Discriminative value of serum adiponectin as a potential adjunctive marker for HOMA-IR-defined insulin resistance in women with unexplained recurrent spontaneous abortion

Insulin resistance (IR) is increasingly recognized as a modifiable contributor to unexplained recurrent spontaneous abortion (uRSA), yet its routine surrogate, the homeostasis model assessment of insulin resistance (HOMA-IR), requires fasting glucose and insulin measurement and therefore an overnight fast, which is inconvenient in routine reproductive-clinic workflows. HOMA-IR is itself a surrogate index rather than the gold standard for insulin sensitivity (the hyperinsulinaemic–euglycaemic clamp), and any downstream marker must be interpreted accordingly. Adiponectin is an insulin-sensitizing adipokine with low biological variability and limited post-prandial fluctuation. We evaluated whether serum adiponectin, measured in this study from fasting oral-glucose-tolerance-test samples, could serve as an adjunctive marker for HOMA-IR-defined IR in women with uRSA. We performed a single-centre retrospective observational study of women managed for uRSA at a tertiary maternal and child health hospital in Dongguan, southern China. All participants were of Chinese Han ethnicity. Serum total adiponectin was quantified by latex-enhanced immunoturbidimetry (Sekisui Medical Co., Tokyo, Japan; catalogue No. 17083) on a Hitachi 7180 automated analyser, with an intra-assay coefficient of variation (CV) of 3.2% and an inter-assay CV of 5.1%. Adiponectin was measured from the fasting blood sample obtained for the oral glucose tolerance test, and IR was defined as HOMA-IR ≥ 2.41. The primary analysis was conducted in the non-pregnant subgroup ( n = 346), with the full cohort ( n = 404, including 58 early-pregnant women) analysed secondarily. The relationship between adiponectin and HOMA-IR was assessed by Spearman correlation; adiponectin was compared between IR and non-IR groups using the Mann–Whitney U test. Discriminative performance was quantified by the area under the receiver operating characteristic curve (AUC), with the optimal cut-off identified by the Youden index. Incremental discrimination of adiponectin beyond age and body mass index (BMI) was evaluated by nested-model AUC comparison using the DeLong test. Sensitivity analyses were conducted with alternative HOMA-IR thresholds (2.0, 2.5 and 3.0). Multivariable logistic regression was used to adjust for age, BMI and sampling status. Among 404 women (346 non-pregnant, 58 early-pregnant; all Chinese Han), 203 (50.2%) met the criterion for HOMA-IR-defined IR. Serum adiponectin was significantly lower in the IR group than in the non-IR group (median 6.40 [interquartile range 5.20–8.70] versus 8.50 [6.90–11.10] µg/mL; P < 0.001) and was inversely correlated with HOMA-IR (Spearman ρ = −0.455, 95% confidence interval [CI] − 0.529 to − 0.374; P < 0.001), with a graded decline across ascending HOMA-IR quartiles (P for trend < 0.001). In the primary non-pregnant analysis, adiponectin yielded an AUC of 0.733 (95% CI 0.680–0.786); in the full cohort the AUC was 0.711 (95% CI 0.661–0.761). Adding adiponectin to a model containing age and BMI increased the AUC from 0.759 to 0.788 (DeLong P = 0.038). In the primary non-pregnant analysis, the optimal cut-off of ≤ 6.60 µg/mL provided a sensitivity of 58.3%, a specificity of 78.7%, a positive predictive value of 72.1% and a negative predictive value of 66.7%, with a positive likelihood ratio of 2.73. Sensitivity analyses with HOMA-IR thresholds of 2.0, 2.5 and 3.0 yielded consistent AUCs (0.697–0.724). Each 1 µg/mL decrease in adiponectin was independently associated with IR (adjusted odds ratio 1.17, 95% CI 1.06–1.29; P = 0.002). In this retrospective single-centre observational study, serum adiponectin measured from fasting samples showed a consistent inverse association with HOMA-IR-defined IR and moderate, specificity-weighted discrimination in women with uRSA. Adiponectin provided modest incremental discrimination beyond age and BMI. Although insufficiently sensitive to replace fasting indices, a low adiponectin concentration may increase clinical suspicion for IR and help triage women for confirmatory fasting assessment. These findings require prospective validation in independent cohorts before adiponectin can be recommended as a clinical triage marker; the study does not demonstrate that adiponectin predicts pregnancy prognosis or that adiponectin-guided management improves reproductive outcomes.