Associations of maternal age with maternity care use and birth outcomes in primiparous women: a comparison of results in 1991 and 2008 in Finland

Authors


  • Correction added on 16 August 2013, after first online publication: A Sainio corrected to S Sainio

Abstract

Objective

To compare birth outcomes and maternity care use in 1991 and 2008 by age among primiparous Finnish women.

Design

Register-based study.

Setting

Nationwide Medical Birth Register.

Population

All primiparous women in 1991 (= 24 765) and 2008 (= 23 511).

Methods

Women aged 35–39 and ≥40 years were compared with women aged 20–34 years in 1991 and 2008, using logistic regression to adjust for women's background.

Main outcome measures

Maternity care: prenatal visits, hospitalisation during pregnancy, labour induction, delivery mode, long postpartum hospital stay; and birth outcomes: birthweight, preterm birth, Apgar scores, intensive/observation unit, respiratory care, perinatal death.

Results

In both years, older women's deliveries were more often induced, instrumental, or by caesarean section. In 2008 compared with 1991, hospitalisations were lower and instrumental deliveries and labour induction were higher in older women. A significant decrease in adjusted odds ratios (OR, 95% confidence intervals) between 1991 and 2008 among women aged 35–39 was found for preterm birth (1.47, 1.18–1.84 versus 0.96, 0.86–1.07) and for intensive/observation unit (1.73, 1.47–2.05 versus 1.21, 1.07–1.37) and, among women aged ≥40 years, for intensive/observation unit (3.14, 2.30–4.29 versus 1.64, 1.31–2.07). The risk for perinatal death (1.66, 0.60–4.60 versus 2.69, 1.07–6.79) was higher in 2008 than in 1991 among women aged ≥40.

Conclusions

In 2008, older primiparous women still used more maternity care, had more interventions, and poorer birth outcomes than younger women, regardless of care advances. Additional risks declined among women aged 35–39 but not among aged ≥40.

Introduction

In 1991 in Finland, older primiparous women had more problems during pregnancy and childbirth, their infant outcomes were worse, and they used more antenatal services than younger primiparous women.[1] Since then, the childbearing population has become older and delayed childbearing has become normalised. In Finland, the average age for the first birth was 26.5 years in 1987 and 28.2 years in 2010, and first births to women aged over 35 increased from 5.3% of all first births in 1987 to 10.1% in 2010.[2] The proportion of women having their first baby at age 40 or older increased from 12% in 1987 to 20% in 2010.[3] Similar trends can be found elsewhere in Europe, the USA and Canada.[4-7]

In Finland, almost all pregnant women (99.7% in 2011) use antenatal care,[2] which is provided in primary health care by public health nurses, midwives and general practitioners, and by obstetricians and midwives at delivery hospitals during outpatient visits. The average number of prenatal visits was 16 in 2011, including both visits to healthcare centres and hospitals.[2] The number of planned homebirths is very low: yearly 8–12 births out of some 60 000 total births.

Many changes in maternity care have occurred in recent decades. Medical technology to achieve pregnancy is in common use. In Finland, assisted reproductive treatments per 1000 women aged 15–49 years increased from 1.8 in 1992 to 7.0 in 2008.[8] In 2010, over 4% of all the children were born following assisted reproductive treatments (including inseminations) in Finland. New methods in prenatal screening and the monitoring of pregnancy, labour and fetal wellbeing as well as to screen for fetal status have been developed, and are now used routinely. New methods to treat perinatal problems have been introduced and the use of technology during labour has increased. All these changes may have changed the service use and improved perinatal health among older primiparous women. Nevertheless, potential changes in the additional risk of older motherhood have not been studied.

The purpose of this study was to compare birth outcomes and use of maternity care in 1991 and in 2008 by maternal age in Finland using national birth register data to establish whether older maternal age at first birth was still an obstetric risk in 2008 regardless of advances and changes in obstetrics and perinatal care.

Materials and methods

Only primiparous women were included, that is, women having their first birth. They may have had previous pregnancies ending in a miscarriage or an induced abortion. To find out whether older maternal age remains an important obstetric risk, we compared results in 1991 with those in 2008. The original data source for both study years was the Finnish Medical Birth Register (MBR),[2] though results for 1991 were taken from a published report.[1]

The MBR was started in 1987 and is currently run by the National Institute for Health and Welfare (THL). The MBR contains information on maternal background, use of maternity care, complications during delivery, and infant outcomes up to the age of 7 days for all infants born in Finland.[2] The data are collected by delivery hospitals and completed through linkage to the Central Population Register and Cause-of-Death statistics (compiled by Statistics Finland). The quality of the MBR has been found to be high for the variables used in this study.[9, 10] No more recent studies on the register's quality exist, though the quality is assumed to be good.

Women aged under 20 years (n = 1608 in 1991 and n = 1249 in 2008) and women with missing information on age (0.1% in 1991 and 0.0% in 2008) and parity (1.8% in 1991 and 0.0% in 2008) were excluded. A total of 24 765 (40%) women delivering were primiparous in 1991 and 23 511 (41%) in 2008. The number of those aged 20–34 years was 23 327 (94%) in 1991 and 21 057 (89%) in 2008; those aged 35–39 years numbered 1206 (5%) in 1991 and 2010 (9%) in 2008, and those aged 40 years or over numbered 232 (1%) in 1991 and 444 (2%) in 2008.

As maternal care use indicators, the following were used: timing of the first prenatal visit (classified as under 8 weeks, 8–15 weeks, and at least 16 weeks of gestation), mean number of prenatal visits (classified as under ten, 10–15 and at least 16 visits), hospitalisation during pregnancy, in hospital 2 days before the delivery, induction of labour, instrumental delivery, mode of delivery, and long postpartum hospital stay (seven or more days).

As birth outcome indicators, the following were used: low birthweight (<2500 g), very low birthweight (<1500 g), preterm birth (<37 weeks), very preterm birth (<32 weeks), low 1-minute and 5-minute Apgar scores (0–6), treatment in an intensive care or observation unit, respiratory care, and perinatal death (from 22 weeks of gestation onwards). Very low birthweight, very preterm birth, 5-minute Apgar scores and respiratory care were not available in 1991. The distinction between an intensive care unit and an observation unit depends on the hospital, and so these were combined. Respiratory care means that the newborn was intubated. Only singleton pregnancies were included in the birth outcome analysis.

Maternal mortality per 100 000 live births by age groups was calculated using birth register data. Due to the low number of deaths, mean rates from the years 1991 to 1995 and 2006 to 2010 were used, while no detailed analyses were possible because of the small number of cases.

In 1991, details of women's education was obtained through record linkages to the 1990 National Education Register and was used as an indicator for socio-economic position. Education level was classified into two categories: women who went to school only until the age of 15 years or less formed the lower educated category, while other women formed the higher educated category. In 2008 socio-economic position was defined by women's occupation at the time of delivery. The occupation in the MBR is classified into eight categories according to the national classification compiled by Statistics Finland. In this study, classification into four categories was used: upper white-collar workers, lower white-collar workers, blue-collar workers, and others (entrepreneurs, students, pensioners, unemployed women and women without a recorded occupation).

Statistical methods

First, background characteristics and outcomes in 1991 and 2008 in total were compared with each other to identify the overall changes. Differences in totals were tested for their relative proportions. Second, relative proportions for age-specific changes from 1991 to 2008 were also calculated. Third, in both time-periods, older primiparous women (35–39 and 40 years and over) were compared with younger women (20–34 years). The differences between age groups were tested with a chi-square test, t test and a test for relative proportions.

Adjusted birth outcomes were studied using logistic regression. Outcomes were adjusted for women's background characteristics: marital status, education/socio-economic position, maternal smoking, area and rural/urban nature of the county of residence. In that analysis only singleton births were included. The differences between the two study years in adjusted birth outcomes were studied by comparing confidence intervals. Statistical analyses were performed in SAS, version 9.1 (SAS Institute, Cary, NC, USA).

The study plan was approved by the STAKES research ethics committee (14 April 2008). The National Data Protection Authority was consulted and permissions were obtained from the registry keepers.

Results

Background characteristics

Overall, single motherhood, urban living, use of in vitro fertilisation and previous miscarriages were higher in 2008 than in 1991 (Table 1). Previous induced abortions and smoking during pregnancy were somewhat lower in 2008.

Table 1. Comparison of background characteristics of primiparous women by age in 1991 and 2008 in Finland (%)
  Year 1991 Year 2008
20–34 35–39 40+ Total 20–34 35–39 40+ Total
( = 23 327) ( = 1206) ( = 232) ( = 24 765) ( = 21 057) ( = 2010) ( = 444) ( n  = 23 511)
  1. *,**,***P value <0.05, <0.01, <0.001 for tests for relative proportions; in each year, groups are compared with each other using 20–34 years as the reference; 2008 total column indicator comparison of 1991 and 2008 (all ages).

  2. ****Includes those quitting smoking during pregnancy.

  3. Highest socio-economic position includes women with education of 13 or more years in 1991 and upper white-collar workers in 2008.

Marital status
Married6359**52***634753***4647***
Cohabiting302932304336***3944***
Single712***16***7101014**10***
Highest socio-economic position 1831***27***191733***37***19
Urban 5761**62577375*7773***
Previous miscarriage 1121***30***121220***20***13***
Previous induced abortion 141818***141113***1711***
In vitro fertilisation0.44***3***1310***15***3***
Smoked during pregnancy **** 171813171710***10***16**
Body Mass Index
25+NANANANA2733***34*28
30+NANANANA911**99

In both study years, older primiparous women were more often highly educated, urban residents, and had experienced previous miscarriage, induced abortions and fertility treatments than primiparous women aged 20–34 years (Table 1). The age differences were more pronounced in 2008 than in 1991 with regard to highest socio-economic position, nonsmoker and having undergone in vitro fertilisation treatment. In 2008, older primiparous women were more often overweight than younger primiparous women (Table 1).

Use of maternity care

Overall, early prenatal care visits (<8 weeks) increased from 32% in 1991 to 55% in 2008 (Table 2). In the age-specific inspection we found a decrease in having at least 16 visits among primiparous women aged 35–39 years (58 versus 50%). In both study years the proportion of women having many visits was higher among primiparous women aged 40 years or older than among those aged 20–34 years.

Table 2. Comparison of maternity care use and interventions among primiparous women by age in 1991 and 2008 in Finland (%)
  Year 1991 Year 2008
20–34 35–39 40+ Total 20–34 35–39 40+ Total
( = 23 327) ( = 1206) ( = 232) ( = 24 765) ( = 21 057) ( = 2010) ( = 444) ( = 23 511)
  1. *,**,***P value <0.05, <0.01, <0.001 for tests for relative proportions; in each year, groups are compared with each other using 20–34 years as the reference; 2008 total column indicator comparison of 1991 and 2008 (all ages).

First visit, weeks of gestation
<83227**303256**5051***55***
≥1634432342***
Number of prenatal visits
≤1067767877***
≥164758***61***484650**60***47*
Inpatient care during pregnancy 2129***39***221113**14*11***
In hospital 2 days before birth 1624***32***171219***19*13***
Induction of labour 1217***18***121723***23**18***
Instrumental delivery 1013***12101517**20**15***
Caesarean section 1735***45***182035***41***22***
Postpartum stay 7+ days 1726***38***1846***9***4***

Hospitalisation during pregnancy declined from 22% in 1991 to 11% in 2008 (Table 2). In 1991 there was a clear difference by age, with more hospitalisations during pregnancy among older women. In 2008 there was only a small difference in prenatal hospitalisations by age. In both years older women were somewhat more often in hospital 2 days before the delivery than the younger women.

Induction of labour and instrumental deliveries increased and long postpartum hospitalisation decreased between 1991 and 2008 (Table 2). In age-specific inspections we found that labour induction and long postpartum hospitalisations increased by age in both study years. Instrumental deliveries increased by age more in 2008 than in 1991. Caesarean section among primiparous women increased from 18% in 1991 to 22% in 2008. In both study years, caesarean section increased by age, but the age difference declined slightly between 1991 and 2008.

Birth outcomes

Overall, the rates of preterm birth in 1991 and 2008 were similar (Table 3). In 1991, the adjusted risk for preterm births increased by age, but in 2008 it increased only among mothers aged 40 years or older. In 2008, the unadjusted rate for very preterm birth was highest among women aged 35–39 years old (Table 3) but compared with women aged 20–34 years the adjusted risk was not higher (Table 4).

Table 3. Raw rates (% or 1/1000) of birth outcomes among primiparous women with singletons by age in 1991 and 2008 in Finland
Outcome Year 1991 Year 2008
20–34 35–39 40+ Total 20–34 35–39 40+ Total
( = 23 327) ( = 1206) ( = 232) ( = 24 765) ( = 21 057) ( = 2010) ( = 444) ( = 23 511)
  1. *,**,***P value <0.05, <0.01, <0.001 for tests for relative proportions; in each year, groups are compared with each other using 20–34 years as the reference; 2008 total column indicator comparison of 1991 and 2008 (all ages).

Very preterm <32 weeks of gestation,%NANANANA0.91.4*0.90.9
Preterm <37 weeks of gestation,%5.68.5***10.8***5.85.88.2***9.0**6.1
Very low birthweight <1500 g,%NANANANA1.01.7**1.71.1
Low birthweight <2500 g,%4.99.6***8.9**5.25.48.7***8.5**5.7*
1-minute Apgar score <7,%5.27.8***4.75.38.09.212.4***8.2***
5-minute Apgar score <7,%NANANANA2.63.35.7***2.7
Intensive care/Observation unit,%8.813.3***11.19.013.616.0**21.0***13.9***
Respiratory care,%NANANANA1.31.9*2.01.4
Perinatal mortality, 1/10006.714.5***8.57.13.96.810.9*4.3***
Table 4. Adjusted odds ratios (OR) and 95% confidence intervals (95% CI) of birth outcomes among primiparous women with singletons by age in 1991 and 2008 in Finland
Outcome Year 1991 Year 2008
35–39 40+ 35–39 40+
( = 1206) ( = 232) ( = 2010) ( = 444)
OR (95% CI) a OR (95% CI) a OR (95% CI) a OR (95% CI) a
  1. NA, not available.

  2. a

    In 1991 adjusted for marital status, education, maternal smoking, previous pregnancies (not ending in birth), and county, and in 2008 adjusted for marital status, socio-economic position, maternal smoking, previous pregnancies (not ending in birth), and urban nature of the residence; reference group mothers aged 20–34 years old (n = 23 327 in 1991, n = 21 057 in 2008).

Very preterm <32 weeks of gestation,%NANA1.00 (0.75–1.32)0.93 (0.34–2.54)
Preterm <37 weeks of gestation,%1.47 (1.18–1.84)2.16 (1.42–3.28)0.96 (0.86–1.07)1.45 (1.04–2.02)
Very low birthweight <1500 g,%NANA1.76 (1.23–2.53)1.69 (0.82–3.48)
Low birthweight <2500 g,%1.96 (1.54–2.50)2.37 (1.46–1.76)1.67 (1.41–1.97)1.59 (1.14–2.23)
1-minute Apgar score <7,%1.50 (1.19–1.89)0.79 (0.58–1.08)1.16 (0.99–1.36)1.63 (1.23–2.17)
5-minute Apgar score <7,%NANA1.28 (0.99–1.66)2.28 (1.51–3.43)
Intensive care/Observation unit,%1.73 (1.47–2.05)3.14 (2.30–4.29)1.21 (1.07–1.37)1.64 (1.31–2.07)
Respiratory care,%NANA1.50 (1.07–2.11)1.49 (0.76–2.93)
Perinatal death, 1/10001.87 (1.16–3.01)1.66 (0.60–4.60)1.76 (0.99–3.12)2.69 (1.07–6.79)

There was a small increase in the rate of low birthweight infants between 1991 and 2008 (Table 3). In both study years, the rate increased by age, but an increase between 1991 and 2008 occurred only among women aged 20–34 years. Older primiparous women still had increased risks in 2008 for low birthweight infants compared with younger women (Table 4). Rates for very low birthweight were also higher among older women in 2008, but significantly increased risk was found only among women aged 35–39 years (Table 4).

Low 1-minute Apgar scores increased from 5.3% in 1991 to 8.3% in 2008 (Table 3). In 1991 there was a small difference by age, but in 2008 the rate increased by age. Likewise in 2008, the rate of low 5-minute Apgar scores increased by age and compared with younger women, the risk was 2.3-fold greater among women aged 40 years or older (Table 4).

The use of an intensive care or observation unit increased from 9.0% in 1991 to 13.9% in 2008 (Table 3). In 2008, the use notably increased by age, whereas adjusted risks were increased among older women compared with younger women (Table 4). In 2008, respiratory care increased by age (Table 3) and the adjusted risk was 1.5-fold among women aged 35–39 years.

The perinatal death rate decreased from 8.5/1000 in 1991 to 4.3/1000 in 2008 (Table 3). In 2008 perinatal death increased linearly by maternal age, and compared with women aged 20–34 years those aged 40 years or older had a 2.7-fold increased risk for perinatal death (Table 4).

When women aged 35–39 years old were compared with women aged 20–34 years, many adjusted risks for poor birth outcomes were lower in 2008 than in 1991 (Table 4). However, a significant decrease between 1991 and 2008 was found only in the risk for preterm birth (odds ratio 1.47, 1.18–1.84 versus 0.96, 0.86–1.07) and for use of an intensive care or observation unit (1.73, 1.47–2.05 versus 1.21, 1.07–1.37). Among women aged 40 years or older, the risk for use of intensive care or an observation unit decreased significantly (3.14, 2.30–4.29 versus 1.64, 1.31–2.07). Risks for low 1-minute Apgar scores (0.79, 0.58–1.08 versus 1.63, 1.23–2.17) and perinatal death (1.66, 0.60–4.60 versus 2.69, 1.07–6.79) were significantly increased in 2008 compared with 1991 among women aged 40 years or older.

There were 15 maternal deaths in the period 1991–95 and 14 in the period 2006–10. In 1991–95, the maternal mortality ratio (per 100 000 live births) was 3.3 for women aged 20–34 years old, 13.2 for women aged 35–39 years old, and 0 for women aged 40 years or older. In 2008, the rates were 2.6, 11.4 and 18.4, respectively.

Discussion

Main findings

Our study shows that the population of older primiparous women has changed between 1991 and 2008, moving more towards being nonsmoking, highly educated women with experience of fertility treatments, but fewer miscarriages and induced abortions than previously. Most pregnancies among older primiparous women proceeded without problems, but compared with younger women, poorer birth outcomes were more common, and deliveries were more often induced, instrumental or by caesarean section. Among women aged 35–39 years the extra risks compared to younger women were the same or had declined from 1991 to 2008. But among women aged 40 years or older, the 1-minute Apgar scores and perinatal death rate suggested a higher additional risk in 2008 than in 1991.

Strengths and limitations

A strength of our study is that it is based on a large nationwide data set including all births in Finland in 1991 and 2008. The data on maternal background, use of maternity services and birth outcomes obtained from the MBR are reliable and of good quality.[9, 10] In the case of birth outcomes, adjustments for important confounders such as marital status, maternal smoking and living area were made similarly in both study years. The only exception was socio-economic position, which in 1991 was based on education and in 2008 on women's occupation. Socio-economic position is one of the most important confounding factors.[11, 12] Two large-scale recent studies were not able to adjust for any confounders describing the socio-economic position.[4, 6]

A limitation of our study is the relatively small number of mothers aged 40 years or older. For some rare outcomes such as placenta praevia, placental abruption, and pre-eclampsia, the 1-year data were too few to enable analyses of these outcomes. However, in terms of birth outcomes the numbers were reasonable. We had no information on 5-minute Apgar scores in 1991 or on chromosomal or congenital anomalies in either study year.

Interpretation

Our results on maternal background among older primiparous women with high level of education, nonsmoking, high body mass index, previous miscarriages and fertility treatments were in line with previous studies.[4-6, 13]

Older mothers in Finland used more antenatal services than younger mothers.[1] This is in line with one study from the UK for the period 1988–97.[14] Similar results on the more common use of induction of labour,[15] Caesarean section[13, 15] and instrumental delivery[13, 15] among older primiparous women have been found in previous studies. In one US study, increased use of caesarean section among primiparous women by maternal age was found even in low-risk pregnancies.[15] We were not able to study low-risk and high-risk pregnancies separately.

Our study confirms earlier results on the increased risk for preterm birth among primiparous women aged 40 years or older.[4-6] We could not confirm in our 2008 data the previous findings of an increased risk for very preterm birth.[13, 15, 16] This might be because multiple pregnancies were excluded from our analysis. However, preterm births in general are not rising in Finland[17] as they are in many other countries.[18]

Our findings on increased risk for low birthweight[5, 13, 16] and very low birthweight infants[13, 16] are in line with previous studies. In 2008, the risk for perinatal death increased by age, as found in other studies.[4-6, 13] Increased risk for use of an intensive care or observation unit by age could also be found in the Belgian study.[13]

Few data on comparisons over time have previously been published. Most published studies report changes overall, and do not give changes by age: a general decrease in hospital days between 1973 and 2000 in Sweden[19] and in Australia in 1999–2004[20]; decreased smoking among pregnant women in Sweden,[19] Australia[20] and France (between 1995 and 2010)[21]; an increase in induction of labour in Australia[20] and France[21]; and an increase in caesarean section in Sweden, Australia and France.[19-21] In our study, instrumental deliveries and induction of labour increased, but smoking decreased only among mothers aged 35–39 years.

Swedish authors have previously suggested an improvement in maternal and child health.[19] This was only partially true in our study. Most of the adjusted birth outcomes improved slightly. However, only preterm birth and use of intensive care or an observation unit among mothers aged 35–39 years and the use of intensive care or an observation unit among mothers aged 40 years or older improved significantly. Risks for perinatal death and low 1-minute Apgar scores, however, increased among mothers aged 40 years or older.

Because older motherhood has become more common, the group of older primiparous women includes more healthy women than previously and this can partly explain the improved outcomes among women aged 35–39 years. Among women aged 40 years or older the risks remained. Up to 7% of pregnant Finnish women have chronic disease such as diabetes and chronic hypertension as calculated by long-term use of medicine during pregnancy.[22] This can partly explain the poorer outcomes among the oldest group of women, who have more chronic diseases (unpublished data).

Conclusions

This study suggests that irrespective of the advances and changes in obstetrics and perinatal care, aging motherhood still affects health and health service use. The general trend in postponing childbearing has somewhat diluted the problems related to older motherhood, but in 2008, older primiparous women and their infants still had poorer health in comparison to that of younger women and their infants. A larger data set and further detailed analyses are required to study the age-effect on rare outcomes and to explore how many of the observed risks are related to age per se and how much to a higher prevalence of diseases such as hypertension, diabetes and pre-eclampsia.

Postponing childbearing is an increasingly important public health concern.[13] Awareness of the special risks in pregnancy at older age is useful for young couples. It is important to remind healthcare professionals about the potential health risks of postponing childbearing and to encourage them to bring this subject up in family planning services. Increased knowledge of the issue will help to diminish false assumptions and beliefs. More social, economic and emotional support is needed for young couples to motivate them to consider having their children at an earlier age.

Disclosure of interests

No conflict of interest.

Contribution to authorship

RK participated in the study design, interpretation of data, drafting the article and modifying the final version. MG participated in the analysis and interpretation of data, planning the article, revising the draft critically and giving the approval of the final version. SS participated in the interpretation of data, revising the draft critically and giving the approval of the final version. EH participated in the conception and design, interpretation of data, revising the draft critically and giving the approval of the final version.

Details of ethics approval

The study plan was approved by the STAKES research ethics committee (14 April 2008). The National Data Protection Authority was consulted and permissions were obtained from the registry keepers.

Funding

The study was financially supported by the Academy of Finland (number 127402) and The National Institute for Health and Welfare (THL).

Ancillary