*Dr O. Fainaru, Department of Obstetrics and Gynecology, Lis Maternity Hospital, Tel-Aviv Sourasky Medical Center, 6 Weizmann Street, Tel Aviv 64239, Israel.
Objective As a first step towards evaluating the role of oxidative stress in the process of labour, we tested whether term labour is associated with increased oxidisibility of maternal serum lipids.
Design A controlled prospective study.
Setting Tertiary care centre.
Population Twenty healthy women in active labour and 20 healthy pregnant women not in labour (controls) matched for maternal and gestational age.
Methods Venous blood was drawn from women in both groups. Serum levels of lipid peroxidation products and the kinetics of copper-induced oxidation ex vivo were monitored spectroscopically at 37°C by continuous recording of absorbance at 245 nm.
Main outcome measures Oxidative stress parameters.
Results The initial optical density (OD) at 245 nm, attributed to preformed dienic hydroperoxides and 7-keto-cholesterol (main products of lipid peroxidation), was higher in the labouring group than in the controls (1.30 ± 0.11 vs 1.18 ± 0.09, OD 245 nm, respectively, P < 0.001). The lag phase, reflecting resistance of serum lipids to oxidation, was significantly shorter in the labouring group than in the controls (43.2 ± 1.4 vs 56.2 ± 4.7 min, respectively, P= 0.01).
Conclusion High levels of serum hydroperoxides and decreased resistance of serum lipids to copper-induced peroxidation ex vivo suggest labour to be associated with high oxidative stress. Whether oxidative stress is involved in initiating the labour process or is consequent awaits further studies.
Little is known about the triggers and mechanisms by which term labour is initiated and by which it progresses. Several recent reports1–5 have suggested that indicators for oxidative stress are higher in women during term labour. Most of these indicators are indirect, however, and are subject to experimental artifact. Preterm labour and preterm prelabour rupture of the membranes are associated with elevated activity of matrix metalloproteinases in the amniotic fluid6 and amniochorionic membranes7. Recently, a report by Buhimschi et al.8 suggested that oxygen free radicals, which are by-products of inflammatory cells, increase the activity of matrix metalloproteinases in human fetal membranes. This may cause degradation of the extracellular matrix, leading to effacement and dilation of the uterine cervix or to rupture of the membranes9.
Much effort has been devoted to the development of ex vivo assays for the evaluation of the oxidative stress responsible for oxygen free radical chain reactions. These include direct quantification of oxidation products in the circulation (‘oxidised low density lipoproteins’, hydroperoxides, thiobarbituric acid-reactive substances) and evaluation of the susceptibility of low density lipoproteins (LDL) to copper-induced oxidation ex vivo10. The clinical relevance of all these analyses of oxidation products and ‘oxidisibility’ is limited by many factors, including possible artifacts that may be caused by fractionation of the LDL and by the continuing peroxidation of blood lipids ex vivo during processing. In relation to ‘oxidative stress’, it is commonly assumed that the most sensitive factor that can be studied ex vivo is the ‘oxidisibility’ of the plasma lipids, namely, their propensity to undergo oxidation upon exposure to various inducers of oxidation. Similar to in vivo‘oxidative stress’, the ‘oxidisibility’ex vivo is a very complex function of the level of prooxidants, including preformed hydroperoxides, and of the concentration of water-soluble and lipid-soluble naturally occurring antioxidants (such as vitamin C and vitamin E, respectively). Our previously developed spectroscopic method, which is capable of monitoring the kinetics of copper-induced peroxidation of serum lipids in unfractionated serum ex vivo11,12, does not involve any processing of the serum, aside from freezing and thawing which we have shown to have no effect on the kinetics of ex vivo oxidation13. Given the possible artifacts that may be caused by LDL fractionation, our assay, which had been previously shown to correlate with the kinetics of copper-induced peroxidation of isolated lipoproteins11, is likely to be more relevant to lipid ‘oxidisibility’in vivo than methods based on fractionated LDL. This method has been shown11–13 to highly correlate with the well established method of determining isolated LDL oxidisibility by monitoring conjugated diene accumulation upon in vitro exposure to copper ions.
In order to study the putative role of oxidative stress in the process of labour, we have used our simplified assay to determine serum lipid oxidisibility in pregnant women in labour compared with pregnant women not in labour.
The study group consisted of 20 healthy women in active labour at term and the control group consisted of 20 healthy pregnant women who were not in labour and matched for maternal and gestational age (Table 1). The parity of the two groups was also similar. Gestational age was determined by the best obstetric estimate based on a combination of the last menstrual period dating and the earliest available ultrasonographic examination.
Table 1. Maternal characteristics [mean (SEM)].
Active labour (n= 20)
Control* (n= 20)
* Pregnant women not in labour.
Gestational age (months)
Venous blood was drawn from each woman and serum was prepared, frozen immediately and stored at −70°C.
The study was approved by the Tel-Aviv Sourasky Medical Center institutional review board and informed consent was obtained from each woman to participate in the study.
Copper-induced oxidation11 was monitored at 37°C by continuous recording of absorbance at 245 nm using a Kontron (Uvikon 933) double-beam spectrophotometer equipped with a 12-position automated sample changer. Measurements were carried out in quartz cuvettes (optical pathway 1 cm) after the addition of CuCl2 (final concentration 100 μM) to a solution containing 720 μM sodium citrate in phosphate-buffered saline (146 mM NaCl, 3.3 mM NaH2PO4, 3.3 mM Na2HPO4, pH = 7.4) and 30 μL serum in a final volume of 1.5 mL12.
The time-dependent increase of absorbance at 245 nm is mainly due to the formation of conjugated dienic hydroperoxides and 7-keto-cholesterol during lipid peroxidation11,14. The absorbance at 245 nm prior to the addition of CuCl2 reflects the concentration of preformed oxidation products (mainly hydroperoxides15). The kinetic profiles obtained after the addition of CuCl2 were analysed in terms of the lag preceding oxidation and the time at which the rate of reaction was maximal (tmax), both reflecting the resistance of serum lipids to oxidation, the maximal rate of accumulation of absorbing products (Vmax) as computed from the first derivative of the time course of absorption and the maximal accumulation of absorbing products (ODmax). These kinetic attributes were evaluated from the kinetic profiles as shown in Fig. 1.
All the data were analysed by ‘Microsoft Excel 7.0’ electronic spreadsheets. The data obtained for the two groups were compared using the two-tailed Student's t test. P < 0.05 was considered significant.
The two groups were similar with respect to the maternal and gestational age as well as parity (Table 1).
Representative kinetic profiles of copper-induced oxidation of unfractionated sera of a woman in labour and of another woman not in labour are shown in Fig. 1, and the kinetic parameters of copper-induced oxidation are listed in Table 2. The data given in this table reveal several differences between the two groups. First, the initial optical density (OD) that was measured at 245 nm (reflecting preformed lipid peroxidation products) was significantly higher in the sera of labouring women than in the sera of the pregnant women not in labour. Second, both tmax and the lag preceding oxidation were significantly shorter in the labouring women. The difference between the women in both groups was similar for the lag (∼13 min) and for tmax (∼12 min). This similarity indicates that the different ‘oxidisibility’ of serum lipids results either from the higher concentration of hydroperoxides and/or from a lower concentration of antioxidants, both known to cause a ‘left shift’ of the kinetic profile15. Lastly, although the maximal accumulation of oxidation products (ODmax) was slightly but significantly higher in the labouring women, the maximal rate of accumulation of absorbing products was similar in the two groups.
Table 2. Kinetic parameters of copper-induced oxidation [mean (SEM)].
Active labour (n= 20)
Controls* (n= 20)
* Pregnant women not in labour.
Initial OD (245 nm)
Vmax (OD 245 nm/minute)
ODmax (OD 245 nm)
The major findings of the present study are that the sera of labouring women contain higher levels of preformed lipid peroxidation products and exhibit shorter lag times preceding copper-induced oxidation compared with pregnant women not in labour (Table 2). These findings, reflecting increased oxidative stress, are in accord with previous observations showing that (i) the activity of xanthine oxidase in placentae of labouring women was found to be higher than in placentae obtained from elective cesarean deliveries1; (ii) the serum levels of several indicators of oxidative stress, including thiobarbituric acid-reactive substances2,3, superoxide dismutase, glutathione peroxidase and catalase3, were all found to be increased during labour; (iii) interestingly, the levels of malonyldialdehyde (MDA) in the neonatal blood from vaginal deliveries were also higher in labouring women than in women undergoing elective cesarean deliveries4; (iv) the MDA levels correlated with the duration of the second stage of labour5. The results of our optimised spectroscopic method, which is less prone to experimental errors than other techniques, lend strong support to the conclusion that labour is associated with oxidative stress.
These findings raise the question as to whether oxidative stress is involved in the labour process. It is possible that superoxide anions, generated by the observed increased oxidative stress, may contribute to the activation of matrix metalloproteinases8, causing enhanced degradation of the extracellular matrix. In turn, the latter may contribute to the propagation of labour. Conversely, antioxidants may possibly act as inhibitors of the activation of matrix metalloproteinases and, by doing so, prevent or inhibit preterm labour or preterm rupture of the membranes.
Another finding of interest is that the maximal accumulation of absorbing products (ODmax) was significantly higher in labouring women. Our earlier results14 had indicated that ODmax correlates with the total cholesterol content of the sera. Hence, the previously reported rise in total cholesterol content during labour16 may explain the higher ODmax that was observed for the group of women in labour. We have not measured the serum levels of antioxidants. Therfore, we cannot attribute the left shift in the kinetic profile solely to the observed higher levels of hydroperoxides, as possible decreased levels of serum antioxidants might have also contributed to this effect. The possibility should be raised that our findings of increased oxidative stress during labour may be affected at least in part by the physical exertion associated with labour, as has been shown previously during aerobic exercise17–19.
Whatever the precise mechanism responsible for the high susceptibility of serum lipids to oxidation in labour, our results indicate that labouring women are subject to high oxidative stress. Although our study included labouring women at term pregnancies, we speculate that our results may reflect a common feature of the process of labour itself, regardless of its dating.