SEARCH

SEARCH BY CITATION

Keywords:

  • Caesarean section;
  • myometrial contractility;
  • obesity

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgement
  9. References

Objective  The aim of the study was to elucidate the reason for the high rate of caesarean section in obese women. We examined the following hypotheses: (1) obese women have a high incidence of complications related to poor uterine contractility—caesarean section for dysfunctional labour and postpartum haemorrhage. 2) The myometrium from obese women has less ability to contract in vitro.

Design  First, a clinical retrospective analysis of data from 3913 completed singleton pregnancies was performed. Secondly, in a prospective study the force, frequency and intracellular [Ca2+] flux of spontaneously contracting myometrium were related to the maternal body mass index.

Setting  Liverpool Women’s Hospital and University of Liverpool.

Population  The clinical study involved all women who delivered in one hospital in 2002. The in vitro study myometrial biopsies were obtained from 73 women who had elective caesarean section at term.

Results  Maternal obesity carried significant risk of caesarean section in labour that was highest for delay in the first stage of labour (OR 3.54). The increased risk of caesarean section in obese women largely occurred in women with normal- and not with high-birthweight infants. Obese women delivering vaginally had increased risk of prolonged first stage of labour and excessive blood loss. Myometrium from obese women contracted with less force and frequency and had less [Ca2+] flux than that from normal-weight women.

Conclusions  We suggest that these findings indicate that obesity may impair the ability of the uterus to contract in labour.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgement
  9. References

Obesity is a serious public health problem, and its prevalence is increasing globally in all age groups.1 The UK Obesity Statistics (2003) state that 32% of women in the UK are overweight and an additional 21% of women are obese. Several published studies have reported an excess risk of caesarean section in obese women.2–6 Despite this significant clinical problem, none of the previous studies have explored the underlying mechanisms and pathogenesis for the association between obesity and raised caesarean section rate.2–6 Most clinicians suspect that the rise in caesarean section rate associated with obesity is due to obstructed labour caused by increased deposition of soft tissues in the maternal pelvis and larger babies.

Body mass index (BMI) and hypercholesterolaemia are well known to be positively associated.7 Cholesterol, an essential component in cell membranes, has recently also been found to have an important role in controlling smooth muscle contractility.8–10 Several components of cellular signalling systems important for smooth muscle signal transduction have been found to localise in cholesterol-rich regions of the cell membrane, known as lipid rafts and caveolae.11–13 In particular, oestrogen and oxytocin receptors on the myometrial membrane are localised to caveolae, and their efficacy is modulated by cholesterol content.14,15 Cholesterol levels are raised in obese pregnant women’s serum16 and myometrial membranes.17 Such alterations of cholesterol may be anticipated to affect contractility in human myometrium, although no studies have investigated this. Therefore, to elucidate the pathophysiology of obesity and increased risk of caesarean section, we tested the hypotheses that the high incidence of caesarean section in obese women is due to dysfunctional labour because of poor myometrial contractility rather than due to obstructed labour.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgement
  9. References

This study was designed as a retrospective study and was conducted using the obstetric records including 3913 women with live births after 20 weeks of gestation in Liverpool Women’s Hospital between 1 January and 30 December 2002. Data were gleaned from the Meditech database and from the labour and delivery record. These data were entered at booking and soon after every delivery.

There were 5672 deliveries in Liverpool Women’s Hospital in 2002. We excluded 576 women who had elective caesarean section or multiple births and 1183 (23%) because of incomplete or faulty data (random midwife input errors). This left 3913 women in whom there were sufficient data left for analysis with respect to BMI and mode of delivery. Maternal heights and weights first recorded at booking in gestational weeks 10–12 were used as the basis for the calculation of BMI (kg/m2).

Based on the WHO classification, BMI was initially divided into four categories: BMI ≤ 19.9 kg/m2 (underweight); BMI 20–24.9 kg/m2 (normal weight); BMI 25–29.9 kg/m2 (overweight); BMI 30–34.9 kg/m2 (obese) and BMI >35 kg/m2.

A midwife entered the indication for emergency caesarean section into the Mediteck hospital database immediately after delivery. Indications were selected from the following list of options: delay in the first stage of labour, delay in the second stage of labour, fetal distress antepartum haemorrhage and breech presentation. Caesarean section for delay in the first stage of labour was performed if there had been no progress in active labour despite 4 hours of uterine contractions augmented by the maximal dose of oxytocin in the hospital protocol (primigravid 2 units/hour, multgravid 1 unit/hour). Caesarean section for delay in the second stage of labour was performed if the fetal head had not descended below the ischial spines despite allowing the head to descend without maternal effort for 2 hours followed by 1-hour active pushing.

To evaluate the risk of high BMI on spontaneous delivery, separate analyses were conducted on a subset with blood loss and length of the first stage. Blood loss was estimated by the midwife who collected the lost blood in a calibrated container. As an uncomplicated vaginal delivery causes 500–600 ml of blood loss,18,19 we defined the blood loss of more than 600 ml as excessive blood loss in spontaneous delivery. High birthweight was classified as a birthweight ≥4000 g and low birthweight was considered as a birthweight ≤2500 g.

We calculated crude odds ratio and 95% CI with the normal weight as standard risk; adjusted odds ratios were determined by using Mantel–Haenszel technique. Maternal age, parity, smoking, birthweight and induction of labour were believed to be potential confounding factors and were included as covariates in the adjusted analyses.

Tissue

Seventy-three biopsies were taken, with informed consent and local Ethical Committee approval, from women undergoing elective caesarean section at term (37–41 weeks). The indications for the caesarean section in the pregnant women were previous caesarean section (24 women), fetal malpresentation (15 women), previous traumatic delivery (6 women), fetal reason (14 women) and maternal reason (14 women). The median age of these women was 29 years (interquartile range 5 years). Exclusion criteria were serious medical complications or the use of medication likely to affect myometrial activity. Full-thickness tissue biopsy specimens were obtained from lower segment of the uterus at caesarean section. Four samples were from underweight, 24 from normal weight, 17 from overweight and 28 from obese women. Small strips (1 × 5 mm) of longitudinal myometrium were isolated from the samples using a dissecting microscope, with care taken during dissection to avoid scar tissue.

Contractility and calcium measurements

The uterus like the heart is made of smooth muscle the nature of which is to contract rhythmically. We studied the spontaneous contractions from the myometrial strips; hence, we studied the intrinsic property of uterine smooth muscle. Measurements of force were made in all preparations. In some experiments, simultaneous Ca2+ recordings were also made, as follows. The longitudinal strips were loaded for 4 hours at 21°C in physiological saline solution containing 25 μM of the membranepermeant form of the calcium-sensitive indicator indo-1 (Molecular Probes, Eugene, OR, USA). The indo-1 was excited at 350 nm and emitted light at 400 and 500 nm detected and used to measure Ca2+, as described in detail elsewhere.10 The myometrial strips were mounted in a 150-μl chamber on an inverted Nikon microscope and viewed with a 10× fluor objective lens. The strips were attached at each end to two metal hooks, one of which was fixed and the other was attached to tension transducer (Grass FT03). Strips were stretched to 1.5 times their slack length and perfused with physiological saline at 5 ml/minute at pH 7.4 and 35°C. The strips were allowed to contract spontaneously, and once a regular pattern of contraction was established, the amplitude and frequency of contractions were measured over a 10-minute period. Student’s t test (P < 0.01) and one-way analysis of variants (P < 0.05) were used to test for significant differences. All chemicals were obtained from Sigma (Poole, Dorset, UK) unless stated otherwise.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgement
  9. References

Of the 3913 completed singleton pregnancies, 1106 (28%) mothers were overweight and 566 (15%) obese. The groups were not significantly different, with regard to age, primiparity and weeks of gestation at delivery (Table 1). We found that there was a significant increase in the risk for caesarean section rate for delay in first and second stage of labour in overweight and obese women (Figure 1A,B, Table 2). Our data also showed that compared with normal-weight women, overweight and obese women were more likely to be induced (Table 2). Therefore, we analysed induced labours separately. Obese women, even after induction of labour, were still more likely to have a caesarean section for delay in the first stage compared with normal-weight women (Table 2). For both spontaneous and induced labours, the risk of caesarean section in the first stage of labour for obese women was greater than in second stage (Table 2).

Table 1.  Demographic characteristic of women with different BMI expressed as frequency (%)
 Total (%)Age (mean ± SD)Parity 0 (%)Gestation (mean ± SD)Smoking (%)
Underweight362 (9)26.1 ± 5.854292.4 ± 2.342.8
Normal1879 (48)28.9 ± 8.350291.7 ± 16.228
Overweight1106 (28)29.5 ± 5.846292.4 ± 2.026.3
Obese566 (15)29.3 ± 5.843292.8 ± 2.529.9
image

Figure 1. The impact of BMI on the proportion of emergency caesarean sections. (A) Proportion of caesarean section. (B) Caesarean section for delay in the first stage of labour. Underweight (under), BMI <19.9 kg/m2; normal (norm), BMI 20–24.9 kg/m2; overweight (over), BMI 25–29.9 kg/m2; obese, BMI >30 kg/m2. Asterisks indicate statistically significant difference from normal-weight group.

Download figure to PowerPoint

Table 2.  The impact of BMI on maternal and infant outcomes in 3913 women
 BMIProportion (%)OR (95% CI) compared to normal weight
  • *

    Statistically significant.

Caesarean sectionUnderweight8.10.74 (0.5–1.1)*
Normal weight13 
Overweight15.81.53 (1.23–1.53)*
Obese17.81.83 (1.41–2.38)*
Caesarean section for delay in first stage after spontaneous onset of labourUnderweight0.60.35 (0.08–1.44)
Normal weight2.2 
Overweight42.49 (1.58–3.94)*
Obese5.73.54 (2.17–5.78)*
Caesarean section for delay in second stage after spontaneous onset of labourUnderweight2.21.43 (0.66–3.11)
Normal weight1.5 
Overweight4.62.99 (1.91–4.68)*
Obese3.42.18 (1.23–3.85)*
Rate of induction of labourUnderweight24.60.89 (0.68–1.15)
Normal weight26.8 
Overweight34.21.41 (1.21–1.66)*
Obese43.52.10 (1.73–2.55)*
Caesarean section for delay in first stage after induction of labourUnderweight2.20.53 (0.12–2.30)
Normal weight4.2 
Overweight6.61.63 (0.90–2.96)
Obese8.92.26 (1.22–4.19)*
Caesarean section for delay in second stage after induction of labourUnderweight7.91.02 (0.44–2.35)
Normal weight7.7 
Overweight10.11.33 (0.83–2.13)
Obese10.61.41 (0.84–2.37)
High birthweightUnderweight5.60.51 (0.32–0.82)*
Normal6.5 
Overweight14.31.40 (1.12–1.77)*
Obese20.52.15 (1.67–2.76)*

We found that birthweight was positively associated with increased maternal BMI (Table 2). Adjusted odds ratios were calculated to account for confounding factors (Table 3). Even after controlling for maternal age, parity and smoking, there was still an increased rate of caesarean section in obese women. When the odds ratio was adjusted for high infant birthweight, the risk of caesarean section was less for obese women compared with normal-weight women (Table 3). In addition, controlling for normal birthweight again found a significantly increased risk of caesarean section in obese women. These findings suggested that the significantly increased caesarean section risk in obese women largely occurred in women with normal-birthweight infants and not in those with high-birthweight infants (Table 3).

Table 3.  The effect of BMI on caesarean rate: OR (95% CI) compared with normal-weight women adjusted for confounding factors
 Confounding factor
AgeSmokingParityHigh birthweightNormal birthweight
  • *

    Statistically significant.

Underweight0.66 (0.44–0.99)*0.57 (0.39–0.84)*0.71 (0.47–1.06)0.08 (0.01–0.62)*0.59 (0.37–0.95)*
Overweight1.15 (0.93–1.43)1.39 (1.12–1.73)*1.61 (1.29–2.02)*0.49 (0.29–0.84)*1.25 (0.97–1.62)
Obese1.37 (1.05–1.77)*1.67 (1.29–2.16)*2.02 (1.54–2.64)*0.52 (0.29–0.84)*1.51 (1.11–2.07)*

Obese women in spontaneous labour, delivering vaginally, had an increased incidence of excessive blood loss (>600 ml)18,19 (Table 4). When we analysed the prolongation of first stage of labour in all women delivering vaginally, there was no significant difference in the duration of labour among these groups (data not shown). However, there was an increased risk of prolonged first stage associated with increasing BMI in primigravid woman in spontaneous labour who delivered vaginally (Table 4).

Table 4.  The effect of BMI on spontaneous vaginal delivery complications
 BMISpontaneous deliveryProportion (%)OR (95% CI)
  • *

    Statistically significant.

Excessive blood loss (>600 ml)Underweight2711.480.55 (0.2–1.57)
Normal weight13362.69 
Overweight7142.81.07 (0.61–1.85)
Obese3125.452.13 (1.18– 3.84)*
Prolongation of first stage of labour in primiparous (>840 minutes)Underweight13914.391.12 (0.66–1.89)
Normal weight62613.1 
Overweight31118.011.46 (1.01–2.11)*
Obese10922.021.87 (1.13–3.12)*

Spontaneous contractions in the myometrial strip obtained from women having elective caesarean section at term show a marked decline in myometrial activity with increasing patient weight (Figure 2). Typical examples of traces obtained from myometrial strips are shown in Figure 2, where there is a clear decrease in the frequency and amplitude of both force and simultaneously measured calcium transients with increasing maternal weight. Analyses of data from all myometrial samples found a significant reduction in amplitude of contractions between normal and overweight (P= 0.046) and normal-weight and obese women (P= 0.049) and in frequency of contractions between normal and overweight (P= 0.0006) and normal-weight and obese women (P= 0.0006) (Figure 2).

image

Figure 2. The force and Ca2+ in spontaneously contracting myometrial strips taken from women having elective caesarean section at term (top). The amplitude and frequency of contractions from underweight (uw), BMI <19.9 kg/m2; normal (n), BMI 20–24.9 kg/m2; overweight (ow), BMI 25–29.9 kg/m2; obese (ob), BMI >30 kg/m2. Asterisks indicate statistically significant difference from normal-weight group.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgement
  9. References

Our data showed that there was a progressive increase in the rate of caesarean section for delay in the first stage with increasing the BMI (Figure 1B), confirming previous studies.2–6 We next asked what was the reason for the emergency caesarean section? We found an increased risk for caesarean section in overweight and in obese women for delay in labour even when labour was induced (Figure 1, Table 2). The greatest risk of caesarean section in obese women was for delay in the first stage of labour, which is likely to be largely contributed to poor uterine contractility. In contrast, the relative risk of obese women having caesarean section for delay in the second stage of labour was less when there had been sufficient uterine activity to fully dilate the cervix but inadequate descent and or rotation of the fetal head; thus, these second-stage caesarean sections are more likely to be due to relative cephalopelvic disproportion. These data suggest that poor contractility may be part of the underlying cause for the raised risk of emergency caesarean section in obese women. However, it is reasonable to question if there are confounding factors in this analysis. Birthweight was positively associated with increased maternal BMI (Table 2), potentially indicating an obstructive cause for the caesarean section. To address the problem of confounding factors, adjusted odds ratios were calculated, an increased risk of caesarean section was persistently found despite adjustment for maternal age, smoking and parity. When we adjusted for the effects of birthweight, the increased risk of caesarean section largely occurred in women with normal- and not with high-birthweight infants. This again suggests that poor myometrial contractility rather than an obstructive reason for the caesarean section rate in obese women.

If myometrial contractility in obese women is poor, does it manifest in other ways? Uterine contractions following delivery are an essential mechanism for stemming postpartum blood loss. If contractility is impaired, then increased blood loss would be predicted and therefore we next examined this in the data set. The obese women in spontaneous labour delivering vaginally did have an increased incidence of excessive blood loss and prolonged first stage of labour if primigravid (Table 3). This is again consistent with the hypothesis posed, namely that inadequate uterine activity is associated with obesity.

The finding of poor uterine contractility from the clinical data was confirmed by the in vitro study. Myometrium obtained from obese women having an elective caesarean section at term contracted with less force and less frequently than that obtained from normal-weight women. Furthermore, the simultaneous measurements of intracellular [Ca2+] indicate that alterations in [Ca2+] are responsible for the changes in contractility. Our finding of poor myometrial contractility in obese women is supported by a previous study that found that overweight and obese women had a significantly slower labour progression before 7-cm dilation and an increased rate of first-stage caesarean section.20 In contrast, other authors have suggested that the increase in caesarean section in obese women is due to cephalopelvic disproportion secondary to increased pelvic soft tissue.6 However, these authors stated that they had not considered the effect of BMI on the quality of contractions and argued that an increase in fetal size of only 142 g in obese women was sufficient to cause sufficient cephalopelvic disproportion to warrant caesarean section, which is intuitively unlikely.6

Recent data from the physiological literature suggest a mechanism by which obesity could inhibit myometrial contractility. Maternal obesity and hypercholesterolaemia are positively associated,7 and obese pregnant women have been found to have high cholesterol, low-density lipoprotein and very low-density lipoprotein levels.21,22 Fasting plasma triglyceride and very low-density lipoprotein cholesterol (VLDL-C) concentrations were significantly higher and high-density lipoprotein cholesterol concentrations lower in obese than in normal-weight pregnant women.16 Increase in VLDL-C has been related to increased free cholesterol/phospholipid ratio, and alterations in membrane viscosity and membrane fluidity.9,23 Reductions in membrane fluidity can affect the function of integral membrane components, such as the translocation of Ca2+ from the extracellular space to the cytoplasm during the contraction–relaxation cycle of smooth muscle cells.24 As mentioned earlier, cholesterol is also crucial to lipid rafts and caveolae, where it is particularly enriched and plays a role in cell signalling pathways; its level can, for example, modulate receptor function, ligand binding and contractility. High cholesterol levels in obese women may, therefore, affect the effectiveness of uterine contractions.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgement
  9. References

This study reiterates the need for pre-pregnancy advice and counselling to women and could be a convincing argument for weight reduction in this group. Pregnancies among overweight and obese women must be classified as high-risk pregnancies, and appropriate antenatal care should be provided, e.g. antenatal anaesthetic referral. Given the significant morbidity associated with dystocia, its frequency in labours and its impact on caesarean section rates, we suggest that there is an urgent need to extend this study.

Acknowledgement

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgement
  9. References

Medical Research Council, Jie Zhang is supported by an ORS award.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgement
  9. References
  • 1
    The World Health Organization. Obesity: Preventing and Managing the Global Epidemic. Geneva: WHO, 1998.
  • 2
    Crane SS, Wojtowycz MA, Dye TD, Aubry RH, Artal R. Association between pre-pregnancy obesity and the risk of cesarean delivery. Obstet Gynecol 1997;89:21316.
  • 3
    Loverro G, Greco P, Vimercati A, Nicolardi V, Varcaccio-Garofalo G, Selvaggi L. Maternal complications associated with cesarean section. J Perinat Med 2001;29:3226.
  • 4
    Weiss JL, Malone FD, Emig D, Ball RH, Nyberg DA, Comstock CH, et al. Obesity, obstetric complications and cesarean delivery rate—a population-based screening study. Am J Obstet Gynecol 2004;190:10917.
  • 5
    Cedergren MI. Maternal morbid obesity and the risk of adverse pregnancy outcome. Obstet Gynecol 2004;103:21924.
  • 6
    Young TK, Woodmansee B. Factors that are associated with cesarean delivery in a large private practice: the importance of prepregnancy body mass index and weight gain. Am J Obstet Gynecol 2002;187:3128.
  • 7
    Gostynski M, Gutzwiller F, Kuulasmaa K, Doring A, Ferrario M, Grafnetter D, et al. Analysis of the relationship between total cholesterol, age, body mass index among males and females in the WHO MONICA Project. Int J Obes Relat Metab Disord 2004;28:108290.
  • 8
    Babiychuk EB, Smith RD, Burdyga TV, Babiychuk VS, Wray S, Draeger A. Membrane cholesterol selectively regulates smooth muscle phasic contraction. J Membr Biol 2004;198:95101.
  • 9
    Dreja K, Voldstedlund M, Vinten J, Tranum-Jensen J, Hellstrand P, Sward K. Cholesterol depletion disrupts caveolae and differentially impairs agonist-induced arterial contraction. Arterioscler Thromb Vasc Biol 2002;22:126772.
  • 10
    Smith RD, Babiychuk EB, Noble K, Draeger A, Wray S. Increased cholesterol decreases uterine activity: functional effects of cholesterol in pregnant rat myometrium. Am J Physiol 2004 (in press).
  • 11
    Simons K, Toomre D. Lipid rafts and signal transduction. Nat Rev Mol Cell Biol 2000;1:319.
  • 12
    Ishizaka N, Griendling KK, Lassegue B, Alexander RW. Angiotensin II type 1 receptor: relationship with caveolae and caveolin after initial agonist stimulation. Hypertension 1998;32:45966.
  • 13
    Nakayama K, Obara K, Tanabe Y, Saito M, Ishikawa T, Nishizawa S. Interactive role of tyrosine kinase, protein kinase C, and Rho/Rho kinase systems in the mechanotransduction of vascular smooth muscles. Biorheology 2003;40:30714.
  • 14
    Gimpl G, Fahrenholz F. Human oxytocin receptors in cholesterol-rich vs. cholesterol-poor microdomains of the plasma membrane. Eur J Biochem 2000;267:248397.
  • 15
    Kim BK, Ozaki H, Hori M, Takahashi K, Karaki H. Increased contractility of rat uterine smooth muscle at the end of pregnancy. Comp Biochem Physiol A Mol Integr Physiol 1998;121:16573.
  • 16
    Ramsay JE, Ferrell W, Crawford L, Wallace M, Greer IE, Sattar NJ. Maternal obesity is associated with dysregulation of metabolic vascular and inflammatory pathways. J Clin Endocrinol Metabol 2002;87:42317.
  • 17
    Pulkkinen MO, Nyman S, Hamalainen MM, Mattinen J. Proton NMR spectroscopy of the phospholipids in human uterine smooth muscle and placenta. Gynecol Obstet Invest 1998;46:2204.
  • 18
    Pritchard J, Baldwin R, Dickey J, Wiggins K. Blood volume changes in pregnancy and the puerperium. II: Red blood cell loss and changes in apparent blood volume during and following vaginal delivery, cesarean section, and cesarean section plus total hysterectomy. Am J Obstet Gynecol 1962;84:127182.
  • 19
    Ueland K. Maternal cardiovascular dynamics. VII: Intrapartum blood volume changes. Am J Obstet Gynecol 1976;126:6717.
  • 20
    Vahratian A, Zhang J, Troendle JF, Savitz DA, Siega-Riz AM. Maternal prepregnancy overweight and obesity and the pattern of labor progression in term nulliparous women. Obstet Gynecol 2004;104:94351.
  • 21
    Piechota W, Staszewski A. Reference ranges of lipids and apolipoproteins in pregnancy. Eur J Obstet Gynecol Reprod Biol 1992;45:2735.
  • 22
    Pobedinskii NM, Chernukha GE, Burlev AA, Shingerei MV. Characteristics of lipid composition of the blood serum in obese pregnant women. Akush Ginekol (Mosk) 1987 Jun: 226.
  • 23
    Wray S, Jones K, Kupittayanant S, Matthew AJG, Monir-Bishty E, Noble K, et al. Calcium signalling and uterine contractility. J Soc Gynecol Invest 2003;10:25264.
  • 24
    Tulenko TN, Bialecki R, Gleason M, D’Angelo G. Ion channels, membrane lipids and cholesterol: a role for membrane lipid domains in arterial function. Prog Clin Biol Res 1990;334:187203.