Reliable detection of atrial fibrillation with a medical wearable under inpatient conditions (CoMMoD-A-fib)

Background

Atrial fibrillation (A-fib) is the most common arrhythmia; however, detection of A-fib is a challenge due to irregular occurrence.

Purpose

Evaluating feasibility and performance of a non-invasive medical wearable for detection of A-fib.

Methods

In the CoMMoD-A-fib trial admitted patients with a high risk for A-fib carried the wearable and an ECG Holter (control) in parallel over a period of 24 hours under not physically restricted conditions. The wearable with a tight-fit upper arm band employs a photoplethysmography (PPG) technology enabling a high sampling rate. Different algorithms (including a deep neural network) were applied to 5 min PPG datasets for detection of A-fib. Proportion of monitoring time automatically interpretable by algorithms (= interpretable time) was analyzed for influencing factors.

Results

In 102 inpatients (age 71.0±11.9 years; 52% male) 2306 hours of parallel recording time could be obtained; 1781 hours (77.2%) of these were automatically interpretable by an algorithm analyzing PPG derived intervals. Detection of A-Fib was possible with a sensitivity of 92.7% and specificity of 92.4% (AUC 0.96). Also during physical activity, detection of A-fib was sufficiently possible (sensitivity 90.1% and specificity 91.2%). Usage of the deep neural network improved detection of A-fib further (sensitivity 95.4% and specificity 96.2%). A higher prevalence of heart failure with reduced ejection fraction was observed in patients with a low interpretable time (p=0.080).

Conclusion

Detection of A-fib by means of an upper arm non-invasive medical wearable with a high resolution is reliably possible under inpatient conditions.

Funding Acknowledgement

Type of funding source: Public Institution(s). Main funding source(s): Internal grant program (PhD and Dr. rer. nat. Program Biomedicine) of the Faculty of Health at Witten/Herdecke University, Germany. HELIOS Kliniken GmbH (Grant-ID 047476), Germany

Euglycemic ketosis in patients with type 2 diabetes on SGLT2-inhibitor therapy-an emerging problem and solutions offered by diabetes technology

Diabetic ketoacidosis is an infrequent but life-threatening acute complication of diabetes, affecting predominantly patients with type 1 diabetes, children, and pregnant women, where ketosis is usually associated with marked hyperglycemia. Recently, an increasing number of cases have been reported of euglycemic diabetic ketoacidosis in patients with type 2 diabetes receiving sodium-glucose cotransporter 2 inhibitor treatment in routine practice. There is a minor, but not negligible diabetic ketoacidosis risk associated with this drug class, which was not seen in randomized clinical trials. However, sodium-glucose cotransporter2 inhibitors increase the risk of ketosis by increasing glucagon secretion in the pancreas and decreasing the renal excretion of 3-hydroxybutyrate and acetoacetate. When used in addition to insulin, any insulin dose reduction required to avoid hypoglycemia may lead to insufficient suppression of lipolysis and ketogenesis. sodium-glucose cotransporter2 inhibitor-induced loss of urinary glucose encourages euglycemia. Normo-glycemic or near-normoglycemic diabetic ketoacidosis represents a major threat to the health and well-being of a patient, because it may occur undetected and without any indicative hyperglycemia. In consequence, patients on sodium-glucose cotransporter2 inhibitors are recommended to perform regular blood ketone tests since they are not alerted to incipient diabetic ketoacidosis by glucose testing alone. This option is offered by several blood glucose meters that can also measure ketones with a separate ketone strip or in one case by an automatic parallel ketone assessment from the same strip. The need for extra testing and the associated costs may be a barrier to patient acceptance of this risk mitigation procedure. However, patients who are at risk for euglycemic diabetic ketoacidosis when being treated with sodium-glucose cotransporter2 inhibitors should be specially advised to monitor blood ketone levels on a regular basis.

HbA1c and Age in Non-Diabetic Subjects: An Ignored Association?

Objective: Target HbA_1c values given in the most National Therapeutic Guidelines for patients with diabetes and cut-off HbA_1c values for diabetes diagnosis are usually not taking the age of the respective patients into account; despite the fact that an increase in HbA_1c in subjects without diabetes with age is known for some time. In order to further quantify the association between age and HbA_1c in non-diabetic subjects an analysis of one German register was performed. Methods: In this cross-sectional study we analyzed data from 7 699 visits of 2 921 patients without diabetes (age 46.6 y [range 18-93 y]; 69.1% women; BMI 27.6±6.4 kg/m²) who had at least one HbA_1c and blood glucose measurement. Data were drawn from an electronic patient record system (EMIL™) in which data were collected between 01/1992 and 01/2014. The patients were divided in 6 age groups (< 30 years [n=1 057];>30-40 years [n=1 160];>40-50 years [n=1 693];>50-60 years [n=1 523];>60-70 years [n=1 310];>70 years [n=956]) and the HbA_1c values of these groups were compared. Patients with: gestational diabetes_, use of systemic glucocorticoids, malignant neoplasm, age<18 y at time of first visit and IGT were excluded. HbA_1c measurements were DCCT adjusted. Results: Patients with age>70 years have a 0.47% [5.14 mmol/mol] higher HbA_1c compared to those<30 years. The mean HbA_1c of the age groups was:<30 4.98% [30.96 mmol/mol],>30-40 5.07% [31.99 mmol/mol],>40-50 5.17% [33.10 mmol/mol],>50-60 5.33% [34.79 mmol/mol],>60-70 5.42% [35.79 mmol/mol] and>70 years 5.45% [36.10 mmol/mol]. In a multiple linear model the regression coefficient for each year of age increase was β=0.0074 (p<0.001); thus age results in an increase of 0.074% in HbA_1c per decade. Conclusion: HbA_1c increases significantly with ageing in people without diabetes. The use of different cut-off values for every age range for diagnosis of diabetes should be discussed.

Home Use of an Artificial Beta Cell in Type 1 Diabetes

BACKGROUND

The feasibility, safety, and efficacy of prolonged use of an artificial beta cell (closed-loop insulin-delivery system) in the home setting have not been established.

METHODS

In two multicenter, crossover, randomized, controlled studies conducted under free-living home conditions, we compared closed-loop insulin delivery with sensor-augmented pump therapy in 58 patients with type 1 diabetes. The closed-loop system was used day and night by 33 adults and overnight by 25 children and adolescents. Participants used the closed-loop system for a 12-week period and sensor-augmented pump therapy (control) for a similar period. The primary end point was the proportion of time that the glucose level was between 70 mg and 180 mg per deciliter for adults and between 70 mg and 145 mg per deciliter for children and adolescents.

RESULTS

Among adults, the proportion of time that the glucose level was in the target range was 11.0 percentage points (95% confidence interval [CI], 8.1 to 13.8) greater with the use of the closed-loop system day and night than with control therapy (P<0.001). The mean glucose level was lower during the closed-loop phase than during the control phase (difference, -11 mg per deciliter; 95% CI, -17 to -6; P<0.001), as were the area under the curve for the period when the glucose level was less than 63 mg per deciliter (39% lower; 95% CI, 24 to 51; P<0.001) and the mean glycated hemoglobin level (difference, -0.3%; 95% CI, -0.5 to -0.1; P=0.002). Among children and adolescents, the proportion of time with the nighttime glucose level in the target range was higher during the closed-loop phase than during the control phase (by 24.7 percentage points; 95% CI, 20.6 to 28.7; P<0.001), and the mean nighttime glucose level was lower (difference, -29 mg per deciliter; 95% CI, -39 to -20; P<0.001). The area under the curve for the period in which the day-and-night glucose levels were less than 63 mg per deciliter was lower by 42% (95% CI, 4 to 65; P=0.03). Three severe hypoglycemic episodes occurred during the closed-loop phase when the closed-loop system was not in use.

CONCLUSIONS

Among patients with type 1 diabetes, 12-week use of a closed-loop system, as compared with sensor-augmented pump therapy, improved glucose control, reduced hypoglycemia, and, in adults, resulted in a lower glycated hemoglobin level. (Funded by the JDRF and others; AP@home04 and APCam08 ClinicalTrials.gov numbers, NCT01961622 and NCT01778348.).

Biosimilar insulins: guidance for data interpretation by clinicians and users

Biosimilar insulins are approved copies of insulins outside patent protection. Advantages may include greater market competition and potential cost reduction, but clinicians and users lack a clear perspective on 'biosimilarity' for insulins. The manufacturing processes for biosimilar insulins are manufacturer-specific and, although these are reviewed by regulators there are few public data available to allow independent assessment or review of issues such as intrinsic quality or batch-to-batch variation. Preclinical measures used to assess biosimilarity, such as tissue and cellular studies of metabolic activity, physico-chemical stability and animal studies of pharmacodynamics, pharmacokinetics and immunogenicity may be insufficiently sensitive to differences, and are often not formally published. Pharmacokinetic and pharmacodynamic studies (glucose clamps) with humans, although core assessments, have problems of precision which are relevant for accurate insulin dosing. Studies that assess clinical efficacy and safety and device compatibility are limited by current outcome measures, such as glycated haemoblobin levels and hypoglycaemia, which are insensitive to differences between insulins. To address these issues, we suggest that all comparative data are put in the public domain, and that systematic clinical studies are performed to address batch-to-batch variability, delivery devices, interchangeability in practice and long-term efficacy and safety. Despite these challenges biosimilar insulins are a welcome addition to diabetes therapy and, with a transparent approach, should provide useful benefit to insulin users.

Biosimilar insulins: a European perspective

Biosimilar insulins are likely to enter clinical practice in Europe in the near future. It is important that clinicians are familiar with and understand the concept of biosimilarity and how a biosimilar drug may differ from its reference product. The present article provides an overview of biosimilars, the European regulatory requirements for biosimilars and safety issues. It also summarizes the current biosimilars approved in Europe and the key clinical issues associated with the use of biosimilar insulins.

Evidence for continuous glucose monitoring: sufficient for reimbursement?

Evidence for continuous glucose monitoring is mounting. Meta-analyses consistently show lowering of HbA1c , and the first trial reporting a reduction in severe hypoglycaemic events in patients with hypoglycaemia unawareness has recently been presented. The more recent trials studied larger numbers of patients and may have shown better results because of improved technology. The various combinations of pump and sensor, with automated bolus calculators and low glucose suspend features, make evaluation more challenging from a reimbursement point of view, but evidence seems convincing enough to justify reimbursement for selected patient groups, including those who have shown a substantial improvement in HbA1c during a trial period, and those with hypoglycaemia unawareness who encountered severe hypoglycaemia in the recent past. More data on cost-efficacy are needed.

Continuous glucose monitoring: quality of hypoglycaemia detection

AIMS

To evaluate the accuracy of a (widely used) continuous glucose monitoring (CGM)-system and its ability to detect hypoglycaemic events.

METHODS

A total of 18 patients with type 1 diabetes mellitus used continuous glucose monitoring (Guardian REAL-Time CGMS) during two 9-day in-house periods. A hypoglycaemic threshold alarm alerted patients to sensor readings <70 mg/dl. Continuous glucose monitoring sensor readings were compared to laboratory reference measurements taken every 4 h and in case of a hypoglycaemic alarm.

RESULTS

A total of 2317 paired data points were evaluated. Overall, the mean absolute relative difference (MARD) was 16.7%. The percentage of data points in the clinically accurate or acceptable Clarke Error Grid zones A + B was 94.6%. In the hypoglycaemic range, accuracy worsened (MARD 38.8%) leading to a failure to detect more than half of the true hypoglycaemic events (sensitivity 37.5%). Furthermore, more than half of the alarms that warn patients for hypoglycaemia were false (false alert rate 53.3%). Above the low alert threshold, the sensor confirmed 2077 of 2182 reference values (specificity 95.2%).

CONCLUSIONS

Patients using continuous glucose monitoring should be aware of its limitation to accurately detect hypoglycaemia.

Insulin aspart has a shorter duration of action than human insulin over a wide dose-range

AIMS

Regular human insulin (RHI) at high doses shows prolongation of its duration of action potentially leading to late postprandial hypoglycaemia. This study compared late metabolic activity (4-12 and 6-12 h post-dosing) and duration of action (time to reach late half-maximal activity) over a range of doses between insulin aspart (IAsp) and RHI.

METHODS

Pharmacokinetic and pharmacodynamic properties of subcutaneous IAsp and RHI (6, 12 and 24 (I)U) were compared in 16 healthy subjects in this double-blind, randomized, six-way crossover glucose clamp study.

RESULTS

With increasing doses of both insulins, metabolic activity, insulin exposure, maximum metabolic effect and maximum serum insulin concentration increased linearly. Late metabolic activity was lower for IAsp than RHI at all doses, reaching statistical significance (p < 0.05) for 12 and 24 (I)U. Likewise, IAsp had a shorter duration of action at all doses (p < 0.01) and reached time to 80% of total metabolic activity earlier at doses of 12 and 24 (I)U (p < 0.05). IAsp, compared with RHI, showed a higher maximum metabolic effect at 12 and 24 (I)U (p < 0.0001) and a stronger early metabolic activity for all three doses (p < 0.05).

CONCLUSIONS

IAsp showed a shorter duration of action and, particularly with doses of 12 and 24 (I)U, less late metabolic activity than RHI. These properties might contribute to the lower incidence of hypoglycaemia observed with IAsp versus RHI in clinical trials as lower late metabolic activity should decrease the risk of late postprandial hypoglycaemia.

U-100, pH-Neutral formulation of VIAject(®) : faster onset of action than insulin lispro in patients with type 1 diabetes

AIMS

VIAject® is a formulation of human insulin with a very fast onset of action. Previous studies used VIAject in a concentration of 25 U/ml and a pH of 4 [VIAject 25 (VJ25)]. Objective of this double blind, three-way crossover study was to compare the pharmacodynamic/pharmacokinetic properties of a novel formulation of VIAject with a concentration of 100 U/ml and a neutral pH [VIAject 7 (VJ7)] with VJ25 and insulin lispro (LIS).

METHODS

Forty-three patients with type 1 diabetes [aged 43 (21-65) years, BMI 24.1 (20-28) kg/m(2) and HbA1c 7.5 (5.7-9.5) %] participated in this study. They received subcutaneous injections of 12 U of each insulin formulation under euglycaemic glucose clamp conditions.

RESULTS

VJ7 was bioequivalent to VJ25 [90% confidence interval (CI) of the ratios for total insulin AUCs and maximum insulin concentration (C(INS max) ) was within 0.80-1.25]. VJ7 showed a faster absorption compared to LIS [time to C(INS max) 23 vs. 60 min; difference (CI) -30 (-35 to -23)] and faster onset of action [time to early half-maximal glucose infusion rate (GIR) 25 vs. 44 min; -18 (-26 to -10)], and a higher AUC of glucose infusion rate (AUC(GIR) ) in the first 60 min after injection [176 vs. 107 mg/kg; ratio 1.65 (1.27 to 2.14)], contributing to a slightly higher value for AUC(GIR 0-480) [1263 vs. 1095 mg/kg; 1.15 (1.06 to 1.26)]. Maximum GIR was similar between VJ7 and LIS [6.1 vs.6.6 mg/kg/min; ratio 0.93 (0.86 to 1.01)], whereas the duration of action (t(GIR50%-late) ) was longer with VJ7 [274 vs. 228 min; 50 (25 to 73)].

CONCLUSIONS

This formulation of VIAject is bioequivalent to the previously used formulation and has a faster absorption/onset of action than LIS.

New ways of insulin delivery

The predominant number of papers published from the middle of 2009 to the middle of 2010 about alternative routes of insulin administration (ARIA) were still about inhaled insulin. Long-term experience with Exubera was the topic of a number of publications that are also of relevance for inhaled insulin in general. The clinical trials performed with AIR insulin by Eli Lilly were published in a supplement issue of one diabetes technology journal and most of these will be presented. A number of other publications (also one in a high ranked journal) about their inhaled insulin were from another company: MannKind. The driving force behind Technosphere insulin (TI) - which is the only one still in clinical development - is Al Mann; he has put a lot of his personal fortune in this development. We will know the opinion of the regulatory authorities about TI in the near future; however, I am personally relatively confident that the Food and Drug Administration will provide TI with market approval. The more critical question for me is: will diabetologists and patients jump on this product once it becomes commercially available? Will it become a commercial success? In view of many negative feelings in the scientific community about inhaled insulin, it might be of help that MannKind publish their studies with TI systematically. Acknowledging being a believer in this route of insulin administration myself, one has to state that Exubera and AIR insulin had not offered profound advantages in terms of pharmacokinetic (PK) and pharmacodynamic (PD) properties in comparison with subcutaneously (SC) applied regular human insulin (RHI) and rapid-acting insulin analogues. The time-action profiles of these inhaled insulins were more or less comparable with that of rapid-acting insulin analogues. This is clearly different with TI which exhibits a strong metabolic effect shortly after application and a rapid decline in the metabolic effect thereafter; probably the duration of action is even too short (see postprandial glycaemic excursions with test meals in the publication by Rosenstock et al. in The Lancet (1)). In the end a number of aspects are of relevance for the success of a given product; one key aspect is clearly the price. However, for patients also practical aspects (handling, need for regular pulmonary function test etc.) are of importance. We shall have to see how creatively MannKind will handle all such questions. Until now Al Mann and his colleagues were able to manage a number of challenges during the clinical development process successfully, so one can have hopes for the market success of TI. However, it is clear that at the same time, if TI fails like Exubera did before, this will be the end for pulmonary insulin in general. Not too many original publications presenting data from clinical trials were published in the last year when it comes to oral insulin (OI), nasal insulin or transdermal insulin developments; simply none with transdermal insulin. Also at the last international congresses not many studies about ARIA were presented. At least in part this might be still a reflection of the shockwaves that the failure of Exubera has sent out to pharmaceutical companies and venture capitalists; they are quite reluctant to invest in any of these developments. However, a considerable number of reviews (in some cases more than original papers!) were published about ARIA. These reviews are listed for completeness, but in most cases are not further commented. OI is still the area of research most companies are active in; however, in some cases it is not clear how active they really are (e.g. Diabetology). Nevertheless, at least some companies are quite active and progressed in their clinical development programme close to market approval, e.g. the large Indian company Biocon is in late phase 3 with IN-105 and the small Israel-based company Oramed is in phase 2b. It appears that other interesting OI developments (e.g. Diasome) were not very active in the last year; at least they have not published new study results. It is clear that for companies that produce insulin themselves (e.g. Biocon) the costs of the good are not of such relevance as for companies that have to buy it commercially. For the latter ones a low bioavailability/biopotency compared with SC insulin administration can be a real hurdle when it comes to the price of their product. Despite some publications about nasal insulin, the overall activity with this route of insulin administration appears to be low; the same holds true for transdermal insulin. Insulin pens have gained more scientific interest in recent years, which is also reflected by an increase in publications, starting from practically nil 10 years ago to a solid number of five to 10 papers per year nowadays. Besides ARIA there are also attempts to increase the speed of insulin absorption after injection into the skin by applying it not into the SC tissue but intradermally or by heating up the skin above the SC insulin depot. Reading a number of papers that were not included in this chapter because they do not present any clinical data but are novel developments tested only in animal experiments so far, the clear message is that there is definitely not a lack of creativity/imagination amongst scientists; each year a plethora of new ideas for insulin application show up. Unfortunately not too many make it towards a full clinical development. As long as there is not a single successful product on the market that is based on a given ARIA approach, this area of research will not mature. For many patients, avoiding the need for SC injections is attractive; however, as long as no clear 'advantage' can be demonstrated, reimbursement will be difficult to achieve. Living in the time of evidence-based medicine it is clear that 'relevant' clinical advantages must be proven. The question is what is relevant. Is it just an improvement in metabolic control (= decrease in HbA1c)? Can this also mean that more patients are willing to start insulin therapy earlier than with conventional SC insulin therapy? With TI we have a product that has improved pharmacological properties (also in comparison to Exubera) for coverage of prandial insulin requirements. Subsequently, in the clinical trials performed, postprandial glycaemic excursions were lower than with SC injection of RHI or rapid-acting insulin analogues. This only in part (if at all) results in an improved metabolic control in general (= lower HbA1c) (see below). The outlook for 2011 is that there are chances that we shall have an inhaled insulin product on the market. Probably also the first OI will be submitted to the regulatory authorities for market approval or will even be available in less regulated markets. In order to select all relevant publications about new ways of insulin delivery I performed a PUBMED search and also checked the table of contents of a number of journals that publish heavily in this area of research as well references in the publications I found for additional references. Selection of the manuscripts from all publications was predominately based on the fact whether they presented data from clinical studies or not. The selected studies were critically reviewed for novelty and appropriate study design etc. In some cases also reviews about a given topic were selected if they provide relevant novel insights.

Short-term metabolic effects of prednisone administration in healthy subjects

AIMS

Supraphysiologic glucocorticoid activity is well established to cause impaired glucose tolerance and insulin resistance, yet no study has evaluated dose-dependent effects of low-dose prednisone during short-term oral administration.

METHODS

The objective of this study was to quantify the effects of daily 10 or 25 mg prednisone administration for one week on insulin sensitivity by employing a two-step hyperinsulinemic euglycemic glucose clamp (Step 1: insulin infusion = 20 mU/m²/min; Step 2: insulin infusion = 80 mU/m²/min) in healthy, lean males. The amount of glucose infused at steady-state to maintain stable blood glucose [90 mg/dl (4.95 mmol/l)] was used to calculate several indices of insulin sensitivity.

RESULTS

During Step 1 of the clamp, whole body glucose disposal (M) was reduced by 35% (p = 0.003) and M/I was reduced by 29% (p = 0.025) for 25 mg prednisone compared to placebo. No appreciable effect of 10 mg prednisone was observed. During Step 2, M was reduced by 33% (p = 0.001) and 15% (p = 0.006) for 25 and 10 mg prednisone compared to placebo; and M/I ratio was reduced by 31% (p < 0.001) and 13% (p = 0.026), respectively. The insulin sensitivity index, Si, calculated as the quotient of augmentation of M/I between Step 1 and 2, was reduced by 35.3% (p < 0.01) and 23.5% (p < 0.05) for 25 and 10 mg prednisone, respectively.

CONCLUSION

Administration of relatively low pharmacological doses of prednisone for one week impaired insulin sensitivity in a dose-dependent manner in healthy males. These observed changes in insulin sensitivity are likely to be clinically relevant, especially in individuals predisposed to develop glucose intolerance.

Oral glucose tolerance test and HbA₁c for diagnosis of diabetes in patients undergoing coronary angiography: [corrected] the Silent Diabetes Study

AIMS/HYPOTHESIS

The primary aim of this study was to compare the results of HbA(1c) measurements with those of an OGTT for early diagnosis of 'silent diabetes' in patients with coronary artery disease (CAD) undergoing angiography without prediagnosed diabetes. A secondary aim was to investigate the correlation between the extent of CAD and the glycaemic status of the patient.

METHODS

Data from 1,015 patients admitted for acute (n = 149) or elective (n = 866) coronary angiography were analysed. Patients with known diabetes were excluded from the study. Using the OGTT results, patients were classified as having normal glucose tolerance (NGT), impaired fasting glucose (IFG), impaired glucose tolerance (IGT) or diabetes. According to the results of the HbA(1c) measurements, patients were classified into three groups: normal (HbA(1c) <5.7% [<39 mmol/mol]), borderline (HbA(1c) 5.7-6.4% [39-47 mmol/mol]) and diabetes (HbA(1c) ≥6.5% [≥48 mmol/mol]).

RESULTS

Based on the OGTT, 513 patients (51%) were classified with NGT, 10 (1%) with IFG, 349 (34%) with IGT and 149 (14%) were diagnosed with diabetes. According to HbA(1c) measurements, 588 patients (58%) were classified as normal, 385 (38%) as borderline and 42 (4%) were diagnosed with diabetes. The proportion of patients with IGT and diabetes increased with the extent of CAD (IGT ρ = 0.14, p < 0.001, diabetes ρ = 0.09, p = 0.01). No differences in HbA(1c) were seen among the groups with different extents of CAD (p = 0.652).

CONCLUSIONS/INTERPRETATION

An OGTT should be performed routinely for diagnosis of diabetes in patients with CAD undergoing coronary angiography, since HbA(1c) measurement alone appears to miss a substantial proportion of patients with silent diabetes. A limitation of the study is that the OGTT was not performed before the angiography.

Dynamics of blood electrolytes in repeated hyper- and/or hypoglycaemic events in patients with type 1 diabetes

AIMS/HYPOTHESIS

Electrolyte disturbances are well-known consequences of the diabetic pathology. However, less is known about the cumulative effects of repeated changes in glycaemia, a characteristic of diabetes, on the electrolyte balance. We therefore investigated the ionic profiles of patients with type 1 diabetes during consecutive hyper- and/or hypoglycaemic events using the glucose clamp.

METHODS

In protocol 1, two successive hyperglycaemic excursions to 18 mmol/l were induced; in protocol 2, a hypoglycaemic excursion (2.5 mmol/l) was followed by a hyperglycaemic excursion (12 mmol/l) and another hypoglycaemic episode (3.0 mmol/l).

RESULTS

Blood osmolarity increased during hyperglycaemia and was unaffected by hypoglycaemia. Hyperglycaemia induced decreases in plasma Na(+) Cl(-) and Ca(2+) concentrations and increases in K(+) concentrations. These changes were faithfully reproduced during a second hyperglycaemia. Hypoglycaemia provoked rapid and rapidly reversible increases in Na(+), Cl(-) and Ca(2+). In sharp contrast, K(+) levels displayed a rapid and substantial fall from which they did not fully recover even 2 h after the re-establishment of euglycaemia. A second hypoglycaemia caused an additional fall.

CONCLUSIONS/INTERPRETATION

Repeated hyperglycaemia events do not lead to any cumulative effects on blood electrolytes. However, repeated hypoglycaemias are cumulative with respect to K(+) levels due to a very slow recovery following hypoglycaemia. These results suggest that recurring hypoglycaemic events may lead to progressively lower K(+) levels despite rapid re-establishment of euglycaemia. This warrants close monitoring of plasma K(+) levels combined with continuous glucose monitoring particularly in patients under intensive insulin therapy who are subject to repeated hypoglycaemic episodes.

TRIAL REGISTRATION

Clinicaltrial.gov NCT01060917.