Download Full Text (471 KB)



Training Level

Resident PGY 3


Henry Ford Hospital


Purpose: The purpose of this study is to assess whether radiologists can detect type 2 diabetes in patients based the on the appearance of the deltoid muscle on shoulder ultrasound from other populations, such as obese non-diabetics. Materials and Methods: This study includes retrospective review of 137 shoulder ultrasounds from type 2 diabetic patients, confirmed by hemoglobin A1c levels and medication review. It also includes 49 shoulder ultrasounds from non-diabetic obese patients based on body mass index. Images of the deltoid muscle from these 186 ultrasounds are blindly reviewed by three musculoskeletal radiologists as to whether the appearance is normal, suspected diabetic or definite diabetic. These results along with the patient’s age, sex, race, hemoglobin A1c level, body mass index, and the presence of insulin use are analyzed. Results: A consensus diagnosis of ‘definite diabetic’ by 3 musculoskeletal radiologists based on increased echogenicity of the deltoid muscle on ultrasound is a powerful predictor of diabetic status. The positive predictive value for the accurate designation of ‘definite diabetic’ is 89% (70 out of 79 diabetic patients). Increased echogenicity of the deltoid muscle is a similarly powerful predictor of pre-diabetes. Of 13 pre-diabetic ultrasounds reviewed, 13 were assigned either ‘suspected diabetic’ (3 of 13, 23%) or ‘definite diabetic’ (10 of 13, 77%) (P= 0.062). Obesity alone [causing adipose infiltration of the deltoid muscle] cannot solely explain the increased echogenicity of the deltoid muscle of diabetic patients. Non-obese diabetics were diagnosed ‘definite diabetic’ with 30% sensitivity (11 of 37 non-obese diabetics). Diabetic patients with higher BMI, though, were more often diagnosed ‘definite diabetic’. Of 137 diabetic patient ultrasounds reviewed, 31(22.6%) were designated ‘normal’ (BMI 30.9 ± 7.3), 36 (26.2%) were designated ‘suspected diabetic’ (BMI 32.6 ± 6.9), and 70 (51.2%) were designated ‘definite diabetic’ (BMI 37.5 ± 8). Conclusion: The more echogenic appearance of the deltoid muscle in type 2 diabetics on shoulder ultrasound is a strong predictor of type 2 diabetes. The more echogenic (bright) deltoid muscle of type 2 diabetics can be differentiated from deltoid muscle of obese non-diabetics. We, therefore, suggest that the brightness of the deltoid muscle in type 2 diabetics is not related to adipose infiltration into the muscle, as is the case for obese individuals. Another mechanism, perhaps decreased intramuscular glycogen secondary to insulin resistance, may be the cause. Importantly, we also conclude that this distinct appearance on shoulder ultrasound may be used to detect early insulin resistance in pre-diabetics and identify under-treated or undiagnosed type 2 diabetics presenting with shoulder pain. This could present a unique opportunity for many physicians across many specialties to improve patient care.

Presentation Date


The Echogenic Appearance of the Diabetic Deltoid Muscle on Shoulder Ultrasound: Is This Simply from Adipose Tissue Infiltration, Can This Appearance Predict Type 2 Diabetes and Be Used to Detect Pre-Diabetes?