Insulin resistance is not one variable. A taxonomy of organ-, pathway-, dose-, time-, and context- specific insulin-resistance phenotypes

Insulin Resistance Is Not One Variable: A Taxonomy of Organ-, Pathway-, Dose-, Time-, and Context-Specific Insulin-Resistance Phenotypes

Insulin resistance (IR) is routinely treated as a single property of a person, often represented by fasting insulin, HOMA-IR, an oral-glucose-tolerance-derived index, or a clamp-derived measure. These quantities are not interchangeable: they capture different projections of a multidimensional physiological response system. This targeted mechanistic review argues that the term “insulin resistance” is scientifically underdetermined unless it specifies the affected tissue, physiological process or signaling branch, insulin exposure, time scale, metabolic context, and measurement method.

The paper develops a taxonomy of core insulin-resistance phenotypes: hepatic resistance of endogenous glucose production to insulin, skeletal-muscle resistance of glucose disposal, adipose resistance of lipolysis to insulin, and adipose storage/adipogenic failure—a process tightly coupled to insulin resistance but not identical to impaired insulin signaling. Regulatory and delivery phenotypes include brain insulin non-response and endothelial/microvascular insulin resistance. Extended and etiologic phenotypes include cardiac, renal, alpha-cell, ovarian, insulin-receptor-related, autoimmune type B, and lipodystrophy-associated syndromes.

The review distinguishes mechanistically independent lesions from flux-mediated propagation, common-cause coupling, compensatory feedback, and measurement-derived coupling. Human physiological, genetic, pharmacological, and disease counterexamples demonstrate why neither a universal “obesity causes IR” sequence nor a universal “IR causes obesity” sequence adequately describes the full phenotype space.

A tiered diagnostic architecture separates routine risk estimation from direct tissue localization. The review explains why universal cutoffs for HOMA-IR, Matsuda index, Adipo-IR, brain IR, or microvascular IR are not defensible without assay-, protocol-, and population-matched reference distributions. Direct subtype confirmation may require isotope-tracer clamps, multistep insulin exposure, glucose-disposal measurements, microvascular imaging, or intranasal-insulin neuroimaging.

Available therapies are mapped according to their primary target lesion and asymmetric transfer across other subtypes. Version 1.1 additionally operationalizes insulin resistance as a dynamic response surface, introduces multiplicity-aware subtype classification and assay-harmonization requirements, and presents a detailed development agenda for missing compartment-selective therapies.

This work is a targeted mechanistic review and proposed phenotyping framework. It is not a registered systematic review, peer-reviewed clinical guideline, or substitute for individualized medical care.