Pregnancy is commonly recognized as a state of physiological, and temporary insulin resistance. This condition is driven by high concentrations of steroid hormones such as progesterone, estrogens, prolactin, cortisol and by placenta derived human placental lactogen – all of them having a diabetogenic action – combined with a decreased sensitivity of insulin receptors within target tissues (1, 2). The main reason for all these changes is to provide a preferential supply of nutrients (mainly glucose) to the fetus. Usually, in maternal compartment low levels of glucose (even lower than these in non-pregnant women) are maintained due to increased maternal production of insulin and accelerated “drainage” of glucose by uteroplacental unit to cover fetal needs (“accelerated starvation”). However, in some percentage of pregnant population (3% up to 9%, according to population and diagnostic methods) a transient form of glucose intolerance develops if a degree of gestational insulin resistance is beyond the compensatory capability of the pancreas (3-5).

Gestational diabetes mellitus (GDM) is defined as a glucose intolerance of any degree diagnosed or first recognized during pregnancy. In most of the cases it is a gestation-related disease and patient’s normal carbohydrates metabolism is restored within a couple of weeks after a delivery. However, a GDM-complicated pregnancy is still associated with a high perinatal risk with increased neonatal mortality and morbidity, mainly as a result of fetal macrosomia as well as operative and instrumental deliveries, birth trauma and metabolic abnormalities in newborn (6). In this group of pregnant women there is still increased incidence of intrauterine fetal death in term pregnancy. Moreover, a history of gestational diabetes is a strong risk factor for serious maternal complications in later life, including metabolic syndrome, type 2 diabetes and cardiovascular diseases that are considered a leading cause of death in the female population (7).

The major reason for poor perinatal outcome is fetal macrosomia (defined as large for gestational age, with a birth weight above the 90th percentile; LGA). Fetal overgrowth induced by fetal hyperinsulinemia can develop as a response to increased placental glucose transfer to the fetus which is secondary to maternal hyperglycemia. Fetal hyperinsulinemia and accelerated fetal growth is also associated with a cluster of abnormalities usually referred to as “diabetic fetopathy” that may be grounds for serious neonatal complications, including hypoglycemia and respiratory distress (6-9).

In recent decades, significant improvements in perinatal care, diagnostic and treatment of GDM complicated pregnancies has been made. Despite that, macrosomia still remains a serious problem, which may complicate up to 30% of diabetic pregnancies (10, 11). Interestingly, the certain percentage of newborns with macrosomia remains high even in women with proper carbohydrate controls, measured using commonly available parameters (fasting and 2-hrs postprandial glycemia, HbA1c concentration) (12, 13). Therefore, further studies are necessary to investigate other factors contributing to fetal overgrowth in diabetic pregnancy. Maternal hyperglycemia is a commonly recognized classical risk factor for fetal overgrowth. However, it is still debatable whether fasting or postprandial hyperglycemia is a pivotal factor for developing fetal overgrowth (11).

In recent studies fetal growth and development is considered a complex process where maternal characteristics, fetal potential and the intrauterine environment plays an important role (2, 13). This concept opens a new horizon for further research; however, there are a limited number of tools and techniques designed for investigating fetal development and the intrauterine milieu. Moreover, the list of factors influencing fetal growth both in normal and in diabetic pregnancy is expanding and still far from being complete, results of different studies are often conflicted and new areas for research emerge (14).

Recently, a great deal of data has accumulated on lipid metabolism in normal and diabetic pregnancies, maternal obesity and gestational insulin resistance, however, data concerning their impact on fetal growth are limited (15-19).

Pregnancy is commonly described as a condition characterized by a rapid increase in all lipids, however, evidence concerning the relationship between lipid metabolism and hormonal changes during fetal gestation is conflicting (20, 21). In the study performed on a small group of 9 women with GDM, Montelongo et al. reported a linear correlation between HDL-cholesterol, TAG and ß-estradiol, progesterone and prolactin – a finding supported only for TAG by more recent study performed on a larger group by Smolarczyk et al. (20, 21). In their study, Knopp et al. reported a significant association between maternal triglyceride levels and neonatal weight both in normal and in GDM-complicated pregnancy (22). Their results were in line with those obtained by Kitaima et al. who reported maternal hypertriglyceridemia as a significant predictor of LGA (23).

In our study we aim to investigate different maternal metabolic characteristics corresponding to particular features of metabolic syndrome and their compounding influence on fetal growth and the incidence of LGA in pregnant women with GDM.


MATERIAL AND METHOD
Our study group consisted of 357 women referred to the Department of Obstetrics and Women Diseases for a tertiary-level, specialistic antenatal care from 1993 to 2005. The protocol of our study involved a retrospective analysis of patients’ records from the database of our Department. The inclusion criteria were as follows: GDM diagnosed following WHO criteria, singleton pregnancy, live birth and no fetal malformation suspected during gestation or detected postpartum. In our research, we investigated the following parameters: patients’ age, pre-pregnancy body mass index (BMI), gestational age when GDM was diagnosed, 75g oral glucose tolerance test (OGTT) (fasting and 2-hrs post-load glycemia), total, and HDL cholesterol (HDL) at booking, triglycerides (TAG) and HbA1c concentration at booking. We also recorded a birth weight of the largest child born in a previous pregnancy (if applicable). Birth weight and the proportion of LGA (defined as a birth weight >90th percentile for local population after adjusting for gestational age and sex) was studied at the end-point (22).

Patients with gestational diabetes diagnosed in primary or secondary-level centres were referred to our Department for further diagnostic and combined perinatal and diabetes-related care. In 72.8% of patients GDM was diagnosed as an abnormal blood glucose level in 2 hours after 75g glucose administered orally (140 mg/dl), whereas in remaining 27.2% of individuals GDM was diagnosed following highly elevated fasting glycemia or highly abnormal result of 50g glucose screening test (1 hour glycemia 200 mg/dl). During their first hospitalization, having assessed metabolic status and a daily glycemia profile (blood samples taken every two hours starting from 8.00 AM), general obstetrical status and fetus well-being, all participants underwent personal training concerning diet (diet low in simple carbohydrates but covering special nutritional needs of pregnant woman), lifestyle and self-monitoring of glycemia. Target glucose levels were as follows: fasting glycemia between 60-90 mg/dl and 2-hrs postprandial glycemia <120 mg/dl. If a diet alone was not sufficient for maintaining glucose levels within recommended values, insulin therapy in the form of multiple injections of short- and long lasting human insulin was introduced, starting from 0.3 IU/ kg of body weight. The insulin doses were adjusted following meal compositions and the current blood glucose level, as well as patients participating in additional training concerning insulin self-administration. Then, all participants were covered with a specialized perinatal care, involving ambulatory visits in our outpatients’ clinic and hospitalization for delivery.

Blood samples for HbA1c and lipid assessment were taken after overnight fasting, centrifuged and assayed according to protocols used in our ISO-certified Central Hospital Laboratory. HbA1c concentration was measured using turbidimetric immunoinhibitory method, with reference values (for nonpregnant individuals) 4.8-6.0%. Total and HDL cholesterol concentrations were measured using quantitative enzymatic colorimetric methods, with reference values (for nonpregnant individuals) 0.57-2.28 and 0.4-0.8 mmol/l, respectively. TAG concentration was measured using a quantitative enzymatic colorimetric method.

For the purposes of this study, we initially investigated bivariate correlations between the maternal parameters and a birth weight, as well as features that correlated significantly with a birth weight were chosen for further calculations. We have analyzed distribution of the following covariates: fasting glycemia in 75g OGTT, HbA1c level, pre-pregnancy BMI, HDL and TAG concentration. Then, we defined values above 75th percentile as altered, except from HDL that was described as altered if below the 25th percentile. As a next step, we retrospectively divided the study group into five subgroups according to the number of altered metabolic parameter found in each participant: from 0 if all of the following: BMI, fasting glycemia, HbA1c concentration, HDL concentration and TAG concentration were within interquartile range (i.e. between 25th and 75th percentile) to 5 if all of them were altered. Then, the birth weight and prevalence of LGA were studied across the subgroups.

Statistical analysis was performed using SPSS 12.0 for Windows software. Results are expressed as median (minimum-maximum value), unless otherwise stated. The significance of the differences between study groups was tested using U Mann-Whitney’s test, or Kruskall-Wallis test, according to a number of groups tested. Differences in categorical variables were tested using a chi-square statistic. Associations between maternal metabolic parameters and birth weight were analysed using Spearman rank correlation coefficients. Linear regression analysis was performed to find independent predictors for birth weight in the study group. p<0.05 was considered statistically significant.


RESULTS
Basic characteristics of the patients enrolled into the study are given in Table 1. Insulin therapy was necessary to achieve a proper metabolic control in 25.2% of individuals. Hypertension (chronic or gestational) was diagnosed in 10.9% of subjects.

Table 1. Characteristics of the study group

We found weak but significant positive linear correlations between birth weight and the following covariates (see Fig. 1A-E): maternal pre-pregnancy BMI (R=0.2, p<0.00001), birth weight of the largest offspring (R=0.34, p<0.00001), HbA1c at booking (R=0.11, p<0.05), fasting glycemia during 75g OGTT (R=0.23, p<0.00001) and TAG concentration at booking (R=0.12, p<0.05). No correlation was found for total and HDL-cholesterol and for maternal age as well as gestational age when GDM was diagnosed.

Fig. 1A. Correlation between maternal BMI before pregnancy and a birth weight (p<0.00001, N=309)

Fig. 1B. Correlation between birth weight of the largest offspring and a birth weight in the study group (p<0.00001, N=170)

Fig. 1C. Correlation between HbA1c concentrations at booking and a birth weight (p<0.05, N=317)

Fig. 1D. Correlation between fasting glycemia during 75g OGTT and a birth weight (p<0.00001, N=282)

Fig. 1E. Correlation between maternal TAG concentrations at booking and a birth weight (p<0.05, N=319)

To determine independent predictors of birth weight in our study group, we performed linear multiple regression analysis with birth weight as an dependent variable and maternal metabolic characteristics as an initial independent variable. Results of the regression analysis are summarized in Table 2. Univariate analysis demonstrated that birth weight of the largest offspring was the strongest independent predictor of a birth weight in our group, accounting for around 19% of the variation. Maternal BMI and fasting glycemia alone predicted 7.5% and 7.7% of the variation in the dependent variable, respectively. Other predictors explained small proportion of the variation in the studied parameter (less than 5%). All predictors, except from TAG concentrations, remained significant when gestational age at diagnosis was entered into models.

Table 2. Independent predictors of birth weight in the study group

Our research also involved an investigation of the combined influence of multiple metabolic alterations on birth weight and the prevalence of macrosomia. There was no association between HDL concentrations and birth weight in our study, however, as a separate analysis showed that HDL-cholesterol concentrations were a significant predictor for LGA (study in progress, data unpublished), we decided to include this parameter in a further analysis. Distribution of maternal metabolic parameters is given in Table 3. Finally, we analysed birth weight and the proportion of LGA newborns in subgroups of individuals with different numbers of altered parameters (i.e. values for particular metabolic characteristics given in Table 3 above 75th percentile/ below 25th percentile for HDL). Results are given in Table 4 and Fig. 2. We found a highly significant difference in birth weight and the prevalence of LGA when comparing pregnant women with the different numbers of altered metabolic features (i.e. values within the highest or the lowest quartile). Proportion of macrosomic newborns varied from approximately10% in individuals with no abnormalities or two altered values to over 80% in patients with 5 out of 6 parameters changed.

Table 3. Metabolic characteristics in the study group

Table 4. Birth weight and % of LGA newborns in relation to maternal metabolic alterations

*) Kruskal-Wallis test; †) Chi2 test; Maternal metabolic characteristics defined as altered: pre-pregnancy BMI above 75th percentile, TAG concentration above 75th percentile, fasting glycemia above 75th percentile, HbA1c at booking (during first hospitalisation) above 75th percentile, HDL concentration below 25th percentile.

Fig. 2. Birth weight in relation to maternal metabolic alterations (p<0.0001)

DISCUSION
Pregnancy is commonly recognised as a period of very intense changes in maternal metabolism. Apart from vast literature regarding gestational diabetes, there is much evidence available on maternal insulin resistance or shifts in lipids/ lipoproteins profiles (18, 19, 25- 27).

In their study, Piechota et al. reported all lipids significantly elevated during uncomplicated gestation in healthy women, with a 2.7-fold increase in the TAG level, 56% increase in total cholesterol and 25% increase in HDL-cholesterol (28). Comparing data from their study with our population, our patients had higher TAG concentrations [95th percentile for our study group: 6.03 mmol/l vs 4.68 mmol/l reported by Piechota et al. (27)], lower total cholesterol concentrations (95th percentile for our study group: 9.03 mmol/l vs 9.83 mmol/l reported by Piechota et al.) and higher HDL-cholesterol concentrations [5th percentile for our study group: 1.20 mmol/l vs 1.04 mmol/l reported by Piechota et al. (27)].

The major difficulty we encountered during our research was the lack of officially recommended reference values for lipids adapted for pregnant women. Therefore, we decided to use the interquartile range to establish subgroups for further analysis and our approach is similar to those of past researchers (23, 25). In their study, Kitajima et al. (23) defined maternal hipertriglyceridemia as serum TAG above the 75th percentile and reported it as a significant risk factor for LGA. They also analysed the maternal fasting plasma glucose, prepregnancy BMI and gestational change in maternal body weight, however, they did not find any association between these covariates and birth weight. The same methodology concerning TAG concentrations was also applied by Di Cianni at al. who reported significant association between elevated maternal TAG levels, pregestational BMI, weight gain during pregnancy and 2-hrs OGTT glycemia (26). The last finding is not supported by our results, as we found a significant association between maternal fasting glycemia and LGA. It seems surprising as a main pathomechanism of gestational diabetes is driven by impaired response to carbohydrate loading. It should be stated that both Kitajima et al. (23) and Di Cianni et al. (26) studies were performed in smaller study groups (146 and 180 individuals, respectively). Some differences in results may also be attributed to different ethnic background (Kitajima et al. (23) investigated Japanese gravidas).

There is an increasing amount of evidence from recent studies that a number of factors appears to have an impact on fetal growth (2, 11, 12). Our finding concerning association between maternal BMI, TAG concentrations and fetal growth is in accordance with results reported by Cianni et al. who investigated pregnant women with different forms of glucose intolerance (impaired fasting glycemia or GDM) (26). Also another study, performed by Schaefer-Graf et al. described an increased incidence of LGA in obese pregnant women with GDM rather than in women with hyperglycemia alone, which is comparable to our findings (18).

The overall proportion of LGA newborns in our study group was approximately 19%, which is similar to findings of past studies (19, 26). However, in our study we report a significant increase (almost eight-fold) in the prevalence of LGA, following the presence of complex, even if mild metabolic alterations. It is important to note that a proportion of LGA in pregnant women from our study group with all analyzed metabolic parameters (maternal BMI, fasting glycemia, HbA1c concentration, HDL concentration and TAG concentration) within an interquartile range is similar to that noted within the normal population (10%). Surprisingly, the same percentage is noted in the group of women with two altered parameters. However, upon closer analysis, we observed that most of these patients had a BMI below the 75th percentile (71.1%), whereas in individual with 3 or more altered metabolic features an elevated BMI was a common finding (up to 100% in the group with 5 altered features), accompanied by a significant increase in birth weight. Our results suggest that in a population of properly controlled and treated women with GDM, fetal overgrowth seems to be driven by other factors, possibly a cluster of metabolic alterations associated with maternal obesity, as we observed an elevated maternal BMI in 36.6% of LGA in our study group. The association between gestational diabetes and metabolic syndrome is a subject of recent studies, and one of these studies reported on the influence of metabolic syndrome on fetal growth in diabetic women regardless of the degree of hyperglycemia (19). It may explain the phenomenon of an increased prevalence of macrosomia in this group of patients, despite improvement in diagnostics and treatment options. However, further studies addressing a variety of factors associated with metabolic syndrome, performed during gestation and their both individualised and combined impact on birth weight are necessary to corroborate our findings. Clinical implications of these findings may include a more active approach to obesity among women in childbearing age as our evidence suggest that maternal obesity is a factor that deteriorates the effectiveness of hypoglycaemic treatment in pregnancy complicated with GDM. Fetal growth in a diabetic pregnancy is a complex process and maternal metabolic parameters other than glucose levels should be addressed to reduce the risk of macrosomia in this group of patients.

Conflict of interest statement: None declared.


REFERENCES

1. Desoye G, Schweditsch MO, Pfeiffer KP, Zechner R, Kostner GM. Correlation of hormones with lipid and lipoprotein levels during normal pregnancy and postpartum. J Clin Endocrinol Metab 1987; 64: 704-712.
2. Zamlynski J, Olejek A, Wiecek A et al. Wplyw zmian metabolicznych w ciazy prawidlowej i powiklanej cukrzyca na wewnatrzmaciczne wzrastanie plodu (The influence of metabolic changes on intrauterine fetal growth in normal gestation and gestation complicated by diabetes). Postepy Hig Med Doœw 2005; 59: 490-495.
3. Lucas MJ. Diabetes complicating pregnancy. Obstet Gynecol Clin North Am 2001; 3: 513-536.
4. Moses RG, Moses M, Russel KG et al. The 75-g glucose tolerance test in pregnancy. Diab Care 1998; 11: 1807-1811.
5. Williams CB, Iqbal S, Zawadzki CM et al. Effect of selective screening for gestational diabetes. Diab Care 1999; 3: 418-421.
6. Casey BM, Lucas MJ, McIntire DD, Leveno KJ. Pregnancy outcome in women with gestational diabetes compared with general obstetric population. Obstet Gynecol 1997; 90: 869-873.
7. Dornhorst A, Rossi M. Risk and prevention of type 2 diabetes in women with gestational diabetes. Diab Care 1998; 21(suppl.2): B43-B49.
8. Svare JA, Hanses BB, Mølsed-Petersen L. Perinatal complications in women with gestational diabetes mellitus. Acta Obstet Gynecol Scand 2001; 80: 899-904.
9. Owens JA, Gatford KL, De Blasio MJ, Edwards LJ, McMillen IC, Fowden AL. Restriction of placental growth in sheep impairs insulin secretion but not sensitivity before birth. J Physiol 2007; 584(Pt 3): 935-949.
10. González-Quintero VH, Istwan NB, Rhea DJ et al. The impact of glycemic control on neonatal outcome in singleton pregnancies complicated by gestational diabetes. Diab Care 2007; 30: 467-470.
11. Langer O. Fetal macrosomia. Etiologic factors. Clin Obstet Gynecol 2000; 43: 283-297
12. Wender-O¿egowska E, Zawiejska A, Kwinecka B, Chmaj K, Biczysko R. Metabolic control and obstetrical results in patients with gestational diabetes mellitus diagnosed before and after 24th week of pregnancy. Diabetol Pol 2002; 4: 165-169.
13. Drake AJ, Walker BR. The intergenerational effects of fetal programming: non-genomic mechanisms for the inheritance of low birth weight and cardiovascular risk. J Endocrinol 2004; 180: 1-16.
14. Rohl J, Huston-Presley L, Amini S, Stepanchak B, Catalano P. Factors associated with fetal growth and body composition as measured by ultrasound. Am J Obstet Gynecol 2001; 185: 1416-1420.
15. Arcos F, Castelo-Branco C, Casals E, Sanllehy C, Cararach V. Normal and Gestational diabetic pregnancies – lipids, lipoproteins and apolipoproteins. J Reprod Med 1998; 43: 144-148.
16. Butte NF. Carbohydrate and lipid metabolism in pregnancy: normal compared with gestational diabetes mellitus. Am J Clin Nutr 2000; 71(suppl): 1256S-1261S.
17. Ramsay JE, Ferrell WR, Crawford L, Wallace MA, Greer IA, Sattar N. Maternal obesity is associated with dysregulation of metabolic, vascular and inflammatory pathways. J Clin Endocrinol Metab 2002; 87: 4231-4237
18. Schaefer-Graf UM, Heuer R, Kilavuz Ö, Pandura A, Henrich W, Vetter K. Maternal obesity not maternal glucose correlates best with high rates of fetal macrosomia in pregnancies complicated by gestational diabetes. J Perinat Med 2002; 30: 313-321.
19. Hultman K, Alexanderson C, Mannerås L, Sandberg M, Holmäng A, Jansson T. Maternal taurine supplementation in the late pregnant rat stimulates postnatal growth and induces obesity and insulin resistance in adult offspring. J Physiol 2007; 579(3): 823-833.
20. Montelongo A, Asuncion MA, Pallado LF, Herdera E. Longitudinal Study of Plasma Lipoproteins and Hormones During Pregnancy in Normal and Diabetic Women. Diabetes 1992; 41: 1651-1659.
21. Smolarczyk R, Czajkowski K, Wojcicka-Jagodzinska J et al. The influence of chosen hormones on lipids and its correlation during normal pregnancy. Gin Pol 2001: 10: 765-771.
22. Knopp RH, Magee MS, Walden CE, Bonet B, Benedetti TJ. Prediction of infant birth weight by GDM screening tests. Importance of plasma triglyceride. Diab Care 1992; 15: 1605-1613.
23. Kitajima M, Satoshi O, Yasuhi I, Fukusa M, Rii Y, Ishimaru T. Maternal serum triglyceride at 24-32 weeks’ gestation and newborn weight in nondiabetic women with positive diabetic screens. Obstet Gynecol 2001; 97: 776-780.
24. Malewski Z, Slomko Z, Klejewski A. Relacje wieku ciazowego i masy urodzeniowej noworodkow z regionu Wielkopolski. Kliniczna Perinatologia i Ginekologia 1995; II (suppl.1): 734.
25. Zwirska-Korczala K, Konturek SJ, Sodowski M et al. Basal and postprandial plasma levels of PYY, ghrelin, cholecystokinin, gastrin and insulin in women with moderate and morbid obesity and metabolic syndrome. J Physiol Pharmacol 2007; 58(Suppl. 1); 13-35.
26. Cianni G, Miccoli R, Volpe L et al. Maternal triglyceride levels and newborn weight in pregnant women with normal glucose tolerance. Diabet Med 2004; 22: 21-25.
27. Piechota W, Staszewski A. Reference ranges of lipids and apolipoproteins in pregnancy. Eur J Obstet Gynecol Reprod Biol 1992; 45: 27-35.
28. Bo S, Menato G, Gallo ML et al. Mild gestational hyperglycemia, the metabolic syndrome and adverse neonatal outcomes. Acta Obstet Gynecol Scand 2004; 83: 335-340.