I was recently asked "What was the significance of insulin antibodies detected in an insulin-treated patient?"
In response to this question, it is worth mentioning that the insulin autoantibody assay does not discriminate antibodies from autoantibodies (IAA = insulin autoantibodies). Insulin antibodies are very common in insulin-treated patients. In children with new-onset type 1 diabetes, up to 60% may be IAA positive. In new-onset type 1 diabetes, IAA levels are fairly low (~<200 nU/mL; reference interval: <126 nU/mL). In insulin-treated patients, IAA may rise above 1000 nU/mL.
Considering all of the people that receive exogenous insulin, it is extremely rare that IAA ever interfere with insulin action. However IAA do interfere with insulin measurements, and therefore, we measure C-peptide as an index of beta-cell function in insulin treated patients.
Recognizing that IAA do not interfere with insulin action, there are rare cases of the "autoimmune hypoglycemia syndrome" where spontaneous IAA cause variable degrees of hypo or hyperglycemia . There is also a rare syndrome (that may NOT exist) where insulin is inactivated subcutaneously after it is injected .
 Winter WE, Hardt NS, Harris NS. Carbohydrate Disorders. In: Contemporary Practice in Clinical Chemistry, 2nd Edition. WA Clarke, DR (ed). AACC Press, Washington, DC, 2011; pp: 343-357.
 Soudan B, Girardot C, Fermon C, Verlet E, Pattou F, Vantyghem MC.Extreme subcutaneous insulin resistance: a misunderstood syndrome.Diabetes Metab. 2003 Nov;29(5):539-46.