manipulating GDF15 action might disturb glucose homeostasis.
After a large postpartum reduction in circulating GDF15, there were no changes between 6 and 18 months after pregnancy; glucose metabolism followed a similar pattern. At 18 months after pregnancy GDF15 was positively associated with HOMA-B for all women and NW and, in OB, with HOMA-IR. The latter finding is in agreement with cross-sectional studies linking insulin resistance and GDF15 in nonpregnant OB17, 28. This association can be explained by the higher release of GDF15 in inflammatory conditions such as insulin resistance and obesity as discussed above.
The major strength of the current study is the longitudinal design comprising three time points during pregnancy and three time points after pregnancy, all involving identical study visits at a single center. Our subjects were well characterized through blood sampling and use of a reference method to measure body composition. However, oral glucose tolerance testing in addition to HOMA calculations would have provided a more extensive evaluation of glucose metabolism and insulin release. As in all such studies, additional limitations include the difficulty of proving causation in relationships, recruitment bias, and generalizability. With a focus on health and anthropometric measurements, recruited women are generally more inclined to keep healthy habits. However, relationships among the various factors measured should not be affected by this potential bias. A final limitation was the reduced statistical power resulting from the number of drop-outs in the OB group for the last visit at 18 months after pregnancy.
In conclusion, we found that the serum level of GDF15 increases throughout pregnancy, and does so to greater extent in NW than OB, and is associated with increased insulin secretory function in normoglycemic pregnancies. Adding to previous research in nonpregnant subjects, our findings suggest that GDF15 has a beneficial effect on beta cell function and may have implications for treatment to overcome insulin resistance . More mechanistic studies will be needed to confirm the effect of GDF15 on beta cells during pregnancy.
Acknowledgements We thank all the women who participated in the study, and registered dietician Evelina Järvinen for expert caretaking and measurements during study visits.
Ethical approval The study was approved by the Regional Ethical Review Board in Gothenburg (Dnr 402-08). All women received oral and written information about the study and gave informed written consent before enrolment.
Author contributions Conception, data acquisition, or data analysis and interpretation; UAH, PS, AH. Drafting or revising article for intellectual content; UAH, LJ, CM, AH. Final appoval of article; UAH, LJ, PS, CM, AH