Associations between GDF15 expression and beta cell function during and after pregnancy
Associations between GDF15 and pregnancy induced increases in insulin resistance and beta cell function from T1-T3 are shown in Table 2. In this model (adjusted for BMI, age and gestational age), absolute levels of GDF15 at T2 and T3 and increases in GDF15 between trimesters associated with the increase in HOMA-B (Table 2). For unadjusted correlations within groups, there were also significant associations between increases in HOMA-B and GDF15, although the timing differed slightly between groups. In NW, HOMA-B increase associated with GDF15 at T2 (R= 0.467, p = 0.016) and with GDF15 change T2 to T3 (R= 0.394, p = 0.046). In OB, HOMA-B increase associated with GDF15 change T1 to T2 (R= 0.664, p < 0.001) and with change in GDF15 T1 toT3 (R= 0.388, p = 0.046).
Fasting blood glucose decreased significantly late in pregnancy (p = 0.008). This decrease between T2 and T3 was inversely associated with changes in GDF15 from T1 to T2 and from T2 to T3 (Fig. 2).
Since GDF15 levels did not change significantly between time points after pregnancy (Fig. 1A), associations between measures of glucose metabolism and changes in GDF15 after pregnancy were not deemed relevant. Only associations for absolute values at 6, 12, and 18 months were investigated. GDF15 did not correlate with HOMA-IR or HOMA-B at 6 or 12 months. At 18 months, however, GDF15 was significantly associated with HOMA-B (beta= 0.301, p = 0.039) and HOMA-IR (beta= 0.304,p = 0.009) among all women after adjustment for maternal age and BMI. In subgroup analyses, GDF15 associated significantly with HOMA-B in NW (beta= 0.424 ,p = 0.031) and with HOMA-IR in OB (beta= 0.587,p = 0.048).