OBJECTIVE—To assess the association of fetal hormones with placental growth and fetal weight–to–placental weight ratio index (FPI) in pregnancies complicated by maternal diabetes.
RESEARCH DESIGN AND METHODS—We conducted a prospective study using umbilical venous blood samples taken at birth from 122 offspring of mothers with type 1 diabetes (OT1D) and 46 control subjects.
RESULTS—Placental weight (P = 0.009) and gestation-adjusted birth weight (P < 0.001) were increased in OT1D, but FPI was unaltered (P = 0.33). Placental weight correlated with birth weight (P < 0.001) and cord leptin (P < 0.001) in control subjects and OT1D, with further relationships with cord insulin, IGF-1, IGF-binding protein-3 (IGFBP-3), and triceps and subscapular thickness in OT1D. FPI was associated with adiponectin in both groups, even after adjustment for confounders.
CONCLUSIONS—Placental and fetal growth show a parallel increase in mothers with type 1 diabetes. The possible role of adiponectin in matching of fetal and placental growth merits further study.
Placental weight is strongly associated with birth weight, with experimental and epidemiological studies demonstrating associations of reduced fetal weight–to–placental weight ratio index (FPI) with later hypertension, glucose intolerance, and coronary heart disease, suggesting in utero programming of adult disease (1). Fetal adiponectin, an adipokine with insulin-sensitizing and anti-inflammatory effects, has been identified as the first biomarker associated with FPI (2). In this study, we examined FPI and its relation to hormonal indexes, in particular those of insulin and adiponectin, in offspring of mothers with type 1 diabetes (OT1D), a group observed to exhibit reduced FPI (3) and adiponectin (4) and to be subject to in utero programming of glucose intolerance (5).
RESEARCH DESIGN AND METHODS—
A comprehensive description of prospective recruitment and exclusion criteria are available in previous publications (4,6). A total of 122 OT1D and 46 control subjects were available for analysis. FPI (birth weight [grams] divided by placental weight [grams]) was calculated for each delivery. Maternal A1C and cord plasma insulin, leptin, IGF-1, IGFBP-3, adiponectin, C-reactive protein (CRP), and intracellular adhesion molecule-1 were assayed centrally (4,6). Hemoglobin and hematocrit (available in 32 control subjects and 81 OT1D) were measured locally by routine clinical hematological analyzers. A1C was included when assessed between weeks 26 and 34 of pregnancy (available in 90 OT1D).
Results are presented as mean ± SD or unadjusted median (interquartile range). Pearson correlations, general linear models, and stepwise logistic regression (P ≤ 0.15 for inclusion of predictors) on log-transformed variables were used to assess relationships.
RESULTS—
Mean ± SD birth weight (control subjects 3,553 ± 520 g, OT1D 3,778 ± 701 g; P < 0.001) and placental weight (control subjects 627 g [551–700], OT1D 700 g [600–800]; P < 0.001) were increased in OT1D, with no difference in FPI (control subjects 5.73 ± 0.9, OT1DM 5.55 ± 1.12; P = 0.33) even after adjustment for gestational age at delivery or sex. Maternal diabetes was associated with increased cord insulin (control subjects 22.5 pmol/l [15.4–37.7], OT1D 111.0 pmol/l [64.9–220.8]; P < 0.001), leptin (control subjects 9.0 ng/ml [4.1–17.2], OT1D 33.3 ng/ml [14.1–58.0]; P < 0.001) and CRP (control subjects 0.14 mg/l [0.12–0.16], OT1D 0.17 mg/l [0.13–0.22]; P < 0.001) and a reduction in adiponectin (control subjects 21.9 ± 5.3 μg/ml, OT1D 19.7 ± 6.2 μg/ml; P = 0.039). All differences remained significant after adjustment for sex, mode of delivery, placental weight, and birth weight.
Placental weight was strongly correlated with birth weight (control subjects r = 0.63, P < 0.001; OT1D r = 0.66, P < 0.001) and fetal leptin (control subjects r = 0.42, P = 0.004; OT1D r = 0.37, P < 0.001). In OT1D, placental weight was also associated with insulin (r = 0.46, P < 0.001), IGF-I (r = 0.54, P < 0.001), IGFBP-3 (r = 0.50, P < 0.001), and maternal A1C (r = 0.45, P < 0.001). FPI was associated with adiponectin (r = 0.43, P = 0.002) and CRP (r = −0.32, P = 0.03) in control subjects and maternal A1C (r = −0.28, P = 0.006), insulin (r = −0.18, P = 0.04), IGF-1 (r = −0.21, P = 0.02), and IGFBP-3 (r = −0.20, P = 0.03) in OT1D. FPI was not related to birth weight in either control subjects or OT1D but showed negative relationships with placental weight in both groups (control subjects r = −0.69, P < 0.001; OT1D: r = −0.66, P < 0.001).
In multivariate analysis (Table 1), birth weight was positively associated with male sex, older gestational age at delivery, insulin, and leptin in both control subjects and OT1D with additional relationships with higher IGF-1 and CRP levels in OT1D. Placental weight showed similar relationships with the fetal cord measures, with the addition of a borderline negative relationship with adiponectin (P = 0.03) in both OT1D and control subjects. By contrast, only adiponectin was significantly related to FPI in both control subjects (contribution to variance [CTV] 20.3%, P = 0.005) and OT1D (CTV 3.9%, P = 0.03) with additional effects of sex, IGF-1, and gestational age at delivery in OT1D.
The contributions of hypoxia and maternal glycemia in the relationship of FPI and adiponectin were assessed by inclusion of cord hematocrit or maternal A1C as predictors. Hematocrit was weakly associated with FPI (CTV 3.5%, P = 0.09) in OT1D independent of other predictors, with the association with adiponectin maintained (CTV 4.4%, P = 0.05). Maternal A1C was a negative predictor of FPI (CTV 7.4%, P = 0.01); however, again, associations of FPI with adiponectin were maintained (CTV 5.1%, P = 0.03).
CONCLUSIONS—
We demonstrate that although both fetal and placental size are increased in the presence of maternal diabetes, their respective weights remain highly correlated, and FPI is not significantly reduced in our series. Lower FPI is driven primarily by higher placental weight in both control subjects and OT1D and is also associated with poorer maternal glycemic control in OT1D. Finally, we confirm recent findings that lower FPI is associated with lower adiponectin concentrations in control subjects (2) and show a similar, albeit weaker, relationship in OT1D.
The contribution of fetal hormonal axes in matching fetal and placental growth in OT1D is largely unknown. Birth and placental weight are positively associated with cord blood levels of IGF-1, IGFBP-3, and IGF-2:IGF-2 receptor ratios (7,8). We observe similar relationships in OT1D and, in addition, a relationship of insulin with birth weight and placental weight. In accordance with this, placental insulin receptors undergo temporospatial shifts to fetal endothelium across gestation, enabling fetal insulin to drive fetal and potentially placental growth (9).
Adiponectin has recently been reported as an associate of FPI independent of fetal adiposity (2), which we speculated might reflect an augmentation of insulin sensitivity by adiponectin and consequent increased fetal growth (2). By contrast, though we confirmed the relationships of adiponectin and FPI, we did not show a correlation of adiponectin to birth weight, which suggests that other mechanisms are responsible. Direct effects of fetal adiponectin on the placenta also appear unlikely given the localization of adiponectin receptors to syncytiotrophoblast (10).
The relationship of adiponectin and FPI might be secondary to the role of hypoxia given the association of hypoxia with reduced FPI and diminished adipocyte adiponectin secretion. However, inclusion of hematocrit into our models did not alter the association between adiponectin and FPI in control subjects or OT1D. Furthermore, the lack of relationship between adiponectin and cord hematological indexes would suggest that the low adiponectin levels of OT1D are unlikely to be due to hypoxia. Similarly, although adiponectin is inversely correlated with systemic indexes of inflammation, the lack of attenuation of the relationship between adiponectin and FPI in control subjects by inclusion of CRP in multivariate analysis and the absence of this relationship in OT1D suggest that FPI and adiponectin are not linked via inflammatory pathways. The role of adiponectin in the linkage of fetal and placental growth merits further study.
Multivariate analysis of predictors of birth weight, placental weight, and FPI
. | Control subjects . | . | . | OT1D . | . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|
. | β . | Percent variance . | P . | β . | Percent variance . | P . | ||||
Birth weight* | ||||||||||
log leptin (ng/ml) | 169 | 24.2 | 0.01 | 140 | 10.0 | 0.001 | ||||
Sex | Male higher | 22.2 | 0.001 | Male higher | 5.4 | 0.001 | ||||
Gestational age (weeks) | 221 | 9.7 | 0.0002 | 173 | 10.3 | <0.001 | ||||
log insulin (pmol/l) | 105 | 6.6 | 0.01 | 128 | 2.8 | 0.008 | ||||
IGF-I (nmol/l) | – | – | – | 89 | 27.6 | <0.001 | ||||
log CRP (mg/ml) | – | – | – | 158 | 1.6 | 0.050 | ||||
Placental weight* | ||||||||||
Adiponectin (μg/ml) | −08 | 6.0 | 0.03 | −4 | 1.6 | 0.06 | ||||
log leptin (ng/ml) | 60 | 15.3 | 0.004 | – | – | – | ||||
Sex | Male higher | 6.5 | 0.07 | – | – | – | ||||
log Insulin (pmol/l) | – | – | – | 45 | 6.0 | 0.002 | ||||
IGF-I (nmol/l) | – | – | – | 27 | 31.1 | <0.001 | ||||
log CRP (mg/ml) | – | – | – | 59 | 3.3 | 0.014 | ||||
FPI* | ||||||||||
Adiponectin (μg/ml) | 0.07 | 20.3 | 0.005 | 0.04 | 3.9 | 0.03 | ||||
Sex | – | – | – | Male higher | 2.7 | 0.001 | ||||
Gestational age (weeks) | – | – | – | 0.18 | 4.7 | 0.02 | ||||
IGF-I (nmol/l) | – | – | – | −0.06 | 5.2 | 0.07 | ||||
log CRP (mg/ml) | −0.56 | 6.0 | 0.08 | – | – | – |
. | Control subjects . | . | . | OT1D . | . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|
. | β . | Percent variance . | P . | β . | Percent variance . | P . | ||||
Birth weight* | ||||||||||
log leptin (ng/ml) | 169 | 24.2 | 0.01 | 140 | 10.0 | 0.001 | ||||
Sex | Male higher | 22.2 | 0.001 | Male higher | 5.4 | 0.001 | ||||
Gestational age (weeks) | 221 | 9.7 | 0.0002 | 173 | 10.3 | <0.001 | ||||
log insulin (pmol/l) | 105 | 6.6 | 0.01 | 128 | 2.8 | 0.008 | ||||
IGF-I (nmol/l) | – | – | – | 89 | 27.6 | <0.001 | ||||
log CRP (mg/ml) | – | – | – | 158 | 1.6 | 0.050 | ||||
Placental weight* | ||||||||||
Adiponectin (μg/ml) | −08 | 6.0 | 0.03 | −4 | 1.6 | 0.06 | ||||
log leptin (ng/ml) | 60 | 15.3 | 0.004 | – | – | – | ||||
Sex | Male higher | 6.5 | 0.07 | – | – | – | ||||
log Insulin (pmol/l) | – | – | – | 45 | 6.0 | 0.002 | ||||
IGF-I (nmol/l) | – | – | – | 27 | 31.1 | <0.001 | ||||
log CRP (mg/ml) | – | – | – | 59 | 3.3 | 0.014 | ||||
FPI* | ||||||||||
Adiponectin (μg/ml) | 0.07 | 20.3 | 0.005 | 0.04 | 3.9 | 0.03 | ||||
Sex | – | – | – | Male higher | 2.7 | 0.001 | ||||
Gestational age (weeks) | – | – | – | 0.18 | 4.7 | 0.02 | ||||
IGF-I (nmol/l) | – | – | – | −0.06 | 5.2 | 0.07 | ||||
log CRP (mg/ml) | −0.56 | 6.0 | 0.08 | – | – | – |
Stepwise regression model for birth weight, placental weight, and FPI included the predictors, log insulin, log leptin, adiponectin, IGF-I, log CRP, sex, and gestational age. Predictors shown where P < 0.10.
Article Information
This study was supported by grants from the Chief Scientist Office (K/MRS/50/C2726) and the GRI Endowment (05REF007).
References
Published ahead of print at http://care.diabetesjournals.org on 13 March 2008. DOI: 10.2337/dc07-2195.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C Section 1734 solely to indicate this fact.