In the present study, we demonstrated that circulating serum ACBP is not significantly different in women with GDM compared to healthy pregnant controls. Furthermore, ACBP levels did not differ between these groups at a postpartum follow-up time point. However, our data show that ACBP is significantly decreased during pregnancy by ~ 50%, irrespective of GDM status. Moreover, HOMA2-B and creatinine remain positive and independent predictors of circulating ACBP after pregnancy.

Previous studies indicate that ACBP is increased in overweight and obese individuals [9, 12, 13]. Furthermore, rs2084202, a specific SNP in the promoter region of the splice variant ACBP1c, has been associated with a decreased risk for T2D [19]. In contrast, another study did not find any difference in circulating ACBP concentrations in individuals with diabetes and prediabetes compared to controls with normal glucose tolerance [20]. Whereas detailed studies on cardiometabolic associations exist in non-pregnant women, data on ACBP regulation during and after pregnancy is lacking.

In our cohort, pregnancy status itself was associated with significantly diminished circulating concentrations of ACBP in both groups, irrespective of GDM status. The exact underlying mechanisms for this, remain unclear, so far. Hypothetically, pregnancy could alter the tissue expression and synthesis of ACBP [21], for instance through placental-secreted factors, which may lead to increased serum concentrations after delivery. Interestingly, a differential regulation of glucose homeostasis-related cytokines between pregnant compared to non-pregnant populations have also been shown for the metabolically active cytokines proneurotensin [22], preadipocyte factor 1 [23], or sclerostin [24]. Importantly, when comparing seven different cytokines in pregnant participants from this cohort (both GDM and non-GDM) to non-diabetic, age and BMI-matched non-pregnant women, most of the investigated seven cytokines can discriminate only pregnancy status, but not GDM status. Thus, it is tempting to speculate whether systemically altered ACBP levels due to the pregnancy status itself is the cause of the non-significant difference between GDM and non-GDM pregnant women.

Apart from this, HOMA2-B as a marker of beta cell function was positively associated with ACBP levels after pregnancy in the total cohort in multivariate correlation analyses, indicating increased activity of beta cells with increasing ACBP concentrations. Furthermore, in our cohort during pregnancy, insulin levels and insulin resistance, quantified by HOMA2-IR, positively correlated with raised ACBP concentrations in univariate analyses. In line with this, in non-pregnant human cohorts, ACBP has been associated with elevated insulin levels in obese individuals, while starved mice injected with anti-ACBP antibodies exhibited a decrease in plasma insulin [9]. Moreover, it is known, that ACBP is linked to an adverse, insulin-resistance favoring lipid profile with increased TG and decreased HDL cholesterol [12,13,14]. Hence, the pathophysiological connection between ACBP and insulin resistance may lead to enhanced insulin secretion of beta cells in order to overcome this insulin resistant state.

In addition, ACBP levels positively and independently correlate with creatinine concentrations in the post-partum cohort, suggesting increased ACBP levels with impaired renal function. This is in accordance with a recent study from us demonstrating significantly increased ACBP concentrations in patients with kidney failure and acute kidney dysfunction [14].

In further correlation analyses, ACBP and BMI in the cohort during pregnancy, as well as in the post-partum cohort, are not associated, which contrasts earlier studies [9, 12, 13]. However, our correlation analyses for BMI and ACBP during pregnancy in fact refer to the BMI value prior to pregnancy. Therefore, it would be interesting to investigate in future studies, whether ACBP levels longitudinally during pregnancy correlate with actual weight gain during pregnancy. Importantly, body weight gain during pregnancy is not only due to fat mass increase, whereas ACBP has been closely linked to biochemical pathways in adipose tissue in non-pregnant cohorts. Thus, associations of ACBP with fat mass (for instance measured by bioelectrical impedance analysis) during pregnancy might detect a link between ACBP and adipose tissue mass more comprehensively than pre-gestational BMI. In accordance with this hypothesis, anti-ACBP antibodies reduce absolute fat mass under a high-fat diet in mice, and periumbilical fat expresses high levels of ACBP mRNA that diminishes upon dietary intervention in patients with obesity [9].

Some limitations of our study need to be emphasized: First, our study has been performed in a prospective cross-sectional design, and, therefore, causality cannot be established. Moreover, follow up investigations were performed only in 82 women out of the 148 initial study participants resulting in a reduced power for longitudinal assessments. To confirm our results, future studies should aim to analyze a full female cohort during and after pregnancy, respectively. Furthermore, it would be interesting to study ACBP levels also at an early postpartum time point in order to validate a potential placental effect on circulating ACBP levels during pregnancy.

In conclusion, ACBP is not a diagnostic marker for GDM, but ACBP is decreased during pregnancy, irrespective of GDM status. Furthermore, ACBP is closely linked to beta cell function and renal markers in post-partum women.