Survey of Hypoglycemia in Elderly People With Type 2 Diabetes Mellitus in Japan

Abstract

Background: The number of elderly type 2 diabetes mellitus (T2DM) patients in Japan is increasing continuously. Hypoglycemia is a significant issue in their treatment. However, the actual situation and related details of their hypoglycemia remain unclear. In order to elucidate them, the Japan Physicians Association conducted a large-scale questionnaire survey for physicians and their outpatients all over Japan.

Methods: Targeted elderly T2DM outpatients were 65 years old or older in 2011. Specialized questionnaire survey forms were distributed to both of physicians and patients. The forms for physicians included questions whether patient had hypoglycemia in the last 1 month or 1 year; those for patients included whether they experienced it in the same durations and any of the 28 symptoms that are suggestive of hypoglycemia or pertaining to geriatric syndrome in the last 1 month, as well as questions about knowledge regarding hypoglycemia. We analyzed associations between hypoglycemia and the symptoms, and between hypoglycemia and medications.

Results: Of 15,892 T2DM patients (age, 74.2 ± 6.3 years; diabetes duration, 12.8 ± 8.9 years; HbA1c, 7.0±1.0%), dipeptidyl peptidase-4 inhibitor (DPP-4i) was the most prescribed medication among all oral hypoglycemic agents (OHAs). The frequencies of hypoglycemia in the last 1 month recognized by physicians and experienced by patients were 7.8% and 10.4% (P < 0.0001), and in the last 1 year were 15.5% and 21.1% respectively (P < 0.0001). The most common symptom was "weakness, fatigue/feeling languid" and the majority of all patients reported neuroglycopenic or autonomic symptoms. Regarding monotherapy, hypoglycemia was observed in 32.7% of the patients with insulin, 4% in sulfonylurea (SU), 3.8% in glinide, and 3.5% in pioglitazone. The questions asking knowledge about hypoglycemia revealed that SU or insulin users had significantly more knowledge of hypoglycemia than others (P < 0.001); however, 63% of patients using insulin, and 31% of patients using SU always carried glucose or a similar medication with them.

Conclusions: The present study suggested two types of "hidden hypoglycemia", one is that physicians did not detect and the other one is that patients were not aware. It is vital that physicians strive to prevent hypoglycemia by paying closer attention to symptoms of "hidden hypoglycemia" in their elderly patients.

Keywords: Countermeasure; Elderly; Hidden hypoglycemia; Hypoglycemia; Hypoglycemic symptom; Knowledge of hypoglycemia; Medication; T2DM.

Figure 1

Questionnaire of hypoglycemia. The items were made in reference to a chapter of hypoglycemia in MEDSi Joslin Diabetes.

Figure 2

Trends in antidiabetic drug prescriptions. (a) Proportions of patients who were taking any one of eight drugs (n = 15,855). Met: metformin; Pio: pioglitazone. (b) Patient SU dosage/day (n = 5,825). Conversion rate: glimepiride 1 mg = gliclazide 40 mg = glibenclamide 0.625 mg.

Figure 3

State of hypoglycemia.

Figure 4

Hypoglycemia in patients with the 28 symptoms. Patients with hypoglycemia commonly presented with weakness, fatigue/feeling languid, dizziness/lightheadedness, cognitive dysfunction/forgetfulness, back pain, and frequent urination, and cold sweats. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 5

Associations between hypoglycemia and the symptoms. Multiple logistic regression analysis of the incidence of hypoglycemia by doctors' records in the last 1 month and the 28 self-reported symptoms by patients is shown above. “Cold sweat” and “hunger” indicated a strong positive relationship with the incidence of hypoglycemia. On the other hand, we found some symptoms like “loss of hunger/fullness” showed negative association with the incidence of hypoglycemia.

Figure 6

Associations between hypoglycemia and drugs. A multiple logistic regression analysis was performed between the incidence of hypoglycemia in the last 1 month, as assessed by doctors, as the dependent variable to analyze the associations with drug use. SU: significantly high OR (odds ratio). Glinide, Met, Pio, αGI: not significant. DPP-4i: the lowest OR (< 1). Insulin: prominently high OR. GLP-1 and other drugs were excluded from this analysis due to the lower number of patients using the drugs.

Figure 7

Hypoglycemia in the patients using SU or insulin. Percent of incidence of hypoglycemia based on doctors’ records.

Figure 8

Knowledge of hypoglycemia. (a) “Are you aware of hypoglycemia?” Hypoglycemia awareness in elderly patients who were taking SU, insulin, or neither. (b) “What do you know about hypoglycemia?” Patients who chose “yes” to the first question answered this question. *P < 0.05, **P < 0.01, ***P < 0.001.