EUROP. J. OBSTET. GYNEC. REPROD. BIOL., 1978,8/5,287-293 0 Elsevier/North-Holland Biomedical Press Solute levels in uterine fluids of patients with normal values of amniotic fluid and wlth hydramnios* K.R. Page, D.R. Abramovich, A.S. Garden and L. Jandial Biophysical Chemistry Unit, Chemistry Department, and Depatiment of Obstetricsand Gynaecology, University of Aberdeen, Aberdeen, U.K. PAGE, K.R., ABRAMOVICH, D.R., GARDEN, AS. and JANDIAL, L. (1978): Solute levels in uterine fluids of patients with normal volumes of amniotic fluid and with hydramnios. Europ. J. Obstet. Gynec. reprod. Biol., 8/S, 281-293. Electrolyte, urea, creatinine and protein levels are determined in the uterine fluids of patients with normal amniotic fluid volumes and those with hydramnios. The results indicate that major osmotic imbalances in the uterine flulds are not associated with hydramnios, and are consistent with the view that the control of solute levels overrides the control of amniotic fluid volume. proteins; osmotic flows; electrolyte levels; urea and creatinine levels Introduction The factors which affect the volume of the amniotic fluid are difficult to identify owing to the multiplicity of ways in which water can in theory enter or leave the amniotic cavity. There is now unequiv- ocal evidence that the fetus both swallows and voids in utero (Pritchard, 1965; Abramovich, 1970; Camp- bell, Wladimiroff and Dewhurst, 1973). There is also the possibility that water can enter or leave the amniotic cavity by osmosis across the fetal mem- branes (Abramovich, Page and Jandial, 1976). Sub- stantial increases in the volume of amniotic fluid occur during the early stages of saline- or glucose- induced mid-trimester abortions and these may be explained in terms of osmotic water fluxes induced by *he injected solute (Wagner, 1966; Kerenyi and * Supported by the Scottish Hospital Endowment Trust. Muzsnai, 1975). There is also a variety of evidence based on work with animals which suggests that osrtrotic water transfers can occur in utero between the amniotic cavity and the fetus, and between the fetus and the mother (Schruefer, Seeds, Behrman, Hellegers and Bruns, 1972; Bruns, Linder, Drose and Battaglia, 1963; Battaglia, Prystowsky, Smis- son, Hellegers and Bruns, 1960). If osmosis is a major factor in the control of amniotic fluid volume, then the magnitude of the volume could be expected to correlate with the levels of at least some solutes in the uterine fluids. Cargnello and D’Andrea (1967) have reported raised albumin levels and lowered globulin and mineral ion levels in the amniotic fluid of patients with hydram- nios. The findings of these workers do not agree with those of Abbas and Tovey (1960) in respect to pro- tein levels, or with Seeds (1965) in respect to mineral ion levels. The present work therefore reexamines the 287
Solute levels in uterine fluids of patients with normal values of amniotic fluid and wlth hydramnios
Apr 10, 2023
The factors which affect the volume of the amniotic fluid are difficult to identify owing to the multiplicity of ways in which water can in theory enter or leave the amniotic cavity. There is now unequivocal evidence that the fetus both swallows and voids in utero (Pritchard, 1965; Abramovich, 1970; Campbell, Wladimiroff and Dewhurst, 1973). There is also the possibility that water can enter or leave the amniotic cavity by osmosis across the fetal membranes (Abramovich, Page and Jandial, 1976). Substantial increases in the volume of amniotic fluid occur during the early stages of saline- or glucoseinduced mid-trimester abortions and these may be explained in terms of osmotic water fluxes induced by *he injected solute (Wagner, 1966; Kerenyi
Welcome message from author
If osmosis is a major factor in the control of
amniotic fluid volume, then the magnitude of the
volume could be expected to correlate with the
levels of at least some solutes in the uterine fluids.
Cargnello and D’Andrea (1967) have reported raised
albumin levels and lowered globulin and mineral ion
levels in the amniotic fluid of patients with hydramnios. T
Transcript
PII: 0028-2243(78)90080-1EUROP. J. OBSTET. GYNEC. REPROD. BIOL., 1978,8/5,287-293 0 Elsevier/North-Holland Biomedical Press
Solute levels in uterine fluids of patients with normal values of amniotic fluid and wlth hydramnios*
K.R. Page, D.R. Abramovich, A.S. Garden and L. Jandial
Biophysical Chemistry Unit, Chemistry Department, and Depatiment of Obstetrics and Gynaecology, University of Aberdeen, Aberdeen, U.K.
PAGE, K.R., ABRAMOVICH, D.R., GARDEN, AS. and JANDIAL, L. (1978): Solute levels in uterine fluids of patients with normal volumes of amniotic fluid and with hydramnios. Europ. J. Obstet. Gynec. reprod. Biol., 8/S, 281-293.
Electrolyte, urea, creatinine and protein levels are determined in the uterine fluids of patients with normal amniotic fluid volumes and those with hydramnios. The results indicate that major osmotic imbalances in the uterine flulds are not associated with hydramnios, and are consistent with the view that the control of solute levels overrides the control of amniotic fluid volume.
proteins; osmotic flows; electrolyte levels; urea and creatinine levels
Introduction
The factors which affect the volume of the amniotic fluid are difficult to identify owing to the multiplicity of ways in which water can in theory enter or leave the amniotic cavity. There is now unequiv- ocal evidence that the fetus both swallows and voids in utero (Pritchard, 1965; Abramovich, 1970; Camp- bell, Wladimiroff and Dewhurst, 1973). There is also the possibility that water can enter or leave the amniotic cavity by osmosis across the fetal mem- branes (Abramovich, Page and Jandial, 1976). Sub- stantial increases in the volume of amniotic fluid occur during the early stages of saline- or glucose- induced mid-trimester abortions and these may be explained in terms of osmotic water fluxes induced by *he injected solute (Wagner, 1966; Kerenyi and
* Supported by the Scottish Hospital Endowment Trust.
Muzsnai, 1975). There is also a variety of evidence based on work with animals which suggests that
osrtrotic water transfers can occur in utero between the amniotic cavity and the fetus, and between the fetus and the mother (Schruefer, Seeds, Behrman, Hellegers and Bruns, 1972; Bruns, Linder, Drose and Battaglia, 1963; Battaglia, Prystowsky, Smis- son, Hellegers and Bruns, 1960).
If osmosis is a major factor in the control of amniotic fluid volume, then the magnitude of the volume could be expected to correlate with the levels of at least some solutes in the uterine fluids. Cargnello and D’Andrea (1967) have reported raised albumin levels and lowered globulin and mineral ion levels in the amniotic fluid of patients with hydram- nios. The findings of these workers do not agree with those of Abbas and Tovey (1960) in respect to pro- tein levels, or with Seeds (1965) in respect to mineral ion levels. The present work therefore reexamines the
287
288
levels of proteins and mineral ions in the uterine fluids of patients with and without hydramnios. The survey additionally includes urea and creatlnine levels. All these solutes have potential osmotic activity depending upon such factors as their membrane permeability and transmembrane concen- tration differential. Levels are measured in the amniotic fluid, and in plasma taken from the cord vein, cord artery and the maternal circulation.
Materials and methods
Samples from 100 patients all within the 38th- 40th wk of gestation were analysed. The patients were divided into three groups. Groups 1 and 2 comprised normal patients (amniotic fluid volume in the range 300-1500 ml), group 1 for spontaneous vaginal deliveries (SVD) and group 2 for deliveries by elective cesarean section (CS). Group 3 comprised patients with hydramnios (amniotic fluid volume in excess of 1500 ml).
Where possible, volumes were measured from fluid collected at rupture of the membranes. In all cases of hydramnios, volumes were also measured by injection of a known quantity of paraamino hippuric acid (PAH, 20% solution; Merck, Sharpe and Dohme).
In the case of vaginal deliveries, samples of amniotic fluid were obtained at the time of rupture
K.R. Page et al.: Solute levels in uterine fluids
of the membranes (rupture being either by spontane- ous or artificial means). In the case of deliveries by elective cesarean section, samples of amniotic fluid were taken prior to the incision of the lower segment, Maternal plasma samples were collected at the time of delivery. Fetal plasma samples were obtained from the umbilical cord after delivery of the infant and before the delivery of the placenta. Whole blood was separated from the samples within 1 h of sampling by centrifugation for 10 min at 10,000 rev./mm. Sam- ples were refrigerated at 4-6°C until required for analyses.
Osmolalities were measured correct to 1% using an Advance freezing point osmometer. Sodium and potassium levels were measured correct to 1% by flame emission spectroscopy using a Pye Unicam SP90 analyser. Chloride levels were measured correct to 1% using a Coming EEL 920 chloride meter. Urea, creatinine, total protein and serum albumin levels were determined correct to 5% using a Technicon AA1 autoanalyser. Urea was estimated by the diacetyl monoxime method and creatinine by the picric acid method. Total proteins were determined using a biuret reagent, and albumin levels were measured both by the bromcresol green (BCG) method and by electrophoresis. The crl, a2 and /I-pro- tein fractions were also determined by electrophore- sis.
Electrophoresis was performed on a Coming EEL
TABLE I Osmolalities, mineral ion and nonelectrolyte levels in uterine fluids
Fluid Group Osmolalities Sodium Potassium Chloride Urea Creatinine
analysed (mOsmol~l-l) (mm01 .I-‘) (mm01 .1-l) (mm01 .1-l ) (mm01 * 1-l) &mol * 1-l )
Amniotic 1 264 f 12 (12) 131 * 15 (21) 4.3 * 0.5 (15) 96 * 6 (19) 5.1 f 0.9 (7) 177 i 42 (8) fluid 2 269 i 11 (14) 138 f 8 (20) 4.3 f 0.4 (20) 99 + 5 (5) 5.1 i 1.7 (15) 162 i 46 (15)
3 269 f 10 (11) 130 f 8 (11) 4.2 f 0.2 (10) 100 + 5 (12) 5.4 i 0.9 (10) 136 * 50 (10)
Cord 1 288 i 8 (16) 142 i 12 (28) 4.2 i 0.5 (12) 99 f 5 (29) 3.6 * 0.9 (23) 56 i 15 (25) vein 2 295 i 8 (18) 144 i 7 (25) 4.4i 0.4(13) 101 k4 (25) 3.9 i 0.9 (22) 52 f 15 (22) plasma 3 290* 7 (10) 139* 9(9) 4.1 f 0.2 (5) 97 f 2 (11) 3.6 i 1.5 (11) 57 f 20 (11)
Cord 1 281 f 17 (7) 158 f 17 (8) 3.7 f 0.7 (2) 96 * 7 (16) 3.2 f 0.6 (9) 60 * 9 (9) artery 2 294 * 10 (12) 146 f 6 (15) 4.3 * 0.4 (9) 101 f 5 (17) 3.9 i 0.9 (18) 52 * 17 (18) plasma 3 295 * 8 (2) 142* 2(4) 4.1 f 0.3 (4) 98 f 1 (6) 3.0 f 0.6 (6) 57 f 17 (7)
Maternal 1 287 f 7 (22) 143 * 12 (31) 4.0 * 0.5 (29) 98 * 5 (30) 3.6 * 1.1 (22) 57 f 15 (24) plasma 2 294 f 6 (18) 141 f 7 (26) 4.2 f 0.4 (24) 102 f 4 (26) 3.6 f 0.9 (22) 50 f 12 (22)
3 294 f 12 (10) 139 * 7 (10) 4.1 * 0.3 (10) 100 f 4 (11) 3.9 f 1.3 (11) 55 f 15 (11)
Levels are shown *SD; number of cases in parentheses.
K.R. Page et al.: Solute levels in uterine fluids 289
electrophoresis apparatus. Samples buffered at pH 8.6 in sodium barbital buffer (Coming EEL) were run on agarose plates against a standard serum (Warner Versitol serum). On completion of each run, protein was stained with amido-black 10B and the fractions separated by cutting the plates into seg- ments and eluting the segments with sodium hydroxide. Concentrations were measured colori- metrically at 580 nm using a Unicam SP30 spectrom- eter. Specimens of amniotic fluid were concen- trated SO-fold using a Mincon-B15 (15,000 MW cut- off) sample concentrator (Amicon Ltd.) prior to electrophoresis. The degree of concentration was checked calorimetrically using a Cecil UV spectro- photometer. IgG and IgM levels were determined by radial immunodiffusion using Tri-Partigen plates (Hoechst Pharmaceuticals). Each plate was run with three standard immunoglobulins and a control serum (Moechst Pharmaceuticals). Maternal, fetal and amniotic fluid samples from the same patient were run on the same plate. All serum samples were diluted lo-fold prior to application. The plates were incubated at room temperature (25’C) for 48 h. IgG and IgM levels were obtained correct to 15%.
Results
Tables I, II and III list the mean solute levels found in this investigation. Each mean is shown with the appropriate standard deviation followed in parentheses by the number of observations
TABLE II Protein levels determined by autoanalyser
Fluid analysed
Albumin
(g*l-‘)
Cord 1 58.4 f 7.6 (44) 27.3 f 4.5 (42) vein 2 54.9 + 4.5 (30) 25.4 f 3.4 (30) plasma 3 51.5 * 7.9 (23) 24.6 f 4.2 (23)
Cotd 1 55.5 * 7.6 (33) 26.0 f 5.3 (27) artery 2 52.4 + 5.4 (27) 23.8 f 4.8 (24) plasma 3 51.4 f 8.3 (17) 25.4 + 2.3 (17)
Maternal 1 64.6 f 6.1 (45) 24.8 f 3.8 (44) plasma 2 63.9 f 4.3 (30) 22.8 r 3.1 (31)
3 60.9 f 5.8 (24) 24.1 ? 5.1 (24)
Levels are shown *SD; number of cases in parentheses.
contributing to the mean. Table I shows osmolalities and solute levels apart from proteins. Table II summarizes protein levels determined using the auto- analyser, and Table III protein levels obtained by electrophoresis and radial immunodiffusion. In all tables the data have been divided into the three groups outlined in the Materials and Methods section, groups 1 and 2 comprising data from patients with normal volumes of amniotic fluid delivered by SCD and CS respectively, and group 3 comprising patients with hydramnios.
Discussion
As indicated in Table IV, the results with a few exceptions show that the solute levels in the cord vein match those in the cord arteries. The only major difference occurs in the group 1 total proteins, but this finding may be questioned as there are no differ- ences in the group 1 protein fractions. The remaining three differences are all very minor, occurring at the 5% level of significance.
Table V gives the results of an analysis of variance conducted between groups 1, 2 and 3. With respect to the solute levels listed In Table I, the maternal chlorides differ at the 1% level of significance whilst cord vein osmolalities and chlorides, and cord artery sodiums, chlorides and ureas, all differ at the 5% level of significance. An examination of Table I shows,
however, that there is no consistent pattern of group 3 means differing from those of groups 1 and 2. It may therefore be concluded that there are no major differences in osmolalities or mineral ion, urea and creatinine concentrations associated with the occurrence of hydramnios. It should be noted that the solute concentrations shown in Table I are in agreement with values reported by Gillibrand (1969), Lind and BIllewicz (1971) and Seeds (1965).
The total protein levels shown in Table II agree well with the values given by Seeds (1965) and Kirsch- baum and De Haven (1968). Table V indicates a highly significant difference between groups 1,2 and 3 in the case of cord vein total proteins and a possibly significant difference between these groups with respect to maternal total proteins.
A paired r-test shows that the albumin levels determined by the BCG method in alI cases agree with
T A
B L
: 5 .
K. R. Page et al. : Solute levels in uterine fluids 291
TABLE IV Paired t-test between solute levels in cord vein and cord artery
Solute Group
1 2 3
Osmolality NS (6) NS (12) NS (2) Sodium NS (7) NS (15) * (4) Potassium _ NS (7) NS (4) Chloride NS (14) NS (17) NS (6) Urea NS (9) NS (18) NS (6) Creatinme NS (10) NS (19) NS (7) Total protein ** (31) * (27) NS (17) Albumin (AA) NS (27) * (24) NS (17) Albumin (EL) NS (24) NS (14) NS (18)
9 NS (24) NS (14) NS (18)
a1 NS (24) NS (14) NS (18)
R NS (24) NS (14) NS (18)
LgC NS (16) NS (4) NS (10) IgM NS (16) NS (9) NS (9)
NS = not significant. * Significant at the 5% level (possibly significant). ** Signikant at the 1% level (significant). Figures in parentheses indicates numbers of pairs examined.
TABLE V Analysis of variance between groups 1,2 and 3
Solute Amniotic Cord Cord Maternal fluid vein artery plasma
plasma plasma
Osmolality Sodium Potassium Chloride Urea Creatinine Total protein Albumin (AA) Albumin (EL)
“1
%I
P
w
NS
* NS NS NS * NS NS NS NS * * **
NS * NS NS NS NS *** NS * * NS *
NS NS NS * ** NS NS NS NS NS NS NS NS NS NS NS NS NS
NS = not significant * Signiticant at the 5% level (possibly significant). ** Significant at the 1% level (significant). *** Significant at the 0.1% level (highly significant).
the results obtained by electrophoresis. In the case of amniotic fluid albumins the levels are about 30% lower than those reported by Abbas and Tovey (1960) and Bonsnes (1966). Cord and maternal albumins are also about 30% lower than Abbas and Tovey’s values but fall within the range of values reported by Kirschbaum and De Haven (1968) De Alvarez, Afonso and Sherrard (1961) and Rebaud, Groulade, Groslambert and Colomb (1963). The last two groups of workers conducted longitudinal studies on maternal albumin levels and it is perhaps noteworthy that there is an almost exact correspon- dence between the levels reported by them for the 38th-40th wk of gestation and the present results. Table V indicates that in the case of albumin there were three minor differences between groups 1, 2 and 3. Two differences were indicated by the BCG method but these were not confirmed by the electro- phoresis results. The remaining difference occurred in amniotic fluid albumins, but Table III shows that the group I means agreed with those of group 3. It would therefore appear that there are no differences in albumin levels that can be associated with hydram- nios. This is a particularly important finding allowing for the fact that the serum albumins are the principal contributors to the colloid osmotic pressure in body fluids.
In all cases the o1 and cllz-protein levels are about twice the values reported by Abbas and Tovey (1960), Bonsnes (1966) and De Alvarez et al. (1961). The level of P-proteins in the amniotic fluid agrees with the fmdings of Bonsnes (1966) whilst the remaining 0 levels are slightly higher than those reported by Abbas and Tovey, De Alvarez et al. and Rebaud et al. (1963). A significant difference occurs between groups 1, 2 and 3 in the case of cord artery or levels and possibly significant differences occur in the cord vein o1 and amniotic fluid fl levels. Table III shows that the group 3 o1 levels in the cord vessels are higher than the corresponding group 1 and 2 levels. The group 3 amniotic fluid /3 level is similar to the group 1 level.
The difference in protein levels determined by electrophoresis in this work, and those reported by other workers, in particular Abbas and Tovey (1960) possibly arises from differences in experimental tech- nique. Abbas and Tovey performed their electro- phoretic separations on paper strips and after staining
292 K.R. Page et al.: Solute levels in uterine fluids
with azocarmine B measured the fractions using a reflecting photodensitometer. In the present work electrophoresis was carried out on agarose plates, and following staining with amido-black lOB, the frac- tions were eluted and measured calorimetrically. Great care was taken to make accurate comparisons between levels in the different groups. All electro- phoretic runs were performed in a standard way against a standard serum.
No significant differences occur between groups with respect to the immunoglobulin levels. Amniotic fluid IgG levels compare welI with those of Cantuaria and Jones (1975) as do cord values with those of Cochrane (1972). Maternal IgC levels are similar to those reported by Studd (1971) and Tatra, Gruber and Breitenecker (1974). Maternal and cord IgM levels agree with those found by Cochrane (1972) whilst the failure to detect IgM in the amniotic fluid is in accord with Cantuaria and Jones (1975).
The present findings do not agree with those of Cargnello and D’Andrea (1967). The amniotic fluid of patients with hydramnios has neither raised albumin nor lowered globulin and mineral ion levels. In a more general context the difference in cord vein total proteins can probably be discounted. The differ- ence is not reflected in the cord vein protein frac- tions, and no abnormalities occur in the cord artery total proteins. The only possible abnormality found to be associated with hydramnios in this work is a raised (Y~ level in the cord blood vessels. It is very unlikely that this difference has any significance concerning the possible involvement of osmosis with hydramnios.
From the present results it therefore appears that major osmotic imbalances in the uterine fluids are not associated with hydramnios. This conclusion does not of course rule out the existence of osmotic flows in the uterus. It is in accord with the view of Seeds (1965) that the overriding factor governing the volume of the amniotic fluid is the need of the fetus to control solute levels. The volume of the amniotic fluid according to this hypothesis is secondary to the maintenance of solute levels, a’major disturbance in the fetal plasma solute levels being immediately corrected by a transfer of water between the fetus and the amniotic cavity.
Acknowledgements
The authors wish to thank the staff of the labor ward and operating theatre, Aberdeen Maternity Hospital, for their co-operation, and Miss Fiona McGregor and Mrs. Loma Michie for their technical assistance.
References
Abbas, T.M. and Tovey, J.E. (1960): Proteins of the liquor amnii. Brit. med. J., I, 416-419.
Abramovich, D.R. (1970): Fetal factors influencing the volume and composition of liquor amnii. J. Obstet. Gynaec. Brit. Cwlth, 77, 856-811.
Abramovich, D.R., Page, K.R. and Jandial, L. (1916): Bulk flows through human fetal membranes. Gynec. Invest., 7, 157-164.
Battaglia, F.C., Prystowsky, H., Smisson, C., Hellegers, A.E. and Bruns, P. (1960): The effect of the administration of fluids intravenously to mothers upon the concentration of water and electrolytes in plasma of human fetuses. Pediatrics, 2S, 2-10.
Bonsnes, R.W. (1966): Composition of amniotic fluid. Clin. Obstet. Gynec., 9, 440-448.
Bruns, P., Linder, R.O., Drose, V.E. and Battaglia, F.C. (1963): The placental transfer of water from fetus to mother following the intravenous infusion of hypertonic mannitol to the maternal rabbit. Amer. J. Obstet. Gynec., 86, 160-167.
Campbell, S., Wladimiroff, J.W. and Dewhurst, C.J. (1973): The ante-natal measurement of fetal urine production. J. Obstet. Gynaec. Brit. Cwlth, 80, 680-686.
Cantuaria, A.A. and Jones, A.L. (1915): Immunoglobulin M in human amniotic fluid and its possible association with neural-tube malformations. Brit. J. Obstet. Gynaec., 82, 262-264.
Cargnello, U. and D’Andrea, A. (1967): Componente elettro- litica e proteica nel liquid0 amniotic0 de polidramnios. Attual. Ostet. Ginec., 12, 99-113.
Cochrane, T.E. (1972): Fetal and maternal immunoglobulm concentrations at delivery and post partum. J. Obstet. Gynaec. Brit. Cwlth, 79, 238-243.
De Alvarez, R.R., Afonso, J.F. and Sherrard, D.J. (1961): Serum protein fractionation in normal pregnancy. Amer. J. Obstet. Gynec., 82, 1096-1111.
Gillibrand, P.N. (1969): Changes in the electrolytes, urea, and osmolality of the amniotic fluid with advancing pregnancy. J. Obstet. Gynaec. Brit. Cwlth, 76,898-905.
Kerenyi, T.D. and Muzsnai, D. (1915): Volume and sodium concentration studies in 300 saline-induced abortions. Amer. J. Obstet. Gynec., 121, 590-596.
Kirschbaum, T.H. and De Haven, J.C.…
Solute levels in uterine fluids of patients with normal values of amniotic fluid and wlth hydramnios*
K.R. Page, D.R. Abramovich, A.S. Garden and L. Jandial
Biophysical Chemistry Unit, Chemistry Department, and Depatiment of Obstetrics and Gynaecology, University of Aberdeen, Aberdeen, U.K.
PAGE, K.R., ABRAMOVICH, D.R., GARDEN, AS. and JANDIAL, L. (1978): Solute levels in uterine fluids of patients with normal volumes of amniotic fluid and with hydramnios. Europ. J. Obstet. Gynec. reprod. Biol., 8/S, 281-293.
Electrolyte, urea, creatinine and protein levels are determined in the uterine fluids of patients with normal amniotic fluid volumes and those with hydramnios. The results indicate that major osmotic imbalances in the uterine flulds are not associated with hydramnios, and are consistent with the view that the control of solute levels overrides the control of amniotic fluid volume.
proteins; osmotic flows; electrolyte levels; urea and creatinine levels
Introduction
The factors which affect the volume of the amniotic fluid are difficult to identify owing to the multiplicity of ways in which water can in theory enter or leave the amniotic cavity. There is now unequiv- ocal evidence that the fetus both swallows and voids in utero (Pritchard, 1965; Abramovich, 1970; Camp- bell, Wladimiroff and Dewhurst, 1973). There is also the possibility that water can enter or leave the amniotic cavity by osmosis across the fetal mem- branes (Abramovich, Page and Jandial, 1976). Sub- stantial increases in the volume of amniotic fluid occur during the early stages of saline- or glucose- induced mid-trimester abortions and these may be explained in terms of osmotic water fluxes induced by *he injected solute (Wagner, 1966; Kerenyi and
* Supported by the Scottish Hospital Endowment Trust.
Muzsnai, 1975). There is also a variety of evidence based on work with animals which suggests that
osrtrotic water transfers can occur in utero between the amniotic cavity and the fetus, and between the fetus and the mother (Schruefer, Seeds, Behrman, Hellegers and Bruns, 1972; Bruns, Linder, Drose and Battaglia, 1963; Battaglia, Prystowsky, Smis- son, Hellegers and Bruns, 1960).
If osmosis is a major factor in the control of amniotic fluid volume, then the magnitude of the volume could be expected to correlate with the levels of at least some solutes in the uterine fluids. Cargnello and D’Andrea (1967) have reported raised albumin levels and lowered globulin and mineral ion levels in the amniotic fluid of patients with hydram- nios. The findings of these workers do not agree with those of Abbas and Tovey (1960) in respect to pro- tein levels, or with Seeds (1965) in respect to mineral ion levels. The present work therefore reexamines the
287
288
levels of proteins and mineral ions in the uterine fluids of patients with and without hydramnios. The survey additionally includes urea and creatlnine levels. All these solutes have potential osmotic activity depending upon such factors as their membrane permeability and transmembrane concen- tration differential. Levels are measured in the amniotic fluid, and in plasma taken from the cord vein, cord artery and the maternal circulation.
Materials and methods
Samples from 100 patients all within the 38th- 40th wk of gestation were analysed. The patients were divided into three groups. Groups 1 and 2 comprised normal patients (amniotic fluid volume in the range 300-1500 ml), group 1 for spontaneous vaginal deliveries (SVD) and group 2 for deliveries by elective cesarean section (CS). Group 3 comprised patients with hydramnios (amniotic fluid volume in excess of 1500 ml).
Where possible, volumes were measured from fluid collected at rupture of the membranes. In all cases of hydramnios, volumes were also measured by injection of a known quantity of paraamino hippuric acid (PAH, 20% solution; Merck, Sharpe and Dohme).
In the case of vaginal deliveries, samples of amniotic fluid were obtained at the time of rupture
K.R. Page et al.: Solute levels in uterine fluids
of the membranes (rupture being either by spontane- ous or artificial means). In the case of deliveries by elective cesarean section, samples of amniotic fluid were taken prior to the incision of the lower segment, Maternal plasma samples were collected at the time of delivery. Fetal plasma samples were obtained from the umbilical cord after delivery of the infant and before the delivery of the placenta. Whole blood was separated from the samples within 1 h of sampling by centrifugation for 10 min at 10,000 rev./mm. Sam- ples were refrigerated at 4-6°C until required for analyses.
Osmolalities were measured correct to 1% using an Advance freezing point osmometer. Sodium and potassium levels were measured correct to 1% by flame emission spectroscopy using a Pye Unicam SP90 analyser. Chloride levels were measured correct to 1% using a Coming EEL 920 chloride meter. Urea, creatinine, total protein and serum albumin levels were determined correct to 5% using a Technicon AA1 autoanalyser. Urea was estimated by the diacetyl monoxime method and creatinine by the picric acid method. Total proteins were determined using a biuret reagent, and albumin levels were measured both by the bromcresol green (BCG) method and by electrophoresis. The crl, a2 and /I-pro- tein fractions were also determined by electrophore- sis.
Electrophoresis was performed on a Coming EEL
TABLE I Osmolalities, mineral ion and nonelectrolyte levels in uterine fluids
Fluid Group Osmolalities Sodium Potassium Chloride Urea Creatinine
analysed (mOsmol~l-l) (mm01 .I-‘) (mm01 .1-l) (mm01 .1-l ) (mm01 * 1-l) &mol * 1-l )
Amniotic 1 264 f 12 (12) 131 * 15 (21) 4.3 * 0.5 (15) 96 * 6 (19) 5.1 f 0.9 (7) 177 i 42 (8) fluid 2 269 i 11 (14) 138 f 8 (20) 4.3 f 0.4 (20) 99 + 5 (5) 5.1 i 1.7 (15) 162 i 46 (15)
3 269 f 10 (11) 130 f 8 (11) 4.2 f 0.2 (10) 100 + 5 (12) 5.4 i 0.9 (10) 136 * 50 (10)
Cord 1 288 i 8 (16) 142 i 12 (28) 4.2 i 0.5 (12) 99 f 5 (29) 3.6 * 0.9 (23) 56 i 15 (25) vein 2 295 i 8 (18) 144 i 7 (25) 4.4i 0.4(13) 101 k4 (25) 3.9 i 0.9 (22) 52 f 15 (22) plasma 3 290* 7 (10) 139* 9(9) 4.1 f 0.2 (5) 97 f 2 (11) 3.6 i 1.5 (11) 57 f 20 (11)
Cord 1 281 f 17 (7) 158 f 17 (8) 3.7 f 0.7 (2) 96 * 7 (16) 3.2 f 0.6 (9) 60 * 9 (9) artery 2 294 * 10 (12) 146 f 6 (15) 4.3 * 0.4 (9) 101 f 5 (17) 3.9 i 0.9 (18) 52 * 17 (18) plasma 3 295 * 8 (2) 142* 2(4) 4.1 f 0.3 (4) 98 f 1 (6) 3.0 f 0.6 (6) 57 f 17 (7)
Maternal 1 287 f 7 (22) 143 * 12 (31) 4.0 * 0.5 (29) 98 * 5 (30) 3.6 * 1.1 (22) 57 f 15 (24) plasma 2 294 f 6 (18) 141 f 7 (26) 4.2 f 0.4 (24) 102 f 4 (26) 3.6 f 0.9 (22) 50 f 12 (22)
3 294 f 12 (10) 139 * 7 (10) 4.1 * 0.3 (10) 100 f 4 (11) 3.9 f 1.3 (11) 55 f 15 (11)
Levels are shown *SD; number of cases in parentheses.
K.R. Page et al.: Solute levels in uterine fluids 289
electrophoresis apparatus. Samples buffered at pH 8.6 in sodium barbital buffer (Coming EEL) were run on agarose plates against a standard serum (Warner Versitol serum). On completion of each run, protein was stained with amido-black 10B and the fractions separated by cutting the plates into seg- ments and eluting the segments with sodium hydroxide. Concentrations were measured colori- metrically at 580 nm using a Unicam SP30 spectrom- eter. Specimens of amniotic fluid were concen- trated SO-fold using a Mincon-B15 (15,000 MW cut- off) sample concentrator (Amicon Ltd.) prior to electrophoresis. The degree of concentration was checked calorimetrically using a Cecil UV spectro- photometer. IgG and IgM levels were determined by radial immunodiffusion using Tri-Partigen plates (Hoechst Pharmaceuticals). Each plate was run with three standard immunoglobulins and a control serum (Moechst Pharmaceuticals). Maternal, fetal and amniotic fluid samples from the same patient were run on the same plate. All serum samples were diluted lo-fold prior to application. The plates were incubated at room temperature (25’C) for 48 h. IgG and IgM levels were obtained correct to 15%.
Results
Tables I, II and III list the mean solute levels found in this investigation. Each mean is shown with the appropriate standard deviation followed in parentheses by the number of observations
TABLE II Protein levels determined by autoanalyser
Fluid analysed
Albumin
(g*l-‘)
Cord 1 58.4 f 7.6 (44) 27.3 f 4.5 (42) vein 2 54.9 + 4.5 (30) 25.4 f 3.4 (30) plasma 3 51.5 * 7.9 (23) 24.6 f 4.2 (23)
Cotd 1 55.5 * 7.6 (33) 26.0 f 5.3 (27) artery 2 52.4 + 5.4 (27) 23.8 f 4.8 (24) plasma 3 51.4 f 8.3 (17) 25.4 + 2.3 (17)
Maternal 1 64.6 f 6.1 (45) 24.8 f 3.8 (44) plasma 2 63.9 f 4.3 (30) 22.8 r 3.1 (31)
3 60.9 f 5.8 (24) 24.1 ? 5.1 (24)
Levels are shown *SD; number of cases in parentheses.
contributing to the mean. Table I shows osmolalities and solute levels apart from proteins. Table II summarizes protein levels determined using the auto- analyser, and Table III protein levels obtained by electrophoresis and radial immunodiffusion. In all tables the data have been divided into the three groups outlined in the Materials and Methods section, groups 1 and 2 comprising data from patients with normal volumes of amniotic fluid delivered by SCD and CS respectively, and group 3 comprising patients with hydramnios.
Discussion
As indicated in Table IV, the results with a few exceptions show that the solute levels in the cord vein match those in the cord arteries. The only major difference occurs in the group 1 total proteins, but this finding may be questioned as there are no differ- ences in the group 1 protein fractions. The remaining three differences are all very minor, occurring at the 5% level of significance.
Table V gives the results of an analysis of variance conducted between groups 1, 2 and 3. With respect to the solute levels listed In Table I, the maternal chlorides differ at the 1% level of significance whilst cord vein osmolalities and chlorides, and cord artery sodiums, chlorides and ureas, all differ at the 5% level of significance. An examination of Table I shows,
however, that there is no consistent pattern of group 3 means differing from those of groups 1 and 2. It may therefore be concluded that there are no major differences in osmolalities or mineral ion, urea and creatinine concentrations associated with the occurrence of hydramnios. It should be noted that the solute concentrations shown in Table I are in agreement with values reported by Gillibrand (1969), Lind and BIllewicz (1971) and Seeds (1965).
The total protein levels shown in Table II agree well with the values given by Seeds (1965) and Kirsch- baum and De Haven (1968). Table V indicates a highly significant difference between groups 1,2 and 3 in the case of cord vein total proteins and a possibly significant difference between these groups with respect to maternal total proteins.
A paired r-test shows that the albumin levels determined by the BCG method in alI cases agree with
T A
B L
: 5 .
K. R. Page et al. : Solute levels in uterine fluids 291
TABLE IV Paired t-test between solute levels in cord vein and cord artery
Solute Group
1 2 3
Osmolality NS (6) NS (12) NS (2) Sodium NS (7) NS (15) * (4) Potassium _ NS (7) NS (4) Chloride NS (14) NS (17) NS (6) Urea NS (9) NS (18) NS (6) Creatinme NS (10) NS (19) NS (7) Total protein ** (31) * (27) NS (17) Albumin (AA) NS (27) * (24) NS (17) Albumin (EL) NS (24) NS (14) NS (18)
9 NS (24) NS (14) NS (18)
a1 NS (24) NS (14) NS (18)
R NS (24) NS (14) NS (18)
LgC NS (16) NS (4) NS (10) IgM NS (16) NS (9) NS (9)
NS = not significant. * Significant at the 5% level (possibly significant). ** Signikant at the 1% level (significant). Figures in parentheses indicates numbers of pairs examined.
TABLE V Analysis of variance between groups 1,2 and 3
Solute Amniotic Cord Cord Maternal fluid vein artery plasma
plasma plasma
Osmolality Sodium Potassium Chloride Urea Creatinine Total protein Albumin (AA) Albumin (EL)
“1
%I
P
w
NS
* NS NS NS * NS NS NS NS * * **
NS * NS NS NS NS *** NS * * NS *
NS NS NS * ** NS NS NS NS NS NS NS NS NS NS NS NS NS
NS = not significant * Signiticant at the 5% level (possibly significant). ** Significant at the 1% level (significant). *** Significant at the 0.1% level (highly significant).
the results obtained by electrophoresis. In the case of amniotic fluid albumins the levels are about 30% lower than those reported by Abbas and Tovey (1960) and Bonsnes (1966). Cord and maternal albumins are also about 30% lower than Abbas and Tovey’s values but fall within the range of values reported by Kirschbaum and De Haven (1968) De Alvarez, Afonso and Sherrard (1961) and Rebaud, Groulade, Groslambert and Colomb (1963). The last two groups of workers conducted longitudinal studies on maternal albumin levels and it is perhaps noteworthy that there is an almost exact correspon- dence between the levels reported by them for the 38th-40th wk of gestation and the present results. Table V indicates that in the case of albumin there were three minor differences between groups 1, 2 and 3. Two differences were indicated by the BCG method but these were not confirmed by the electro- phoresis results. The remaining difference occurred in amniotic fluid albumins, but Table III shows that the group I means agreed with those of group 3. It would therefore appear that there are no differences in albumin levels that can be associated with hydram- nios. This is a particularly important finding allowing for the fact that the serum albumins are the principal contributors to the colloid osmotic pressure in body fluids.
In all cases the o1 and cllz-protein levels are about twice the values reported by Abbas and Tovey (1960), Bonsnes (1966) and De Alvarez et al. (1961). The level of P-proteins in the amniotic fluid agrees with the fmdings of Bonsnes (1966) whilst the remaining 0 levels are slightly higher than those reported by Abbas and Tovey, De Alvarez et al. and Rebaud et al. (1963). A significant difference occurs between groups 1, 2 and 3 in the case of cord artery or levels and possibly significant differences occur in the cord vein o1 and amniotic fluid fl levels. Table III shows that the group 3 o1 levels in the cord vessels are higher than the corresponding group 1 and 2 levels. The group 3 amniotic fluid /3 level is similar to the group 1 level.
The difference in protein levels determined by electrophoresis in this work, and those reported by other workers, in particular Abbas and Tovey (1960) possibly arises from differences in experimental tech- nique. Abbas and Tovey performed their electro- phoretic separations on paper strips and after staining
292 K.R. Page et al.: Solute levels in uterine fluids
with azocarmine B measured the fractions using a reflecting photodensitometer. In the present work electrophoresis was carried out on agarose plates, and following staining with amido-black lOB, the frac- tions were eluted and measured calorimetrically. Great care was taken to make accurate comparisons between levels in the different groups. All electro- phoretic runs were performed in a standard way against a standard serum.
No significant differences occur between groups with respect to the immunoglobulin levels. Amniotic fluid IgG levels compare welI with those of Cantuaria and Jones (1975) as do cord values with those of Cochrane (1972). Maternal IgC levels are similar to those reported by Studd (1971) and Tatra, Gruber and Breitenecker (1974). Maternal and cord IgM levels agree with those found by Cochrane (1972) whilst the failure to detect IgM in the amniotic fluid is in accord with Cantuaria and Jones (1975).
The present findings do not agree with those of Cargnello and D’Andrea (1967). The amniotic fluid of patients with hydramnios has neither raised albumin nor lowered globulin and mineral ion levels. In a more general context the difference in cord vein total proteins can probably be discounted. The differ- ence is not reflected in the cord vein protein frac- tions, and no abnormalities occur in the cord artery total proteins. The only possible abnormality found to be associated with hydramnios in this work is a raised (Y~ level in the cord blood vessels. It is very unlikely that this difference has any significance concerning the possible involvement of osmosis with hydramnios.
From the present results it therefore appears that major osmotic imbalances in the uterine fluids are not associated with hydramnios. This conclusion does not of course rule out the existence of osmotic flows in the uterus. It is in accord with the view of Seeds (1965) that the overriding factor governing the volume of the amniotic fluid is the need of the fetus to control solute levels. The volume of the amniotic fluid according to this hypothesis is secondary to the maintenance of solute levels, a’major disturbance in the fetal plasma solute levels being immediately corrected by a transfer of water between the fetus and the amniotic cavity.
Acknowledgements
The authors wish to thank the staff of the labor ward and operating theatre, Aberdeen Maternity Hospital, for their co-operation, and Miss Fiona McGregor and Mrs. Loma Michie for their technical assistance.
References
Abbas, T.M. and Tovey, J.E. (1960): Proteins of the liquor amnii. Brit. med. J., I, 416-419.
Abramovich, D.R. (1970): Fetal factors influencing the volume and composition of liquor amnii. J. Obstet. Gynaec. Brit. Cwlth, 77, 856-811.
Abramovich, D.R., Page, K.R. and Jandial, L. (1916): Bulk flows through human fetal membranes. Gynec. Invest., 7, 157-164.
Battaglia, F.C., Prystowsky, H., Smisson, C., Hellegers, A.E. and Bruns, P. (1960): The effect of the administration of fluids intravenously to mothers upon the concentration of water and electrolytes in plasma of human fetuses. Pediatrics, 2S, 2-10.
Bonsnes, R.W. (1966): Composition of amniotic fluid. Clin. Obstet. Gynec., 9, 440-448.
Bruns, P., Linder, R.O., Drose, V.E. and Battaglia, F.C. (1963): The placental transfer of water from fetus to mother following the intravenous infusion of hypertonic mannitol to the maternal rabbit. Amer. J. Obstet. Gynec., 86, 160-167.
Campbell, S., Wladimiroff, J.W. and Dewhurst, C.J. (1973): The ante-natal measurement of fetal urine production. J. Obstet. Gynaec. Brit. Cwlth, 80, 680-686.
Cantuaria, A.A. and Jones, A.L. (1915): Immunoglobulin M in human amniotic fluid and its possible association with neural-tube malformations. Brit. J. Obstet. Gynaec., 82, 262-264.
Cargnello, U. and D’Andrea, A. (1967): Componente elettro- litica e proteica nel liquid0 amniotic0 de polidramnios. Attual. Ostet. Ginec., 12, 99-113.
Cochrane, T.E. (1972): Fetal and maternal immunoglobulm concentrations at delivery and post partum. J. Obstet. Gynaec. Brit. Cwlth, 79, 238-243.
De Alvarez, R.R., Afonso, J.F. and Sherrard, D.J. (1961): Serum protein fractionation in normal pregnancy. Amer. J. Obstet. Gynec., 82, 1096-1111.
Gillibrand, P.N. (1969): Changes in the electrolytes, urea, and osmolality of the amniotic fluid with advancing pregnancy. J. Obstet. Gynaec. Brit. Cwlth, 76,898-905.
Kerenyi, T.D. and Muzsnai, D. (1915): Volume and sodium concentration studies in 300 saline-induced abortions. Amer. J. Obstet. Gynec., 121, 590-596.
Kirschbaum, T.H. and De Haven, J.C.…
Related Documents
