Overall, 1,134 women were approached to participate in MAMAS. 585 of these mothers (51%) either did not meet the inclusion criteria or declined to participate (Fig. 1). 549 mothers consented before 28 weeks gestation. A further 40 participants were recruited between 28 weeks’ gestation and delivery. 62 mothers (11%) were either lost to follow up or withdrew from the study between 28 weeks’ gestation and delivery. Therefore, demographic, medical and pregnancy data was collected on a final cohort of 527 participants (n = 158 20–30, n = 212 35–39 and n = 157 ≥ 40 year olds).
There were no differences between ethnicity, employment and deprivation between participants in different age groups (Table 1). Paternal age was higher in pregnancies to mothers 35–39 and ≥ 40 years, and mothers ≥ 40 years had higher BMI (p = 0.0025) compared to 20–30 year olds (p < 0.0001). More mothers aged 35–39 years were married, and more mothers ≥ 40 years were in partnerships than the other maternal age groups (p = 0.003). More AMA mothers were non-smokers (p = 0.047), homeowners (p < 0.001), multiparous (p = 0.0005), previous FGR (p = 0.022, 35–39 years only) and had previous or current fertility treatment (predominantly in vitro fertilisation, p < 0.001) compared to controls. Of those who conceived using assisted reproductive technologies, 47% of those ≥ 40 years used egg and/or sperm donors, compared to 8% of 35–39 year olds.
Table 1
Demographic data for total MAMAS participants.
Demographics
|
20–30 Years
(n = 158)
|
35–39 Years
(n = 212)
|
≥ 40 Years
(n = 157)
|
p value
Overall
|
p value
Multiple Comparisons
|
Maternal age
|
26 (20–30)
|
37 (35–39)
|
42 (40–49)
|
|
|
Paternal Age a
|
29 (18–48)
|
38 (21–50)
|
43 (25–60)
|
< 0.0001
|
†¶Δ<0.0001
|
BMI (kg/m2) a
|
23.8 (18.7–29.9)
|
24.1 (18.5–29.9)
|
25.0 (19.0-29.8)
|
0.0024
|
†0.95, ¶0.0025, Δ0.027
|
Ethnicity b
European
|
88% (144)
|
79% (168)
|
92% (144)
|
0.23
|
|
Marital Status b
Single
Married/CP
Partner
Other
|
4% (7)
41% (65)
47% (74)
8% (12)
|
2% (5)
62% (131)
31% (66)
5% (10)
|
4% (6)
37% (58)
52% (81)
7% (12)
|
0.003
|
†<0.001, ¶0.18, Δ0.21
|
Occupation b
Employed
Unemployed
Unknown
|
76% (120)
17% (27)
7% (11)
|
79% (167)
15% (31)
6% (14)
|
75% (117)
16% (25)
9% (15)
|
0.65
|
|
Smoking Status b
Non-smoker
Current
Ex-smoker
|
72% (113)
11% (17)
17% (28)
|
77% (163)
6% (11)
17% (37)
|
83% (130)
3% (5)
14% (21)
|
0.047
|
†0.12, ¶0.021, Δ0.25
|
Housing b
Owns
|
40% (63)
|
40% (107)
|
78% (122)
|
< 0.001
|
†0.043, ¶Δp < 0.001
|
IMD a
Score
|
18.89
(2.28–71.95)
|
16.11
(1.94–76.09)
|
15.54
(1.94–73.53)
|
0.25
|
|
Parity b
Primiparous
Parous
Grandmultiparous
|
50% (79)
50% (79)
0% (0)
|
29% (61)
70% (148)
1% (3)
|
32% (52)
63% (99)
4% (6)
|
< 0.001
|
†0.0001, ¶0.37, Δ0.0024
|
Previous APO b
FGR
Stillbirth
|
16% (26)
6% (11)
< 1% (1)
|
24% (50)
15% (31)
2% (5)
|
22% (34)
8% (13)
2% (3)
|
0.24
0.046
0.63
|
†0.022, ¶0.66, Δ0.063
|
Previous ART b
|
< 1% (1)
|
9% (20)
|
20% (31)
|
< 0.001
|
†0.0003, ¶0.0001, Δ0.0045
|
ART b
Hormonal
IVF
IVF - Donor
IUI
|
1% (2)
1% (2)
0% (0)
0% (0)
0% (0)
|
10% (22)
3% (7)
6% 12
< 1%(1)
1% (2)
|
14% (22)
2% (3)
12% (19)
6% (9)
0% (0)
|
< 0.001
|
†0.0004, ¶0.001, Δ0.29
|
Data are mean (range) or percentage (number). BMI = body mass index, CP = civil partnership, IMD = index of multiple deprivation, APO = adverse pregnancy outcome, ART = assisted reproductive therapies, IVF = in vitro fertilisation, Donor = egg and/or sperm, IUI = intrauterine insemination. Statistical differences are aKruskal-Wallis with Dunn’s multiple comparisons or b Fishers exact test. When overall p > 0.05, multiple comparisons p values are reported († 20–30 vs. 35–39 years, ¶ 20–30 vs ≥ 40 years, Δ 35–39 vs. ≥40 years). Significant differences are highlighted with bold p values |
AMA women delivered moderately earlier than women aged 20–30 years (39 weeks + 1 day vs. 39 weeks + 5 days; p < 0.001; Table 2). Women ≥ 40 years had a 38% rate of IOL compared to 25–30% in the younger groups (p = 0.0065). Fewer AMA mothers had NVDs (p = 0.0012) and more had elective or emergency caesarean sections (p = 0.013 and 0.037 respectively). Birthweight, IBC and the incidence of pregnancy-related maternal disease did not differ between the maternal age groups. AMA women had lower incidence of normal pregnancy outcome (81% of 20–30 year olds vs only 71% and 65% of women aged 35–39 and ≥ 40 years respectively (p = 0.006)). Four women in MAMAS had a stillbirths (between 38+ 1 and 40+ 3 weeks gestation), all were ≥ 40 year old (p = 0.014).
Table 2
Pregnancy outcome of MAMAS participants
Pregnancy Outcome
|
20–30 Years
(n = 158)
|
35–39 Years
(n = 212)
|
≥ 40 Years
(n = 157)
|
p value Overall
|
p Value
Multiple Comparisons
|
Gestation at Delivery a
Weeks + days
|
39 + 5
(34 + 1–42 + 2)
|
39 + 3
(30 + 5–43 + 1)
|
39 + 1
(30 + 1–42 + 4)
|
< 0.001
|
†0.043, ¶<0.0001, Δ0.12
|
Birthweight (g) a
|
3401 (1880–4680)
|
3387 (1300–4900)
|
3375 (1480–4560)
|
0.72
|
---
|
Individualised Birthweight
Centile a
|
45.1 (0.1–98.7)
|
47.3 (0.0-99.6)
|
47.7 (0.0-99.8)
|
0.66
|
---
|
Induction of Labour b
|
30% (48)
|
25% (53)
|
38% (60)
|
0.024
|
†0.25, ¶0.14, Δ 0.0065
|
Mode of Delivery b
Normal Vaginal Delivery
Elective Caesarean Section
Emergency Caesarean Section
Instrumental Vaginal Delivery
|
60% (94)
13% (21)
8% (11)
19% (32)
|
50% (105)
25% (52)
10% (22)
15% (33)
|
39% (61)
25% (40)
17% (25)
19% (31)
|
0.001
|
† 0.009,¶ <0.001, Δ 0.12
|
Maternal Disease b
Preeclampsia
Gestational Diabetes Mellitus
|
4% (7)
< 1% (1)
|
1% (2)
2% (5)
|
4% (6)
2% (3)
|
0.23
0.87
|
|
NPO b
|
81% (128)
|
71% (150)
|
65% (102)
|
0.006
|
†0.024, ¶0.0013, Δ 0.24
|
APO b
Pre-Term Birth
Small for gestational Age
Fetal Growth Restriction
Large for Gestational Age
Neonatal Intensive Care Unit admission
Stillbirth
|
13% (21)
3% (5)
6% (9)
4% (6)
4% (7)
4% (7)
0% (0)
|
14% (30)
4% (9)
8% (17)
5% (11)
4% (9)
7% (14)
0% (0)
|
17% (26)
4% (6)
6% (10)
6% (9)
8% (13)
7% (11)
3% (4)
|
0.69
0.87
0.66
0.71
0.19
0.58
0.014
|
†1.00, ¶0.0044, Δ 0.020
|
Data are mean (range) or percentage (number). Statistical differences are aKruskal-Wallis with Dunn’s multiple comparisons or b Fishers exact test. When overall p > 0.05, multiple comparisons p values are reported († 20–30 vs. 35–39 years, ¶ 20–30 vs ≥ 40 years, Δ 35–39 vs. ≥40 years). |
Demographic predictors of APO: Univariate logistic regression demonstrated that ex- and current smokers had higher odds of APO than non-smokers (OR 2.00 (95%CI 1.20–3.33) and 4.14 (95%CI 2.01–8.55), p = 0.008 and < 0.001 respectively; Table 3). Multiparity was protective against APO (OR 0.65 (95%CI 0.43-1.00), p = 0.048) compared to primiparous women. Maternal ethnicity did not independently affect the risk of APO, although there was a trend towards a protective effect of non-European or Asian ethnicity (“Other” OR of 0.27 (95%CI 0.06–1.16) p = 0.08). ART had no detectable effect on outcome (OR 1.08 (95%CI 0.52–2.24) p = 0.84). Paternal age had a protective effect against APO (OR 0.54 (95%CI 0.31–0.94) p=-0.03). Only the associations between APO and maternal parity or cigarette smoking remained statistically significant after adjusting for maternal ethnicity, parity, smoking status, housing, and paternal age (Table 3).
Table 3
Unadjusted and adjusted odds ratios for prediction of adverse pregnancy outcome of MAMAS participants.
|
Unadjusted
|
Model 1
|
Model 2
|
|
OR
|
95% CI
|
P value
|
AOR
|
95% CI
|
P value
|
AOR
|
95% CI
|
P value
|
Ethnicity
European
Asian
Other
|
Reference
0.71
0.27
|
0.27–1.88
0.06–1.16
|
0.49
0.079
|
Reference
0.77
0.26
|
0.27–2.20
0.06–1.18
|
0.63
0.081
|
Reference
0.63
0.28
|
0.20-2.00
0.06–1.40
|
0.43
0.12
|
Parity
Nulliparous
Multiparous
|
Reference
0.65
|
0.43-1.00
|
0.048
|
Reference
0.57
|
0.35–0.91
|
0.019
|
Reference
0.56
|
0.34–0.91
|
0.019
|
ART
None
ART
|
Reference
1.08
|
0.52–2.238
|
0.84
|
Reference
1.20
|
0.55–2.66
|
0.64
|
Reference
0.99
|
0.43–2.30
|
0.99
|
Smoking
No
Ex
Current
|
Reference
2.00
4.14
|
1.20–3.33
2.01–8.55
|
0.008
< 0.001
|
Reference
1.92
4.76
|
1.12–3.31
2.13–10.60
|
0.018
< 0.001
|
Reference
1.82
4.34
|
1.04–3.17
1.88–9.99
|
0.036
0.001
|
Housing
Rented
Owns
|
Reference
0.72
|
0.47–1.10
|
0.12
|
Reference
0.72
|
0.43–1.18
|
0.19
|
Reference
0.65
|
0.38–1.11
|
0.12
|
IMD Quintile
1st
2nd
3rd
4th
5th
|
Reference
1.31
0.48
1.15
0.97
|
0.69–2.49
0.22-1.00
0.58–2.26
0.50–1.88
|
0.40
0.052
0.69
0.93
|
Reference
1.23
0.42
0.94
0.91
|
0.64–2.39
0.64–0.91
0.46–1.92
0.44–1.87
|
0.53
0.028
0.86
0.80
|
Reference
1.36
0.48
1.07
1.02
|
0.69–2.71
0.21–1.07
0.51–2.24
0.48–2.17
|
0.38
0.073
0.86
0.96
|
Paternal Age
< 30 years
30–34 years
35–39 years
≥ 40 years
|
Reference
0.77
0.54
0.54
|
0.39–1.51
0.30–0.98
0.31–0.94
|
0.45
0.043
0.029
|
|
|
|
Reference
0.86
0.54
0.46
|
0.37–1.98
0.23–1.28
0.19–1.12
|
0.72
0.16
0.08
|
Model 1 – multivariable logistic regression including all variables except paternal age, Model 2 – multivariable regression additionally adjusting for paternal age. ART = Assistive Reproductive Techniques; IMD = Index of multiple deprivation; IMD Quintile 1st = least deprived, 5th = most deprived. Statistics performed were univariate and multivariable meta-regression. Adjustments included maternal ethnicity, parity, smoking status, housing, and paternal age. |
Nested case-control study to determine whether AMA associated with elevated circulating biomarkers of inflammation and oxidative stress
The characteristics of participants (n = 40/group) in this nested study are shown in Supplementary Table 2. The three groups differed for factors strongly associated with older age: higher home-ownership in 35–39 and ≥ 40 year olds (95% of ≥ 40 vs. 90% of 35–39 vs. 58% of 20–30 years) and multiparous (78% of ≥ 40 vs. 80% of 35–39 vs. 53% of 20–30 yeas) and mothers ≥ 40 years had higher previous miscarriage rates (53% vs. 25% of 20–30 years). AMA mothers delivered earlier (by 5–7 days) than 20–30 year olds (p < 0.001; Supplementary Table 3).
Large variation was seen in inflammatory biomarker concentrations in maternal serum. At 28 weeks’ gestation, lower circulating concentrations of IL-1α and IL-1RA (p = 0.04 and 0.005, Fig. 2A, E) were measured in mothers 35–39 years compared to 20–30 years, but no significant differences were present between 20–30 and ≥ 40 year olds. Similar trends were apparent at 36 weeks gestation but were not statistically significant. Anti-inflammatory IL-10 was lower in mothers ≥ 40 years at 28 weeks’ (p = 0.05), with a stepwise age-related decrease apparent at 36 weeks (p = 0.03, Figs. 2G-H). No differences were seen in IL-1β or TNF-α at 28 weeks or 36 weeks between age groups (Fig. 2). There were no differences in maternal circulating markers of oxidative stress at 28 weeks’ gestation (Fig. 3A, C, E). TAC was increased at 36 weeks’ gestation in mothers ≥ 40 years compared to controls (p = 0.015; Fig. 3B). No differences of other oxidative stress markers were seen (Fig. 3D and F, Supplementary Fig. 1). Gestational age affected TAC in mothers ≥ 40 s (p = 0.004), whereas 8-isoprostane was positively related to gestational age in all groups (p < 0.001; data not shown). When assessing change in oxidative status over time, TAC levels fell across the third trimester in controls but increased in AMA women (p = 0.005; Fig. 3G). In contrast, elevated lipid peroxidation (8-isoprostane) was apparent in mothers 20–30 and 35–39 years across the third trimester but decreased in women ≥ 40 (p = 0.04; Fig. 3H). There was a positive relationship between TAC and 8-isoprostane in all participants at 28 weeks’ gestation, strongest in women 35–39 years (r = 0.61 vs 0.45 in women aged 20–30, Fig. 3I). At 36 weeks a negative correlation existed between TAC and 8-isoprostane in women 35–39 and ≥ 40 years (r=-0.42 and − 0.39 respectively; p = 0.005 and p = 0.01 Fig. 3J), whereas no relationship was seen in women aged 20–30 (r=-0.17, p = 0.32).
Nested case-control study to determine whether maternal biomarkers associated with adverse pregnancy outcome in women of AMA
Women ≥ 35 years of age who had APO (n = 43) were well matched for demographic variables to women ≥ 35 years of age who had a normal pregnancy outcome (Supplementary Table 4). The majority of infants in the APO group were classified as SGA (84%), with 44% under the 5th centile (FGR) (Supplementary Table 5). There were fewer NVDs (39% vs. 65%, p = 0.02) and more EMCS (23% vs. 7%, p = 0.04) in the APO group compared to NPOs. A quarter of mothers in the APO group delivered preterm (< 37 weeks, p < 0.001) and a quarter of infants were admitted to NICU (p < 0.001). There were 3 stillbirths included in the APO group (one stillbirth was excluded due to congenital abnormality identified as the cause of death).
There were no differences in circulating cytokines at 28 weeks’ or 36 weeks’ gestation between AMA women with normal and adverse pregnancy outcome (Supplementary Fig. 2). TAC was higher in AMA mothers with APO compared to normal outcomes at both 28 and 36 weeks’ gestation (p = 0.002 and 0.006 respectively, Fig. 4A, B). 8-isoprostane increased between 28–36 weeks’ gestation in both NPO and APO groups (data not shown). 8-isoprostane also significantly increased in AMA mothers with APO at 28 weeks but was not elevated at 36 weeks’ gestation (p = 0.01 and p = 0.82 respectively, Fig. 4E-F). No differences were detected in markers of DNA/RNA damage or protein carbonyl between groups at either 28 weeks (p = 0.66 and 0.34 respectively) or 36 weeks’ gestation (p = 0.57 and 0.60 respectively; Fig. 4C, D, G, H). A positive correlation between TAC and 8-isoprostane at 28 (r = 0.46 for APO and 0.31 for NPO) and at 36 weeks with normal outcomes (r=-0.39; p = 0.01 Fig. 4J), whereas there was no relationship seen in women with APO (r=-0.33, p = 0.07, Fig. 4I-J).
Concentrations of hPL, PlGF and PlGF:sFlt ratio were unchanged at 28 weeks’ gestation (Fig. 5A, C, G). hPL was lower at 36 weeks’ gestation in women with APO (p = 0.007; Fig. 5B). Similarly, PlGF concentrations were lower at 36 weeks’ when measured alone (p < 0.001, Fig. 5D) or adjusted for sFlt-1 (human VEGF R1/Flt-1) concentrations (p = 0.03, Fig. 5H). sFlt-1 was lower in women with APO at 28 weeks’ with a similar trend at 36 weeks’ gestation (p = 0.05 and 0.07, Fig. 5E,F). No differences were detected in circulating hCG, PAPP-A or progesterone with APO (Supplementary Fig. 3). ROC curves were created for all biomarkers that reached a statistical significance of p < 0.01 between NPO and APO. TAC and 8-isoprostane had predictive area under the curve values of 0.69 and 0.66 respectively (ranked as a poor prognostic markers, Fig. 6A-B), whilst hPL and PlGF had predictive values of 0.68 (poor) and 0.74 (fair), respectively (Fig. 6C-D).