In considering the subject matter for my next blog, my mind wandered a bit and settled on a patient with type 2 diabetes I had seen last week. While riding her bicycle, she had suddenly felt dizzy and disoriented and side swiped a parked car. She was lucky–as a result of the fall she had some serious abrasions and a few stitches, but no broken bones. Her husband, who was riding beside, immediately noted her confusion, realized she was hypoglycemic, and gave her  juice and crackers he had in his knapsack.

Now I must say I felt partly to blame given that her A1C had been riding around 5.9-6.3% recently. We had several frank discussions in the previous few visits about backing off her insulin given the risks involved, but she had refused to significantly back off given she was ”doing just fine.“ I had touched on some of the new study results such as ACCORD and ADVANCE and how they may or may not apply to her, but at the end of the day she did not want to back down significantly.

As I considered this patient, I wondered how many other episodes of hypoglycemia she had suffered in the past few years. Clearly she did not complain of them and did not pick them up on her faithful self monitoring of blood glucose. But she must have had them. One does not get down to such a low A1C without them.

So my mind wandered back to this blog. Hypoglycemia. What adverse events are clearly associated with it? How solid is the evidence about these associations? How can we translate this evidence into better practice at the bedside? Can pharmacotherapy of the future eliminate hypoglycemia (or at least significantly reduce its incidence)? A beta cell website is incomplete without keeping our eye focused on this issue!

Awareness of hypoglycemia as an important consideration was increased this year in light of landmark trials such as ACCORD. The putative mechanism (and theories) for the surprising negative outcomes found with ACCORD revolve around the causal role of hypoglycemic episodes. I do not intend to rehash ACCORD—as this has been done quite frequently, and well, in the past year. Instead I wondered: What other insights can be gleaned from the literature of the past year relative to the effects, symptoms, and treatment of hypoglycemia? What new evidence is out there that may not be all that familiar to the average primary care doctor? So, to address these questions I decided to go on a literature search and present to you some interesting findings of my journey through the literature of the past year (in no particular order of importance or interest)….

1. Hypoglycemia may lead to increased inflammation and platelet aggregation. In a well-done study,[1] serum markers of inflammation (CD40 expression) were significantly elevated in patients with type 1 diabetes during hypoglycemia and returned to normal with euglycemia. Additionally, using flow cytometry, researchers found increased platelet-monocyte aggregation with hypoglycemia. Is this one of the links between hypoglycemia and increased cardiac endpoints?
2. Repeated episodes of hypoglycemia can increase intimal thickening, markers of inflammation, and risk of hypoglycemia unawareness. In a small, case-control study by Gimenez et al.[2] patients with type 1 diabetes were separated into those who had self-reported histories of recurrent hypoglycemic episodes and controls (without reported hypoglycemia). Patients with recurrent hypoglycemia not only displayed hypoglycemia unawareness but had less flow-mediated brachial dilation and increased femoral intima-media thickness (both markers of vascular disease) as well as increased baseline markers of inflammation (leuokocytes, von Willebrand factor, fibrinogen, and soluble intercellular adhesion molecule—all building on point #1 above.
3. Administrative data show the link between hypoglycemic events and cardiovascular endpoints. In an elegant look at HEDIS administrative data for over 860,000 patients with type 2 diabetes,[3] researchers found an independent association between ICD-9-coded hypoglycemic outpatient events and subsequent admission for cardiovascular event (acute myocardial infarction, coronary artery bypass grafting, revascularization, percutaneous coronary intervention, and incident unstable angina). Coded hypoglycemic events occurred in 3.1% of patients. The odds ratio of 1.79 (79% more likely) held up after adjustment for important confounding variables, including age, sex, geography, insurance type, comorbidity scores, cardiovascular risk factors, diabetes complications, total baseline medical expenditures, and prior acute cardiovascular events.
4. Caffeine intake may increase the symptoms of hypoglycemia. So if hypoglycemia has serious adverse affects on health, is there anything we can do to increase the warning signs? A study by Watson et al.[4] randomized type 1 diabetic patients to placebo or daily 200 mg caffeine tablets. Patients given caffeine tablets experienced more warning signs and symptoms of hypoglycemia than those on placebo.
5. Caffeine may decrease risk of exercise-induced hypoglycemia. In a small study,[5] athletes with type 1 diabetes  on basal bolus insulin were given either placebo or caffeine. Those given caffeine were less likely to develop hypoglycemia requiring glucose rescue during stationary bicycle riding up to a VO2 max (maximal oxygen uptake) of 70%. This proof of concept study needs to be replicated with lower doses of caffeine and more patients, but watch for this concept in the future as a possible strategy for your athletic type 1s.
6. Hypoglycemia is common! It has become conventional wisdom to liberalize A1C targets for elderly patients with type 2 diabetes. The intent of this paradigm is to protect the patient from hypoglycemia given the risk/benefit ratio. A recent study[6] of elderly (mean age 75) type 2 diabetic patients (with 93% using insulin) found that despite higher A1Cs (mean 9.3%) elderly patients still had a high incidence of hypoglycemia. After continuous glucose monitoring for 3 days, investigators found 65% of patients had a blood sugar less than 70mg/dl. Almost half of those patients had blood sugars less than 50mg/dl.   Alarmingly, 93% of hypoglycemic events were not detected by finger stick glucose monitoring performed 4 times a day or with symptoms. We may therefore need to do more than liberalize the A1C target to protect the elderly from hypoglycemia.
7. Skittles to the rescue! On the other end of the age spectrum, researchers found that sucrose was equally effective as glucose at increasing blood glucose.[7]  A cohort of children with type 1 diabetes were given BD Glucose tablets, sucrose (Skittles), or fructose (Fruit to Go). The Skittles increased blood sugar after hypoglycemic events to the same extent as the more expensive BD glucose product. Fruit-based treatment did not perform as well. Skittles thus may represent a more economical way to manage hypoglycemic episodes.
8. Severe hypoglycemia may permanently affect cognitive function. In a study by Asvoid[8] a cohort of children with type 1 diabetes were followed for 16 years. Patients who had a significant episode of hypoglycemia (defined as seizure or loss of consciousness) scored worse with regard to problem solving, verbal functioning, and psychomotor efficiency. This builds on our understanding of the neural affects of hypoglycemia. Separately it is also apparent that for adults even one episode of hypoglycemia severe enough to cause an ER visit or hospitalization is associated with an increased risk of dementia.[9]
9. Surprise! Insulin causes more hypoglycemia than other diabetes medications. In a network analysis[10]  involving 18 trials and over 4500 patients, potential options for add-on therapy to metformin and a sulfonylurea were compared with respect to relative impact on A1C, weight, and hypoglycemia. The trials were short (all 52 weeks or less). All agents reduced A1C to the same extent, but insulin caused the most weight gain and hypoglycemia relative to other agents (acarbose, DPP4 inhibitors, GLP1 agonists, and TZDs). The only surprise: insulin was not more effective in lowering A1C—perhaps due to hypoglycemia limiting up titration of dose?
10. Hypoglycemic symptoms present with poor intra-observer consistency. In a novel study,[11] researchers had 59 patients with diabetes (mean age 65, 77% Type 1) record the symptoms associated with every hypoglycemic event during a 9-12 month period. While men had more consistency than woman, all subjects showed tremendous amount of intra-subject variability in hypoglycemic symptoms recorded (the presentations often varied from episode to episode). This important observation should be kept in mind when teaching patients or counseling patients about hypoglycemic symptoms.  

As I move forward and prepare for my next diabetic patient, I must look to balance the clear benefits of A1C reduction with an increasing understanding of the risk of hypoglycemia. I owe it to my patients to individualize therapy in an effort to reduce harm and maximize gains. We all need to search for this sweet spot between efficacy and safety.