Leptin, an adipocyte-derived hormone that is also produced by a number of other tissues including the stomach, intestine, and the placenta in humans,1–5 acts on hypothalamic receptors to decrease food intake and increase energy expenditure.1 Emerging evidence suggests that leptin also has a number of specific effects on the regulation of whole-body glucose homeostasis.6,7 Results from numerous metabolic studies document a positive association between direct and indirect measures of adiposity with plasma leptin concentrations.8,9 Weight loss, fasting, and starvation are known to induce reductions in leptin concentrations, whereas concentrations are increased with weight gain and hyperinsulinemia,.8–12 Changes in circulating leptin concentrations in pregnant women are generally consistent with changes in maternal fat stores and glucose metabolism.11 Maternal leptin concentrations are known to increase 2- to 3-fold above nonpregnant concentrations, with the peak occurring around 28 weeks of gestation.11 Results from clinical studies suggest that pregnancy-associated increases in maternal plasma leptin may result from an upregulation of adipocyte leptin synthesis in the presence of increasing insulin resistance and hyperinsulinemia in the second half of pregnancy.12 Importantly, investigators have shown that leptin directly affects whole body insulin sensitivity by regulating the efficiency of insulin-mediated glucose metabolism by skeletal muscle13 and by hepatic regulation of gluconeogenesis.14 Moreover, available evidence suggests that leptin exerts an acute inhibitory effect on insulin secretion.6
In animal models, leptin deficiency (ob/ob mice) and leptin resistance (db/db mice having a defective leptin receptor) lead to hyperphagia and decreased energy expenditure. Consequently, affected animals become obese and develop insulin resistance.15,16 However, in humans, as recently reviewed by Ceddia et al,6 investigations of whether leptin is a diabetogenic or an antidiabetogenic hormone has produced conflicting results. Data from a large epidemiological study showed that plasma leptin concentrations were positively associated with insulin resistance in men and nonpregnant women.17 Additionally, investigators have noted that chronic hyperglycemia is associated with reductions in leptin concentrations in the peripheral circulation.18 On balance, available data suggest a complex relation between leptin and glucose homeostasis in humans.
To date, only 2 teams of investigators19,20 have studied maternal leptin concentrations in women with pregnancies complicated by gestational diabetes mellitus (GDM), and these published results are conflicting. In addition to the conflicting results, available data do not allow for determining the extent to which possible alterations in leptin concentrations are the cause or consequence of the metabolic perturbations, such as hyperglycemia, that are intrinsic to GDM. Additionally, the magnitude of any possible association of GDM risk with varying concentrations of leptin was not assessed in either study.
We therefore used available information and plasma specimens from an ongoing prospective cohort study of women receiving prenatal care before 16 weeks of gestation to examine whether elevated maternal leptin concentrations, measured in early pregnancy, are independently associated with an increased risk of GDM.
MATERIALS AND METHODS
The population for the present analysis was drawn from participants of the ongoing Omega study, which is designed to examine maternal dietary and lifestyle risk factors of preeclampsia, gestational diabetes, and other adverse pregnancy outcomes. The study population is composed of women attending prenatal care clinics affiliated with Swedish Medical Center and Tacoma General Hospital in Seattle and Tacoma, Washington, respectively. Women who initiated prenatal care before 16 weeks of gestation were eligible for the study. Participants were invited to provide blood and urine samples and to participate in an in-person interview that took place during the 13th week of gestation, on average. Maternal and infant records were reviewed and data were abstracted. The procedures used in this study were in agreement with the protocols approved by the Institutional Review Boards of Swedish Medical Center and Tacoma General Hospital, respectively. All participants provided written, informed consent.
The study population for this report is from the first 1,000 participants who were enrolled in the Omega study during the period 1996–2000. During this period, 1,219 eligible women were approached and 1,000 (approximately 82%) agreed to participate. A total of 968 participants provided blood samples. Women found to have chronic hypertension (n = 45) and pregestational diabetes mellitus (n = 4) were excluded. Women who experienced a spontaneous abortion or who had an induced abortion were excluded (n = 22). Also excluded were those women for whom the outcome of pregnancy was unknown for any of the following reasons: moved, delivered elsewhere, or missing medical records (n = 46). Hence, a cohort of 851 women completed their pregnancies and was available for analysis.
The diagnosis of GDM was made by using the recently revised guidelines set forth by the American Diabetes Association.21 In our study settings, all patients are screened at 26–28 weeks of gestation by using a 50-g, 1-hour oral glucose screening test. Those patients who failed this screening test (postload glucose concentrations of 140 ng/dL or higher) were then followed-up within 1 or 2 weeks with a 100-g, 3-hour oral glucose tolerance test. Women were diagnosed with GDM if 2 or more of the 4 diagnostic 100-g, 3-hour glucose concentrations exceeded the following criteria21: fasting glucose exceeding 95 mg/dL, 1-hour postchallenge exceeding 180 mg/dL, 2-hour postchallenge greater than 155 mg/dL, and 3-hour postchallenge greater than 140 mg/dL. Plasma glucose concentrations were determined in certified clinical laboratories that used standardized enzymatic procedures. From this cohort, we identified 47 confirmed GDM cases and 776 women in whom GDM did not develop. Women with insufficient or illegible laboratory results (n = 28) were excluded from these analyses because a diagnosis of GDM could not be ruled out with certainty. Hence, a total cohort of 823 women was available for study.
From structured questionnaire and medical records, we obtained covariate information including maternal age, educational attainment, height, prepregnancy weight, reproductive and medical histories, and medical histories of first-degree family members. We also collected information on annual household income and maternal smoking before and during pregnancy. Prepregnancy body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared. Using information from maternal medical records, we also calculated maternal BMI at the time of blood collection (13 weeks of gestational age, on average).
Maternal nonfasting blood samples, collected in 10-mL Vacutainer tubes (Becton, Dickinson, and Company, Franklin Lakes, NJ) at 13 weeks of gestation, were frozen at −80°C until analysis. Plasma leptin concentrations were measured by using a enzyme immunoassay (Diagnostic Systems Laboratory, Inc, Webster, TX) with the intraassay and interassay coefficients of variation both less than 8%. All assays were performed without knowledge of pregnancy outcome.
Women were grouped according to tertiles determined by the distribution of plasma leptin concentrations among the entire cohort. We examined frequency distributions of maternal sociodemographic characteristics and medical and reproductive histories according to plasma leptin tertiles. The Spearman correlation coefficient was used to measure the closeness of a linear relationship between maternal plasma leptin concentrations and selected characteristics. We estimated the relative association between varying tertiles of plasma leptin and risk of GDM by using Stata 7.0 (Stata, College Station, TX). We fitted generalized linear models,22,23 using a log-link function, to derive risk ratios (RR) and 95% confidence intervals (CIs). Briefly, the log-link function (ie, log of the probability of GDM) allowed for the calculation of relative RRs, rather the calculations of odds ratios.22,23 We also explored the possibility of a nonlinear relation between plasma leptin concentrations and GDM risk by fitting a multivariable logistic regression model that implemented the generalized additive model method as previously described.24 We used S-Plus 6.1 (release 2) to perform these analyses (Insightful, Inc, Seattle, WA). In multivariable analyses, we evaluated linear trends in risk by treating the tertiles as a continuous variable.25 To assess confounding, we entered variables into a generalized linear model one at a time and then compared the adjusted and unadjusted RRs. Final generalized linear models included covariates that altered unadjusted RRs by at least 10%, as well as those covariates of a priori interest (eg, parity, prepregnancy BMI, BMI at blood collection, and family history of non–insulin-dependent diabetes mellitus). In an effort to examine the independent relationship between maternal plasma leptin concentration and GDM risk, we calculated the ratio of leptin to early-pregnancy BMI as a means of expressing individual leptin concentration per percent body fat,9 and repeated all analyses. All reported P values are 2-tailed, and all CIs were calculated at the 95% level.
The characteristics of women in the study cohort according to tertiles of plasma leptin concentrations are summarized in Table 1. Overall, participants were primarily non-Hispanic whites, well educated, and employed during pregnancy. Women with high plasma leptin concentrations (highest tertile 31 ng/mL or higher) were more likely to be multiparous and obese and to have a family history of non–insulin-dependent diabetes mellitus, as compared with women with low plasma leptin concentrations (lowest tertile 14.3 ng/mL or lower). Maternal age, marital status, educational attainment, smoking status during the index pregnancy, and gestational age at blood collection did not vary appreciably across tertiles of plasma leptin concentrations.
Maternal plasma leptin concentrations were highly correlated with prepregnancy BMI, and BMI in early pregnancy (Spearman correlation coefficients: r = 0.55, P < .001 and r = 0.61, P < .001, respectively). Plasma leptin concentrations were not correlated with maternal age (r = 0.01, P = .73), gestational age at blood collection (r = 0.03, P = .48), or infant birth weight (r = 0. 01, P = .76). Maternal leptin concentrations were approximately 48% higher in women in whom GDM subsequently developed (39.1 ± 3.0 ng/mL) as compared with women in whom the complication did not develop (26.4 ± 0.7 ng/mL) (t test: P ≤ .001).
As seen in Table 2, the risk of GDM increased across successively higher tertiles of leptin concentrations (unadjusted RRs: 1.0, 5.9, and 8.6; with the lowest tertile as the referent group, P value for trend < .001). Women with plasma concentrations of 31.0 ng/mL or higher (highest tertile) experienced a 8.6-fold increased risk of GDM (95% CI 2.6, 28.4) as compared with women whose plasma concentrations were 14.3 ng/mL or lower (lowest tertile). The risk remained, although attenuated somewhat, after adjusting for parity, and family history of non–insulin-dependent diabetes mellitus (RR 7.8; 95% CI 2.4, 25.8). When we included maternal prepregnancy BMI in the multivariable model, the leptin-GDM association was further attenuated (RR 4.9; 95% CI 1.4, 17.5), although it remained statistically significant. A similar association was seen when we adjusted for maternal adiposity by using the BMI at blood collection (ie, in early pregnancy) variable (RR 4.7; 95% CI 1.2, 18.0).
We next modeled the risk of GDM in relation to maternal plasma leptin concentrations expressed as a continuous variable using a generalized additive model. From these analyses, we noted an approximately linear relation between GDM risk and plasma leptin concentrations. Therefore, on the basis of this observation, we modeled plasma leptin concentrations expressed as a continuous variable in a generalized linear model. In this analysis, we noted that a 10-ng/mL increase in plasma leptin concentration was associated with a 20% increase in GDM risk (RR 1.2; 95% CI 1.0, 1.3) even after adjusting for parity, first-degree family history of non–insulin-dependent diabetes mellitus, and early pregnancy BMI.
Because the risk of GDM and plasma leptin concentrations both increase with increasing adiposity, it was necessary to more thoroughly evaluate the extent to which adiposity-adjusted leptin concentrations (ie, the ratio of plasma leptin to maternal BMI at blood collection, in nanograms per milliliter per kilogram per meter squared) were elevated among women in whom GDM subsequently developed as compared with those in whom GDM did not develop. We therefore repeated analyses using an alternate strategy to assess the independent association (independent of maternal adiposity) of maternal plasma leptin concentrations with subsequent GDM risk. As seen in Table 3, women with the highest adiposity-adjusted leptin concentrations (expressed as per unit BMI) experienced a 4.2-fold increased risk of GDM, as compared with women whose values that fell within the lowest third of the distribution (RR 4.2; 95% CI 1.7, 10.1). These results were consistent with those from multivariable logistic regression models that included maternal BMI (Table 2). Hence, using different strategies, we noted that the association between maternal leptin and GDM risk was independent of maternal adiposity.
In this prospective cohort study of pregnant women, we observed a statistically significant positive relation between plasma leptin concentrations in early pregnancy and incidence of GDM. Women with high plasma leptin concentrations experienced a 4.7-fold increased risk of GDM (95% CI 1.2, 18.0) as compared with women with concentrations of 14.3 ng/mL or lower. Women experienced a 20% increase in GDM risk for every 10-ng/mL increase in leptin. This positive association was independent of established risk factors of GDM such as prepregnancy BMI, BMI in early pregnancy, parity, and family history of non–insulin-dependent diabetes mellitus.
Our results are generally consistent with 1 of the 2 published studies that assessed maternal plasma leptin concentrations in the third trimester among women with pregnancies complicated by GDM. In their case-control study, Kautzky-Willer et al20 reported that maternal third-trimester plasma leptin concentrations were higher in cases than in controls (24.9 versus 18.2 ng/mL; P < .001). However, Festa et al,19 in an earlier case-control study, noted that maternal third-trimester leptin concentrations were statistically significantly lower in GDM cases as compared with controls after adjusting for possible confounding by BMI and insulin concentrations (19.4 versus 26.9 ng/mL; P = .001). There are several possible explanations why the existing studies are not in complete agreement, the most likely being differences in study design and the timing of maternal blood collection. Differences in the timing of maternal blood collection (ie, gestational age) may likely account for some of the variability seen across studies. Distortions from uncontrolled confounding secondary to whether blood samples were collected before, after, or during labor, and other maternal factors including whether women were receiving pharmacologic and dietary therapy after diagnosis and before blood was collected for leptin determination, may also have been present in many of the previous studies. Moreover, variations in population characteristics, such as the proportion of subjects with poor glycemic control after the GDM diagnosis, as well as dissimilar distributions of the severity of GDM could also account for some of the observed differences in study results. An important limitation of the two available studies merits discussion and consideration. Because of the cross-sectional case-control design of the studies (leptin concentrations were determined after the diagnosis of GDM), one cannot determine whether any observed alterations in plasma leptin concentrations preceded GDM, or whether the differences may be attributed to disease-related alterations in glucose metabolism.
In the current study, determination of maternal leptin concentrations that used plasma collected in early pregnancy served to clarify the temporal relationship between elevated maternal plasma leptin concentrations and subsequent risk of GDM. The high follow-up rate (greater than 95%) also minimized possible selection bias. Several limitations of the present study, however, merit discussion. First, a single measurement of leptin concentration is not likely to provide a time-integrated measure of maternal status during the index pregnancy. Longitudinal studies with serial measurement of maternal plasma leptin concentrations are needed to elucidate the mechanisms and pathophysiologic consequences of excess leptin synthesis and release during pregnancy. Studies are also needed to help determine the sources of excess leptin synthesis and release. Second, nondifferential errors in reporting of maternal prepregnancy BMI are likely to have occurred. Weights reported by participants in other studies, however, have been shown to be valid.26,27 Troy et al27 reported previously that women's self-reports of recalled weights at age 18 years were highly correlated (r = 0.87) with weights at age 17–21 years that were abstracted from nursing school records. Third, although we have adjusted for established and suspected risk factors of GDM, we cannot exclude the possibility of residual confounding from unmeasured covariates. Forth, our relatively small number of GDM cases hindered inferences from some of our analyses. Notably, the 95% CIs were wide reflecting the statistical imprecision of our RR estimates. Larger studies are needed to provide more precise RR estimates. Finally, the current study focused primarily on nulliparous women, and cohort members were predominantly well educated, white women. Therefore, caution must be taken when generalizing our results to other populations.
The increased risk of GDM with increasing maternal plasma concentrations of leptin is biologically plausible and is likely accounted for by diverse molecular and biochemical pathways in multiple tissues. Leptin has been shown to regulate peripheral glucose homeostasis through its actions in skeletal muscle and its effects on hepatic gene expression of the gluconeogenic enzyme, phosphoenolpyruvate carboxykinase.14 Additionally, leptin has been shown to directly modulate glucose handling in skeletal muscle, the most important target tissue for insulin in glucose metabolism, by promoting fatty acid oxidation.28–30 Apart from the insulin-adiposity link, investigators have postulated that leptin-induced insulin resistance may be secondary to a glucose flux via the hexosamine pathway.31
Hence, although the pathophysiology of hyperleptinemia in GDM and glucose intolerance in men and nonpregnant women is presently unknown, it is clear that leptin has numerous actions on target tissues and is involved in the regulation of several endocrine pathways.4,7 Although the biologic actions of leptin are primarily mediated through interactions with receptors expressed in the hypothalamus;15 leptin receptors are widely distributed across other tissues including the lungs, liver, kidney, pancreas, heart, and the placenta.32–35 This wide distribution of leptin receptors portends the peptide's diverse influence on neuroendocrine, cardiovascular, and reproductive functions. As reviewed in detail recently,7 leptin is correlated with a series of endocrine parameters including insulin, insulin-like growth factor, hemoglobin A1c, and sex-hormone–binding globulin. Further investigations should focus on the association of increased leptin with other early pregnancy markers and mediators of GDM risk. Increased knowledge from such studies may yield strategies for identifying women at highest risk of developing GDM. Moreover, results from such studies may help identify strategies for reducing the occurrence of GDM.
We have shown that leptin predicts the development of GDM independent of maternal BMI and other risk factors. The findings from this prospective study are generally consistent with the report by Kautzky-Willer et al.20 Our findings are also consistent with a much larger body of evidence from experimental, clinical, and epidemiological investigations which suggest that leptin is an important mediator of glucose homeostasis in humans.6 Taken together with the available literature, our results suggest that leptin may also play a role in mediating glucose metabolism in pregnancy. Measurement of leptin alone, or combined with the assessment of other risk factors, may help identify women at high risk of developing GDM.
1. Zhang Y, Proenca R, Maffei M, Barone M, Leopold L, Friedman JM. Positional cloning of the mouse obese gene and its human homologue. Nature 1994;372:425–32.
2. Halaas JL, Gajiwala KS, Maffei M, Cohen SL, Chait BT, Rabinowitz D, et al. Weight-reducing effects of the plasma protein encoded by the obese gene. Science 1995;269:543–6.
3. Masuzaki H, Ogawa Y, Sagawa N, Hosoda K, Matsumoto T, Mise H, et al. Nonadipose tissue production of leptin: leptin as a novel placenta-derived hormone in humans. Nat Med 1997;3:1029–33.
4. Considine RV, Caro JF. Pleiotropic cellular effects of leptin. Curr Opin Endocrinol Diabetes 1999;6:163–9.
5. Friedman JM. Leptin, leptin receptors, and the control of body weight. Nutr Rev 1998;56:38–46.
6. Ceddia RB, Koistinen HA, Zierath JR, Sweeney G. Analysis of paradoxical observations on the association between leptin and insulin resistance. FASEB J 2002;16:1163–76.
7. Al-Dahhri N, Bartlett WA, Jones AF, Kumar S. Role of leptin in glucose metabolism in type 2 diatetes. Diabetes Obes Metabol 2002;4:147–55.
8. Grunfeld C, Zhao C, Fuller J, Pollack A, Moser A, Friedman J, et al. Endotoxin and cytokines induce expression of leptin, the ob gene product, in hamsters: a role of leptin in the anorexia of infection. J Clin Invest 1996;97:2152–7.
9. Havel PJ, Kasim-Karakas S, Mueller W, Johnson PR, Gingerich RL, Stern JS. Relationship of plasma leptin to plasma insulin and adiposity in normal weight and overweight women: effects of dietary fat content and sustained weight loss. J Clin Endocrinol Metab 1996;81:4406–13.
10. Jequier E. Leptin signaling, adiposity and energy balance. Ann NY Acad Sci 2002;967:379–88.
11. Schubring C, Englaro P, Siebler T, Blum WF, Demirakca T, Kratzsch J, et al. Longitudinal analysis of maternal plasma leptin levels during pregnancy, at birth and up to six weeks after birth: relation to body mass index, skinfolds, sex steroids and umbilical cord blood leptin levels. Hormone Res 1998;50:276–83.
12. Laivuori H, Kaaja R, Koistinen H, Karonen S-L, Andersson S, Koivisto V, et al. Leptin during and after preeclamptic or normal pregnancy: its relation to serum insulin and insulin sensitivity. Metabolism 2000;49:259–63.
13. Cohen B, Novick D, Rubinstein M. Modulation of insulin activities by leptin. Science 1996;274:1185–8.
14. Rossetti L, Massillon D, Barzilai N, Vuguin P, Chen W, Hawkins M, et al. Short term effects of leptin on hepatic gluconeogenesis and in vivo
insulin action. J Biol Chem 1997;272:27758–63.
15. Tartaglia LA, Dembski M, Weng X, Deng N, Culpepper J, Devos R, et al. Identification and expression cloning of a leptin receptor, OB-R. Cell 1995;83:1263–71.
16. Maffei M, Halaas J, Ravussin E, Pratley RE, Lee GH, Zhang Y, et al. Leptin levels in human and rodent: measurements of plasma leptin and ob RNA in obese and weight-reduced subjects. Nat Med 1995;1:1155–61.
17. Donahue RP, Prineas RJ, Donahue RD, Zimmet P, Bean JA, De Courten M, et al. Is fasting leptin associated with insulin resistance among nondiabetic individuals? The Miami Community Health Study. Diabetes Care 1999;22:1092–6.
18. Moriya M, Okumura T, Takahashi N, Yamagata K, Motomura W, Kohgo Y. An inverse correlation between serum leptin and hemoglobin A1c in patients with non–insulin-dependent diabetes mellitus. Diabetes Res Clin Pract 1999;43:187–91.
19. Festa A, Shnawa N, Krugluger W, Hopmeier P, Schernthaner G, Haffner SM. Relative hypoleptinaemia in women with mild gestational diabetes mellitus. Diabetes Med 1999;16:656–62.
20. Kautzky-Willer A, Pacini G, Tura A, Bieglmayer C, Schneider B, Ludvik B, et al. Increased plasma leptin in gestational diabetes. Diabetologia 2001;44:164–72.
21. Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 2003;26(suppl 1):2002;S5–20.
22. Baker RJ, Nelder JA. The generalized linear interactive modeling system, release 3.77. Oxford, England: Numerical Algorithms Group; 1985.
23. Hardin JW, Hilbe J. Generalized linear models and extensions. College Station (TX): Stata Press; 2001.
24. Hastie TJ, Tibshirani RJ. Generalized additive models. London, England: Chapman-Hall; 1990.
25. Rothman KJ, Greenland S. Modern epidemiology. 2nd ed. Philadelphia (PA): Lippincott-Raven Publishers; 1998.
26. Willett WC, Stampfer MJ, Bain C, Lipnick R, Speizer FE, Rosner B, et al. Cigarette smoking, relative weight, and menopause. Am J Epidemiol 1983;117:651–8.
27. Troy LM, Hunter DJ, Manson JE, Colditz GA, Stampfer MJ, Willett WC. The validity of recalled weight among younger women. Int J Obes Relat Metab Discord 1995;19:570–2.
28. Ceddia RB, William WN, Curi R. Leptin increases glucose transport and utilization in skeletal muscle in vitro
. Gen Pharmacol 1998;31:799–801.
29. Muoio DM, Dohm GL, Tapscott EB, Coleman RA. Leptin opposes insulin's effects on fatty acids partitioning in muscles isolated from obese ob/ob
mice. Am J Physiol 1999;276:E913–21.
30. Ceddia RB, William W, Curi R. Comparing effects of leptin and insulin on glucose metabolism in skeletal muscle. Evidence for an effect of leptin on glucose uptake and decarboxylation. Int J Obes Relat Metab Disord 1999;23:75–82.
31. Mueller WM, Gregoire FM, Stanhope KL, Mobbs CV, Mizuno TM, Warden CH, et al. Evidence that glucose metabolism regulates leptin secretion from cultured rat adipocytes. Endocrinology 1998;139:551–8.
32. Hoggard N, Hunter L, Duncan JS, Williams LM, Trayhurn P, Mercer JG. Leptin and leptin receptor mRNA and protein expression in the murine fetus and placenta. Proc Natl Acad Sci U S A 1997;94:11073–8.
33. Chen SC, Kochan JP, Campfield LA, Burn P, Smeyne RJ. Splice variants of the OB receptor gene are differentially expressed in brain and peripheral tissues of mice. J Recept Signal Transduct Res 1999;19:245–66.
34. Kieffer TJ, Heller RS, Habener JF. Leptin receptors expressed on pancreatic beta-cells. Biochem Biophys Res Commun 1996;244:522–7.
35. Schulz S, Hackel C, Weise W. Hormonal regulation of neonatal weight: placental leptin and leptin receptors. Br J Obstet Gynaecol 2000;107:1486–91.