A 71-year-old woman with type 2 diabetes was referred to our hospital because of severe lipoatrophy throughout the whole abdomen induced by insulin therapy. The patient was diagnosed with type 2 diabetes when she was 63 years old and was treated with diet therapy only. When she was 69, her glycemic control became poor and insulin therapy was introduced (before breakfast, 20 units biphasic insulin Novolin 30R). Just after treatment began, she noticed that her abdomen gradually became atrophic, but she left it as it was. HbA1c levels were ∼8–9% (64–75 mmol/mol). She was treated only with biphasic insulin, and other antidiabetes agents were not used. Since her understanding about diabetes was poor, it was possible that she forgot to rotate the insulin injection site. When she was 71 years old, her glycemic control became worse and she was referred to our department and hospitalized. Marked subcutaneous fat atrophy was observed throughout the whole abdomen (Fig. 1A). Its appearance was similar to an ice cube tray. Abdomen computerized tomography showed a very severe loss of adipose tissue in the abdomen (Fig. 1B). There was no lipoatrophy elsewhere in the body. We diagnosed this subject with insulin-induced lipoatrophy because it appeared just after the introduction of insulin therapy and it started spreading from the site of insulin injection. HbA1c was 8.7% (72 mmol/mol), and the blood glucose level was 391 mg/dL. Insulin antibody was positive (73.6%), and total insulin level was 840 μU/mL. She had no diabetes complications. We stopped long-acting insulin and started ultra-rapid insulin NovoRapid (6 units before each meal). After the amelioration of glycemic control, we stopped insulin therapy and started an oral antidiabetes drug (metformin hydrochloride and nateglinide). Insulin lipoatrophy then gradually diminished (Fig. 1C). After the disappearance of lipoatrophy, HbA1c levels were gradually decreased to ∼7–8% (53–64 mmol/mol). Then glucagon-like peptide 1 receptor activator liraglutide was introduced, but lipoatrophy was not observed at all, suggesting that insulin lipoatrophy was not induced by the injection process.
A: Marked and ice cube tray–shaped insulin lipoatrophy throughout the whole abdomen in a subject with type 2 diabetes. B: Very severe loss of adipose tissue in the entire abdomen is shown with computerized tomography. C: Photograph of patient′s abdomen after the recovery of insulin lipoatrophy.
A: Marked and ice cube tray–shaped insulin lipoatrophy throughout the whole abdomen in a subject with type 2 diabetes. B: Very severe loss of adipose tissue in the entire abdomen is shown with computerized tomography. C: Photograph of patient′s abdomen after the recovery of insulin lipoatrophy.
Insulin-induced lipoatrophy is a subcutaneous fat atrophy at the site of insulin injection and is a rare side effect of insulin therapy. Lipoatrophy is not only a cosmetic problem but also extremely important because of the variability of insulin absorption. Once insulin lipoatrophy becomes apparent, it is very difficult to maintain good glycemic control. After the introduction of recombinant human insulin, such insulin lipoatrophy, as well as lipohypertrophy, has become extremely rare. Indeed, there are only a few reports about these conditions in association with human insulin (1–3). Therefore, it remains unclear in which situations insulin lipoatrophy is more easily induced, including type of diabetes, sex, ethnicity, and type of insulin.
To the best of our knowledge, insulin lipoatrophy in this subject is one of the most severe cases after the introduction of recombinant human insulin. Lipoatrophy is a loss of subcutaneous fat caused by insulin injection. The precise pathogenesis remains unclear, but possible mechanisms include immune reaction to insulin or its injection solution and injury from repeated local injections. Once insulin lipoatrophy is observed, it would be better to stop insulin injection or inject insulin elsewhere.
In conclusion, we should be aware of the possibility of such severe insulin lipoatrophy upon the initiation of insulin therapy even with recombinant human insulin.
Article Information
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. K.Ko. researched data and wrote the manuscript. H.K. and K.Ka. wrote the manuscript and reviewed and edited the manuscript. S.K., M.S., S.H., K.T., M.M., and T.M. researched data and contributed to discussion. H.K. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.