First-trimester SPISE mediates the association between early-pregnancy body fat percentage change and gestational diabetes mellitus: an exploratory secondary analysis of a prospective cohort.

BACKGROUND

Gestational diabetes mellitus (GDM) is closely related to maternal adiposity and insulin resistance. The rate of body fat percentage (BFP) change in early pregnancy may describe estimated BFP gain before routine glucose screening. The Single-Point Insulin Sensitivity Estimator (SPISE) offers a fasting estimate of insulin sensitivity without insulin testing. It is still uncertain whether SPISE explains the link between early BFP change and later GDM.

METHODS

We analyzed public, de-identified data from a prospective Korean cohort of singleton pregnancies. BFP was estimated from age and body mass index. Early-pregnancy estimated BFP change rate was defined as BFP at 10-14 weeks minus pre-pregnancy BFP, divided by gestational week at assessment, and rescaled to per mille units. SPISE was calculated from fasting triglycerides, high-density lipoprotein cholesterol, and body mass index. GDM was assessed at 24-28 weeks with a two-step diagnostic approach. The analyses used restricted cubic splines, segmented regression, mediation models, and propensity score matching.

RESULTS

Among 586 women in the analytic cohort, 37 (6.3%) developed GDM. Higher estimated BFP change rate was linked to greater GDM risk (P overall = 0.006; P non-linearity = 0.140). SPISE showed a non-linear inverse association with GDM (P overall < 0.001; P non-linearity = 0.003), with an estimated threshold of 7.089. Below this threshold, each one-unit increase in SPISE was linked to lower GDM odds (OR 0.127, 95% CI 0.049-0.328; P < 0.001). Above the threshold, the association was weaker and not significant. estimated BFP change rate was also inversely associated with SPISE. In mediation analysis, the model-based estimated proportion mediated was 43.65%. After propensity score matching, available-case outcomes showed more GDM events in the low-SPISE group than in the high-SPISE group (14.5% vs 0.0%; P = 0.003).

CONCLUSIONS

Faster estimated BFP change in early pregnancy was linked to higher later GDM risk, and lower SPISE explained part of this relationship in statistical mediation models. The proposed SPISE threshold and mediation pattern need validation in independent cohorts before clinical use.