Which growth criteria better predict fetal programming?
- Sandra S Mattos1,2,
- Maria Elizabeth C Chaves1,
- Suzana Maria Ramos Costa1,
- Ana Catarina Matos Ishigami1,2,
- Sarah Bezerra Rêgo1,2,
- Vinicius Souto Maior1,
- Rossana Severi2,
- José Luiz de Lima Filho1
- 1Laboratório de Imunopatologia Keizo-Asami—LIKA, Universidade Federal de Pernambuco—UFPE, Recife, Brazil
- 2Unidade de Cardiologia Materno-Fetal—UCMF, Real Hospital Português de Beneficência em Pernambuco—RHP, Recife, Brazil
- Correspondence to Dr Sandra S Mattos, Maternal-Fetal Cardiac Unit, Royal Portuguese Hospital, Av Portugal 163, Recife, PE 50090-900, Brazil;
- Accepted 21 February 2011
- Published Online First 27 March 2011
Objective To test whether customised (ct) growth criteria are more reliable than standard (st) ones to predict intrauterine insult.
Patients 32 mothers and their singleton term neonates selected as small for gestational age (st-SGA=8) or appropriate for gestational age (st-AGA=24).
Main outcome measures Nitric oxide, high-sensitive C reactive protein (hs-CRP), uric acid, blood lipids and protein levels were analysed in maternal and cord blood.
Results Applying customised criteria yielded 16 ct-AGA, 13 ct-SGA and 3 ct-LGA (large for gestational age) babies. Both st-SGA and ct-SGA babies had higher nitric oxide and hs-CRP levels. Their mothers had lower albumin fractions. st-SGA babies also had higher triglyceride and cholesterol levels. ct-LGA babies and mothers had higher uric acid levels, and the mothers had higher triglyceride levels.
Conclusions Customised growth criteria better identify babies submitted to unfavourable intrauterine environments. The authors suggest that combined with maternal biochemistry, these growth criteria can be used to screen for adverse fetal programming.
Low birth weight is considered the hallmark of an adverse intrauterine environment. This paradigm, however, has been disputed. Whether fetal weight gain is not always affected by an adverse intrauterine environment or standard (st) criteria fail to evaluate true growth potential is unclear. Customised (ct) growth criteria have been proposed to better differentiate between constitutional smallness and true growth restriction. Based on the above, we hypothesised that customised growth criteria are more reliable to predict intrauterine insult in small for gestational age (SGA) babies. Therefore, we measured biochemical parameters previously associated with intrauterine growth restriction, such as nitric oxide, high-sensitive C reactive protein (hs-CRP), uric acid, blood lipids and protein levels,1,–,4 in both st-SGA and ct-SGA babies and their mothers.
Patients and methods
Approval was obtained from the Ethics Committee and all participants signed consent forms. Exclusion criteria were: gemelarity, genetic anomaly, infection, collagen disease, maternal drug abuse, prematurity and pre-eclampsia.
Thirty-two neonates of either sex born between 37 and 42 weeks, during the period August 2009 to March 2010, fulfilled the selection criteria and were enrolled. They included 8 st-SGA and 24 appropriate for gestational age (st-AGA) control babies. No large for gestational age (st-LGA) babies were selected.
Customised growth criteria were applied using the Gestation Related Optimal Weight (GROW) software (available at http://www.gestation.net). Because the dataset for Latin America is not included in the software, birthweight percentiles were plotted against all available ethnic groups and mean values below 10% and above 90% were used to classify ct-SGA and ct-LGA babies, respectively.
Maternal and cord blood samples were collect and stored, and placentas were weighed, according to established protocols. The DINO 250 kit from BioAssay Systems (Hayward, CA, USA) was used for determination of nitric oxide production. Biochemical analyses followed routine laboratory procedures.
Statistical differences were calculated using the Kruskal–Wallis' non-parametric test in the software R-v.2.10.0, available at http://www.r-project.org. Hypothesis tests were performed by considering a level of significance of 5% (p<0.05). Values are presented with 95% CIs.
Maternal and fetal anthropometric data are shown in table 1.
Based on customised criteria, one-third of the st-AGA babies were moved to different categories, five to ct-SGA and three to ct-LGA.
Besides being smaller (p=0.003) and lighter (p<0.001) than their st-AGA counterparts, st-SGA babies had smaller placentas (p=0.002) and were born at an earlier gestational age (p=0.024). ct-SGA babies were also smaller (p=0.013) and lighter (p<0.001) than ct-AGA babies, but exhibited no differences in gestational age or placental size. The ratio of placental weight to fetal weight was not different between st-SGA versus st-AGA (p=0.459) but was increased between ct-SGA versus ct-AGA (p=0.011). No anthropometric differences were found between ct-LGA versus ct-AGA babies.
Statistical analysis confirmed the differences among nitrite/nitrate and hs-CRP levels as follows: nitrate levels, st-AGA versus st-SGA (p=0.015) and ct-AGA versus ct-SGA babies (p=0.001); hs-CRP levels, st-AGA versus st-SGA (p=0.004) and ct-AGA versus ct-SGA (p=0.002). Mothers of ct-SGA (p=0.054) and ct-LGA (p=0.057) babies also had elevated nitrate levels.
Uric acid levels were elevated in both maternal (p=0.029) and umbilical cord blood (p=0.019) only in the ct-LGA group.
Total cholesterol and its fractions were elevated in both st-SGA and ct-SGA neonates. Statistical analyses showed differences between total cholesterol (p=0.041) and very low-density lipoprotein (p=0.052) fraction for st-AGA versus st-SGA. Low-density lipoprotein fraction differences were present in st-AGA versus st-SGA (p=0.028) and in ct-AGA versus ct-SGA (p=0.041). The ratio of total cholesterol to high-density lipoprotein fraction is reduced in both st-SGA (p=0.016) and ct-SGA (p=0.030). Mothers of ct-LGA babies had significantly higher triglyceride levels (p=0.019).
Lower percentages of albumin with elevation of all other protein fractions were observed in the mothers of st-SGA and ct-SGA babies. Statistical analyses showed differences between the proportions of albumin and α-1-globulin for both st-AGA versus st-SGA (p=0.013 and p=0.010) and ct-AGA versus ct-SGA babies (p=0.016 and p=0.002), respectively. For α-2–globulin, percentage differences were noted only for ct-AGA versus ct-SGA babies (p=0.017). Neonatal protein profiles were not different between groups.
We chose Gardosi's customised growth criteria5 because it incorporates the mother's height, weight, ethnicity and parity—variables that are known to account for 20–35% of the variability of birth weight at term. The dataset for the Latin American ethnicity is not included in the GROW software. As Brazilians have diverse ethnic backgrounds, we plotted birth weights against all ethnic groups available in the database and used mean values.
Unexpectedly, three babies were classified as ct-LGA. Although this is a very small number, they were included in the study because they were enrolled as st-AGA. These babies and their mothers presented with higher uric acid levels. Whether this could act as a peroxynitrate scavenger protecting babies from nitrative stress, or impair nitric oxide release from endothelial cells and further promote an oxidative stress, needs to be further evaluated.1 Interestingly, these mothers started gestation with lower body mass indexes and put on more weight than ct-SGA or ct-AGA mothers. They also had higher triglyceride levels. The literature has documented that the concentration of triglycerides in the third trimester of pregnancy is a stronger predictor of birth weight.
This study confirmed many biochemical findings reported in association with intrauterine growth restriction, such as an altered lipid profile, the presence of an inflammatory state and disturbed endothelial function. It further expanded knowledge by demonstrating that neonates classified as AGA by standard growth criteria, but who fall short of their true growth potential by customised criteria, show biochemical abnormalities which are compatible with an adverse intrauterine environment. In childhood, these biochemical abnormalities are predictive of adult health. It is not clear as to what could be the long-term effects of these abnormalities when they occur at such an early and sensitive period of life.
Elevated nitrite/nitrate levels were observed in both st-AGA and ct-AGA babies. Although this would suggest vasodilatation and better perfusion, a recent study by Hracsko et al2 demonstrated reduced superoxide dismutase activity in SGA babies, leading to an increase in superoxide anions and nitrative stress.
Interestingly, only in ct-SGA babies the placenta:birthweight ratios were increased. This probably reflects an attempt by the placenta, by increasing in size, to adequately supply the fetuses in an adverse intrauterine environment.
It has been suggested that, compared with AGA babies, SGA babies have increased visceral fat stores, which would link the adipose depot to aseptic inflammation and explain the elevated hs-CRP levels found in both st-SGA and ct-SGA babies.3
Finally, mothers of st-SGA and ct-SGA babies were not clinically malnourished and had normal total protein levels. However, they had significantly lower albumin fractions, which is in agreement with the general knowledge that protein deficiency leads to growth restriction.4
In conclusion, we established a biochemical profile for an adverse intrauterine environment and showed that customised growth criteria are better to detect which babies are at risk from it. Our findings also indicate that the effects of such an adverse environment follow a continuum from which some babies will show little clinical evidence while others will develop significant growth restriction or, possibly, excessive weight gain.
We suggest that customised growth criteria, combined with maternal biochemistry, can be used as a screening approach for individuals at risk from adverse fetal programming.
This work has been supported by FACEPE (Science and Technology Foundation /Pernambuco) and CNPq (National Research Council) as part of a PhD programme at RENORBIO (Northeast Biotechnology Network), Brazil. The authors wish to express their gratitude to the many researches at LIKA and UCMF for their support.
Funding Research grants from FACEPE and CAPES/Brazil.
Competing interests None.
Patient consent Obtained.
Ethics approval This study was conducted with the approval of the National Ethics Committee through Hospital Agamenon Magalhães/Recife.
Provenance and peer review Not commissioned; externally peer reviewed.