maternal diabetes and neonatal respiratory distress. i. maturation of fetal surfactant

British Journal of Obsietrics and Gynaecology April 1984, Vol. 91, pp. 316-324

Maternal diabetes and neonatal respiratory distress. I. Maturation of fetal surfactant

D. K. JAMES Senior Registrar in Obstetrics, M. L. CHISWICK Consultant Neonatal Paediatrician, A. HARKES Research Biochemist, M. WILLIAMS Research Technician & V. R. TINDALL Professor of Obstetrics, North Western Regional Perinatal Centre, S t Mary’s Hospital, Manchester

Summary. The phospholipid composition of amniotic fluid samples from 30 normal patients and 44 diabetic patients over the last 10 weeks of pregnancy was studied. Higher levels of phosphatidylcholine (PC) and phosphatidylinositol (PI) were found in diabetic pregnancies where there was excellent glucose control. These differences were statistically significant at 34-36 weeks. Phosphatidylglycerol (PG) appeared significantly earlier in the well controlled diabetic pregnancies. but even in the poorly controlled diabetics the levels of PC, PI and PG were comparable to those in normal pregnancies. There was no evidence of delayed appearance of fetal surfactant phospholipids in either the well or poorly controlled diabetic pregnancies. The absolute lecithin (PC)/sphingomyelin (SM) ratio in diabetic pregnancies was generally greater for any given gestational age than those in normal pregnancies. Whilst in most cases this was due to a higher PC concentration, in a few poorly controlled diabetics it was the result of a lower concentration of SM.

The association between maternal diabetes in pregnancy and neonatal respiratory distress is well established (Gellis & Hsia 1959; Tsang et al. 198 1); the conditions most commonly seen are transient tachypnoea of the newborn (TTN) and respiratory distress syndrome (RDS) or hyaline membrane disease (Lemons et al. 1981). Never- theless the influence of maternal diabetes on the maturation of fetal lung surfactant remains con- troversial, and it is difficult to assess the contribution of elective preterm delivery and delivery by caesarean section in such pregnancies to the subsequent development of neonatal respiratory problems. Similarly the amniotic fluid lecithin/sphingomyelin (L/S) ratio seems to be less reliable in diabetic pregnancies at predicting

Correspondence: D. K. James, Department of Obstetrics and Gynaecology, Southmead Hospital, Westbury-on-Trym, Bristol BSlO 5NB.

316

fetal lung maturity (Dahlenburg et al. 1977; Mueller-Heubach et al. 1978).

We studied the maturation of fetal lung surfactant in normal and diabetic pregnancies by analysing the phospholipids in amniotic fluid over the last 10 weeks of pregnancy. We measured the amounts of phosphatidylcholine (PC), phosphatidylinositol (PI) and phosphatidylglycerol (PG) (Hallman et al. 1976; Obladen et al. 1979) and the fatty acid composition of PG (Shelley et al. 1979).

Materials and methods

Patients

Forty-eight specimens of amniotic fluid were obtained by amniocentesis from 44 diabetic patients. Gestational age, confirmed by ultrasonic scan in the first half of pregnancy, ranged from 30 to 39 weeks. When subdivided according

Diabetes and RDS 3 17

to White’s classification (White 1974), there were six class A diabetics, 12 class B, 14 class C and 12 class D; 41 of the 44 patients were controlled with insulin for two or more trimesters. The diabetic patients were further subdivided according to the quality of diabetic control. Seventeen patients had excellent control with an average peak value of 24-h blood glucose series over the last two trimesters of G8.5 (mean 7.39, SD 1.00) mmol/l. Twenty-seven patients had suboptimal control with an average peak value of 24-h blood glucose series over the last two trimesters of >8.5 (mean 11.10, SD 1.68) mmol/l.

The phospholipid biochemistry of the amniotic fluid from these diabetic patients was compared with that of 33 specimens obtained from 30 non- diabetic patients. These samples were also obtained by amniocentesis or at caesarean section to determine the L/S ratio (Table 1). The non- diabetic patients had normal pregnancies with no conditions known to influence fetal lung maturation (Gluck 1980). Gestational age, confirmed by ultrasonic scan in the first half of pregnancy, ranged between 30 and 40 weeks.

Comparison of amniotic fluid samples from the non-diabetic and diabetic patients was made in three gestational subclasses: <33 weeks, 34-36 weeks and >37 weeks. Whilst most of the overall data taken over the last 10 weeks of pregnancy tended to follow a log-normal distribution, this did not apply within each gestational subclass. Thus results were compared non-parametrically using the Kruskal-Wallis one-way analysis of variance (more than two groups) or the two tailed Mann-Whitney U test (two groups).

Method

A 1- or 2-ml sample of amniotic fluid super- natant was routinely used for lipid extraction with methanol and chloroform according to the standard method of Folch et al. (1957). The lipid extract was halved, the first half was used for quantitative assay of PC, PI, PG and sphingomyelin (SM) and the second half was used to analyse the fatty acid composition of PC.

Both halves of the extract were subjected to two-dimensional thin-layer chromatography (TLC) using 100 mmx 100 mm silica gel plates. The phospholipid spots on one plate (first half of extract) were visualized in iodine vapour before spectrophotometric assay of phosphorus content. The details of TLC and phospholipid assay have been described previously (James et al. 1984a).

Table 1. Source of amniotic fluid samples from non- diabetic patients

Source of sample No.

Amniocentesis Poor obstetric history, elective

Clinically growth-retarded fetus,

Elderly mother, elective preterm

Elective caesarean section for breech presentation or cephalopelvic disproportion 18

Total 30

preterm delivery contemplated 4

not confirmed by ultrasound and birthweight 5

delivery contemplated 3

The second plate (second half of extract) was visualized by spraying with 0.1% dichloro- fluorescein in methanol and observing under ultraviolet light. The PC spot of this plate was scraped off separately and the phospholipid was eluted from the silica with chloroform: methanol: water (65:25:4; v:v:v). After removal of the solvent (by evaporation under nitrogen at 65OC), antioxidant solution (butylated hydroxylated anisole) was added. Methylation and esterifica- tion at room temperature used 0 . 1 N potassium hydroxide in methanol and 1 N aqueous hydro- chloric acid. The methyl esters were extracted using distilled petroleum spirit (b.p. 40-60OC). After evaporation at 50°C under nitrogen, the methyl esters of PC were separated by gas-liquid chromatography (GLC) on a 2-m glass column packed with 50% DEGS-PS run isothermally at 165OC using a C A R L 0 ERBA Series 4200 instrument.

Results

Biochemistry of phospholipids in amniotic fluid during normal pregnancy

The mean concentration of four phospholipids (PC, PI, PG and SM) in the amniotic fluid samples from the non-diabetic patients in shown in Fig. 1. Whilst the mean concentration of SM remained low and constant (at 2-4 mg/l) over the last 10 weeks of pregnancy, the mean concentration of PC rose progressively after 33 weeks from <15 mg/l to nearly 150 mg/l at 39 weeks. The mean concentration of PI also rose after 33 weeks from < 1 .O mg/l to 25 mg/l after 35 weeks. The mean concentration of PI fell

3 18 D. K . James et al.

150r 115

0 a Gestation (weeks) c

Mean concentrations of four phospholipids in amniotic fluid over the last 10 weeks of non-diabetic pregnancies.

Gestation (weeks)

Fig. 2. Mean ratio of phosphatidylcholine (PC), phosphatidylinositol (PI) and phosphatidylglycerol (PG) to sphingomyelin (SM) in amniotic fluid over the last 10 weeks of non-diabetic pregnancies. The arrow (left) indicates a ratio equivalent to an L/S ratio of 2.0.

below 5 mg/l at 40 weeks. The mean concentration of PG did not start to rise until 36 weeks and exceeded the concentration of PI only after 37 weeks.

Fig. 2 shows the mean concentrations of PC, PI and PG with respect to SM in the amniotic fluid samples from the non-diabetic patients. After 34 weeks there was a simultaneous rise of the mean PC/SM and PI/SM ratios whilst the mean PG/SM ratio did not rise dramatically until after 37 weeks. There was a mathematical relation between the absolute ratio of PC/SM obtained by this method and the ‘conventional’

L/S ratio (Gluck et al. 1971) obtained by single- dimension TLC and measured planimetrically [log (L/S)=0.321. log (PC/SM)+l. 1; correlation coefficient 0.881. An absolute PC/SM ratio of 4.2 corresponded to a ‘conventional’ L/S ratio of 2.0. The mean PC/SM ratio exceeded this value after 33 weeks.

The mean percentage of methyl palmitate (C16:O) and methyl myristate (C14:O) rose progressively over the last 10 weeks of pregnancy, whilst the mean percent of methyl stearate (C18:0), methyl oleate (C18:l) and methyl linoleate (C18:2) fell (Fig. 3).

Diabetes and RDS 3 19

Gestation (weeks)

Fig. 3. Mean relative precentages of methyl esters of fatty acids in phosphatidylcholine (PC) in amniotic fluid over the last 8 weeks of non-diabetic pregnancies. A, C14:0, methyl myristate; 0, C16:0, methyl palmitate; 0, C18:0, methyl stearate; m, C18:1, methyl oleate; 0, C 18:2, methyl lioleate.

Concentration ofphosphatidylcholine (Table 2) The median PC levels in all three gestational periods of the well controlled diabetic patients were higher than the median PC concentrations in the non-diabetic and the poorly controlled diabetic patients. The differences for 34-36 weeks were statistically significant (P(O.05).

The median PC concentrations in the poorly controlled diabetics were similar to the levels found in the non-diabetic patients.

controlled diabetics were constant and similar to those in the non-diabetic samples for all three gestational periods. In the poorly controlled diabetics the median SM concentrations for 34-36 weeks and >31 weeks were also similar to those values in normal pregnancy.

However, in those of <33 weeks the median SM concentration was significantly lower than in normal pregnancy (P<0,05) and the well controlled pregnancy (P<0.05).

Concentration of sphingomyelin (Table 3) The median concentrations of SM in the well

Ratio of PCISM (Table 4)

The higher levels of PC in the well controlled

Table 2. Concentration of phosphatidylcholine (PC) in amniotic fluid

Concentration of PC (mgh)

Gestational age Poorly controlled Well controlled (weeks) Non-diabetics diabetics diabetics

< 33 5.53 (2.96-9.97) 7.48 (4.73-1 3.3) 20.73 (5.45-36.01)* (n = 7) (n = 3) ( n = 2)

( n = 8) (n = 21) (n = 12) 34-36 24.74 (8.81-79.41) 25.87 (2.29-137.02) 60.06 (17.18-208.83)**

2 37 55-45 (41.47-287.56) 85.71 (18.65-380.31) 110.04 (54.08-202.13)* ( n = 18) (n = 6) (n = 4)

Results are median (range). *Not significant. **Well controlled vs non-diabetics P = 0.04; well controlled vs poorly controlled P = 0.02.

320 D. K. James et al.

Table 3. Concentration of sphingomyelin (SM) in amniotic fluid

Concentration of SM (mg/l)


< 33 3.26 (2.88-4-12) 0.79 (0.70-1.10)* 3.57 (2-78-4-36) ( n = 7) ( n = 3) ( n = 2)

( n = 8) ( n = 21) ( n = 12) 34-36 2.37 (0.96-5-25) 2.99 (0.75-14.99) 2.43 (1.61-6 70)

> 37 3.20 (1.40-1 7.04) 3.08(1.35-7.18) 4.02 (3.54-5.64) ( n = 18) (n = 6) ( n = 4)

Results are median (range). *Poorly controlled vs non-diabetic P = 0.04; poorly controlled vs well controlled P = 0.04.

Table 4. Ratio of phosphatidylcholine/sphingomyelin (PC/SM) in amniotic fluid

PC/SM ratio


< 33 1.88 (0.81-2.67) 6 .SO (5.98-19 -OO)* 5.11 (1 96-8.26)** ( n = 7) ( n = 3) ( n = 2)

( n = 8) (n = 21) ( n = 12) 34-36 11.38 (5-01-21.46) 9.76 (2.59-98.03) 22.63 (7.19-50-49)***

> 37 18.00 (9.17-68.5 1) 23.10 (6.3 9-5 2.97) 30-92 (13-03-42-05) (n = 18) (n = 6) (n = 4)

Results are median (range). *Poorly controlled vs non-diabetic P = 0.02. **Well controlled vs non-diabetic not significant (P = 0.14). ***Well controlled 1's non-diabetic P = 0.02; well controlled vs poorly controlled P = 0.002.

diabetic patients resulted in higher median PC/SM ratios in all three gestational periods compared with those in the non-diabetic patients. This was statistically significant for the 34-36 weeks period (P<0.05). The exceptionally low SM concentrations in the poorly controlled diabetics of <33 weeks gestation was responsible for the median PC/SM ratio in this group being significantly higher than that in the corresponding preterm non-diabetic patients (P<O,O5). The median PC/SM ratios after 33 weeks in the poorly controlled diabetic pregnancies were similar to those in the normal pregnancies.

It should be noted that in both the poorly and well controlled preterm diabetic patients the median PC/SM ratios, as well as the majority of the individual values were >4.21 (that ratio corresponding to a conventional L/S ratio of 2.0). This observation is especially relevant in those diabetics who were <33 weeks. In those well

controlled diabetics ,<33 weeks the elevation of the PC/SM ratio was due to a real elevation of PC levels; whilst in the poorly controlled diabetics of comparable gestational age, the elevation of the PC/SM ratio was mainly the result of a lowering of SM concentration.

Concentration of phosphatidvlinositol (Table 5) In all three gestational periods the median PI concentrations in the well controlled diabetics were greater than in both the non-diabetic and the poorly controlled diabetic patients. This was statistically significant at 34-36 weeks (P<0.05). The concentrations of PI were similar in the poorly controlled diabetic and normal patients for each gestational period.

Concentralion of phosphatidylglycerol (Table 6) There were very low concentrations of PG at <33

Diabetes and RDS 321

Table 5. Concentration of phosphatidylinositol (PI) in amniotic fluid

Concentration of PI (mg/l)


< 33 0.41 (0.00-2.05) 1.15 (0.90-1.48) 2.71 (0.92-4.50)* (n = 7) (n = 3) ( n = 2)

(n = 8) (n = 21) ( n = 12)

(n = 18) (n = 6) (n = 4)

34-36 1.50(0.64-11.61) 2.95 (0.00-6.26) 5.25 (2.46-22.39)**

> 37 4.50 (0.37-1 3.40) 3~10(1.52-11.83) 8.08 (5.17-10.49)*

Results are median (range). *Not significant. **Well controlled vs non-diabetic P = 0.03; well controlled vs poorly controlled P = 0.01.

Table 6. Concentration of phosphatidylglycerol (PG) in amniotic fluid

Concentrations of PG (mg/l)


< 33 0.33 (0.00-0.60) 0.00 (0.00) 0.22 (0.00-0.43) (n = 7) (n = 3) (n = 2)

(n = 8) (n = 21) (n= 12)

(n = 8) (n = 6) (n = 4)

34-36 0.30 (0.00-1.81) 0.59 (0.00-7.15) 1.14 (0.00-17.01)*

> 37 3.06 (0.00-15.40) 4.86 (0.25-25.17) 2.83 (1.09-13.03)

Results are median (range). *Well controlled vs non-diabetics P = 0.01; well controlled vs poorly controlled P = 0.03.

weeks in all three groups of patients, but at 34-36 weeks, whilst the median concentrations of P G remained low in the normal and poorly controlled diabetic patients, there was a significant rise in median PG concentration in the well controlled diabetics (P<0.05). At term (>37 weeks) there were similar levels of PG in all three groups of patients.

Fatt.v acid content of phosphatidylcholine

There were no significant differences between the fatty acid composition of PC in non-diabetics and either poorly or well controlled diabetics over the last 10 weeks of pregnancy.

Discussion

In normal pregnancy, after 33 weeks the mean concentrations of PC and PI rose progressively.

By 34 weeks, the mean absolute PC/SM concentration was >4.2 (Fig. 2)-a value corresponding to a conventional L/S ratio of 2.0, the level indicating mature fetal lungs in most pregnancies. Phosphatidylglycerol on the other hand did not appear in significant amounts until 36 or 37 weeks. These findings support the view that, in terms of phospholipids, two types of surfactant are produced by the developing fetus during normal pregnancy. The ‘early’ type comprises PC and PI and is able to stabilize relatively immature alveoli, up to 36 weeks gestation. The ‘late’ type comprises PC, PI and PG and is able to stabilize mature alveoli, from 37 weeks onwards (Gluck 1980).

The association between neonatal and maternal diabetes is well established (Gellis & Hsia 1959; Tsang et al. 1981; Lemons et al. 1981), and we expected to find a delay in the appearance of one or more of the three surfactant phospholipids

322

studied. In fact the reverse was demonstrated. In the well controlled diabetics, the median levels of PC and PI were higher than in normal pregnancies and PG was present in significant amounts by 34 weeks. Perhaps more surprisingly, in the poorly controlled diabetics the values for all three phospholipids were no worse than in normal pregnancy. The paradox, furthermore, was not explained in terms of differences in the fatty acid content of PC.

Other workers have claimed that maternal diabetes may be associated with enhanced phospholipid production. A preliminary report of a small number of diabetic patients by Cunningham (1981) suggested a premature elevation of PC, PI and PG, but the numbers were too small to permit further analysis in terms of class of diabetes. Kulovich & Gluck (1979) reported an earlier appearance of PG in classes C, D, F and R diabetic patients, but claimed that classes A and B were associated with a delayed production of the phospholipid. The disadvant- age of assessing diabetic pregnancies in terms of White’s (1974) classification is that it does not bear any relation to the metabolic enviromment to which the developing fetus is exposed. We feel that analysing the data in terms of glucose homeostasis has more meaning and more practical value.

There is evidence from in-vitro and in-vivo experiments that under certain circumstances insulin drives the synthesis of phosphatidylcholine (Smith et al. 1975; Epstein et al. 1976). This action appears to be competitively inhibited by cortisol (Smith et al. 1975). The effect of insulin on surfactant synthesis may, in addition, be one of degree since very high insulin levels produce a reduction of lecithin synthesis in vitro (Neufield et al. 1979). It is speculative that in the well controlled diabetics a moderate degree of fetal hyperinsulinaemia may exist sufficient to stimulate PC synthesis; in the poorly controlled diabetic patients, however, there may be a severe degree of fetal hyperinsulinaemia which would fail to stimulate and even retard PC synthesis.

It is possible that the effect on surfactant maturation demonstrated in this paper may be mediated by mechanisms other than by fetal insulin production, (Sosenko et al. 1979) such as by alteration of glucose concentration directly (Hallman et al. 1982) or even fetal lipid metabolism (Baritussio et al. 1980).

If fetal lung maturation in diabetic pregnancies is at worst comparable to that in normal

D. K. James et al.

pregnancies, and advanced in the well controlled diabetics, what is the explanation for the association between maternal diabetes and RDS? It is possible, for example, that whilst surfactant production in utero may be enhanced, the influence of perinatal hypoxia or delayed clearance of lung fluid after caesarean section (both more common in infants of diabetic mothers) may reduce neonatal surfactant production or release. Similarly, it may be that ‘early surfactant’ (PI and PC) is not able to maintain alveolar stability in infants who are large for their gestational age. In our opinion, however, an important cause for the association of RDS and maternal diabetes is iatrogenic. For over 20 years preterm delivery has been advocated in diabetic pregnancies in order to prevent the rise in stillbirths noted towards the end of the last trimester (Gellis & Hsia 1959). Whilst this practice is perhaps on the decline, it is still likely to be contemplated especially where obstetric complications exist. Diabetic pregnancies with complications are likely to be those in which the control has been poor (Coustan 1980) and unfortunately fetal lung maturity may not be enhanced. Furthermore, the L/S ratio may be falsely reassuring to the clinician.

We demonstrated in both poorly and well controlled diabetic pregnancies that the absolute PC/SM ratio was higher than in normal pregnancies at most gestations. In the well controlled pregnancies this was due to higher levels of PC, but in the very preterm ( 2 3 3 weeks) poorly controlled patients it was the result of a significantly lower concentration of SM. The majority of SM in amniotic fluid comes from the lipoprotein fraction, most of which, in turn, is transported through the placenta (Gebhardt et al. 1979). Gebhardt (1982) has argued that placental pathology in diabetic pregnancies (Jones & Fox 1976) should reduce the lipoprotein and hence sphingomyelin content of the amniotic fluid. The present study is the first report to support this hypothesis.

The conventional L/S ratio is likely to be mis- leading in diabetic pregnancies for reasons other than altered PC and SM concentrations. The assay is performed using single dimension TLC. The ‘S’ spot, the denominator of the L/S ratio, comprises only SM. This is not true for the ‘L‘ or lecithin spot which supplies the numerator. PI and phosphatidylserine (PS) co-chromatogram with PC to form this spot. We have shown that PI levels may be elevated in diabetes; in addition we

Diabetes and R D S 323

have unpublished evidence that in both poorly and well controlled diabetes higher concentrations of PS may be found.

I n summary, therefore, in the individual diabetic pregnancy there is no guarantee that an amniotic fluid L/S ratio of >2.0 represents a fetus that is producing surfactant phospholipids in adequate amounts. The clinician who undertakes elective preterm delivery in a diabetic pregnancy erroneously confident that the L/S ratio is ‘mature’, runs the risk of generating an increased incidence of neonatal RDS (James et al. 19846).

Acknowledgments

We are grateful to the following for financial help: University of Manchester, William Walter Will Trust, North Western Regional Health Authority and Birthright. We thank Mrs Patricia Taylor for secretarial assistance. We appreciate the statistical advice and guidance given by Miss L Hunt and Mr B Faragher.

References

Baritussio, A,, Engi, G., Inelmen, E. M., Schiavon, M., de Biasi, F., Allegra, L., Ursini, F. & Baldo, G. ( 1 980) Altered surfactant synthesis and function in rats with diet induced hyperlipidaemia. Metabolism

Coustan, D. R. (1980) Recent advances in the manage- ment of diabetic pregnant women. Clin Perinatol7, 299-3 1 1.

Cunningham, M. D. (1981) Determination of fetal maturity in diabetic pregnancy. Clin Obstet Gjwecol24,73-89.

Dahlenburg, G. W., Martin, F. I. R., Jeffrey, P. E. & Horacek, I. (1977) Amniotic fluid lecithin/ sphingomyelin ratio in pregnancy complicated by diabetes. Br J Obstet Gvnaecol84,294-299.

Epstein, M. F., Farrell, P. M. & Chez, R. A. (1976) Fetal lung lecithin metabolism in the glucose intolerant rhesus monkey pregnancy. Pediatrics 57, 722-728.

Farrell, P. M. & Hamosh, M. (1978) The biochemistry of fetal lung development. Clin Perinatol 5, 197- 229.

Folch, J., Lees, M. SC Sloane-Stanley, G. H. (1957) A simple method for the isolation and purification of total lipids from animal tissues. J Bid Chem 226,

Gebhardt, D. 0. E. (1982) Explanation for lack of agreement between lecithin/spingomyelin ratio of aminotic fluid in diabetic pregnancies and occurrence of respiratory distress. Am J Obstet Gvnecol 142, 1068-1069.

29,503-5 10.

497-509.

Gebhardt, D. 0. E., Beintema, A,, Reman, F. C. & van Gent, C. M. (1979) The lipoprotein composition of amniotic fluid. Clin Chim Acta 94,93-100.

Gellis, S. S. & Hsia, D. Y. Y. (1959) The infant of the diabetic mother. Am J Dis Child 97, 1-4 1.

Cluck, L. (1980) Fetal lung development. Mead Johnson Svmp Perinat Dev Med 40-49.

Cluck, L., Kulovich, M. V., Borer, R. C., Brenner, P. H., Anderson, G. G. & Spellacy, W. N. (1971) Diagnosis of the respiratory distress syndrome by amniocentesis. Am J Obstel Gynecol 109,440-445.

Hallman, M., Kulovich, M., Kirkpatrick, E., Sugar- man, R. G. & Cluck, L. (1976) Phosphatidyl- inositol and phosphatidylglycerol in aminotic fluid: indices of lung maturity. Am J Obstet Gynecol 125,

Hallman, M., Werner, D., Epstein, B. L. & Cluck, L. (1982) Effects of maternal insulin or glucose infusion on the fetus: study on lung surfactant phospolipids, plasma myoinositol, and fetal growth in the rabbit. Am J Obstet Gynecol 142,877-882.

James, D. K., Harkes, A., Williams, M., Chiswick, M. L., Tindall, V. R., Richardson, T. & Gowen- lock, A. (1984~) Amniotic fluid phosphatidylglycerol and prediction of fetal lung maturity in diabetic pregnancies. J Obstet Gynaecol4, 166- 169.

James, D. K., Chiswick, M. L., Harknes, A., Williams, M. & Tindall, V. R. (19846) Maternal diabetes and neonatal respiratory distress. 11. Prediction of fetal lung maturity. Br J Obstet Gynaecol91,325-329.

Jones, C. J. P. & Fox, H. (1976) Placental changes in gestational diabetes. Obstet Gynecol48, 274-280.

Kulovich, M. V. & Cluck, L. (1979) The lung profile. 11. Complicated pregnancy. Am J Obstet Gynecol 135, 57-63.

Lemons, J. A., Vargas, P. & Delaney, J. J. (1981) Infant of the diabetic mother: review of 225 cases. Obstet G.vnecol57, 187-192.

Mueller-Heubach, E., Caritas, S. N., Edelstone, D. I. & Turner, J. H. (1978) Lecithin/sphingomyelin ratio in amniotic fluid and its value for the prediction of neonatal respiratory distress syndrome in pregnant diabetic women. A m J Obstet Gynecol 130,28-34.

Neufield, N. D., Sevanian, A,, Barrett, C . T. & Kaplan, S. A. (1979) Inhibition of surfactant production by insulin in fetal rabbit lung slices. Pediat Res 13, 753-754.

Obladen, M., Merritt, T. A. & Gluck, L. (1979) Acceleration of pulmonary surfactant maturation in stressed pregnancies. Am J Obstet Gynecol 135,

Shelley, S. A., Kovacevic, M., Piciga, J. E. & Balis, J. U. (1979) Sequential changes of surfactant phosphatidylcholine in hyaline membrane disease of the newborn. New EnglJ Med 300, 112-1 16.

Smith, B. T., Giroud, C. J. P., Robert, M. & Avery, M. E. (1975) Insulin antagonism of cortisol action on lecithin synthesis by cultured fetal lung cells. J Pediatr 87,953-955.

613-61 7.

1079-1085.

324 D. K. James et al.

Sosenko, I. R., Kitzmiller, J. L. Loo, S . W., Blix, P., Rubenstein, A. H. & Gabbay, K. H. (1979) The infant of the diabetic mother. Correlation of Perinatol 1,33 1-347. increased cord C-peptide levels with macrosomia and hypoglycaemia. N Engl JMed 301,859-862.

Tsang, R. C., Ballard, J. & Bram, C. (1981) The infant of the diabetic mother: today and tomorrow. Clin

Obstet Gynecol24, 125-47. White, P. (1974) Diabetes mellitus in pregnancy. Clin

Received 7Apri l1983 Accepted 19 August 1983

maternal diabetes and neonatal respiratory distress. i. maturation of fetal surfactant

Documents