Many women in high-income countries plan to have their first child after the age of 30 years, and this postponement of childbirth compared with the previous generation has caused concern among health care providers as a result of the risk of infertility and adverse pregnancy outcomes.1 Along with increasing demands for assisted reproductive techniques, advanced maternal age has been associated with preterm birth,2 fetal growth restriction leading to neonates who are small for gestational age (SGA),2 chromosomal abnormality,3 low Apgar score,4 stillbirth,5 and neonatal death.6 Advanced maternal age is often defined as 35 years or older, and in this study, the question posed was whether the risk of some of the most serious pregnancy outcomes started earlier.
Several of the outcomes mentioned are also associated with smoking and overweight or obesity. Smoking increases the risk of preterm birth,7 SGA,8 stillbirth,5 and neonatal death7; and overweight or obesity during pregnancy is associated with preterm birth,9 low Apgar score,4 stillbirth,5 and neonatal death.10 The relative significance of the respective lifestyle factors during pregnancy is less explored, except regarding the risk of stillbirth. A systematic review reported that maternal overweight and obesity contributed to approximately 8,000 stillbirths (22 weeks of gestation or greater) annually across all high-income countries and that advanced maternal age (older than 35 years) and smoking contributed to 4,200 and 2,800 stillbirths, respectively.5 In this study we investigated whether advanced maternal age was a risk factor comparable to smoking and overweight or obesity regarding the outcomes listed subsequently.
The specific aims of the study were: 1) to investigate associations between advanced maternal age and very preterm birth, moderately preterm birth, SGA, low Apgar score, stillbirth, and neonatal death in nulliparous women aged 30 years and older with a singleton pregnancy compared with women aged 25–29 years who gave birth in Sweden and Norway during 1990–2010; and 2) to compare the risks associated with advanced maternal age with those related to smoking and being overweight or obese in the Swedish sample.
MATERIALS AND METHODS
For the first aim of the study, we obtained data from the Swedish and Norwegian Medical Birth Registers, and for the second aim, we used Swedish data only. The Swedish and Norwegian Medical Birth Registers include information from the standardized medical records used by all antenatal clinics and delivery units and cover 98% or greater of all births in the respective country. Several variables used in the present study have previously been validated with satisfactory results such as maternal age, gestational age, and infant survival.11–14
To study a potential increase of risk in any of the outcome factors at a younger age than 35 years, which is a commonly used cutoff for advanced maternal age, we defined the reference group as maternal age 25–29 years and compared the outcomes in this group with those in the age groups of 30–34 years, 35–39 years, and 40 years or older, respectively. For simplicity, the three age groups of 30 years and older are referred to as advanced maternal age in this study. In contrast to our Norwegian data set, in which maternal age was given as a 5-year interval for each individual, the Swedish data included information about the exact maternal age, which allowed calculation of the mean age within each of the four age groups. These were (standard deviation): 27.0 years (±1.4), 31.6 (±1.37), 36.4 (±1.33), and 41.3 (±1.59).
Smoking referred to any smoking at the first antenatal booking in early pregnancy regardless of the number of cigarettes smoked. Information about maternal weight and height was collected on the same occasion, and body mass index (BMI, calculated as weight (kg)/[height (m)]2) was defined as underweight, less than 18.5; normal weight, 18.5–24.9; overweight, 25–29.9; and obese, 30 or greater. Information about smoking and BMI was not available for the entire observation period in the Norwegian sample, which precluded adjustment for these variables in the statistical analyses.
In addition, adjustment was made for the following factors: year of birth (continuous variable), civil status (Sweden: living with the child's father compared with not; Norway: married or cohabiting compared with not), chronic hypertension, and diabetes reported at the antenatal booking in both countries. The Swedish data were also adjusted for country of birth (Nordic=Sweden, Norway, Denmark, Finland, and Iceland compared with other country), smoking, and BMI.
The six outcomes of the study were defined as follows: very preterm birth: 22–31 gestational weeks; moderately preterm birth: 32–36 gestational weeks; SGA: greater than 2 standard deviations under normal weight for gestational age adjusted for the sex of the neonate; Apgar score: less than 7 at 5 minutes after the birth; fetal death: from gestational week 22; and neonatal death: within 28 days after delivery.
To investigate the association between maternal age and each outcome, we estimated the crude and adjusted odds ratios with 95% confidence intervals (CIs) in the two national samples (Table 1).
For the second aim of the study we used the Swedish population only. As a first step, we estimated the association between each pregnancy outcome, one at a time, and maternal age, smoking, and BMI and adjusted for five risk factors: year of birth, civil status, country of birth, diabetes, and chronic hypertension. Second, we investigated whether the effect of age was similar regardless of whether the woman was a smoker, and regardless of BMI, and tested two-way (age×smoking; age×BMI) and three-way interactions (age×smoking×BMI) by adding each of the interactions, one at a time, to the model in the first step, which included eight factors. The Wald test was used to assess the effect of each factor and the interaction; P values <.05 were defined as statistically significant. Goodness of fit for each model was tested by the Hosmer and Lemeshow test,15 and P values >.05 were regarded as adequate-fitted models. To detect influential points, which may distort the outcome and accuracy of the regression analysis, we calculated the Cooke's distance, and values greater than 1 were interpreted as high.16 No such points were found.
To estimate the attributable risk of advanced maternal age, compared with smoking, overweight, and obesity, respectively, we calculated the population rate of each outcome in a low-risk group of women corresponding to the reference levels for the risk factors (nonsmokers, normal weight, age 25–29 years) and multiplied this rate with the adjusted odds ratios (ORs) (in Table 2) for the respective outcomes related to each of the three risk factors in the estimated models. The absolute population rates for each outcome are presented along with these relative adjusted rates. Finally, we estimated the number of additional cases of each outcome, which might be explained by advanced maternal age (30–34 years; 35–39 years; 40 years or older; and 30 years or older), smoking, overweight, and obesity (and overweight or greater) based on the differences between the adjusted rates and the absolute risk in the low-risk group. For example, the rate of fetal death among the 162,464 neonates in the low-risk group was 0.21%, corresponding to 341 cases during the years 1990–2010. Among woman aged 40 years or older, the adjusted risk of fetal death was 2.33-fold higher compared with woman aged 25–29 years, corresponding to an estimated absolute ratio of 0.49% (2.33×0.21%). The difference in risk attributable to age is therefore 0.28 percentage units, which implies that if all the 11,430 woman older than 40 years had given birth at the age of 25–29 years, and if the relationship between maternal age and fetal death was completely causal, the number of fetal deaths could have been reduced by 32 cases (0.28%×11,430).
To address the issue of multiple hypothesis testing, a stricter threshold at P<.001 was used. We found that all the statistically significant findings at level P<.05 were also significant at P<.001 (indicated as footnotes Table 1 and 2).
As a result of the long timespan of the study, all regression models were adjusted for year of birth, and the larger Swedish data set was split into two decades to compare outcomes from 1990 to 1999 with those from 2000 to 2010. To compare the adjusted ORs between 1990–1999 and 2000–2010, we calculated the difference between the estimated ln(adjusted OR) divided by the pooled standard error ln(adjusted OR) for the respective outcome. Two-sided P values were calculated based on the standard normal distribution.
The analyses were conducted using SPSS 20. The study was approved by the appropriate Regional Committees for Ethics in Medical Research in Sweden and Norway.
The adjusted OR for each outcome increased by maternal age in a similar way in Sweden and Norway (Table 1). Exceptions were the adjusted OR for fetal and neonatal deaths in the oldest age group of 40 years or older in the Norwegian sample, where observations were too few (28 and 8, respectively) to allow valid conclusions.
In both populations, there was an increased risk in the 30- to 34-year-old age group for the following outcomes: very preterm birth, SGA, low Apgar score, and fetal death. In this age group in the Swedish population, the risk of neonatal death also increased, and in the Norwegian sample, moderately preterm birth increased. Table 2 presents the adjusted associations of advanced maternal age, smoking, and BMI for each outcome in the Swedish sample. We found no statistically significant interactions between maternal age and smoking, maternal age and BMI, or among all three lifestyle factors. Advanced maternal age, smoking, overweight, and obesity were each associated with all the outcomes of the study with the following exceptions: age 30–34 years was not associated with moderately preterm birth, smoking was not associated with low Apgar score, and overweight was not associated with SGA. Goodness of fit was low in the SGA model (P<.001) and also in the moderately preterm model (P=.04).
We found no statistically significant differences between the adjusted ORs for adverse outcomes in relation to maternal age, smoking, or BMI when comparing findings from 1990–1999 and 2000–2001, with the exception of moderately preterm birth at age 35–39 years (adjusted ORs 1.24, 95% CI 1.17–1.32 compared with adjusted OR 1.05, 95% CI 1.00–1.11).
Table 3 shows the rates of the respective outcome in a low-risk group of 162,464 women with none of the three lifestyle factors and in women of advanced age, in smokers, and in women who were overweight or obese. The adjusted rates were generally lower than the population rates (Table 2). The estimated additional cases associated with each risk factor compared with the low-risk group are indicated in bold. Maternal age 30 years or older was associated with a larger number of additional cases of very preterm birth (693) and neonates with SGA (2,749) than smoking (very preterm birth: 158; SGA: 1,739) and overweight or obesity (very preterm birth: 470; SGA: −81) and with the same numbers of fetal deaths (251) as overweight or obesity (251). Of the three lifestyle factors, overweight or obesity was associated with the largest number of additional cases of moderately preterm births (1,255), neonates with low Apgar score (883), and neonatal deaths (92).
Based on independent nationwide samples of nulliparous women from two high-income countries, this study confirms that the risk of very preterm birth, moderately preterm birth, SGA, low Apgar score, stillbirth, and neonatal death all increases with advancing maternal age. The estimated risks related to advancing age were approximately the same in both countries despite the adjustment for slightly different variables (no adjustment for smoking, BMI, and country of birth in the Norwegian sample). The study shows that risks may increase before the age of 35 years or older, which is the commonly used definition of advanced maternal age. For the individual woman aged 30–34 years, the absolute risks for very preterm birth, SGA, low Apgar score, and fetal death, respectively, were small, but for society, they may be significant as a result of the large number of women who give birth in this age range.
Compared with the low-risk group of normal-weight nonsmokers aged 25–29 years, we found that higher maternal age was associated with a larger number of additional fetal deaths than overweight plus obesity. In contrast, Flenady and colleagues5 reported in a systematic review of stillbirths that maternal overweight or obesity had a higher effect than advanced maternal age. This difference in findings could be related to the different definitions of advanced maternal age (30 years or older compared with 35 years or older). The choice of age cutoff as well as the definition of the reference group is crucial when investigating effects of maternal aging. If, for example, women 35 years or older are compared with women younger than 35 years, the effect of aging may be underestimated because of the U-shaped distribution of some of the adverse pregnancy outcomes such as the rates of preterm birth, low Apgar score, and neonatal mortality, which are higher in teenagers than in women in their 20s.17
Estimations of additional cases of adverse pregnancy outcomes related to advanced maternal age, smoking, or overweight or obesity are dependent on the prevalence of these factors, and the findings are therefore influenced by lifestyle changes in the study population such as the reduction of smoking in high-income countries,18 the size of the obesity epidemic,19 and the phenomenon of childbirth postponement.1 In this study, data in Table 3 were from a country with a low rate of smokers and comparatively few women who were obese.
The reasons why advanced maternal age has negative effects on pregnancy outcomes are not fully understood. The ORs for very preterm birth, SGA, and fetal death in this study indicate factors in the uterine environment such as aging processes in the uterus and placenta. Placental pathology has been discussed in relation to the increased risk of stillbirth in older women,20 and this is estimated to explain one in four stillbirths.21 Sclerotic lesions in the myometrial arteries could cause underperfusion, and such lesions increase by age.20 Placental pathology may also contribute to SGA and preterm birth,22 and the placental dysfunction syndrome, including preeclampsia, SGA, placental abruption, and preterm birth,23 is associated with advanced maternal age.24 Neonatal death and low Apgar score may be the result of many different causes including complications during labor. Aging processes in the myometrium may explain the higher rates of uterine dysfunction in older nulliparas,25,26 and different theories have been presented such as hormonal effects on the uterus,27 decreased number of oxytocin receptors,28,29 and upregulation of ATP-sensitive (KATP) channels in the myometrium by increasing age.30
Similar to advanced maternal age, smoking may have negative effects on the uterine environment by increasing the risk of placental complications, fetal growth restriction, and preterm birth,7 and the biological effects might be related to the vasoconstrictive effect of nicotine and the reduced prostacyclin synthesis.31–33 Overweight or obesity may also increase the risk of adverse pregnancy outcomes such as gestational hypertension, gestational diabetes, cesarean delivery and stillbirth.34 The biological mechanisms underlying these associations are probably related to the metabolic and inflammatory disorders seen in obese persons and also documented in overweight pregnant women.35,36
We found no interaction among advanced maternal age, smoking, and BMI and, if the effects were causal and independent, the multiplied effect on adverse pregnancy outcomes would be considerable.
The findings of this study should be interpreted in light of the observational design, which does not allow definite conclusions about causality. Analyses were not adjusted for smoking and BMI in the Norwegian data set, and information about education was missing. For the second aim of the study, differences in women's socioeconomic background were therefore only controlled for indirectly by the variables civil status, country of birth, and smoking, which is more prevalent in women with a low level of education in Sweden.18 Nevertheless, the risk estimates were relatively similar to those of other population-based studies, although comparisons are difficult as a result of varying definitions of advanced maternal age and reference groups. However, the systematic review of stillbirth previously referred to5 allowed comparison with our Swedish findings (Table 2), and the risks associated with all three lifestyle factors were fairly similar.
Another limitation of our study was that the SGA model did not fit well with the data and should therefore be interpreted with some caution. One explanation could be lack of differentiation between severe and moderate SGA. Severe SGA is more related to fetal growth restriction and advanced maternal age, whereas moderate SGA may be a mixture of growth restriction and biologically small neonates.8
Strengths of this study were the size and quality of the two data sets retrieved from the Swedish and Norwegian Medical Birth Registers. Analyses from two independent populations strengthened the validity of the findings related to the risk of adverse pregnancy outcomes as a result of advanced maternal age. The definition of the reference group as an age group when pregnancy outcomes are more or less optimal (25–29 years) made it possible to discover an increased risk for some of the outcomes already at the age of 30–34 years.
This study confirms that advanced maternal age is associated with an increased risk of adverse pregnancy outcomes. It also shows that the risk may increase by the age of 30–35 years, and although the individual risk is small, it may be significant for society as a result of the large number of women who give birth at this age.
Advanced maternal age is an independent risk factor in relation to smoking and overweight, and the combination of the three factors is associated with a substantially increased risk of negative pregnancy outcomes.
1. Bewley S, Ledger W, Nikolaou D, editors. Reproductive ageing. London (UK): RCOG Press; 2009.
2. Newburn-Cook CV, Onyskiw JE. Is older maternal age a risk factor for preterm birth and fetal growth restriction? A systematic review. Health Care Women Int 2005;26:852–75.
3. Park IY, Kwon JY, Kim YH, Kim M, Shin JC. Maternal age-specific rates of fetal chromosomal abnormalities at 16–20 weeks' gestation in Korean pregnant women >or=35 years of age. Fetal Diagn Ther 2010;27:214–21.
4. Straube S, Voigt M, Jorch G, Hallier E, Briese V, Borchardt U. Investigation of the association of Apgar score with maternal socio-economic and biological factors: an analysis of German perinatal statistics. Arch Gynecol Obstet 2010;282:135–41.
5. Flenady V, Koopmans L, Middleton P, Froen JF, Smith GC, Gibbons K, et al.. Major risk factors for stillbirth in high-income countries: a systematic review and meta-analysis. Lancet 2011;377:1331–40.
6. Jacobsson B, Ladfors L, Milsom I. Advanced maternal age and adverse perinatal outcome. Obstet Gynecol 2004;104:727–33.
7. Cnattingius S. The epidemiology of smoking during pregnancy: smoking prevalence, maternal characteristics, and pregnancy outcomes. Nicotine Tob Res 2004;6(suppl 2):S125–40.
8. Campbell MK, Cartier S, Xie B, Kouniakis G, Huang W, Han V. Determinants of small for gestational age birth at term. Paediatr Perinat Epidemiol 2012;26:525–33.
9. McDonald SD, Han Z, Mulla S, Beyene J; Knowledge Synthesis Group. Overweight and obesity in mothers and risk of preterm birth and low birth weight infants: systematic review and meta-analyses. BMJ 2010;341:c3428.
10. Nohr EA, Villamor E, Vaeth M, Olsen J, Cnattingius S. Mortality in infants of obese mothers: is risk modified by mode of delivery? Acta Obstet Gynecol Scand 2012;91:363–71.
11. Cnattingius S, Ericson A, Gunnarskog J, Källén B. A quality study of a medical birth registry. Scand J Soc Med 1990;18:143–8.
12. The Swedish Medical Birth Register. A summary of content and quality. Stockholm (Sweden): Centre for Epidemiology, The National Board of Health and Welfare; 2003.
13. Engeland A, Bjorge T, Daltveit AK, Vollset SE, Furu K. Validation of disease registration in pregnant women in the Medical Birth Registry of Norway. Acta Obstet Gynecol Scand 2009;88:1083–9.
14. Irgens LM. The Medical Birth Registry of Norway. Epidemiological research and surveillance throughout 30 years. Acta Obstet Gynecol Scand 2000;79:435–9.
15. Bagley SC, White H, Golomb BA. Logistic regression in the medical literature: standards for use and reporting, with particular attention to one medical domain. J Clin Epidemiol 2001;54:979–85.
16. Howell DC. Statistical methods for psychology. 7th ed. Belmont (CA): Wadsworth; 2010.
17. Chen XK, Wen SW, Fleming N, Demissie K, Rhoads GG, Walker M. Teenage pregnancy and adverse birth outcomes: a large population based retrospective cohort study. Int J Epidemiol 2007;36:368–73.
18. Cnattingius S, Lambe M. Trends in smoking and overweight during pregnancy: prevalence, risks of pregnancy complications, and adverse pregnancy outcomes. Semin Perinatol 2002;26:286–95.
19. World Health Organization. Controlling the global obesity epidemic. Geneva (Switzerland): World Health Organization; 2003.
20. Naeye RL. Maternal age, obstetric complications, and the outcome of pregnancy. Obstet Gynecol 1983;61:210–6.
21. Flenady V, Middleton P, Smith GC, Duke W, Erwich JJ, Khong TY, et al.. Stillbirths: the way forward in high-income countries. Lancet 2011;377:1703–17.
22. Froen JF, Gardosi JO, Thurmann A, Francis A, Stray-Pedersen B. Restricted fetal growth in sudden intrauterine unexplained death. Acta Obstet Gynecol Scand 2004;83:801–7.
23. Rasmussen S, Irgens LM, Dalaker K. A history of placental dysfunction and risk of placental abruption. Paediatr Perinat Epidemiol 1999;13:9–21.
24. Ananth CV, Vintzileos AM. Ischemic placental disease: epidemiology and risk factors. Eur J Obstet Gynecol Reprod Biol 2011;159:77–82.
25. Main DM, Main EK, Moore DH 2nd. The relationship between maternal age and uterine dysfunction: a continuous effect throughout reproductive life. Am J Obstet Gynecol 2000;182:1312–20.
26. Smith G The science of the ageing uterus and placenta. London (UK): RCOG Press; 2009.
27. Smith GC. Age at menarche and the risk of operative first delivery. BJOG 2009;116:1613–21.
28. Gareen IF, Morgenstern H, Greenland S, Gifford DS. Explaining the association of maternal age with Cesarean delivery for nulliparous and parous women. J Clin Epidemiol 2003;56:1100–10.
29. Adashek JA, Peaceman AM, Lopez-Zeno JA, Minogue JP, Socol ML. Factors contributing to the increased cesarean birth rate in older parturient women. Am J Obstet Gynecol 1993;169:936–40.
30. Du Q, Jovanović S, Tulić L, Sljivančanin D, Jack DW, Zižić V, et al.. KATP channels are up-regulated with increasing age in human myometrium. Mech Ageing Dev 2013;134:98–102.
31. Ahlsten G, Ewald U, Tuvemo T. Maternal smoking reduces prostacyclin formation in human umbilical arteries. A study on strictly selected pregnancies. Acta Obstet Gynecol Scand 1986;65:645–9.
32. Longo LD. Carbon monoxide: effects on oxygenation of the fetus in utero. Science 1976;194:523–5.
33. Morrow RJ, Ritchie JW, Bull SB. Maternal cigarette smoking: the effects on umbilical and uterine blood flow velocity. Am J Obstet Gynecol 1988;159:1069–71.
34. Villamor E, Cnattingius S. Interpregnancy weight change and risk of adverse pregnancy outcomes: a population-based study. Lancet 2006;368:1164–70.
35. Ramsay JE, Ferrell WR, Crawford L, Wallace AM, Greer IA, Sattar N. Maternal obesity is associated with dysregulation of metabolic, vascular, and inflammatory pathways. J Clin Endocrinol Metab 2002;87:4231–7.
© 2014 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
36. Retnakaran R, Hanley AJ, Raif N, Connelly PW, Sermer M, Zinman B. C-reactive protein and gestational diabetes: the central role of maternal obesity. J Clin Endocrinol Metab 2003;88:3507–12.