A case of an elderly patient with insulin-dependent diabetes and dementia receiving one basal insulin plus one bolus insulin injections a day for 6 months

Abstract

Multiple daily injections of insulin, referred to basal–bolus regimen, are generally essential in achieving glycemic control and preventing ketosis in insulin-dependent diabetes, such as type 1 diabetes (T1D). A 75-year-old man with T1D receiving basal–bolus insulin therapy exhibited symptoms of dementia after hospitalization due to pyelonephritis and failed to continue insulin self-injection. Given that his social and familial circumstances allowed insulin injection once a day during the morning, bolus insulin injections needed to be discontinued. Ketonuria was observed the day following discontinuation of bolus insulin. Although increasing the basal insulin dose (degludec) from 10 to 15 units improved ketonuria, his preprandial glucose levels increased to ≥ 500 mg/dL before lunch and ≥ 400 mg/dL before dinner. Hence, another bolus insulin injection was simultaneously added to the basal insulin dose before breakfast, which, subsequently, decreased his preprandial glucose levels to ≤ 220 mg/dL before lunch and ≤ 350 mg/dL before dinner. For half a year after discharge, ketonuria or hypoglycemia had not been detected. After 6 months, he was able to restart intensive insulin therapy with familial support. Hence, in cases where elderly patients with diabetes exhibit symptoms of dementia and can receive insulin injection once a day due to their social circumstances, short-term one basal plus one bolus insulin injections a day might be considered to prevent life-threatening diabetes complications among those who are insulin-dependent.

Introduction

Basal and bolus insulin therapy is required for patients with insulin-dependent diabetes, such as type 1 diabetes (T1D). Given that diabetes is an aging-associated disease, basal and bolus insulin therapy is sometimes hard to maintain due to cognitive decline [1]. Accordingly, the Japan Diabetes Society does not endorse the use of insulin or sulfonylurea among elderly patients with diabetes suffering from dementia when an effective alternative therapy exists [2]. Given that severe hypoglycemia causes cognitive dysfunction and vascular damages, target HbA1c levels for older subjects with diabetes, which range from < 7 to < 8.5%, are set according to risk for hyperglycemia, cognitive function, or activities of daily living [2]. However, no alternative treatment for subjects with insulin-dependent T1D suffering from dementia has yet been established. Here, we report a potential treatment for a patient with insulin-dependent diabetes who had developed dementia.

Case report

A 75-year-old male with T1D was hospitalized due to acute pyelonephritis. T1D was diagnosed at the age of 65 with his laboratory data showing an HbA1c of 10%, C-peptide index (C-peptide/fasting plasma glucose) of 0.38, and insulin autoantibody of 4.6% (normal range: < 0.4%), although his glutamic acid decarboxylase (GAD) antibody was undetectably low. He subsequently received insulin lispro, 13 units before breakfast, 9 units before lunch, and 4 units before supper, and 10 units of insulin degludec before bed. Ceftriaxone (4 g/day) was administered intravenously for 10 days as treatment for pyelonephritis while continuing the same dose of basal and bolus insulin as before hospitalization to maintain glycemic control. Preprandial glucose levels before admission ranged from 200 to 300 mg/dL. The patient was then transferred to the diabetes department after recovering from pyelonephritis at the emergency department to optimize his insulin regimen.

Examination revealed a body temperature of 36.6 °C, pulse rate of 67/min, respiratory rate of 18/min, and blood pressure of 122/67 mmHg. The patient was 156 cm tall and weighed 51.3 kg (body mass index: 21.1 kg/m²). Laboratory studies showed normal neutrophil and C-reactive protein levels for 8 days, while urinalysis showed negative ketone bodies. Serum and urine C-peptide levels were undetectably low during the glucagon stimulation test and 24-h urine test, respectively. The patient had no comorbidities associated with diabetes, such as retinopathy and kidney and liver dysfunction (serum creatinine, 0.81 mg/dL; aspartate aminotransferase, 18 U/L; alanine aminotransferase, 23 U/L). His blood glucose readings ranged from 125 to 248 mg/dL under the insulin regimen [lispro (13-9-4-0) and degludec (0-0-0-10)] (Table 1).

Table 1.

Time course of insulin therapy and blood glucose levels

	Day after admission	X	X + 1	X + 2	X + 7	X + 8	X + 9	X + 10	X + 11
Insulin therapy	Lispro (units) at each meal	13-9-4	None	None	None	None	13-0-0	13-0-0	13-0-0
	Degludec (units) before breakfast	10	10	10	15	15	15	15	15
		Blood glucose level (mg/dL)
Breakfast	Before	212	194a	278a	119a	220a	136a	191	204a
	3 h later	125	368a					189
Lunch	Before	128	519a	401a	586a	538a	179a	206	215a
	3 h later	193	≥ 600a					457
Dinner	Before	232	≥ 600a	411a	486a	438a	210a	340	349a
	3 h later	248	≥ 600a	≥ 600a	593a	488a	176a	416	410a
Ketone body on the next day
Urea ketone body		–		±	–			–
Total ketone body (µmol/L) (normal range < 130)		61			56			69
3-Hydroxybutyric acids (µmol/L) (normal range < 85)		32			43			53

^aBlood glucose levels were examined using Accu-Chek Aviva Nano; no mark, blood glucose levels were examined using the glucose oxidation method (chemical reagent and Glucose AUTO, and STAT GA-1160 analyzer; Arkray Inc.)

Apart from his symptoms of dementia, the patient was unable to perform insulin self-injection. His activities of daily living tended to decline before hospitalization, while his Mini-Mental State Examination (MMSE) score was 20 (cut-off, 24), 3 months after discharge. His wife could support his insulin self-injection only during the morning considering that she had to take care of their sick children during the day and was always exhausted at night. Moreover, they had no other family members that could provide support. Despite careful explanation to the patient and his wife about the risk for hyperglycemic coma when withdrawing insulin lispro injections before meals, she strongly wished to discontinue insulin injections except during the morning. Therefore, we attempted to discontinue insulin lispro injections before every meal, and only ten units of insulin degludec was administered subcutaneously in the morning. The following day, the patient’s fasting and preprandial glucose levels increased to ≥ 250 mg/dL and ≥ 400 mg/dL, respectively, and he developed ketonuria. Hence, the insulin degludec dose was increased to 15 units, which subsequently addressed ketonuria. However, his preprandial glucose levels increased to ≥ 500 mg/dL before lunch and ≥ 400 mg/dL before dinner. We, therefore, decided to simultaneously administer a combination of insulin lispro (13 units, the same dose provided at breakfast before admission) and degludec (15 units) before breakfast. As a result, his preprandial glucose levels decreased to ≤ 220 mg/dL before lunch and ≤ 350 mg/dL before dinner (Table 1). Given the absence of ketosis or ketonuria under the one basal plus one bolus insulin injections a day for 4 days and concerns regarding the decline in activities of daily living with longer hospitalization, the patient was subsequently discharged on day 13.

After discharge, his fasting blood glucose profile ranged from 125 to 390 mg/dL, perhaps due to irregular snacking and/or exercise. The patient would be unable to undergo diet and exercise therapy given that he forgets when he eats meals and showed behavioral symptoms, such as eating disorders, due to dementia. Assessment of his glucose profile using flash glucose monitoring 6 months after discharge (196–200 days after starting one basal plus one bolus insulin injections a day, Fig. 1) showed relatively low glucose levels (< 100 mg/dL) after breakfast. At that time, the patient’s HbA1c levels ranged from 9.5 to 11.0%, although no signs of hypoglycemia or ketonuria had been observed. Six months after discharge, his wife allowed us to resume intensive therapy instead of one basal plus one bolus insulin injections a day considering that she gradually overcame the difficulty in performing multiple daily insulin injections to her husband while caring for their sick children. After a year, the patient’s HbA1c levels remained under 9.4% with intensive insulin regimen [lispro (13-9-4-0), degludec (10-0-0-0); the same dose before hospitalization]. During the same period, the patient’s Dementia Assessment Sheet for Community-based Integrated Care System-8 items (DASC-8) score was 23, for which he was diagnosed with moderate or severe dementia.

Fig. 1.

Glucose levels measured using flash glucose monitoring 6 months after discharge. Insulin degludec (15 units) and lispro (6 units, optimized after discharge) were administered before breakfast. Minimum glucose levels were measured after the timing of insulin injection (breakfast)

Discussion

We herein described a case involving a patient with insulin-dependent T1D who could not perform insulin self-injection because of dementia. Moreover, his wife was only able to help him inject insulin once a day in the morning. Although one basal insulin injection a day could initially prevent ketosis, his glucose levels indicated severe hyperglycemia even before meals. Hence, we decided to initiate one basal plus one bolus insulin injections a day, which was continued for at least 6 months after discharge, to prevent hyperglycemia complications.

With the increasing number of elderly patients with diabetes, more patients with dementia who cannot perform insulin self-injection are expected. Diabetes is a significant risk factor for both vascular dementia and Alzheimer’s disease [3]. Patients with dementia experience hypoglycemic episodes more frequently than those without dementia, while severe hypoglycemic episodes result in the development of dementia [4]. Most elderly couples in Japan live their lives separate from other family members [5]. As such, many patients with diabetes have no family members to support them. Moreover, social support services, such as daily nurse visitations, impose substantial economic burden to the family. Administration and maintenance of insulin pump therapy for elderly patients with diabetes and dementia have often been difficult. Accordingly, one basal plus one bolus insulin injections a day to prevent hyper- and hypoglycemia complications is occasionally unavoidable among insulin-dependent patients with dementia, as with our patient, due to limited availability of social support.

Given an MMSE score below the cut-off value and his inability to recall insulin self-injections for 3 months after discharge, the patient was thought to have progressive dementia based on the fifth edition of Diagnostic and Statistical Manual of Mental Disorders [6]. In general, diagnosing subtypes of dementia at an early stage is often difficult [7] given that he might exhibit typical symptoms of vascular or Alzheimer’s dementia in the future. His brain magnetic resonance image did not show apparent abnormalities indicative of brain atrophy or cerebral infarction.

Given that the patient had a DASC-8 score of 23, he was placed into category III for patient background according to the glycemic targets for elderly patients with diabetes recommended by the Japan Diabetes Society/Japan Geriatrics Society Joint Committee [8, 9]. Accordingly, HbA1c levels < 8.5% were recommended for such patients with diabetes suffering from dementia. Although the current patient had shown poor glycemic control (HbA1c levels ≥ 9.0% under intensive insulin therapy), he had no microangiopathy or macroangiopathy except for dementia. Patients with T1D who have extremely poor glycemic control (HbA1c levels ≥ 9.7%) are at increased risk for all-cause mortality [10]. However, among patients with diabetes who had HbA1c levels ≥ 9.7%, glycemic control was not significantly correlated with mortality [11]. In this report, the patient had shown HbA1c levels ranging from 9.5 to 11% with the one basal plus one bolus insulin injections a day. Although the patient and his wife were advised to not eat snacks between meals, he sometimes indulged in them given his eating behavior disturbance due to vascular dementia [12]. During daytime, his wife had to bring her sick children outside, leaving the patient alone at home without any support. Such circumstances made us feel that further diet and exercise therapy would be difficult. Despite the relatively poor glycemic control, he had not experienced any severe hyperglycemia or hypoglycemia for 6 months. The basal plus bolus insulin injections once a day proved to be a convenient and safe therapy that improved the quality of life of the patient and his wife for 6 months. Eventually, his wife decided to reinitiate basal and bolus insulin therapy.

Post-breakfast hyperglycemia is generally observed in patients with diabetes, partly due to the overnight fast [13]. Therefore, one basal plus one bolus insulin injections a day before breakfast was able to prevent extreme hyperglycemia in our patient. The patient ingested almost the same amount of carbohydrates (56 g) during each meal, while a lunch box was prepared by his wife, while she was away during the daytime. Insulin therapy should be tailored according to each patient’s carbohydrate intake [14]. Some patients might need one-time insulin injection with the largest carbohydrate meal. Otherwise, such insulin therapy was inevitably tried before lunch or supper, not breakfast, according to the social support.

Moreover, co-formulation insulin degludec/aspart 70/30 (Ryzodeg^®) has occasionally been recommended for convenience [15] if patients need fewer doses of bolus insulin than this patient. This patient required 15 units of insulin degludec and 13 units of lispro; therefore, Ryzodeg^® would lead to more postprandial hyperglycemia and was thought to be inappropriate. The other types of insulin should be taken into consideration; short- or intermediate-acting insulin might have been tried in this patient, instead of rapid-acting insulin.

This patient experienced no episodes of liver and kidney dysfunction, while his endogenous insulin secretion could not be detected. Insulin therapy should be planned according to the patient’s endogenous insulin secretion and diabetes complications. In the present case, one basal plus one bolus insulin injections a day prevented ketosis for at least half a year. Nonetheless, we needed longer safety profile data from the patient. More case reports investigating the management of patients with insulin-dependent diabetes who are unable to perform multiple daily insulin self-injection are warranted. In general, patients with T1D must receive basal and bolus insulin therapy. It should be noted that that one basal and one bolus insulin injections a day for TID has yet to be established, and might contribute to the development of severe complications, such as diabetic ketoacidosis, stroke, dementia, or sarcopenia due to poor glycemic control [16–18]. Therefore, bolus and basal insulin injections once a day should only be employed when the condition of the patient permits such a treatment.

Moreover, as shown in Fig. 1, one basal plus one bolus insulin injections a day will lead to severe hypoglycemia if this patient fails to have a breakfast for any reason. When this insulin regimen is tried to patients, they will be informed on the risk of hypoglycemia and those around them should be taught how to deal with hypoglycemic episodes.

Conclusion

Daily one basal plus one bolus insulin injections could be an alternative short-term option for preventing ketosis. Considering the increase in the number of elderly patients with T1D, some of whom would be unable to perform insulin self-injection due to cognitive dysfunction, and bolus and basal insulin injections once a day may be unavoidable for such patients.

Compliance with ethical standards

Conflict of interest

YT has received honoraria for lectures from MSD K.K.; Ono Pharmaceutical Co., Ltd; Nippon Boehringer Ingelheim Co., Ltd; Takeda Pharmaceutical Co., Ltd; Mitsubishi Tanabe Pharma Corp.; Daiichi Sankyo Co., Ltd; Sanwa Kagaku Kenkyusho Co., Ltd; Eli Lilly Japan K.K.; Sanofi K.K.; Dainippon Sumitomo Pharma Co., Ltd; Novo Nordisk Pharma Ltd; and Astellas Pharma, Inc., and has obtained research support from MSD K.K.; Ono Pharmaceutical Co., Ltd; Nippon Boehringer Ingelheim Co., Ltd; Takeda Pharmaceutical Co., Ltd; Daiichi Sankyo Co., Ltd; Novo Nordisk Pharma Ltd; Eli Lilly Japan K.K.; Sanofi K.K.; and Dainippon Sumitomo Pharma Co., Ltd.

Informed consent

Informed consent was obtained from him and his wife for the case report.