The Meaning of Autoantibody Titers, Exercise, and Alcohol in a Thin 65-Year-Old Man Hospitalized for Cholangitis and Coincidental New Diabetes

T.R. is a thin 65-year-old white man who was admitted for acute choledocholithiasis and ascending cholangitis. The patient, who had not seen a physician for many years, also had symptoms of polyuria, polydipsia, and nocturia for 1 week before admission. While hospitalized, his blood glucose levels were 123-223 mg/dl. His weight was 154 lb (BMI 22.1 kg/m²). Inpatient blood pressure readings were 180-190/100-110 mmHg. He had no family history of diabetes, took no medications except occasional nonprescription analgesics, and exercised regularly. An eye examination 2 years previously was normal. He had no known kidney disease, peripheral neuropathy symptoms, or foot lesions.

Outpatient follow-up revealed random capillary blood glucose (CBG) levels of 227 and 139 mg/dl and a hemoglobin A_1c (A1C) of 6.5%. His normal weight and regular exercise regimen suggested the possibility of type 1 diabetes. However, islet-cell antibodies (ICAs) and GAD antibodies were negative. He also had hypercholesterolemia, with an LDL of 154 mg/dl. To avoid having to take diabetes medications, the patient increased his exercise and modified his diet by decreasing his carbohydrate intake.

T.R. monitored his CBG levels at home and returned to clinic after 2 weeks with a log and graph of his results. He noticed a significant increase in his blood glucose levels to > 200 mg/dl ∼ 1.5 hours after breakfast. He varied his breakfast meal from carbohydrates to protein but observed no change in his CBG levels. He did not have substantial postprandial hyperglycemia after other meals. He noticed that exercising in the morning before breakfast lowered his CBG levels. Additionally, he reported that his CBG would consistently fall after he drank a glass of white wine.

The patient's symptoms resolved after he controlled his diabetes with diet and exercise. He has no evidence of retinopathy or neuropathy.

1. Which type of diabetes is suspected in a thin older man with new-onset hyperglycemia in a metabolically stressful situation?

2. How would new-onset type 1 diabetes be ruled out in this patient?

3. What is the effect of cortisol on insulin resistance?

4. How does exercise affect hyperglycemia?

5. What effect does alcohol consumption have on blood glucose levels?

Diabetes is a metabolic disorder in which hyperglycemia is caused by abnormalities in insulin secretion or resistance. Insulin resistance in peripheral tissues and decreasing pancreatic β-cell insulin production characterizes type 2 diabetes. Autoimmune destruction of β-cells leads to complete absence of insulin secretion in type 1 diabetes and in latent autoimmune diabetes of adults (LADA), a slower autoimmune process.¹ 

This patient clearly met the American Diabetes Association criteria for diabetes, with two separate fasting plasma blood glucose tests ≥ 126 mg/dl or one casual plasma glucose level ≥ 200 mg/dl plus symptoms.²  However,it is difficult to distinguish between type 1 and type 2 diabetes in patients such as this.¹ Because T.R. is athletic and thin, he lacked the body habitus and lifestyle typically encountered in type 2 diabetes and had no specific risk factors for type 2 diabetes (race or family history).

In adult-onset type 1 diabetes, autoantibodies are directed against pancreatic β-cell antigens. Three clinically useful serum autoantibodies detected in type 1 diabetes are ICAs, insulin autoantibodies (IAAs), and antibodies to GAD. Patients may have any combination of ICAs, IAAs, and GAD antibodies. T.R. was negative for ICAs and GAD antibodies; he was not tested for IAAs.

Antibodies to GAD are predictive of progression to hyperglycemia in the absence of ICAs or IAAs and have been shown to have a positive predictive value of 50% and a negative predictive value of 97% for late insulin deficiency.³ Moreover, the presence of two or more autoantibodies is highly predictive of the development of type 1 diabetes.⁴ Nonetheless, some type 1 diabetic patients remain negative for all antibody types. In this case, because many adults with type 1 diabetes or LADA have a prolonged phase of declining β-cell function, only time will tell whether his disease progression will be more consistent with type 1 or type 2 diabetes. But in thin adults, clinicians should follow patients carefully,educate them about the symptoms of diabetic ketoacidosis, and be ready to start insulin if oral agents appear to be ineffective.⁵ 

The postbreakfast hyperglycemia reported by T.R. is probably related to increased insulin resistance in the morning (the “dawn phenomenon”). Cortisol is one of several counterregulatory hormones that increase glucose production and ultimately lead to a transient worsening of glycemic control.⁶ Cortisol levels are highest in the early morning and reach their lowest point in the late afternoon and evening. Growth hormone (also counterregulatory)also peaks in the early morning.

This patient found that exercising before breakfast decreased his hyperglycemia. Exercise improves insulin sensitivity in patients with insulin resistance. Increased insulin sensitivity associated with physical exercise was found to be especially beneficial for type 2 diabetic patients with impaired glucose tolerance.⁷ Additionally, there is a cumulative effect of transient improvements in glucose tolerance associated with each individual period of exercise. These long-term benefits are maintained if patients exercise at least once every 2 or 3 days.⁸ 

T.R. consistently reported hypoglycemia after drinking wine. Alcohol intake has been associated with hypoglycemia in type 1 diabetes. A randomized controlled study investigating delayed hypoglycemia found that moderate consumption of alcohol in the evening may predispose patients with type 1 diabetes to hypoglycemia after breakfast the next morning.⁹ 

Finally, given his BMI and presentation, it is clinically likely that T.R. may still have LADA or type 1 diabetes in “honeymoon” phase. Patients such as this should have regular (every 3-6 months) evaluations for rapidly progressive insulin deficiency.

• Autoantibody tests are helpful but not definitive methods of detecting type 1 diabetes.

• Early-morning insulin resistance (the “dawn phenomenon”) may affect blood glucose levels and is likely related to counterregulatory hormone secretion.

• Regular physical exercise is associated with increased insulin sensitivity and thus enhanced glycemic control in patients with type 2 diabetes.

• Moderate intake of alcohol may precipitate hypoglycemic episodes in patients with type 1 diabetes.