. - To examine associations of quantitative computed tomography-derived hepatic, visceral, subcutaneous, and pancreatic adiposity with an oral glucose tolerance test-derived insulin clearance surrogate in adults with type 2 diabetes.

. - This retrospective cross-sectional study included 166 adults not receiving exogenous insulin before admission. The molar C-peptide-to-insulin area-under-the-curve ratio from 0 to 120 minutes was used as the clearance surrogate. Multivariable regression was performed, with hepatic fat and visceral fat area prespecified as primary exposures and Holm correction applied to primary-model P values.

. - Each one-standard-deviation higher hepatic fat was associated with an 8.44% lower clearance surrogate (95% confidence interval, −12.69% to −3.98%; Holm-adjusted P < 0.001). Visceral fat area showed a directionally inverse association that did not meet the multiplicity-adjusted threshold (−6.13%; 95% confidence interval, −12.29% to 0.46%; Holm-adjusted P  = 0.067). The unconditional hepatic-fat association with insulin area under the curve did not meet the false-discovery-rate-adjusted threshold (adjusted P  = 0.066); the C-peptide-conditioned analysis was treated as exploratory. Restricted cubic spline analysis showed no evidence of nonlinearity ( P  = 0.380).

. - Higher quantitative computed tomography-derived hepatic fat was associated with a lower oral glucose tolerance test-derived insulin clearance surrogate in adults with type 2 diabetes. The observed pattern across adiposity measures is hypothesis-generating and does not establish that hepatic fat is a stronger correlate than other depots; prospective confirmation using direct or model-based measures of insulin clearance is needed.