2% of all births [
28]. In this study, 0.16% of women giving birth in Victoria, during the study period, were in this age group. However, rates of very advanced maternal age are increasing in Australia and globally and are likely to continue to increase as more sophisticated reproductive technologies become available [
4]. These factors make it an important area of future study. The principal aim of this study was to determine maternal and perinatal outcomes of pregnancies in women aged 45 years or older, giving birth in Victoria, Australia, compared to women in the median age group of 30–34 years. We found that despite considerable emphasis on chronic illness in other studies, women in our study had low levels of pre-existing disease and most (72.2%) conceived naturally. These features suggest a generally healthy cohort. Nonetheless, women aged 45 years or older were at higher odds of experiencing a number of maternal and perinatal complications. These findings are discussed below.
Pre-existing conditions
Pre-existing hypertension occurred at more than twice the rate among women aged 45 years or older, compared to women aged 30–34 years. However, there were too few older women with pre-existing diabetes to be confident of an increase in the older group. Nonetheless, recorded rates of 2.8% (hypertension) and 1.4% (diabetes) are modest compared to other studies examining pregnancy at 45 years and older [
9,
11,
19]. Yogev et al., [
11], for example, found rates of 6.8% (hypertension) and 4.5% (diabetes) among women aged more than 45 years in Israel, while Dildy et al., [
9] found equivalent rates of 3.8% and 3.8% in the US, and Abu Heija et al., [
19] found rates of 9.6% and 4.4% respectively in Jordan. Only one study was located that had lower rates of these two chronic conditions. Callaway et al., [
6], in a study of Australian women, found very low rates, with just one woman out of 77 presenting with pre-existing hypertension and none with pre-existing diabetes. The low incidence of pre-existing disease seen in both our study and Callaway et al., [
6] may relate to a tendency in Australia, for older maternal age to be associated with higher socio-economic status and generally healthy women [
29,
30]. Although we could not establish the degree of multiparity in our study, subjects aged 45 years or older in studies from Hong Kong, Jordan and Utah, US, tended to be of very high parity, and this factor may have contributed to the increased rates of pre-existing disease seen in those studies [
9,
19,
31].
Pregnancy complications
Similar to other studies in this area [
8,
9,
11,
19], very advanced maternal age in our study was characterised by an increased incidence of pregnancy complications such as pre-eclampsia, gestational diabetes, antepartum haemorrhage and caesarean section. However, although the rate for pre-eclampsia among women 45 years or older was almost double the rate for younger women, at 4.6%, it was significantly lower than several comparable studies which reported rates of 10–12% [
6,
9‐
11,
19,
32] and very much lower than Glasser et al.’s study [
8], which reported rates of pre-eclampsia at 18.3%. Glasser et al., [
8] examined pregnancy outcomes among primiparous ART recipients in Israel and reported a number of extreme findings, most likely related to primiparity, extremely advanced maternal age (45–65 years), and the unscreened nature of the sample. In contrast to Glasser et al., [
8], other studies that examined pregnancy morbidity among women receiving ART, generally reported on women who had been screened for pre-existing disease prior to acceptance into fertility programs [
16,
33]. Moreover, higher rates of pre-existing hypertension found among other study populations aged 45 years or older [
9,
11,
19] may have influenced subsequent higher rates of pre-eclampsia, as this link is well established [
34]. Gestational diabetes mellitus (GDM) rates were high in our study group, at 9.7% compared to the Australian national rate of 4.6% [
35], but lower than Australia-wide rates of 13% for women aged 45–49 years [
35]. Our findings are also consistent with other studies of very advanced maternal age where GDM presents at rates of 8–12% [
6,
9]. In contrast to these more general findings, Yogev et al., [
11] and Glasser et al., [
8] each reported extremely high rates of GDM at 17% and 42.7% respectively. Both these Israeli studies reported significantly higher rates of a number of outcome variables which may in part be explained by the extreme age of participants [
8] and high rates of ART, which have been previously been linked to poorer pregnancy outcomes [
36].
Antepartum haemorrhage was reported in our study at 5.5%, however, comparison with other studies was limited as only one other study was found to have examined this variable. Abu Heija et al., [
19] found rates of antepartum haemorrhage of 10.5% in Jordan, which likely relates to trends of very high parity and continued childbearing until menopause, reported in that study. Compared to women aged 30–34 years, placenta praevia (4.6%) was more than four times as common in our study population and similar rates were found in other studies at 4.4–5.6% [
11,
19]. Rates for postpartum haemorrhage (PPH) were unaffected by maternal age 45 years or older (9.7% vs 9.5%) in our study, although rates for both age groups were higher than comparable studies which reported rates of 4.0%–6.35% [
11,
32]. It is not immediately clear why this disparity exists although differences in estimation and reporting of PPH may contribute [
37]. It may also relate to the year of study as rates of postpartum haemorrhage are increasing over time [
38].
Caesarean rates of 55.3% are reported in our study and this is consistent with caesarean rates in Australia overall where 45–57% of women aged 40 years and over will have a caesarean section [
39,
40]. Elsewhere, caesarean section for women aged 45 years or older showed little consensus. At the lower end of the scale, Dildy et al., [
9], and Dulitzki et al., [
10] reported rates of 31% to 40% while Yogev et al., [
11] and Glasser et al., [
8] reported much higher rates of 78.5% and 93.9% respectively. The lower rates reported by Dildy et al., [
10] and Dulitzki et al., [
11] may relate to the age of these publications which report on data from 1985–1994 and 1996 respectively. Caesarean section rates have increased significantly since this era and women in both these studies were predominantly multiparous, a feature which is also associated with lesser rates of caesarean section. In contrast, high rates of caesarean section were reported by Yogev et al., [
11] (78.5%) and Glasser et al., [
8] (93.9%) and may reflect increasing trends of caesarean section overall, higher numbers of primiparous women in these studies [
8,
11], very extreme maternal age [
8] and high use of ART. Numbers of women with a previous caesarean section may also have contributed in Yogev et al’s study [
11], as previous caesarean is a strong predictor for repeat caesarean [
40].
Perinatal outcomes
Adverse perinatal outcomes such as preterm birth at 32–36 weeks, low birth-weight, and birth-weight below the 10th centile were all increased 1.5–2 fold for women aged 45 years or more in this study and this increase was most obvious among primiparous women, although preterm birth was also more common in older multiparous women. These findings are consistent with trends of poorer perinatal outcomes reported in other studies [
6,
9,
11,
19,
32]. However, direct comparison of specific variables was not always possible due to different study approaches and categorisation of variables. For pre-term births, there was a tendency in the literature, to report all preterm births as births at < 37 weeks gestation and therefore our total rate of 17.5% (32–36 weeks and before 32 weeks) is presented here for comparison. Our study rates were less than rates of 18.3%–29.8% reported by Dulitzki et al., [
10], Glasser et al., [
8] and Yogev et al., [
11] and slightly higher than rates of 11.6–15.2% found by Callaway et al., [
6] and Dildy et al., [
9]. Only one study, Yogev et al., [
11] reported on pre-term birth at less than 32 weeks gestation and found rates of 2.3% which are very similar to our rates of 2.2%. For low birth-weight, our rates of 12.6% are consistent with Callaway et al’s [
6] and Dulitzki et al’s [
10] findings of 10–11% and lower than rates reported by Dildy et al., [
9] and Glasser et al., [
8] at 17.3%–27.8%. Rates of macrosomia, on the other hand, were very similar among our study population, compared to the comparison group, but rates overall were very low. Comparison with the literature was hampered by the tendency, in other studies, to consider macrosomia at 4.0 kg compared to 4.5 kg in our study [
9,
10].
Small for gestational age was the most common weight measurement reported by other studies of women aged 45 years or older [
6,
8,
11,
32,
41,
42] and rates ranged from 4.8% to 30%. Our rates of 13.8% are slightly higher that the lower range reported in this literature, and 1.5–2 fold higher than the 5–9% rate found in the general Australian population [
39]. At the lower end of the scale, Shrim et al., [
32], Jacobsson et al., [
41], Callaway et al., [
6], and Yogev et al., [
11], reported SGA rates of 4.8%–11.3%. At the higher end of the scale, Israeli studies by Glasser et al., [
8] and Simchen et al., [
42] reported rates of 25% and 30% respectively. The current study found increased rates of PTB for older women having first births as well as subsequent births, for singleton but not multiple births, and for those who did not use ART but not for those who did use it. As would be expected, rates of PTB were substantially higher for multiple rather than singleton births, and for those who used ART regadless of maternal age.This association between preterm birth, very advanced maternal age and primiparity is found in other studies [
8,
15,
43] as is the link with SGA [
8,
11]. Cano et al., [
44] suggest that SGA in this maternal age group is likely related to poorer placental development and generally poorer performance of the aging uterus. We found that LBW, but not SGA was associated with multiple birth in our population and this is in contrast with the literature generally, where older maternal age is more commonly associated with higher infant birthweight among multiple pregnancies [
45‐
51]. We also found that babies born after ART and multiples had higher rates of low birthweight regardless of maternal age, but VAMA increased the risk of PTB only for first births.
There is however some difficulty with direct comparison, as very few studies examined multiple pregnancy in women of very advanced maternal age. Most considered advanced maternal age in a single group, ≥35 years [
46,
47,
49], or in two groups ≥35 and ≥40 years [
48,
51]. Studies explicitly examining outcomes among women ≥ 45 years typically report on very small numbers of women with multiple births and do not analyse outcomes separately for singletons and multiples [
6,
42]. In general, women of very advanced maternal age are considerably more likely to have multiple births, and in this study, multiple birth rates were approximately 6% for the study group, which is three fold the rate reported among younger women. This rate is also considerably in excess of rates for multiple births occurring in the gereral population, in Australia of approximately 2.5% to 3% [
39].
The absolute number of perinatal deaths to women of 45 years or older was very low in this study, and therefore valid comparison with younger women was not possible. Rates are also subject to random fluctuation from year to year [
42] and for this reason, continued monitoring is recommended to determine if a pattern of age related increase in perinatal death is occurring. Our low rates of perinatal death are similar to recorded rates of 0–1% [
6,
9] and considerably less than findings in Yogev et al.’s [
11] study of 7.4%. However, direct comparison with other studies was difficult as infant death was frequently reported in other studies as stillbirth [
13] or neonatal death (0–28 days) alone [
19,
41], while we used perinatal deaths (stillbirths and neonatal deaths). This combined approach was used because perinatal death is a more meaningful outcome than stillbirth or neonatal death alone, and because of the small number of deaths to women in the older age group.
Outcomes were generally favourable
Overall, there was evidence, in this study, of increased rates of pre-existing chronic conditions such as hypertension and diabetes; higher rates of pregnancy complications, such as pre-eclampsia, gestational diabetes and placenta praevia; and; higher rates of perinatal complications such as preterm birth, low birth-weight and SGA. These findings have ramifications for women and their families, and for clinicians and health services providing maternity care. They foreshadow an increasing demand on maternity services and resources as trends of very advanced maternal age continue. Nonetheless, this study also provides some reassuring findings and maternal and perinatal outcomes were favourable for the vast majority of women and babies. Moreover, the absolute rate of perinatal death remains low, at generally less than 10 per thousand births in high income countries such as Australia [
12]. This suggests that with careful prenatal care, most women in this age group will achieve a live birth. This information may assist with risk counselling for contemporary mothers of very advanced maternal age.
Potential areas for future study include the investigation of long-term outcomes for infants born to women of this age group, as high rates of preterm birth, LBW and SGA may impact on the future health of this group of children.