HOMA-IR (revision 23)
Old revision·19:36, 25 Nov 2025·ArchiveBot
| HOMA-IRDerived index | |
|---|---|
| Full name | Homeostatic model assessment of insulin resistance |
| Inputs | Fasting glucose and fasting insulin |
| Reference method | Hyperinsulinaemic-euglycaemic clamp |
| Topic infobox · conventions | |
HOMA-IR is an index of insulin resistance derived from a single fasting sample. In its original linear approximation it is the product of fasting glucose and fasting insulin divided by a constant, and it correlates moderately with clamp-derived measures of insulin sensitivity across populations.[1]
Its appeal is that it requires one blood draw rather than a several-hour clamp. Its weakness is that it is a model-derived approximation from two values, inherits the poor standardisation of insulin assays, and describes hepatic rather than peripheral insulin resistance most directly.[2]
The authors of the model have themselves written about its misuse, and the practical conclusion is that it tracks change within an individual on a consistent assay better than it classifies individuals against published thresholds.[2]
Calculation and the two forms
[edit]The original linear approximation is
\text{HOMA-IR} = (\text{glucose}\text{mmol/L} \times \text{insulin}\text{mU/L}) / 22.5
A revised non-linear computer model, HOMA2, was published later and gives different values, particularly at the extremes. The two are not interchangeable, and a figure quoted without saying which was used is ambiguous.[2]
Unit conventions add a second ambiguity: glucose in mg/dL requires a different divisor, and insulin in pmol/L a further conversion. A published cut-off carries its unit convention with it.[1]
Both forms assume steady-state fasting conditions, and neither has a meaning in anyone receiving exogenous insulin, since the insulin measured is then not the beta cell's output.
What it does and does not measure
[edit]| Property | Assessment |
|---|---|
| Correlation with clamp | Moderate at population level |
| Individual classification | Poor; wide overlap between groups |
| Tracking change within a person | Better, on a consistent assay |
| Compartment | Reflects hepatic insulin resistance most directly |
| Comparability between studies | Limited by assay and model version |
The population-versus-individual distinction is the crux. A moderate population correlation is compatible with substantial misclassification of individuals, and an index that performs acceptably as a group descriptor may perform poorly as a personal one.[2]
For that reason it appears widely in epidemiology and rarely in individual clinical decision-making.[1]
Behaviour under weight loss and incretin therapy
[edit]HOMA-IR falls with weight loss by any means, and the fall tracks the improvement in insulin sensitivity that accompanies reduced adiposity. This is a well-replicated finding and is not specific to any pharmacological class.[3]
Under incretin therapy, both inputs move: insulin secretion is amplified while the drug is present, and sensitivity improves with weight loss. The resulting index change is therefore a composite, and attributing it to a change in sensitivity alone is unsound.[2]
The index appears on many of the laboratory panels discussed in this field; see Baseline laboratory panel. Nothing on this wiki is medical advice.[4]
See also
References
- ^ a b c Matthews DR, Hosker JP, Rudenski AS, et al. "Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man." Diabetologia 28(7):412–419 (1985). PMID 3899825.
- ^ a b c d e Wallace TM, Levy JC, Matthews DR. "Use and abuse of HOMA modeling." Diabetes Care 27(6):1487–1495 (2004). PMID 15161807.
- ^ Drucker DJ. "Mechanisms of action and therapeutic application of glucagon-like peptide-1." Cell Metabolism 27(4):740–756 (2018). PMID 29617641.
- ^ American Diabetes Association. "Standards of Care in Diabetes." Diabetes Care 47(Suppl 1) (2024).