Usage of continuous glucose monitoring (CGM) for detecting an unrecognized hypoglycemia and management of glucocorticoid replacement therapy in adult patients with central hypoadrenalism

Abstract

Patients with adrenal insufficiency require appropriate glucocorticoid replacement therapy; however, reliable biological parameters for optimizing glucocorticoid supplementation are limited. The physician has to rely primarily on clinical judgment, carefully taking into account signs and symptoms potentially suggestive of over- or under-replacement. We have found that some patients who are viewed as receiving sufficient doses of glucocorticoids occasionally exhibit morning headache or morning discomfort, which may be caused by unrecognized nocturnal hypoglycemia. Our aim in this study was to evaluate the usefulness of continuous glucose monitoring (CGM) for detecting unrecognized hypoglycemia and optimizing glucocorticoid replacement therapy in adult patients with central hypoadrenalism. Six patients with central hypoadrenalism of various etiologies were included in this study. All patients exhibited occasional morning headache or discomfort. We performed CGM to measure plasma glucose levels in all patients, and CGM identified unrecognized hypoglycemia episodes at midnight and early in the morning in five patients (83%). The CGM findings were used to fine-tune the dosing and regimens of glucocorticoid replacement and to re-evaluate glucose levels to avoid further unrecognized hypoglycemic events. This optimization of hydrocortisone supplementation prevented additional nocturnal hypoglycemia incidences in all cases. The addition of L-thyroxine with hydrocortisone continued to provide favorable glycemic control. Occasional symptoms also improved after maintenance in all patients. These findings demonstrated that CGM may represent a powerful tool for identifying unrecognized hypoglycemia and for optimizing supplementary hormones in patients with central hypoadrenalism, thereby improving their quality of life.

ADRENAL INSUFFICIENCY is a life-threatening endocrine emergency caused by primary adrenal failure or hypothalamic-pituitary impairments in the corticotropic axis. Both types of adrenal insufficiency require daily glucocorticoid replacement therapy. The main symptoms of adrenal insufficiency include general fatigue, extreme weakness, anorexia, body weight loss, nausea, and vomiting. Children, but rarely adult patients, are also prone to hypoglycemia [1-3]. Guidelines and consensus statements for the diagnosis and treatment of primary adrenal insufficiency have recently been reported [1, 2, 4]. The diagnosis and management of adrenal insufficiency secondary to hypothalamic-pituitary failure have also been discussed in an expert review [3]. The typical management treatment consists of administration of a short-acting hydrocortisone replenished two or three times daily, with a half or two-thirds of the total daily dose given in the daytime to mimic the physiological diurnal secretion of cortisol [1-4].

The management of optimal glucocorticoid replacement therapy in adrenal insufficiency, and particularly in secondary adrenal insufficiency, continues to be a challenge because of the lack of reliable biological parameters for its assessment. Consequently, assessment is based primarily on clinical judgment, with careful accounting of signs and symptoms potentially suggestive of over- or under-replacement [1-3]. Overreplacement has recently been implicated as the cause of the increased mortality recently reported in hypoadrenal patients [5, 6], mainly due to metabolic problems [7]. Other adverse effects may disrupt bone metabolism [8] and quality of life [9, 10]. Therefore, the administration of lower glucocorticoid doses (15–25 mg/day hydrocortisone) is currently recommended [1-4]. However, these glucocorticoid replacement regimens still do not eliminate nocturnal hypocortisolemia [11].

Fasting hypoglycemia is a common feature in children with adrenocorticotropin (ACTH)/adrenal insufficiency and can affect neurocognitive development. It may be caused by the lack of glycogenic effects of cortisol [12, 13]. Nocturnal hypoglycemia in patients who also have type 1 diabetes mellitus (DM) can result in physical fatigue, morning headache, and extreme weakness the next day [14]. Asymptomatic hypoglycemia events can be detected in patients with DM by continuous glucose monitoring (CGM), a sophisticated method for measuring interstitial glucose concentrations day and night [15]. A consensus statement has recently supported the usefulness of CGM for DM patients [16], and its usefulness was also examined in newborn babies at risk of hypoglycemia [17] and in children with other hypoglycemic disorders [18]. CGM was recently used successfully to identify nocturnal hypoglycemia in adult patients with Addison’s disease [19]. However, the usefulness of CGM for identifying unrecognized hypoglycemia in patients with central hypoadrenalism has yet to be evaluated.

In our previous study, we reported that some patients with central hypoadrenalism who were considered to be receiving appropriate hydrocortisone replacement were occasionally exhibiting morning headache and/or discomfort in routine clinical visits in the absence of acute stressful conditions. We speculated that unrecognized nocturnal hypoglycemia could be causing these occasional symptoms in these patients. Therefore, in the present study, we investigated the usefulness of CGM for identifying unrecognized hypoglycemia in patients with central hypoadrenalism who characteristically exhibit occasional morning headache and/or discomfort. Our overall aim was to optimize hormone replacement therapy to reduce their hypoglycemic events and improve their quality of life.

Subjects and Methods

This study was approved by the Gunma University Institutional Review Board and all patients provided written informed consent. Patients with complications like DM or impaired glucose tolerance were excluded. Eleven adult patients with central hypoadrenalism and exhibiting occasional morning headache or discomfort were enrolled from our outpatient clinic between June 2014 and December 2016. The diagnosis of adrenal insufficiency was confirmed by an ACTH stimulation test, as described [3-5]. Following an intravenous administration of synthetic ACTH, 91% of these patients (10 out of 11) showed peak plasma cortisol levels of less than 20 μg/dL. On admission, symptoms and physical findings, such as body mass index (BMI) and blood pressure (BP), were assessed. Blood tests and urinalyses were performed, including plasma ACTH (at fasting status, before taking daily hydrocortisone supplementation), serum cortisol (at fasting status, before taking daily hydrocortisone supplementation), fasting plasma glucose (FPG), hematocrit (Hct), hemoglobin (Hb), white blood cell count (WBC), serum Na (s-Na), serum K (s-K), creatinine kinase (CK), total cholesterol, triglycerides (TG), high density lipoprotein (HDL)-cholesterol, low density lipoprotein (LDL)-cholesterol, and daily urinary free cortisol concentrations. The CGM systems used were MiniMed CGMS-Gold (Medtronic, Northridge, CA) and iPro2 (Medtronic, Northridge, CA). Patients underwent CGM for 24–48 h, including at least one night before and after the modification of glucocorticoid replacement regimens in their hospital stay with regular meals (2,000 kcal/day). We served breakfast at 0730 h, lunch at 1200 h, and dinner at 1800 h daily. Hypoglycemia was defined in this study as glucose levels less than 70 mg/dL. Among all 11 patients who underwent CGM analysis, several cases experienced hypoglycemic events that were not caused by adrenal insufficiency but for other reasons, such as reactive hypoglycemia. We excluded those cases from our analysis and ultimately analyzed six cases. All group data were expressed as the mean ± standard deviation (SD). Group comparisons in standard or non-standard distributions were performed by ANOVA, followed by the Student’s t-test or the Wilcoxon rank-sum test, respectively, using JMP 5.1.2 software (SAS Institute Inc.). All tests for significance and the resulting p values were two-sided, with a level of significance of 5%.

Results

Clinical features of patients

Table 1 shows the clinical characteristics of the enrolled patients. The six index cases had various etiologies of central hypoadrenalism: one patient had lymphocytic hypophysitis, one had hypopituitarism caused by pituitary apoplexy, two had hypopituitarism after diencephalon surgery of craniopharyngioma, one had panhypopituitarism due to De Morsier syndrome, and one had isolated ACTH deficiency (IAD). All examined cases were enrolled in this study because they characteristically exhibited occasional headache and/or discomfort in the morning on their routine clinical visits in the absence of any acute stressful conditions. The mean age of the six cases was 42 ± 23 years and the age of disease onset varied from 2 to 67 years. Three patients were treated with hydrocortisone at a dose ranging from 5 to 30 mg/day.

Table 1 Baseline characteristics of patients

Case	Sex	Age (yrs)	Diagnosis	Disease Onset (yrs)	Symptoms (occasionally)	BMI
1	M	71	Lymphocytic hypophysitis	67	morning headache	27.8
2	M	63	Hypopituitarism after pituitary apoplexy	62	morning headache	20.9
3	M	29	Hypopituitarism after diencephalon surgery	5	morning headache	27.0
4	F	18	Hypopituitarism due to De Morsier syndrome	2	morning headache	19.1
5	M	20	Hypopituitarisym after surgery for pineal tumor	13	body weight loss morning discomfort	15.0
6	F	53	Isolated ACTH deficiency	45	morning discomfort	17.8
Mean ± SD		42 ± 23				21.3 ± 5.1

BMI, body mass index; F, female; M, male

Table 2 shows the physical and laboratory findings of these patients, including systolic/diastolic BP, FPG, plasma ACTH (at fasting status, before taking daily hydrocortisone supplementation), serum cortisol (at fasting status, before taking daily hydrocortisone supplementation), WBC, percentage of neutrophils (Neu) and eosinophils (Eo), s-Na, and s-K before and after modifying hormonal supplementation. The daily urinary free cortisol level, which is regarded as another referable clinical marker useful for hydrocortisone supplementation [20], was 50 ± 71 μg/day (range: 7.1–173 μg/day). Other laboratory data, including Hct, Hb, CK, total cholesterol, TG, HDL-cholesterol, and LDL-cholesterol, were within normal ranges in these patients (data not shown).

Table 2 Physiological findings and laboratory data of patients

Case		Systolic BP (mmHg)	Diastolic BP (mmHg)	FPG (mg/dL)	free cortisol (μg/day)	ACTH (pg/mL)	cortisol (μg/dL)	WBC (/μL)	Neu (%)	Eo (%)	s-Na (mEq/L)	s-K (mEq/L)
1	pre	105	66	95	10.1	11.8	0.9	4,300	39.7	5.2	142	4.7
	post	115	74	91	51.9	33.4	0.8	4,600	56.5	3.5	141	3.8
2	pre	94	54	76	20.9	7.3	2.0	4,100	42.0	7.4	143	3.7
	post	100	58	118	57.2	2.1	2.9	5,700	67.8	3.9	143	4.4
3	pre	102	71	84	173	7.6	1.2	7,800	38.4	4.9	139	4.0
	post	107	52	83	44.8	5.6	0.4	7,000	43.9	3.8	140	3.7
4	pre	84	44	81	7.1	13.1	0.6	4,400	32.9	3.4	140	4.3
	post	99	59	85	88	n.d	1.3	6,200	68.9	0.6	139	4.0
5	pre	85	42	88	37	26.2	5.4	4,500	51.0	4.5	141	4.2
	post	89	61	90	63	23.2	6.3	4,500	52.2	6.2	140	4.0
6	pre	107	68	72	n.d	1.0	3.5	5,700	72.0	1.8	137	4.5
	post	109	67	92	n.d	n.d	n.d	5,500	74.1	1.6	139	4.4
mean ± SD	pre	96.2 ± 10.1	57.5 ± 12.6	82.7 ± 8.3	49.6 ± 70.7	11.2 ± 8.5	2.3 ± 1.9	5,133.3 ± 1,423.6	46.0 ± 14.0	4.5 ± 1.9	140.3 ± 2.3	4.2 ± 0.4
	post	103.2 ± 9.1	61.8 ± 7.7	93.2 ± 12.7	61.0 ± 16.5	16.1 ± 14.8	2.3 ± 2.4	5,583.3 ± 953.8	50.2 ± 23.8	3.3 ± 2.0	140.3 ± 1.5	4.1 ± 0.3

All results are expressed as the mean ± SD for continuous variables and as absolute values and percentages for categorical variables.

BP, blood pressure; FPG, fasting plasma glucose; WBC, white blood cell; Neu, neutrophils; Eo, eosinophils; s-Na, serum sodium; s-K, serum potassium; n.d, not determined

pre, status before modifying the regimen of the hormone supplement;

post, status after modifying the regimen of the hormone supplement

CGM analysis in tested cases

All patients consumed regular meals in their hospital stay during CGM. Table 3 shows a summary of CGM results before and after the modification of hormone replacement therapy in individual cases, including the levels of mean PG, minimum PG, and maximum PG, as well as the frequency of hypoglycemic events observed during 24–48 h of CGM. The mean amplitude of glycemic excursion (MAGE) was also listed. As anticipated, CGM identified hypoglycemic events in five out of six patients (83%) (Table 3, and the left panels of Fig. 1). Hypoglycemia events were observed at midnight and early in the morning, suggesting that the doses and timing of hydrocortisone supplementation were not appropriate to maintain normal nocturnal glucose levels. Additionally, although case 1 did not have hypoglycemia, the minimum PG levels fell to 71 mg/dL. The usual gap occurred between this measurement and the actual plasma glucose levels, so the blood glucose level of 71 mg/dL during CGM might have been lower than 70 mg/dL.

Table 3 Changes in the glycemic status detected using CGM in each of the tested cases before and after modifications to the regimen of the hormone supplement

Case		hormonal supplement	CGM mean PG (mg/dL)	min PG (mg/dL)	max PG (mg/dL)	MAGE	Frequency of hypoglycemia
1	pre	hydrocortisone 5-0-0-0 mg (2.89 mg/m2)	111	71	168	36.7	0
	post	hydrocortisone 10-0-5-0 mg (8.67 mg/m2) levothyroxine 25 μg	121	75	146	37.3	0
2	pre	no supply	111	60	178	57.3	2
	post	hydrocortisone 10-0-5-0 mg (9.32 mg/m2) levothyroxine 25 μg/day	111	98	173	20.5	0
3	pre	hydrocortisone 20-0-10-0 mg (14.69 mg/m2)	105	67	128	12	2
	post	hydrocortisone 10-0-0-5 mg (7.34 mg/m2)	101	78	119	17.3	0
4	pre	no supply	95	59	124	22.3	1
	post	hydrocortisone 10-0-5-0 mg (11.32 mg/m2)	90	76	107	21	0
5	pre	no supply	76	59	111	42.7	5
	post	hydrocortisone 5-0-0-0 mg (3.79 mg/m2)	94	75	122	16	0
6	pre	hydrocortisone 7.5-0-2.5-0 mg (7.13 mg/m2)	98	58	161	50.7	2
	post	hydrocortisone 5-0-5-0 mg (7.13 mg/m2)	103	90	123	25	0
Mean ± SD	pre		99.3 ± 13.2	62.3 ± 5.4	145.0 ± 27.4	37.0 ± 17.2	2.0 ± 1.7
	post		103.3 ± 11.3	82.0 ± 9.7	131.7 ± 23.9	22.9 ± 7.7	0
p-value			0.585	0.0014	0.3902	0.096

All results are expressed as the mean ± SD for continuous variables and as absolute values and percentages for categorical variables.

pre, status before modifying the regimen of the hormone supplement; post, status after modifying the regimen of the hormone supplement; CGM, continuous glucose monitoring; PG, plasma glucose; MAGE, the mean amplitude of glycemic excursion

Fig. 1

Continuous glucose monitoring of plasma glucose profiles in the tested subjects.

The contents of the hormonal supplement are presented at the right upper corner of each box. The black arrows indicate hypoglycemia. The black bar indicate continuous hypoglycemic period. The white arrows indicate the time of hydrocortisone supplementation. The bullets indicate the time of each meal.

After modifications to hormone supplementation, no hypoglycemia was detected.

The amplitude of glycemic excursions was also improved (see also Table 3).

Pre: status before modifying the hormone supplement regimen.

Post: status after modifying the hormone supplement regimen.

We aimed to reduce hypoglycemic events in these cases by maintaining glucocorticoid replacement therapy for each case as described below and then re-examining their glucose levels with CGM (Table 3). As shown in the right panels in Fig. 1 and Table 3, all subjects became free of hypoglycemia during the daily CGM. Comparisons of pre- vs. post-modified CGM data revealed a significant improvement in the mean minimum PG (62 ± 5 mg/dL vs. 82 ± 10 mg/dL, p < 0.005). On the other hand, the mean PG during CGM (99 ± 13 mg/dL vs. 103 ± 11 mg/dL) and MAGE (37.0 ± 17.2 vs. 22.9 ± 7.7) were not significantly altered (p = 0.585 and p = 0.096). The mean maximum PG (145 ± 27 mg/dL) during pre-modification did not differ from the post-modified status (132 ± 24 mg/dL, p = 0.390).

The following are case-by-case descriptions.

Hydrocortisones were administered 1–4 times daily (on awakening, after lunch, late afternoon (before dinner) and after dinner). We used the expression ‘5-5-0-5 mg’ if the patient took a 5-mg pill on awakening, after lunch, and after supper.

Case 1: A male case of lymphocytic hypophysitis.

His disease onset was at the age of 67. He had secondary adrenal insufficiency with diabetes insipidus and central hypogonadism. He took 5 mg hydrocortisone on awakening before enrolment in this study. He did not exhibit any symptoms suggestive of underoptimized glucocorticoid supplementation except for an occasional morning headache. Because his daily urinary free cortisol level was low (10.1 μg/day) (Table 2) and CGM found that his minimum PG levels fell to 71 mg/dL at around 0300 h, we increased the supplemental doses of hydrocortisone to 15 mg/day (10-0-5-0 mg). After this modification, daily urinary free cortisol levels were normalized to 51.9 μg/day. At the time of his diagnosis, central hypothyroidism was not detected. During this hospital stay, his anterior pituitary functions were re-tested. Although he did not have adult GH deficiency (AGHD), he was diagnosed with central hypothyroidism. Supplementation with 25 μg L-T4 for central hypothyroidism was initiated after the modification to hydrocortisone supplementation. The addition of L-thyroxine with hydrocortisone still resulted in favorable glycemic control in a follow-up CGM. After the modification of supplementation, he never exhibited further morning headaches.

Case 2: A male case of panhypopituitarism after pituitary apoplexy.

The patient was diagnosed with pituitary apoplexy several month prior at another hospital. He had not taken any hormonal supplementation on admission to our hospital. He had no symptoms suggesting adrenal insufficiency, including fatigue, weight loss, anorexia, and weakness. He only exhibited occasional morning headache. During this hospital stay, his anterior pituitary functions were re-tested and he was diagnosed with central hypoadrenalism, central hypothyroidism, central hypogonadism, and AGHD. CGM identified two nocturnal hypoglycemic events at around 0200 h and 0600 h, and the minimum PG level was 60 mg/dL. He was supplemented with 15 mg hydrocortisone (10-0-5-0 mg) and a subsequent CGM revealed no nocturnal hypoglycemia. The addition of L-thyroxine with hydrocortisone still resulted in favorable glycemic control in a follow-up CGM. He rejected supplementation with somatropin and testosterone enanthate. With these supplementations, his occasional headache was improved after long-term follow-up.

Case 3: A male case of panhypopituitarism after surgery for craniopharyngioma.

The patient was diagnosed with craniopharyngioma at the age of 5 and administered 30 mg of hydrocortisone per day (20-0-10-0 mg), in addition to 50 μg L-T4 per day and 125 mg testosterone enanthate every 4 weeks. He did not meet the diagnostic criteria for AGHD. Although CGM revealed two hypoglycemia events between 2400 h and 0200 h (Fig. 1), his daily urinary free cortisol level was above the normal range (173 μg/ day). We decreased his daily dose of hydrocortisone to 15 mg per day and modified the regimen (10-0-0-5 mg). His daily urinary free cortisol level was normalized thereafter (44.8 μg/day), and a subsequent CGM revealed no hypoglycemia (Table 3 and Fig. 1). This case demonstrated that a change in the timing of administration could improve glycemic profile.

Case 4: A female case of hypopituitarism due to De Morsier syndrome.

The patient was diagnosed with septo-optic dysplasia/‌De Morsier syndrome at the age of 2. She had taken 100 μg L-T4, 10 mg dydrogesterone, and 0.72 mg estradiol per day, but had received no hydrocortisone supplementation on admission to our hospital. She did not meet the diagnostic criteria for AGHD. She also exhibited occasional morning discomfort, without any other symptoms related to adrenal insufficiency. During this hospital stay, her anterior pituitary functions were re-tested and she was diagnosed with secondary adrenal insufficiency. CGM revealed continuous nocturnal hypoglycemic events between 0430 h and 0830 h and the minimum PG level was 59 mg/dL. Hydrocortisone supplementation at 15 mg/day (10-0-5-0 mg) revealed no hypoglycemic events and improved her symptoms.

Case 5: A male case of hypopituitarism after surgery for a pineal tumor.

He was treated with surgery and sequential radiation therapies at the age of 13. He had taken 0.4 mg somatropin per day, but had received no hydrocortisone supplementation on admission to our hospital. He exhibited occasional morning discomfort, without any other symptoms related to adrenal insufficiency. During this hospital stay, his anterior pituitary functions were re-tested and he was diagnosed with secondary adrenal insufficiency. CGM revealed frequent hypoglycemic events between 0500 h and 0800 h. Hydrocortisone supplementation with only 5 mg/day (5-0-0-0 mg) revealed no hypoglycemic events and improved his symptoms.

Case 6: A female case of IAD.

The patient was diagnosed at the age of 45 and received 10 mg of hydrocortisone per day (7.5-0-2.5-0 mg). CGM revealed continuous hypoglycemic events between 0300 h and 0600 h with 58 mg/dL of the minimum BG. We modified the regimen of supplementation hydrocortisone (5-0-5-0 mg). No hypoglycemic events were observed in a follow-up CGM. This case demonstrated that a change of the proportion of hydrocortisone administration could improve the glycemic profile without increasing the total amount of hydrocortisone administered.

Discussion

The aim of long-term therapy for adrenal insufficiency is to provide appropriate replacement doses of glucocorticoids to mimic the physiological circadian secretion of serum cortisol, as well as to improve the QOL and reduce the mortality of patients by preventing serious adverse events [9, 10]. Despite the use of synthetic glucocorticoids, premature mortality is still reported in patients with adrenal insufficiency.

A quantitative assessment of QOL previously revealed significant impairments in patients with primary or secondary adrenal insufficiency [21]. Excessive supplementation with glucocorticoids also resulted in an impaired subjective health status in patients with adrenal insufficiency [5-10]. Decisions regarding the optimal dosing of glucocorticoid replacement therapy in individual patients are mainly based on crude but important physical findings, such as body weight, well-being, and blood pressure [1-4, 22]. The signs and laboratory tests suggesting under-replacement are general fatigue, extreme weakness, depression, lethargy, anorexia, body weight loss, nausea, vomiting, low blood pressure, hyponatremia, and hyperkalemia. Hypoglycemia is reported to be most relevant in children, but is rarely seen in adults [1-3].

All patients included in the present study showed normal fasting glucose levels in the morning, along with normal BMI (except case 5), systolic and diastolic BP, Na, K, and cholesterol (Table 2). Along with these physical signs, laboratory tests and the fact that none of them complained of any symptoms suggestive of adrenal insufficiency, indicated that these patients had taken appropriate doses of hydrocortisone along with other hormones. However, these patients characteristically exhibited the occasional headache and/or discomfort, especially in the morning, in their routine clinical visits, and this may have been caused by unrecognized nocturnal hypoglycemia. Although no patients showed glucose levels less than 72 mg/dL in routine blood tests performed under morning fasting conditions, we detected unrecognized hypoglycemia in 5 out of 6 patients (83%) by CGM. Glucose levels less than 3.89 mM (70 mg/dL) on CGM are rarely observed in normal adult populations [23, 24]; therefore, we set the cutoff point for hypoglycemia at glucose levels of less than 70 mg/dL in this study. These hypoglycemic events occurred mainly in the early hours of the day and may have reflected a nadir in serum cortisol levels. Therefore, they suggested that the supplementary doses of glucocorticoids in the late afternoon or evening may have been insufficient to maintain normal nocturnal glucose levels.

We used the results of the CGM to fine-tune the doses and regimens of glucocorticoid replacement in each individual patient. We then re-evaluated daily glucose levels with a second CGM. Modifications to hormone replacement therapy successfully diminished the nocturnal hypoglycemia and morning headache/discomfort in all cases. Furthermore, the result of MAGE revealed that the daily profiles of plasma glucose levels became stable in some cases after the modifications. This reduction of the amplitude of glucose excursion should aid in lessening the risk of cardiovascular diseases [25].

Regarding the timing of hydrocortisone administration, multiple dosing is recommended to mimic physiological conditions. The first and largest dose is suggested to be given upon awakening, the second dose after lunch, and, in the case of a three-dose regimen, the last and smallest dose not later than 4–6 hours before bedtime [1-3]. The rationale for this regimen is to endeavor to mimic the circadian rhythm, as well as to avoid high doses in the evening, as these may compromise sleep and insulin sensitivity [26, 27]. In the present study, we administered the last dose of hydrocortisone after dinner to avoid their nocturnal hypoglycemia, as in case 3. This patient must be evaluated carefully in routine clinical visits, especially for the occurrence of insomnia or impaired glucose tolerance.

Recently, new guidelines were provided for the evaluation and management of adults with hypoglycemic disorders. In the guidelines, clinical hypoglycemia was defined as a plasma (or serum) glucose concentration sufficiently low to cause symptoms and/or signs, including impairment of brain function. The clinical manifestations of hypoglycemia are nonspecific, so a single plasma glucose concentration cannot be identified that categorically defines hypoglycemia, and a low measured plasma glucose concentration can be artifactual. Therefore, hypoglycemia is confirmed by documentation of Whipple’s triad: symptoms, signs, or both consistent with hypoglycemia, a low plasma glucose concentration, and resolution of those symptoms or signs after the plasma glucose is raised [25]. Several reports have also indicated that a hypoglycemia unawareness may result from a direct impairing influence of hypoglycemia on neurocognitive function [28, 29]. Taken together, our results suggested that morning headache or discomfort in patients with central hypoadrenalism was caused, at least in part, by unrecognized nocturnal hypoglycemia, similar to the condition occurring in DM patients.

A recent study reported that modified release preparations of hydrocortisone may stimulate overnight increases in cortisol release [30, 31]. Once-daily oral hydrocortisone dual release was shown to achieve a better circadian-based serum cortisol profile in adult patients with adrenal insufficiency, as well as lower body weights and blood pressure, and better glucose metabolism and QOL when compared to conventional hydrocortisone treatment [32, 33]. However, persistent low cortisol levels were still observed in the late evening and at midnight.

Today, much of the physiology of the hypothalamo-pituitary-adrenal axis is well understood, but its clinical assessment and the diagnostic procedures for establishing a need for replacement are still far from perfect. Thus, to a certain extent, clinical judgement is still vital (3). In the present study, since the duration period of CGM was limited to 24 to 48 hours, reproducibility could not be confirmed. Nevertheless, this study revealed that CGM may represent a powerful tool for detecting an unrecognized hypoglycemia in patients with no typical symptoms suggestive of adrenal insufficiency (especially in cases 1, 2, 4, and 5). CGM could also provide a useful index for assessing and optimizing the doses and regimens of conventional glucocorticoid replacement in patients with central hypoadrenalism. For example, we were able to eliminate unrecognized hypoglycemic events without increasing the total amount of hydrocortisone supplementation (cases 3 and 6).

The long-term usefulness of CGM-based modifications to glucocorticoid replacement therapy requires further prospective evaluations. Another issue worth mentioning is that thyroid hormone must be maintained to obtain a stable glycemic pattern. GH is considered as the counter-regulatory hormone of hypoglycemia and an insufficient GH level could affect glycemic control. Some cases in the present study demonstrated AGHD (cases 2 and 5), but the dose of GH supplementation was not changed during the CGM series in any of our cases.

Central hypoadrenalism is characterized by maintenance of the adrenal medullary function, so the frequency of hypoglycemic onset is assumed to be lower than occurs with primary hypoadrenalism. Nevertheless, even in primary hypoadrenalism, the onset of hypoglycemia is reported as rare in adult cases (1–3). However, in these reports, no stipulation was made regarding how many of the blood glucose levels should viewed as representing hypoglycemia. In our study, glucose levels less than 70 mg/dL were regarded as hypoglycemia. Symptoms that seems to be due to hypoglycemia, such as morning headache and/or morning discomfort, completely disappeared, along with the free of glucose level less than 70 mg/dL, after the revision of steroid replacement therapy regimens. Therefore, we considered that unrecognized hypoglycemic events frequently occurred in the central hypoadrenalism cases tested in the present study. In some cases, the level of urinary free cortisol was below the standard value, suggesting that the total amount of supplemented hydrocortisone was insufficient (case 1 and 4). Conversely, we succeeded in suppressing the onset of hypoglycemia by changing the administration time (case 3), or changing the proportion of administration (case 6) without increasing the total dose of hydrocortisone.

In summary, the present study demonstrated that a nocturnal hypoglycemia commonly occurred in adult patients with central hypoadrenalism, more than previously reported in Europeans and Americans [1-3], even though apparently appropriate hydrocortisone replacement therapies were administered. CGM may represent a powerful tool for detecting nocturnal hypoglycemia episodes, with or without associated symptoms, as well as for optimizing therapeutic regimens to avoid overdosing with glucocorticoids. To the best of our knowledge, this is the first study to employ CGM as a dynamic tool to improve the QOL of patients with central hypoadrenalism.

Acknowledgments

We thank the patients, referring physicians, and staff of the Department of Medicine and Molecular Science, Gunma University Graduate School of Medicine.

Disclosure

None of the authors have any potential conflicts of interest associated with this manuscript.

Grants

This work was supported by Japan Society for the Promotion of Science (JSPS) KAKENHI (Grants-in-aid for Scientific Research) Grant Number 26461374 and 23591346 to A.O.

References

• 1   Bornstein  SR,  Allolio  B,  Arlt  W,  Barthel  A,  Don-Wauchope  A, et al. (2016) Diagnosis and treatment of primary adrenal insufficiency: an Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab 101: 364–389.
• 2   Husebye  ES,  Allolio  B,  Arlt  W,  Badenhoop  K,  Bensing  S, et al. (2014) Consensus statement on the diagnosis, treatment and follow-up of patients with primary adrenal insufficiency. J Intern Med 275: 104–115.
• 3   Grossman  AB (2010) The diagnosis and management of central hypoadrenalism. J Clin Endocrinol Metab 95: 4855–4863.
• 4   Yanase  T,  Tajima  T,  Katabami  T,  Iwasaki  Y,  Tanahashi  Y, et al. (2016) Diagnosis and treatment of adrenal insufficiency including adrenal crisis: a Japan Endocrine Society clinical practice guideline. Endocr J 63: 765–784.
• 5   Bensing  S,  Brandt  L,  Tabaroj  F,  Sjöberg  O,  Nilsson  B, et al. (2014) Increased death risk and altered cancer incidence pattern in patients with isolated or combined autoimmune primary adrenocortical insufficiency. Clin Endocrinol (Oxf) 69: 697–704.
• 6   Bergthorsdottir  R,  Leonsson-Zachrisson  M,  Odén  A,  Johannsson  G (2006) Premature mortality in patients with Addison’s disease: a population-based study. J Clin Endocrinol Metab 91: 4849–4853.
• 7   Filipsson  H,  Monson  JP,  Koltowska-Haggstrom  M,  Mattsson  A,  Johannsson  G (2006) The impact of glucocorticoid replacement regimens on metabolic outcome and comorbidity in hypopituitary patients. J Clin Endocrinol Metab 91: 3954–3961.
• 8   Zelissen  P,  Croughs  R,  van Rijk  P,  Raymakers  J (1994) Effect of glucocorticoid replacement therapy on bone mineral density in patients with Addison disease. Ann Intern Med 120: 207–210.
• 9   Bleicken  B,  Hahner  S,  Loeffler  M,  Ventz  M,  Decker  O, et al. (2010) Influence of hydrocortisone dosage scheme on health-related quality of life in patients with adrenal insufficiency. Clin Endocrinol (Oxf) 72: 297–304.
• 10   Hahner  S,  Loeffler  M,  Fassnacht  M,  Weismann  D,  Koschker  AC, et al. (2007) Impaired subjective health status in 256 patients with adrenal insufficiency on standard therapy based on cross-sectional analysis. J Clin Endocrinol Metab 97: 3912–3922.
• 11   Feek  CM,  Ratcliffe  JG,  Seth  J,  Gray  CE,  Toft  AD, et al. (1981) Patterns of plasma cortisol and ACTH concentrations in patients with Addison’s disease treated with conventional corticosteroid replacement. Clin Endocrinol (Oxf) 14: 451–458.
• 12   Ten  S,  New  M,  Maclaren  N (2001) Clinical review 130: Addison’s disease. J Clin Endocrinol Metab 86: 2909–2922.
• 13   Christiansen  JJ,  Djurhuus  CB,  Gravholt  CH,  Iverson  P,  Christiansen  JS, et al. (2007) Effects of cortisol on carbohydrate, lipid, and protein metabolism: studies of acute cortisol withdrawal in adrenocortical failure. J Clin Endocrinol Metab 92: 3553–3559.
• 14   King  P,  Kong  MF,  Parkin  H,  Macdonald  IA,  Tattersall  RB (1998) Well-being, cerebral function, and physical fatigue after nocturnal hypoglycemia in IDDM. Diabetes Care 21: 341–345.
• 15   Phillip  M,  Danne  T,  Shalitin  S,  Buckingham  B,  Laffel  L, et al. (2012) Use of continuous glucose monitoring in children and adolescents. Pediatr Diabetes 13: 215–228.
• 16   Gross  TM,  Bode  BW,  Einhorn  D,  Kayne  DM,  Reed  JH, et al. (2000) Performance evaluation of the Mini-Med continuous glucose monitoring system during patient home use. Diabetes Technol Ther 2: 49–56.
• 17   Conrad  SC,  Mastrototaro  JJ,  Gitelman  SE (2004) The Use of the continuous glucose monitoring system in hypoglycemic disorders. J Pediatr Endocrinol Metab 17: 281–288.
• 18   Maran  A,  Crepaldi  C,  Avogaro  A,  Catuogno  S,  Burlina  A, et al. (2004) Continuous glucose monitoring in conditions other than diabetes. Diabetes Metab Res Rev 20: S50–S55.
• 19   Meyer  G,  Hackemann  A,  Reusch  J,  Badenhoop  K (2012) Nocturnal hypoglycemia identified by a continuous glucose monitoring system in patients with primary adrenal insufficiency (Addison’s disease). Diabetes Technol Ther 14: 386–388.
• 20   Burch  WM (1982) Urine free-cortisol determination. A useful tool in the management of chronic hypoadrenal states. JAMA 247: 2002–2004.
• 21   Hahner  S,  Loeffler  M,  Fassnacht  M,  Weismann  D,  Koschker  AC, et al. (2007) Impaired subjective health status in 256 patients with adrenal insufficiency on standard therapy based on cross-sectional analysis. J Clin Endocrinol Metab 92: 3912–3922.
• 22   Arlt  W,  Rosenthal  C,  Hahner  S,  Allolio  B (2006) Quality of glucocorticoid replacement in adrenal insufficiency: clinical assessment vs. timed serum cortisol measurements. Clin Endocrionol (Oxf) 64: 384–389.
• 23  Mazze, RS,  Strock  E,  Wesley  D,  Borgman  S,  Morgan  B, et al. (2008) Characterizing glucose exposure for individuals with normal glucose tolerance using continuous glucose monitoring and ambulatory glucose profile analysis. Diabetes Technol Ther 10: 149–159.
• 24   Zhou  J,  Li  H,  Ran  X,  Yang  W,  Li  Q, et al. (2009) Reference values for continuous glucose monitoring in Chinese subjects. Diabetes Care 32: 1188–1193.
• 25   Wei  L,  Macdonald  TM,  Walker  BR (2004) Taking glucocorticoids by prescription is associated with subsequent cardiovascular disease. Ann Intern Med 141: 764–770.
• 26   Plat  L,  Leproult  R,  LʼHermite-Baleriaux  M,  Fery  F,  Mockel  J, et al. (1999) Metabolic effects of short-term elevations of plasma cortisol are more pronounced in the evening than in the morning. J Clin Endocrinol Metab 84: 3082–3092.
• 27   Simon  N,  Castinetti  F,  Ouliac  F,  Lesavre  N,  Brue  T, et al. (2010) Pharmacokinetic evidence for suboptimal treatment of adrenal insufficiency with currently available hydrocortisone tablets. Clin Pharmacokinet 49: 45–63.
• 28   Deary  IJ,  Crawford  JR,  Hepburn  DA,  Langan  SJ,  Blackmore  LM, et al. (1993) Severe hypoglycemia and intelligence in adult patients with insulin-treated diabetes. Diabetes 42: 341–344.
• 29   Gonder-Frederick  L,  Cox  D,  Koyatchev  B,  Schlundt  D,  Clarke  WA (1997) Biopsychobehavioral model of risk of severe hypoglycemia. Diabetes Care 20: 661–669.
• 30   Newell-Price  J,  Whiteman  M,  Rostami-Hodjegan  A,  Darzy  K,  Shalet  S, et al. (2008) Modified-release hydrocortisone for circadian therapy: a proof-of-principle study in dexamethasone-suppressed normal volunteers. Clin Endocrinol (Oxf) 68: 130–135.
• 31   Debono  M,  Ghobadi  C,  Rostami-Hodjegan  A,  Huatan  H,  Campbell  MJ, et al. (2009) Modified-release hydrocortisone to provide circadian cortisol profiles. J Clin Endocrinol Metab 94: 1548–1554.
• 32   Johannsson  G,  Bergthorsdottir  R,  Nilsson  AG,  Lennernas  H,  Hedner  T, et al. (2009) Improving glucocorticoid replacement therapy using a novel modified-release hydrocortisone tablet: a pharmacokinetic study. Eur J Endocrinol 161: 119–130.
• 33   Johannsson  G,  Nilsson  AG,  Bergthorsdottir  R,  Burman  P,  Dahlqvist  P, et al. (2012) Improved cortisol exposure-time profile and outcome in patients with adrenal insufficiency: a prospective randomized trial of a novel hydrocortisone dual-release formulation. J Clin Endocrinol Metab 97: 473–481.