- Email: [email protected]
Can plasma glucose and HbA1c predict fetal growth in mothers with different glucose tolerance levels?
Diabetes Research and Clinical Practice 77 (2007) 465–470 www.elsevier.com/locate/diabres
Can plasma glucose and HbA1c predict fetal growth in mothers with different glucose tolerance levels? A. Lapolla a,*, M.G. Dalfra` a, M. Bonomo b, M.T. Castiglioni c, G. Di Cianni d, M. Masin a, E. Mion b, R. Paleari e, C. Schievano f, M. Songini g, G. Tocco g, L. Volpe d, A. Mosca e a
Dipartimento di Scienze Mediche e Chirurgiche-Cattedra di Malattie del Metabolismo, Universita` di Padova, Via Giustinuani n 2, 35100 Padova, Italy b Centro Antidiabetico, H Maggiore Ca` Granda, Milano, Italy c Divisione di Ostetricia e Ginecologia, H San Raffaele, Milano, Italy d Dipartimento di Endocrinologia e Metabolismo, Universita` di Pisa, Pisa, Italy e Dipartimento di Scienze e Tecnologie Biomediche, Universita` di Milano, Segrate (Milano), Italy f Universita` di Padova, Italy g Centro Antidiabetico, Ospedale Brotzu, Cagliari, Italy Received 8 September 2006; accepted 24 January 2007 Available online 9 March 2007
Abstract To assess whether HbA1c and plasma glucose predicts abnormal fetal growth, 758 pregnant women attending 5 Diabetic Centers were screened for gestational diabetes mellitus (GDM). On glucose challenge (GCT) at 24–27 weeks of gestation (g.w.), negative cases formed the normal control group (N1). Positive cases took an oral glucose tolerance test (OGTT): those found negative were classed as false positives screening test (N2); if they had an OGTT result at least as high as their normal glucose levels, they were classed as having one abnormal glucose value (OAV) at OGTT; two values as GDM. HbA1c was assayed on the day of GCT. We considered fetal macrosomia, large for gestational age (LGA), ponderal index and mean growth percentile. Mean age, pre-pregnancy BMI, fasting plasma glucose (FPG) and HbA1c were progressively higher from N1 to GDM patients. The newborn of N2 mothers were heavier than those with N1 or GDM. The mean growth percentile was significantly higher in N2 than in N1. More LGA babies were born to OAV than to N1 or N2 women. Macrosomia and ponderal index did not differ significantly in the four groups. At logistic regression only plasma glucose at GCT could predict LGA babies and a ponderal index above 2.85. At risk analysis, GDM and OAV significantly predicted LGA babies, and GDM a ponderal index >2.85. In conclusion, FPG at GCT could predict fetal overgrowth and plasma glucose >85 mg/dl doubles the risk of LGA infants. HbA1c at 24–27 g.w. does not predict fetal overgrowth. Mild alterations in glucose tolerance correlate with fetal overgrowth and needs monitoring and treatment. # 2007 Elsevier Ireland Ltd. All rights reserved. Keywords: Diabetes; Pregnancy; Fetal growth; Plasma glucose; HbA1c
1. Introduction
* Corresponding author. Tel.: +39 049 8216857; fax: +39 049 8216838. E-mail address: [email protected] (A. Lapolla).
Fetal growth depends on a series of mechanisms, transplacental fuel transport, uteroplacental flow, hormones of the feto-placental unit and genetic factors being considered the most important [1].
0168-8227/$ – see front matter # 2007 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.diabres.2007.01.022
466
A. Lapolla et al. / Diabetes Research and Clinical Practice 77 (2007) 465–470
In normal pregnant women, characteristic changes in maternal metabolism, e.g. ‘‘facilitated anabolism’’ and ‘‘accelerated starvation’’, provide nutrients for fetal growth in addition to maternal and fetal energy requirements [2]. When pregnancy is complicated by diabetes, a series of maternal–fetal complications can occur, the most common being macrosomia, defined as a birth weight higher than 4000 g [3,4]. Macrosomia is frequently associated with birth trauma, neonatal hypoglycemia, and other neonatal complications [3,5]. The Pedersen hypothesis suggests that maternal hyperglycemia, which is characteristic of diabetes, stimulates the fetal pancreas to produce insulin, which causes excessive growth, or macrosomia [6]. The direct effect of abnormal maternal carbohydrate metabolism has so far been difficult to assess, however. Reports showing that macrosomia occurs despite ‘‘normoglycemia’’ [7,8] support the hypothesis that other fuels are important in determining normal or fetal overgrowth [9,10]. On the other hand, studies reporting that a good metabolic control in diabetic pregnant women is able to normalize birth weight in infants of diabetic mothers confirm the importance of the mother’s glucose levels [11,12]. It is also worth emphasizing that recent studies have reported a higher risk of macrosomia and abnormal fetal growth in women with only mild alterations in glucose tolerance [13,14]. So a simple, inexpensive and sufficiently specific and sensitive test capable of predicting abnormal fetal growth early in pregnancy would be helpful. HbA1c is considered a good indicator of mean plasma glucose level in the 4 weeks preceding the test [15] and it is currently used to assess glucose control in diabetic pregnancies too [16]. Contrasting results have been reported on the ability of HbA1c and plasma glucose levels, either fasting or after oral glucose challenge, to predict macrosomia in pregnancy [17–25]. Hence our interest in evaluating the ability of HbA1c and glucose levels on traditional screening at 24th to 27th weeks of gestation to predict abnormal fetal growth in a large number of pregnant women, both normal and with varying degrees of glucose tolerance. We also chose to assess fetal growth not only by calculating the macrosomia rate, but also considering the proportion of babies large for gestational age (LGA), their ponderal index and their mean growth percentile.
2. Materials and methods 2.1. Subjects Pregnant women attending the Diabetic Centers of Cagliari, Milano (Ospedale Niguarda and S Raffaele), Padova and Pisa for routine prenatal clinical care were screened for gestational diabetes mellitus (GDM). Subjects who met the following inclusion criteria were invited to take part in this study: no smoking, no chronic hypertension or specific conditions known to affect glucose metabolism (conditions found in about 4%, 2% and 2% of our pregnant women, respectively). Seven hundred and fifty-eight pregnant women were screened, 611 of them were evaluated and 147 were not included in the analysis because data were not available on the outcome of the mothers and/or babies. The women took a glucose challenge test (GCT) [26] between the 24th and 27th weeks of gestation: those who had a plasma glucose level lower than 140 mg/day an hour after ingesting 50 g of glucose (negative GCT) [26] were enrolled as normal controls (N1); the women with plasma glucose levels of 140 mg/dl or more (positive GCT) took an oral glucose tolerance test (OGTT) with 100 g of glucose, assaying plasma glucose at time 0, 60, 120 and 180 min. Women found positive at GCT and negative at OGTT (<95, <180, <155, <140 mg/dl at 0, 60, 120, 180 min, respectively [27]) were considered as having false positive screening test results (N2). Women with one OGTT value at least as high as their normal glucose levels were diagnosed as having one abnormal glucose value (OAV) at OGTT, while those with two plasma glucose values at least as high as normal were diagnosed as having GDM. All patients gave their informed consent to participation in the study, which was approved by the Local Ethical Committee. HbA1c was measured on the same day as the GCT in all cases. GDM patients were put on a diet and checked by evaluating their fasting and post-prandial glucose levels. If, after 2 weeks, plasma glucose levels were not maintained within an acceptable range (i.e., FPG > 95 mg/dl and/or 2-h post-prandial plasma glucose > 120 mg/dl), insulin treatment was started. All GDM women were monitored for metabolic and obstetric control up until delivery. In general, the women delivered at term unless additional obstetric complications set in. The other women received only conventional care. The maternal parameters considered were age, pre-pregnancy BMI (kg/m2), timing and mode of delivery. For the fetal parameters, we recorded length and weight at birth. We considered macrosomia as a birth weight of 4000 g or more. Babies were large for gestational age (LGA) if their birth weight was >90th percentile and small for gestational age (SGA) if their birth weight was <10th percentile on the basis of the standard growth and development tables for our population [28]. Then we evaluated the growth percentile, i.e. the percentile of the weight in relation to gestational age
A. Lapolla et al. / Diabetes Research and Clinical Practice 77 (2007) 465–470
467
Table 1 Clinical and metabolic parameters evaluated in 611 pregnant women, of which 334 were normal (N1), 128 had false positive screening test results (N2), 48 had one abnormal glucose value (OAV) at OGTT and 101 had gestational diabetes mellitus (GDM)
Age (years) Pre-pregnancy BMI (kg/m2) FPG (mg/dl) HbA1c (%)
N1
N2
OAV
GDM
ANOVA, p
30.9 4.7 22.4 4.2 77.9 6.7 4.8 0.4
31.7 4.9 22.8 3.9 81.8 6.9 4.9 0.4
32.5 4.4 23.7 4.7 87.3 8.6 5.1 0.5
33.4 4.4 24.7 4.8 87.9 11.6 5.0 0.4
<0.0001 <0.0001 <0.0001 <0.0001
ANOVA p is reported for each parameter. The t-test was used to evaluate effects between group means. Age: N1 vs. OAV p = 0.0067, N1 vs. GDM p < 0.0001, N2 vs. GDM p = 0.009; BMI: N1 vs. OAV p = 0.042, N1 vs. GDM p < 0.0001, N2 vs. GDM p = 0.002; FPG: N1 vs. N2–OAV–GDM p < 0.0001, N2 vs. OAV–GDM p < 0.0001, OAV vs. GDM p = ns; HbA1c: N1 vs. OAV–GDM p < 0.0001, N2 vs. OAV p < 0.0001, N2 vs. GDM p = 0.02, OAV vs. GDM p = 0.03.
at birth, calculated as an interval between category (i.e. <10, 10–20, 20–30, . . ., >90) and the ponderal index (PI) as the ratio of weight to length cubed (g/cm3), considering a PI higher than 2.85 as excessive growth [29]. 2.2. Methods Plasma glucose was tested with a glucose-oxidase method [30]. HbA1c was measured on EDTA-anticoagulated fresh blood samples. Both glucose and HbA1c measurements were performed locally at the laboratories of the centers involved in the study [31]. Two labs (Cagliari and Padova) used the Menarini HA 8140 HPLC analyzer, while the other three used the Menarini HA 8160 system. The quality of glucose and HbA1c measurements was assessed by means of an appropriate external quality assessment scheme, using control materials of proven commutability and target values assigned by reference methods, as recently described [32]. The glucose measurements were reproducible (mean CV 3.2%) and accurate (bias from target values 0.2 0.7 mg/dl, mean S.E.) The HbA1c measurements were also found reproducible, with a CV of 2.0%, and accurate (bias from target values, 0.10 0.06%, mean S.E.).
2.3. Statistical evaluation Data are expressed as mean standard deviation (S.D.) and %. ANOVA and Student’s t-test were used on all the variables to evaluate differences between diagnostic groups. On LGA, macrosomia and PI > 2.85 stepwise logistic regression was applied to identify the most significant predictors among mother’s age, pre-pregnancy BMI, HbA1c, plasma glucose at time 0 min and 60 min and diagnostic group. Multiple risk analysis was performed to evaluate differences between diagnosis groups in predicting fetal overgrowth. Calculations were performed using the SAS v8 system (SAS Institute, Inc.).
3. Results Of the 758 patients screened, 611 were evaluated, including 334 normal controls (N1), 128 with a false positive screening test result (N2), 48 OAV and 101 GDM. The clinical and metabolic characteristics of the women are shown in Table 1. Mean age and prepregnancy BMI increased progressively from the N1 to the GDM groups. Mean fasting plasma glucose (FPG)
Table 2 Pregnancy outcome evaluated in 611 pregnant women, of which 334 were normal (N1), 128 had false positive screening test findings (N2), 48 had one abnormal glucose value (OAV) at OGTT and 101 had gestational diabetes mellitus (GDM)
Delivery (g.w.) Newborn weight (g) Ponderal index (g/cm3) Average percentile of growth LGA (%) Macrosomia (%) PI > 2.85 (%) SGA (%)
N1
N2
OAV
GDM
ANOVA
39.3 1.4 3223 458 2.62 0.25 56.2 28.9 10.5 2.4 16.5 9.0
39.1 1.5 3384 490 2.65 0.32 66.7 24.7 13.1 6.1 29.1 2.0
38.5 1.9 3271 550 2.63 0.39 63.9 29.4 28.6 7.1 20.7 7.1
38.3 2.2 3202 524 2.63 0.29 61.5 30.1 18.8 5.0 28.7 6.0
p < 0.0001 p = 0.0262 ns p = 0.0110 p = 0.0081 ns ns ns
ANOVA p is reported for each parameter. The t-test was used to evaluate effects between group means. g.w.: N1 vs. OAV p = 0.0002, N1 vs. GDM p < 0.0001, N2 vs. OAV p = 0.02, N2 vs. GDM p = 0.007; newborn weight: N1 vs. N2 p = 0.0037, N2 vs. GDM p = 0.012; % growth: N1 vs. N2 p = 0.0015; LGA: N1 vs. OAV p = 0.001, N1 vs. GDM p = 0.03, N2 vs. OAV p = 0.028.
468
A. Lapolla et al. / Diabetes Research and Clinical Practice 77 (2007) 465–470
Table 3 Multivariate logistic regression predicting fetal overgrowth in pregnant women under study Variables entered
PI > 2.85
LGA Full model
After stepwise
Full model
After stepwise
OR (CI)
p
OR (CI)
p
OR (CI)
p
OR (CI)
p
GC T0
1.033 (0.972–1.097)
0.3001
1.061 (1.029–1.093)
<0.0001
1.069 (1.004–1.140)
0.0383
1.032 (0.999–1.065)
0.05
GC T60 HbA1c GDM vs. N1 OAV vs. N1 N2 vs. N1 Mother’s BMI Mother’s age
0.999 2.757 0.998 2.530 0.637 1.026 1.049
0.9134 0.0995 0.9980 0.2214 0.5980 0.6249 0.2990
1.021 0.315 0.390 0.576 1.264 1.036 0.943
0.0802 0.0926 0.2772 0.5361 0.7588 0.4970 0.1921
(0.978–1.020) (0.825–9.215) (0.180–5.539) (0.572–11.198) (0.119–3.409) (0.927–1.135) (0.959–1.148)
and HbA1c values were significantly higher in GDM and OAV women than in N1 and N2 ( p < 0.0001); N2 women also had significantly higher FPG levels than N1 ( p < 0.0001). Women with GDM and OAV delivered earlier than women with N1 and N2 (38.3 2.2 g.w. versus 38.5 1.9 g.w. versus 39.3 1.4 g.w. versus 39.1 1.5 g.w., respectively; p < 0.0001). As for mode of delivery, the frequency of cesarean section was higher in GDM and OAV women than in women with N2 and N1 (43% versus 57% versus 35% versus 30%; p = 0.002). The pregnancy outcomes and other related data are shown in Table 2. As for fetal growth, the newborn of N2 mothers were heavier than those born of N1 or GDM mothers ( p = 0.0037 and p = 0.012) and did not differ significantly from those born to OAV mothers. The mean growth percentile was significantly higher in N2 than in N1 ( p = 0.0015), but no different from OAV or GDM. LGA babies were born more frequently to OAV women than to N1, N2 ( p = 0.001 and p = 0.028), while there was no significant difference vis-a`-vis those born to GDM mothers. Macrosomia, PI and the rate of babies with PI > 2.85 did not differ significantly in the four groups. The frequency of babies SGA was no different in the groups examined. In the full model of regression analysis mother’s age, pre-pregnancy BMI, HbA1c, plasma glucose at time 0 min and 60 min and diagnostic group were evaluated to predict LGA babies, with a PI > 2.85 and macrosomia. After stepwise regression only plasma glucose at time 0 min of the GCT test can predict LGA and a PI > 2.85 (LGA OR = 1.061, p < 0.0001; PI > 2.85 OR = 1.032, p = 0.05) (Table 3). A glycemia value higher than 85 mg/dl at GCT test time 0 min was also a strong predictor of LGA babies (OR 2.08 [CI 95% 1.52–2.63]; p < 0.01).
(0.998–1.044) (0.082–1.210) (0.071–2.131) (0.100–3.313) (0.284–5.634) (0.936–1.146) (0.864–1.030)
Risk analysis was performed to evaluate the differences in fetal outcome related to diagnosis and showed that GDM and OAV significantly predict LGA babies (GDM OR 1.99 [CI 95% 1.35–2.63] p = 0.036; OAV OR 3.39 [CI 95% 2.61–4.17]; p < 0.002) and GDM also predicts a ponderal index higher than 2.85 (GDM OR 2.04 [CI 95% 1.36–2.71]; p < 0.041); N1 was used as a base group for risk evaluation. 4. Discussion First of all, our study showed that older age and higher pre-pregnancy BMI coincide with greater degrees of glucose tolerance, and that women with OAV and GDM deliver earlier than normal controls or women who only have false positive screening test results. As for fetal outcome, some interesting aspects emerged from the analysis of the four classes of women we studied. By comparison with normal controls, the newborn of women with false positive screening test results have a higher birth weight and mean growth percentile. Babies born from women with OAVor GDM have a higher mean growth percentile and higher chances of being LGA and having macrosomia (though the difference was not statistically significant). It is worth emphasizing that pinpointing babies that are LGA (i.e. evaluating birth weight in relation to gestational age) describes fetal overgrowth better than macrosomia, which is simply a matter of birth weight. In our study, using a strict external quality assessment scheme to reduce the variability of plasma glucose and HbA1c measurement and applying the same procedure for screening and the diagnosis of GDM at the different centers, we were able to evaluate a large number of subjects using standardized tests. The results of this study show that HbA1c values measured
A. Lapolla et al. / Diabetes Research and Clinical Practice 77 (2007) 465–470
at routine 24th to 27th gestational week screening did not correlate with fetal macrosomia or the other fetal growth measurements that we considered. This finding is in contrast with the report from Morris et al. [18] of glycated hemoglobin values in early gestation being related to macrosomia in 48 women with a normal glucose tolerance and 21 GDM patients. Our results are also not consistent with those of Dielmis [20], who found that HbA1c levels measured at 20th to 28th gestational weeks in 46 normal pregnant women, 51 with impaired glucose tolerance (IGT), 43 GDM and 150 type 1 diabetic pregnant women correlated positively with macrosomia. These discrepancies may be due to the different sample sizes, the different methods used to measure HbA1c and, in the case of the Dielmis paper, the inclusion in the study of patients with type 1 diabetes. The other two papers reporting a positive correlation between HbA1c and macrosomia cannot be compared with our study because of their small sample size and because it is not clear when the patients were evaluated [17,19]. So our study confirms Schrader’s demonstration [21] that HbA1c evaluated at routine 24th to 27th gestational week screening is not sufficiently sensitive in predicting macrosomia. We also show, for the first time, that HbA1c is also unable to predict the other fetal growth parameters that we considered. It is conceivable that, at this point in the pregnancy, HbA1c values reflect the daily glycemic fluctuations determined by the accelerated starvation and insulin resistance of the 4–6 weeks preceding the test and are unable to predict an abnormal fetal growth. Recently, in a very elegant paper on type 1 diabetic patients, Kerssen et al. [33] showed that the risk of delivering a macrosomic baby is not predicted by HbA1c levels during pregnancy, which explained only 3–7% of the variation in birth weight of these mothers’ offspring. The results of logistic regression analysis of plasma glucose levels on GCT showed that, in all groups, fasting plasma glucose predicts LGA newborn and a ponderal index higher than 2.85. Glycemia values higher than 85 mg/dl also coincide with a two-fold risk of having a baby LGA in all the women (OR = 2.08; p < 0.01). These results are partially consistent with Schrader’s [21], showing that a fasting plasma glucose level higher than 90 mg/dl at OGTT between the 24th and 28th weeks of gestation is a good predictor of macrosomia, and Skyler [23] reported similar findings. To our knowledge, our results are the first to suggest that fasting plasma glucose at GCT is able to predict the
469
other fetal growth parameters we considered and may describe fetal overgrowth better than macrosomia. We also found that, irrespective of plasma glucose and HbA1c levels, the diagnosis of GDM is a strong predictor of LGA and a PI higher than 2.85, and the diagnosis of OAV are strong predictors of LGA. These results prompt a few considerations: once a diagnosis of GDM has been made, it is generally agreed that this condition should be monitored and treated to contain negative maternal and fetal outcomes [12,34], but how women with mild alterations in glucose tolerance i.e. false positive screening test results and/or only one abnormal glucose value at OGTT remains a matter of debate [13,14,35,36]. The results of our study emphasize the importance of following up pregnant women with mild alterations in glucose tolerance too [13,14,37]. In conclusion, our study – conducted on a large number of women, utilizing standardized measurements and considering different fetal growth parameters – shows that fasting plasma glucose at GCT between the 24th and 27th gestational weeks can predict fetal overgrowth. HbA1c levels assessed at routine 24th to 27th gestational week screening are unsuitable as a marker for predicting fetal overgrowth. Even a mild alteration in glucose tolerance correlates with fetal overgrowth and consequently deserves monitoring. Therapeutic trials comparing treatment versus no treatment for such a condition are needed to establish whether treatment coincides with a better outcome. Acknowledgements We wish to thank the following investigators for taking part in the present study: Ferruccio Ceriotti (Diagnostica e Ricerca S. Raffaele spa, Milano), Simona Granata (Laboratorio di Biochimica, H. Maggiore Ca` Granda, Milano), Giovanni Pellegrini and Lucia Malloggi (Laboratorio Analisi, Azienda Ospedaliera-Universitaria Pisa), Mario Plebani (Servizio Medicina di Laboratorio, Azienda OspedalieraUniversita` di Padova). References [1] M. Ounsted, C. Ounsted, On fetal growth rate (its variation and their consequences), in: Clinics in Developmental Medicine n 46 Suffolk, Lavenham Press, United Kingdom, 1973, pp. 1–26. [2] N. Freinkel, R. Phelps, B.E. Metzger, The mother in pregnancies complicated by diabetes, in: H. Rifkin, D.J. Porte (Eds.), Diabetes Mellitus Theory and Practice, 4th ed., Elsevier, New York, 1990, pp. 634–650. [3] S.G. Gabbe, Gestational diabetes mellitus, N. Engl. J. Med. 315 (1986) 1025–1030.
470
A. Lapolla et al. / Diabetes Research and Clinical Practice 77 (2007) 465–470
[4] J.L. Kitzmiller, Macrosomia in the infant of diabetic mothers: characteristics, causes, prevention, in: L. Jovanovic, C.M. Peterson, K. Fuhrmann (Eds.), Diabetes and Pregnancy: Teratology, Toxicology, and Treatment, Praeger Publishers, New York, 1986, pp. 85–120. [5] G.S. Berkowitz, S.H. Roman, R.H. Lapinski, M. Alvarez, Maternal characteristics, neonatal outcome, and the time of diagnosis of gestational diabetes, Am. J. Obstet. Gynecol. 167 (1992) 976–982. [6] J. Pedersen, Weight and length at birth of infants of diabetic mothers, Acta Endocrinol. 16 (1954) 330–342. [7] P. Dandona, H.S. Besterman, D.P. Freedman, F. Boag, A.M. Taylor, A.G. Beckett, Macrosomia despite well-controlled diabetic pregnancy, Lancet 1 (1984) 737. [8] G.H.A. Visser, S. vanBallegooie, W.J. Sluiter, Macrosomy despite well-controlled diabetic pregnancy, Lancet (i) (1984) 284–285. [9] R.K. Kalkhoff, E. Kandaraki, P.G. Morrow, T.H. Mitchell, S. Kelber, H.I. Borkowf, Relationship between neonatal birth weight and maternal plasma amino acid in lean and obese nondiabetic women and in type 1 diabetic pregnant women, Metabolism 37 (1988) 234–239. [10] G. Di Cianni, R. Miccoli, L. Volpe, C. Lencioni, S. DelPrato, Intermediate metabolism in normal pregnancy and in gestational diabetes, Diabetes Metab. Res. Rev. 19 (2003) 259–270. [11] L. Jovanovic-Peterson, C.M. Peterson, G.F. Reed, B.E. Metzger, J.L. Mills, R.H. Knopp, et al., Maternal postprandial glucose levels and infant birth weight: the diabetes in early pregnancy study, J. Obstet. Gynecol. 164 (1991) 103–111. [12] O. Langher, Y. Yogev, O. Most, E.M. Xenakis, Gestational diabetes: the consequence of not treating, Am. J. Obstet. Gynecol. 192 (2005) 989–997. [13] A. Aberg, H. Rydhsttroem, A. Frid, Impaired glucose tolerance associated with adverse pregnancy outcome: a population based study in southern Sweden, Am. J. Obstet. Gynecol. 184 (2001) 77–83. [14] X. Yang, B. Hsu-Hage, H. Zhang, C. Zhang, Y. Zhang, C. Zhang, Women with impaired glucose tolerance during pregnancy have significantly poor pregnancy outcome, Diabetes Care 25 (2002) 1619–1624. [15] W.G. John, Glycated haemoglobin analysis, Ann. Clin. Biochem. 34 (1997) 17–31. [16] G.C. Gabbe, Management of diabetes mellitus complicating pregnancy, Obstet. Gynecol. 102 (2003) 857–868. [17] L. Baxi, D. Barad, A. Reece, E.A. Farber, Use of glycosylated hemoglobin as a screen for macrosomia in gestational diabetes, Obstet. Gynecol. 64 (1984) 347–350. [18] M.A. Morris, A.S. Grandis, J.C. Litton, Glycosylated hemoglobin concentration in early gestation associated with neonatal outcome, Am. J. Obstet. Gynecol. 153 (1985) 651–654. [19] L.J. Wyse, M. Jones, F. Mandel, Relationship of glycosylated hemoglobin, fetal macrosomia and birth weight macrosomia, Am. J. Perinatol. 11 (1994) 260–262. [20] J. Dielmis, D. Blajic, D. Bukovic, D. Pfeifer, M.I. Ivanisevic, S. Kendic, et al., Glycosylated hemoglobin and fetal growth in normal, gestational and insulin dependent diabetes mellitus pregnancy, Coll. Antropol. 21 (1997) 621–629.
[21] H.M. Schrader, L. Jovanovic-Peterson, W.C. Bevier, C.M. Peterson, Fasting plasma glucose and glycosylated plasma protein at 24–28 weeks of gestation predict macrosomia in the general population, Am. J. Perinatol. 12 (1995) 247–251. [22] F. Andrelli, I. Plotton, P. Arnould, C. Thivolet, Are conventional targets for metabolic control sufficient to prevent fetal macrosomia during diabetic pregnancy? Diabetes Metab. 25 (1999) 341–343. [23] J.S. Skyler, M.J.O. O’Sullivan, E.G. Robertson, D.L. Skyler, K.K. Holsinger, I.A. Lasky, et al., Blood glucose control during pregnancy, Diabetes Care 3 (1980) 69–76. [24] M.D. Berkus, O.D. Langer, Glucose tolerance test: degree of glucose abnormality correlates with neonatal outcome, Obstet. Gynecol. 81 (1993) 344–348. [25] R.R. Little, E.M. Mc Kenzie, J.M. Shyken, S.E. Winkelmann, L.M. Ramsey, R.W. Madsn, et al., Lack of relationship between glucose tolerance and complications of pregnancy in nondiabetic women, Diabetes Care 13 (1990) 483–487. [26] L. Jovanovic, C.M. Peterson, Screening for gestational diabetes. Optimum timing and criteria for retesting, Diabetes 34 (Suppl. 2) (1985) 21–23. [27] M.W. Carpenter, D.R. Coustan, Criteria for screening tests for gestational diabetes, J. Obstet. Gynecol. 144 (1982) 768–773. [28] F. Parazzini, I. Cortinovis, R. Bortulus, L. Fedele, A. Recarli, Weight at birth by gestational age in Italy, Hum. Reprod. 10 (1995) 1852–1863. [29] O. Langer, Y. Yogev, O. Most, E.M.J. Xenakis, Gestational diabetes: the consequences of not treating, Am. J. Obstet. Gynecol. 192 (2005) 989–997. [30] A.S.T. Huggett, D.A. Nixon, Use of glucose oxidase peroxidase and O-dianisine in the determination of blood and urine glucose, Lancet II (1957) 368–370. [31] A. Mosca, R. Paleari, A. Trapolino, F. Capani, G. Pagano, M. Plebani, A re-evaluation of glycohaemoglobin standardisation: the Italian experience with 119 laboratories and 12 methods, Eur. J. Clin. Chem. Clin. Biochem. 35 (1997) 243–248. [32] A. Mosca, R. Paleari, M.G. Dalfra`, G. DiCianni, I. Cuccuru, I. Pellegrini, et al., Reference intervals for hemoglobin A1c in pregnant women: data from an Italian multicenter study, Clin. Chem. 52 (2006) 1138–1143. [33] A. Kerssen, H.W. deWalk, G.H.A. Visser, Sibling birth weight as a predictor of macrosomia in women with type 1 diabetes, Diabetologia 48 (2005) 1743–1748. [34] B.E. Metzger, D.R. Coustan, Summary and recommendations of the fourth international workshop-conference on gestational diabetes, Diabetes Care 21 (Suppl. 2) (1998) B161–B167. [35] A. Verma, B.F. Mitchell, N. Demianczuk, G. Flowerdew, N.B. Okun, Relationship between plasma glucose levels in glucoseintolerant women and newborn macrosomia, J. Mat. Fetal. Med. 6 (1997) 187–193. [36] S. Ramtoola, F. Home, H. Damry, A. Husnoo, S. Ah-Kion, Gestational impaired glucose tolerance does not increase perinatal mortality in a developing country: cohort study, BMJ 322 (2001) 1025–1026. [37] M. Bonomo, D. Corica, E. Mion, G. Goncalves, G. Motta, R. Merati, et al., Evaluating the therapeutic approach in pregnancies complicated by borderline glucose intolerance: a randomized clinical trial, Diabetic Med. 22 (2005) 1536–1541.
Can plasma glucose and HbA1c predict fetal growth in mothers with different glucose tolerance levels? A. Lapolla a,*, M.G. Dalfra` a, M. Bonomo b, M.T. Castiglioni c, G. Di Cianni d, M. Masin a, E. Mion b, R. Paleari e, C. Schievano f, M. Songini g, G. Tocco g, L. Volpe d, A. Mosca e a
Dipartimento di Scienze Mediche e Chirurgiche-Cattedra di Malattie del Metabolismo, Universita` di Padova, Via Giustinuani n 2, 35100 Padova, Italy b Centro Antidiabetico, H Maggiore Ca` Granda, Milano, Italy c Divisione di Ostetricia e Ginecologia, H San Raffaele, Milano, Italy d Dipartimento di Endocrinologia e Metabolismo, Universita` di Pisa, Pisa, Italy e Dipartimento di Scienze e Tecnologie Biomediche, Universita` di Milano, Segrate (Milano), Italy f Universita` di Padova, Italy g Centro Antidiabetico, Ospedale Brotzu, Cagliari, Italy Received 8 September 2006; accepted 24 January 2007 Available online 9 March 2007
Abstract To assess whether HbA1c and plasma glucose predicts abnormal fetal growth, 758 pregnant women attending 5 Diabetic Centers were screened for gestational diabetes mellitus (GDM). On glucose challenge (GCT) at 24–27 weeks of gestation (g.w.), negative cases formed the normal control group (N1). Positive cases took an oral glucose tolerance test (OGTT): those found negative were classed as false positives screening test (N2); if they had an OGTT result at least as high as their normal glucose levels, they were classed as having one abnormal glucose value (OAV) at OGTT; two values as GDM. HbA1c was assayed on the day of GCT. We considered fetal macrosomia, large for gestational age (LGA), ponderal index and mean growth percentile. Mean age, pre-pregnancy BMI, fasting plasma glucose (FPG) and HbA1c were progressively higher from N1 to GDM patients. The newborn of N2 mothers were heavier than those with N1 or GDM. The mean growth percentile was significantly higher in N2 than in N1. More LGA babies were born to OAV than to N1 or N2 women. Macrosomia and ponderal index did not differ significantly in the four groups. At logistic regression only plasma glucose at GCT could predict LGA babies and a ponderal index above 2.85. At risk analysis, GDM and OAV significantly predicted LGA babies, and GDM a ponderal index >2.85. In conclusion, FPG at GCT could predict fetal overgrowth and plasma glucose >85 mg/dl doubles the risk of LGA infants. HbA1c at 24–27 g.w. does not predict fetal overgrowth. Mild alterations in glucose tolerance correlate with fetal overgrowth and needs monitoring and treatment. # 2007 Elsevier Ireland Ltd. All rights reserved. Keywords: Diabetes; Pregnancy; Fetal growth; Plasma glucose; HbA1c
1. Introduction
* Corresponding author. Tel.: +39 049 8216857; fax: +39 049 8216838. E-mail address: [email protected] (A. Lapolla).
Fetal growth depends on a series of mechanisms, transplacental fuel transport, uteroplacental flow, hormones of the feto-placental unit and genetic factors being considered the most important [1].
0168-8227/$ – see front matter # 2007 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.diabres.2007.01.022
466
A. Lapolla et al. / Diabetes Research and Clinical Practice 77 (2007) 465–470
In normal pregnant women, characteristic changes in maternal metabolism, e.g. ‘‘facilitated anabolism’’ and ‘‘accelerated starvation’’, provide nutrients for fetal growth in addition to maternal and fetal energy requirements [2]. When pregnancy is complicated by diabetes, a series of maternal–fetal complications can occur, the most common being macrosomia, defined as a birth weight higher than 4000 g [3,4]. Macrosomia is frequently associated with birth trauma, neonatal hypoglycemia, and other neonatal complications [3,5]. The Pedersen hypothesis suggests that maternal hyperglycemia, which is characteristic of diabetes, stimulates the fetal pancreas to produce insulin, which causes excessive growth, or macrosomia [6]. The direct effect of abnormal maternal carbohydrate metabolism has so far been difficult to assess, however. Reports showing that macrosomia occurs despite ‘‘normoglycemia’’ [7,8] support the hypothesis that other fuels are important in determining normal or fetal overgrowth [9,10]. On the other hand, studies reporting that a good metabolic control in diabetic pregnant women is able to normalize birth weight in infants of diabetic mothers confirm the importance of the mother’s glucose levels [11,12]. It is also worth emphasizing that recent studies have reported a higher risk of macrosomia and abnormal fetal growth in women with only mild alterations in glucose tolerance [13,14]. So a simple, inexpensive and sufficiently specific and sensitive test capable of predicting abnormal fetal growth early in pregnancy would be helpful. HbA1c is considered a good indicator of mean plasma glucose level in the 4 weeks preceding the test [15] and it is currently used to assess glucose control in diabetic pregnancies too [16]. Contrasting results have been reported on the ability of HbA1c and plasma glucose levels, either fasting or after oral glucose challenge, to predict macrosomia in pregnancy [17–25]. Hence our interest in evaluating the ability of HbA1c and glucose levels on traditional screening at 24th to 27th weeks of gestation to predict abnormal fetal growth in a large number of pregnant women, both normal and with varying degrees of glucose tolerance. We also chose to assess fetal growth not only by calculating the macrosomia rate, but also considering the proportion of babies large for gestational age (LGA), their ponderal index and their mean growth percentile.
2. Materials and methods 2.1. Subjects Pregnant women attending the Diabetic Centers of Cagliari, Milano (Ospedale Niguarda and S Raffaele), Padova and Pisa for routine prenatal clinical care were screened for gestational diabetes mellitus (GDM). Subjects who met the following inclusion criteria were invited to take part in this study: no smoking, no chronic hypertension or specific conditions known to affect glucose metabolism (conditions found in about 4%, 2% and 2% of our pregnant women, respectively). Seven hundred and fifty-eight pregnant women were screened, 611 of them were evaluated and 147 were not included in the analysis because data were not available on the outcome of the mothers and/or babies. The women took a glucose challenge test (GCT) [26] between the 24th and 27th weeks of gestation: those who had a plasma glucose level lower than 140 mg/day an hour after ingesting 50 g of glucose (negative GCT) [26] were enrolled as normal controls (N1); the women with plasma glucose levels of 140 mg/dl or more (positive GCT) took an oral glucose tolerance test (OGTT) with 100 g of glucose, assaying plasma glucose at time 0, 60, 120 and 180 min. Women found positive at GCT and negative at OGTT (<95, <180, <155, <140 mg/dl at 0, 60, 120, 180 min, respectively [27]) were considered as having false positive screening test results (N2). Women with one OGTT value at least as high as their normal glucose levels were diagnosed as having one abnormal glucose value (OAV) at OGTT, while those with two plasma glucose values at least as high as normal were diagnosed as having GDM. All patients gave their informed consent to participation in the study, which was approved by the Local Ethical Committee. HbA1c was measured on the same day as the GCT in all cases. GDM patients were put on a diet and checked by evaluating their fasting and post-prandial glucose levels. If, after 2 weeks, plasma glucose levels were not maintained within an acceptable range (i.e., FPG > 95 mg/dl and/or 2-h post-prandial plasma glucose > 120 mg/dl), insulin treatment was started. All GDM women were monitored for metabolic and obstetric control up until delivery. In general, the women delivered at term unless additional obstetric complications set in. The other women received only conventional care. The maternal parameters considered were age, pre-pregnancy BMI (kg/m2), timing and mode of delivery. For the fetal parameters, we recorded length and weight at birth. We considered macrosomia as a birth weight of 4000 g or more. Babies were large for gestational age (LGA) if their birth weight was >90th percentile and small for gestational age (SGA) if their birth weight was <10th percentile on the basis of the standard growth and development tables for our population [28]. Then we evaluated the growth percentile, i.e. the percentile of the weight in relation to gestational age
A. Lapolla et al. / Diabetes Research and Clinical Practice 77 (2007) 465–470
467
Table 1 Clinical and metabolic parameters evaluated in 611 pregnant women, of which 334 were normal (N1), 128 had false positive screening test results (N2), 48 had one abnormal glucose value (OAV) at OGTT and 101 had gestational diabetes mellitus (GDM)
Age (years) Pre-pregnancy BMI (kg/m2) FPG (mg/dl) HbA1c (%)
N1
N2
OAV
GDM
ANOVA, p
30.9 4.7 22.4 4.2 77.9 6.7 4.8 0.4
31.7 4.9 22.8 3.9 81.8 6.9 4.9 0.4
32.5 4.4 23.7 4.7 87.3 8.6 5.1 0.5
33.4 4.4 24.7 4.8 87.9 11.6 5.0 0.4
<0.0001 <0.0001 <0.0001 <0.0001
ANOVA p is reported for each parameter. The t-test was used to evaluate effects between group means. Age: N1 vs. OAV p = 0.0067, N1 vs. GDM p < 0.0001, N2 vs. GDM p = 0.009; BMI: N1 vs. OAV p = 0.042, N1 vs. GDM p < 0.0001, N2 vs. GDM p = 0.002; FPG: N1 vs. N2–OAV–GDM p < 0.0001, N2 vs. OAV–GDM p < 0.0001, OAV vs. GDM p = ns; HbA1c: N1 vs. OAV–GDM p < 0.0001, N2 vs. OAV p < 0.0001, N2 vs. GDM p = 0.02, OAV vs. GDM p = 0.03.
at birth, calculated as an interval between category (i.e. <10, 10–20, 20–30, . . ., >90) and the ponderal index (PI) as the ratio of weight to length cubed (g/cm3), considering a PI higher than 2.85 as excessive growth [29]. 2.2. Methods Plasma glucose was tested with a glucose-oxidase method [30]. HbA1c was measured on EDTA-anticoagulated fresh blood samples. Both glucose and HbA1c measurements were performed locally at the laboratories of the centers involved in the study [31]. Two labs (Cagliari and Padova) used the Menarini HA 8140 HPLC analyzer, while the other three used the Menarini HA 8160 system. The quality of glucose and HbA1c measurements was assessed by means of an appropriate external quality assessment scheme, using control materials of proven commutability and target values assigned by reference methods, as recently described [32]. The glucose measurements were reproducible (mean CV 3.2%) and accurate (bias from target values 0.2 0.7 mg/dl, mean S.E.) The HbA1c measurements were also found reproducible, with a CV of 2.0%, and accurate (bias from target values, 0.10 0.06%, mean S.E.).
2.3. Statistical evaluation Data are expressed as mean standard deviation (S.D.) and %. ANOVA and Student’s t-test were used on all the variables to evaluate differences between diagnostic groups. On LGA, macrosomia and PI > 2.85 stepwise logistic regression was applied to identify the most significant predictors among mother’s age, pre-pregnancy BMI, HbA1c, plasma glucose at time 0 min and 60 min and diagnostic group. Multiple risk analysis was performed to evaluate differences between diagnosis groups in predicting fetal overgrowth. Calculations were performed using the SAS v8 system (SAS Institute, Inc.).
3. Results Of the 758 patients screened, 611 were evaluated, including 334 normal controls (N1), 128 with a false positive screening test result (N2), 48 OAV and 101 GDM. The clinical and metabolic characteristics of the women are shown in Table 1. Mean age and prepregnancy BMI increased progressively from the N1 to the GDM groups. Mean fasting plasma glucose (FPG)
Table 2 Pregnancy outcome evaluated in 611 pregnant women, of which 334 were normal (N1), 128 had false positive screening test findings (N2), 48 had one abnormal glucose value (OAV) at OGTT and 101 had gestational diabetes mellitus (GDM)
Delivery (g.w.) Newborn weight (g) Ponderal index (g/cm3) Average percentile of growth LGA (%) Macrosomia (%) PI > 2.85 (%) SGA (%)
N1
N2
OAV
GDM
ANOVA
39.3 1.4 3223 458 2.62 0.25 56.2 28.9 10.5 2.4 16.5 9.0
39.1 1.5 3384 490 2.65 0.32 66.7 24.7 13.1 6.1 29.1 2.0
38.5 1.9 3271 550 2.63 0.39 63.9 29.4 28.6 7.1 20.7 7.1
38.3 2.2 3202 524 2.63 0.29 61.5 30.1 18.8 5.0 28.7 6.0
p < 0.0001 p = 0.0262 ns p = 0.0110 p = 0.0081 ns ns ns
ANOVA p is reported for each parameter. The t-test was used to evaluate effects between group means. g.w.: N1 vs. OAV p = 0.0002, N1 vs. GDM p < 0.0001, N2 vs. OAV p = 0.02, N2 vs. GDM p = 0.007; newborn weight: N1 vs. N2 p = 0.0037, N2 vs. GDM p = 0.012; % growth: N1 vs. N2 p = 0.0015; LGA: N1 vs. OAV p = 0.001, N1 vs. GDM p = 0.03, N2 vs. OAV p = 0.028.
468
A. Lapolla et al. / Diabetes Research and Clinical Practice 77 (2007) 465–470
Table 3 Multivariate logistic regression predicting fetal overgrowth in pregnant women under study Variables entered
PI > 2.85
LGA Full model
After stepwise
Full model
After stepwise
OR (CI)
p
OR (CI)
p
OR (CI)
p
OR (CI)
p
GC T0
1.033 (0.972–1.097)
0.3001
1.061 (1.029–1.093)
<0.0001
1.069 (1.004–1.140)
0.0383
1.032 (0.999–1.065)
0.05
GC T60 HbA1c GDM vs. N1 OAV vs. N1 N2 vs. N1 Mother’s BMI Mother’s age
0.999 2.757 0.998 2.530 0.637 1.026 1.049
0.9134 0.0995 0.9980 0.2214 0.5980 0.6249 0.2990
1.021 0.315 0.390 0.576 1.264 1.036 0.943
0.0802 0.0926 0.2772 0.5361 0.7588 0.4970 0.1921
(0.978–1.020) (0.825–9.215) (0.180–5.539) (0.572–11.198) (0.119–3.409) (0.927–1.135) (0.959–1.148)
and HbA1c values were significantly higher in GDM and OAV women than in N1 and N2 ( p < 0.0001); N2 women also had significantly higher FPG levels than N1 ( p < 0.0001). Women with GDM and OAV delivered earlier than women with N1 and N2 (38.3 2.2 g.w. versus 38.5 1.9 g.w. versus 39.3 1.4 g.w. versus 39.1 1.5 g.w., respectively; p < 0.0001). As for mode of delivery, the frequency of cesarean section was higher in GDM and OAV women than in women with N2 and N1 (43% versus 57% versus 35% versus 30%; p = 0.002). The pregnancy outcomes and other related data are shown in Table 2. As for fetal growth, the newborn of N2 mothers were heavier than those born of N1 or GDM mothers ( p = 0.0037 and p = 0.012) and did not differ significantly from those born to OAV mothers. The mean growth percentile was significantly higher in N2 than in N1 ( p = 0.0015), but no different from OAV or GDM. LGA babies were born more frequently to OAV women than to N1, N2 ( p = 0.001 and p = 0.028), while there was no significant difference vis-a`-vis those born to GDM mothers. Macrosomia, PI and the rate of babies with PI > 2.85 did not differ significantly in the four groups. The frequency of babies SGA was no different in the groups examined. In the full model of regression analysis mother’s age, pre-pregnancy BMI, HbA1c, plasma glucose at time 0 min and 60 min and diagnostic group were evaluated to predict LGA babies, with a PI > 2.85 and macrosomia. After stepwise regression only plasma glucose at time 0 min of the GCT test can predict LGA and a PI > 2.85 (LGA OR = 1.061, p < 0.0001; PI > 2.85 OR = 1.032, p = 0.05) (Table 3). A glycemia value higher than 85 mg/dl at GCT test time 0 min was also a strong predictor of LGA babies (OR 2.08 [CI 95% 1.52–2.63]; p < 0.01).
(0.998–1.044) (0.082–1.210) (0.071–2.131) (0.100–3.313) (0.284–5.634) (0.936–1.146) (0.864–1.030)
Risk analysis was performed to evaluate the differences in fetal outcome related to diagnosis and showed that GDM and OAV significantly predict LGA babies (GDM OR 1.99 [CI 95% 1.35–2.63] p = 0.036; OAV OR 3.39 [CI 95% 2.61–4.17]; p < 0.002) and GDM also predicts a ponderal index higher than 2.85 (GDM OR 2.04 [CI 95% 1.36–2.71]; p < 0.041); N1 was used as a base group for risk evaluation. 4. Discussion First of all, our study showed that older age and higher pre-pregnancy BMI coincide with greater degrees of glucose tolerance, and that women with OAV and GDM deliver earlier than normal controls or women who only have false positive screening test results. As for fetal outcome, some interesting aspects emerged from the analysis of the four classes of women we studied. By comparison with normal controls, the newborn of women with false positive screening test results have a higher birth weight and mean growth percentile. Babies born from women with OAVor GDM have a higher mean growth percentile and higher chances of being LGA and having macrosomia (though the difference was not statistically significant). It is worth emphasizing that pinpointing babies that are LGA (i.e. evaluating birth weight in relation to gestational age) describes fetal overgrowth better than macrosomia, which is simply a matter of birth weight. In our study, using a strict external quality assessment scheme to reduce the variability of plasma glucose and HbA1c measurement and applying the same procedure for screening and the diagnosis of GDM at the different centers, we were able to evaluate a large number of subjects using standardized tests. The results of this study show that HbA1c values measured
A. Lapolla et al. / Diabetes Research and Clinical Practice 77 (2007) 465–470
at routine 24th to 27th gestational week screening did not correlate with fetal macrosomia or the other fetal growth measurements that we considered. This finding is in contrast with the report from Morris et al. [18] of glycated hemoglobin values in early gestation being related to macrosomia in 48 women with a normal glucose tolerance and 21 GDM patients. Our results are also not consistent with those of Dielmis [20], who found that HbA1c levels measured at 20th to 28th gestational weeks in 46 normal pregnant women, 51 with impaired glucose tolerance (IGT), 43 GDM and 150 type 1 diabetic pregnant women correlated positively with macrosomia. These discrepancies may be due to the different sample sizes, the different methods used to measure HbA1c and, in the case of the Dielmis paper, the inclusion in the study of patients with type 1 diabetes. The other two papers reporting a positive correlation between HbA1c and macrosomia cannot be compared with our study because of their small sample size and because it is not clear when the patients were evaluated [17,19]. So our study confirms Schrader’s demonstration [21] that HbA1c evaluated at routine 24th to 27th gestational week screening is not sufficiently sensitive in predicting macrosomia. We also show, for the first time, that HbA1c is also unable to predict the other fetal growth parameters that we considered. It is conceivable that, at this point in the pregnancy, HbA1c values reflect the daily glycemic fluctuations determined by the accelerated starvation and insulin resistance of the 4–6 weeks preceding the test and are unable to predict an abnormal fetal growth. Recently, in a very elegant paper on type 1 diabetic patients, Kerssen et al. [33] showed that the risk of delivering a macrosomic baby is not predicted by HbA1c levels during pregnancy, which explained only 3–7% of the variation in birth weight of these mothers’ offspring. The results of logistic regression analysis of plasma glucose levels on GCT showed that, in all groups, fasting plasma glucose predicts LGA newborn and a ponderal index higher than 2.85. Glycemia values higher than 85 mg/dl also coincide with a two-fold risk of having a baby LGA in all the women (OR = 2.08; p < 0.01). These results are partially consistent with Schrader’s [21], showing that a fasting plasma glucose level higher than 90 mg/dl at OGTT between the 24th and 28th weeks of gestation is a good predictor of macrosomia, and Skyler [23] reported similar findings. To our knowledge, our results are the first to suggest that fasting plasma glucose at GCT is able to predict the
469
other fetal growth parameters we considered and may describe fetal overgrowth better than macrosomia. We also found that, irrespective of plasma glucose and HbA1c levels, the diagnosis of GDM is a strong predictor of LGA and a PI higher than 2.85, and the diagnosis of OAV are strong predictors of LGA. These results prompt a few considerations: once a diagnosis of GDM has been made, it is generally agreed that this condition should be monitored and treated to contain negative maternal and fetal outcomes [12,34], but how women with mild alterations in glucose tolerance i.e. false positive screening test results and/or only one abnormal glucose value at OGTT remains a matter of debate [13,14,35,36]. The results of our study emphasize the importance of following up pregnant women with mild alterations in glucose tolerance too [13,14,37]. In conclusion, our study – conducted on a large number of women, utilizing standardized measurements and considering different fetal growth parameters – shows that fasting plasma glucose at GCT between the 24th and 27th gestational weeks can predict fetal overgrowth. HbA1c levels assessed at routine 24th to 27th gestational week screening are unsuitable as a marker for predicting fetal overgrowth. Even a mild alteration in glucose tolerance correlates with fetal overgrowth and consequently deserves monitoring. Therapeutic trials comparing treatment versus no treatment for such a condition are needed to establish whether treatment coincides with a better outcome. Acknowledgements We wish to thank the following investigators for taking part in the present study: Ferruccio Ceriotti (Diagnostica e Ricerca S. Raffaele spa, Milano), Simona Granata (Laboratorio di Biochimica, H. Maggiore Ca` Granda, Milano), Giovanni Pellegrini and Lucia Malloggi (Laboratorio Analisi, Azienda Ospedaliera-Universitaria Pisa), Mario Plebani (Servizio Medicina di Laboratorio, Azienda OspedalieraUniversita` di Padova). References [1] M. Ounsted, C. Ounsted, On fetal growth rate (its variation and their consequences), in: Clinics in Developmental Medicine n 46 Suffolk, Lavenham Press, United Kingdom, 1973, pp. 1–26. [2] N. Freinkel, R. Phelps, B.E. Metzger, The mother in pregnancies complicated by diabetes, in: H. Rifkin, D.J. Porte (Eds.), Diabetes Mellitus Theory and Practice, 4th ed., Elsevier, New York, 1990, pp. 634–650. [3] S.G. Gabbe, Gestational diabetes mellitus, N. Engl. J. Med. 315 (1986) 1025–1030.
470
A. Lapolla et al. / Diabetes Research and Clinical Practice 77 (2007) 465–470
[4] J.L. Kitzmiller, Macrosomia in the infant of diabetic mothers: characteristics, causes, prevention, in: L. Jovanovic, C.M. Peterson, K. Fuhrmann (Eds.), Diabetes and Pregnancy: Teratology, Toxicology, and Treatment, Praeger Publishers, New York, 1986, pp. 85–120. [5] G.S. Berkowitz, S.H. Roman, R.H. Lapinski, M. Alvarez, Maternal characteristics, neonatal outcome, and the time of diagnosis of gestational diabetes, Am. J. Obstet. Gynecol. 167 (1992) 976–982. [6] J. Pedersen, Weight and length at birth of infants of diabetic mothers, Acta Endocrinol. 16 (1954) 330–342. [7] P. Dandona, H.S. Besterman, D.P. Freedman, F. Boag, A.M. Taylor, A.G. Beckett, Macrosomia despite well-controlled diabetic pregnancy, Lancet 1 (1984) 737. [8] G.H.A. Visser, S. vanBallegooie, W.J. Sluiter, Macrosomy despite well-controlled diabetic pregnancy, Lancet (i) (1984) 284–285. [9] R.K. Kalkhoff, E. Kandaraki, P.G. Morrow, T.H. Mitchell, S. Kelber, H.I. Borkowf, Relationship between neonatal birth weight and maternal plasma amino acid in lean and obese nondiabetic women and in type 1 diabetic pregnant women, Metabolism 37 (1988) 234–239. [10] G. Di Cianni, R. Miccoli, L. Volpe, C. Lencioni, S. DelPrato, Intermediate metabolism in normal pregnancy and in gestational diabetes, Diabetes Metab. Res. Rev. 19 (2003) 259–270. [11] L. Jovanovic-Peterson, C.M. Peterson, G.F. Reed, B.E. Metzger, J.L. Mills, R.H. Knopp, et al., Maternal postprandial glucose levels and infant birth weight: the diabetes in early pregnancy study, J. Obstet. Gynecol. 164 (1991) 103–111. [12] O. Langher, Y. Yogev, O. Most, E.M. Xenakis, Gestational diabetes: the consequence of not treating, Am. J. Obstet. Gynecol. 192 (2005) 989–997. [13] A. Aberg, H. Rydhsttroem, A. Frid, Impaired glucose tolerance associated with adverse pregnancy outcome: a population based study in southern Sweden, Am. J. Obstet. Gynecol. 184 (2001) 77–83. [14] X. Yang, B. Hsu-Hage, H. Zhang, C. Zhang, Y. Zhang, C. Zhang, Women with impaired glucose tolerance during pregnancy have significantly poor pregnancy outcome, Diabetes Care 25 (2002) 1619–1624. [15] W.G. John, Glycated haemoglobin analysis, Ann. Clin. Biochem. 34 (1997) 17–31. [16] G.C. Gabbe, Management of diabetes mellitus complicating pregnancy, Obstet. Gynecol. 102 (2003) 857–868. [17] L. Baxi, D. Barad, A. Reece, E.A. Farber, Use of glycosylated hemoglobin as a screen for macrosomia in gestational diabetes, Obstet. Gynecol. 64 (1984) 347–350. [18] M.A. Morris, A.S. Grandis, J.C. Litton, Glycosylated hemoglobin concentration in early gestation associated with neonatal outcome, Am. J. Obstet. Gynecol. 153 (1985) 651–654. [19] L.J. Wyse, M. Jones, F. Mandel, Relationship of glycosylated hemoglobin, fetal macrosomia and birth weight macrosomia, Am. J. Perinatol. 11 (1994) 260–262. [20] J. Dielmis, D. Blajic, D. Bukovic, D. Pfeifer, M.I. Ivanisevic, S. Kendic, et al., Glycosylated hemoglobin and fetal growth in normal, gestational and insulin dependent diabetes mellitus pregnancy, Coll. Antropol. 21 (1997) 621–629.
[21] H.M. Schrader, L. Jovanovic-Peterson, W.C. Bevier, C.M. Peterson, Fasting plasma glucose and glycosylated plasma protein at 24–28 weeks of gestation predict macrosomia in the general population, Am. J. Perinatol. 12 (1995) 247–251. [22] F. Andrelli, I. Plotton, P. Arnould, C. Thivolet, Are conventional targets for metabolic control sufficient to prevent fetal macrosomia during diabetic pregnancy? Diabetes Metab. 25 (1999) 341–343. [23] J.S. Skyler, M.J.O. O’Sullivan, E.G. Robertson, D.L. Skyler, K.K. Holsinger, I.A. Lasky, et al., Blood glucose control during pregnancy, Diabetes Care 3 (1980) 69–76. [24] M.D. Berkus, O.D. Langer, Glucose tolerance test: degree of glucose abnormality correlates with neonatal outcome, Obstet. Gynecol. 81 (1993) 344–348. [25] R.R. Little, E.M. Mc Kenzie, J.M. Shyken, S.E. Winkelmann, L.M. Ramsey, R.W. Madsn, et al., Lack of relationship between glucose tolerance and complications of pregnancy in nondiabetic women, Diabetes Care 13 (1990) 483–487. [26] L. Jovanovic, C.M. Peterson, Screening for gestational diabetes. Optimum timing and criteria for retesting, Diabetes 34 (Suppl. 2) (1985) 21–23. [27] M.W. Carpenter, D.R. Coustan, Criteria for screening tests for gestational diabetes, J. Obstet. Gynecol. 144 (1982) 768–773. [28] F. Parazzini, I. Cortinovis, R. Bortulus, L. Fedele, A. Recarli, Weight at birth by gestational age in Italy, Hum. Reprod. 10 (1995) 1852–1863. [29] O. Langer, Y. Yogev, O. Most, E.M.J. Xenakis, Gestational diabetes: the consequences of not treating, Am. J. Obstet. Gynecol. 192 (2005) 989–997. [30] A.S.T. Huggett, D.A. Nixon, Use of glucose oxidase peroxidase and O-dianisine in the determination of blood and urine glucose, Lancet II (1957) 368–370. [31] A. Mosca, R. Paleari, A. Trapolino, F. Capani, G. Pagano, M. Plebani, A re-evaluation of glycohaemoglobin standardisation: the Italian experience with 119 laboratories and 12 methods, Eur. J. Clin. Chem. Clin. Biochem. 35 (1997) 243–248. [32] A. Mosca, R. Paleari, M.G. Dalfra`, G. DiCianni, I. Cuccuru, I. Pellegrini, et al., Reference intervals for hemoglobin A1c in pregnant women: data from an Italian multicenter study, Clin. Chem. 52 (2006) 1138–1143. [33] A. Kerssen, H.W. deWalk, G.H.A. Visser, Sibling birth weight as a predictor of macrosomia in women with type 1 diabetes, Diabetologia 48 (2005) 1743–1748. [34] B.E. Metzger, D.R. Coustan, Summary and recommendations of the fourth international workshop-conference on gestational diabetes, Diabetes Care 21 (Suppl. 2) (1998) B161–B167. [35] A. Verma, B.F. Mitchell, N. Demianczuk, G. Flowerdew, N.B. Okun, Relationship between plasma glucose levels in glucoseintolerant women and newborn macrosomia, J. Mat. Fetal. Med. 6 (1997) 187–193. [36] S. Ramtoola, F. Home, H. Damry, A. Husnoo, S. Ah-Kion, Gestational impaired glucose tolerance does not increase perinatal mortality in a developing country: cohort study, BMJ 322 (2001) 1025–1026. [37] M. Bonomo, D. Corica, E. Mion, G. Goncalves, G. Motta, R. Merati, et al., Evaluating the therapeutic approach in pregnancies complicated by borderline glucose intolerance: a randomized clinical trial, Diabetic Med. 22 (2005) 1536–1541.