Accuracy and reliability of continuous glucose monitoring in individuals with type 1 diabetes during recreational diving.(Original Article)(Clinical r

Background: This study evaluated the accuracy and function of the Continuous Glucose Monitoring System (CGMS[R], Minneapolis, MN) during recreational scuba diving in individuals with type 1 diabetes.

Methods: Twenty-four adults, 12 with type 1 diabetes and 12 healthy controls, were studied during five recreational scuba dives performed on three consecutive days. All the participants used the CGMS on all the days and during all the dives. Comparisons were made between plasma glucose at specific time intervals and the CGMS.

Results: The recording by the CGMS was robust, with few sensor problems. The mean sensor survival time was >48 h. Eighty-five percent of the individuals used one sensor during the entire length of the trial. The overall mean absolute difference (MAD) within the group with diabetes was 14.4 [+ or -] 6%, and the corresponding daily figures were 23.2 [+ or -] 19.3% on day 1, 11.6 [+ or -] 4.5% on day 2, and 11.2 [+ or -] 5.7% on day 3. A significant improvement regarding MAD when day 1 was compared with day 2 and 3 (P < 0.05). With a limit set at 70 mg/dL, hypoglycemia pre- and post-dive was detected with a positive predictive value of 0.39, negative predictive value of 0.98, sensitivity of 0.64, and specificity of 0.94.

Conclusions: We demonstrate that the CGMS was used with accuracy in such difficult conditions as scuba diving and provided robust information on glucose variations.

Full Text :COPYRIGHT 2009 Mary Ann Liebert, Inc.
Introduction

WE ESTIMATE THAT 100,000 individuals from 10 million (1%) active divers are on insulin treatment. The Divers Alert Network found 1.5% with diabetes in a group of 1,180 divers in Project Dive Exploration. (1) This is a surprisingly high number because insulin-treated diabetes, in many countries, is a contraindication to diving. The risk of hypoglycemia is a major concern for divers with diabetes. A severe hypoglycemic event could result in major problems for the diver, as well as the diving partner. As a consequence, insulin treatment is an absolute contraindication for recreational diving in many countries.

The Continuous Glucose Monitoring System (CGMS[R], Medtronic, Minneapolis, MN) monitors subcutaneous glucose continuously and provides values, as well as the rate and direction of these values, every 5 min throughout the day and night.

The value of continuous glucose monitoring in the management of type 1 diabetes was recently evaluated and revealed that the use of continuous glucose monitoring was associated with improved metabolic control in motivated adults. (2)

In another recently published article, (3) we reported that the use of downloaded self-monitored blood glucose, CGMS, and repetitive plasma glucose (PG) in a monitoring schedule could be a tool to identify those subjects who are suitable for diving.

The aim of the present study was to evaluate the accuracy and reliability of the CGMS in diving conditions.

Research Design and Methods

Participants

Twelve individuals with type 1 diabetes and 12 healthy controls were included. Characteristics and study eligibility have previously been published. (3) Key characteristics are given in Table 1.

All the subjects gave written informed consent prior to participation in the study. The ethics committee at Uppsala University, Uppsala, Sweden, approved the study protocol.

The study was performed in 2007.

Design

The 24 participants were divided into 12 pairs of divers, with one diabetes subject and one control subject in each pair. Each diving pair performed five recreational scuba dives 50 min in duration on three consecutive days in a water temperature of 8-11[degrees]C. All the divers wore dry suits on every dive.

Throughout the study, the individuals with diabetes made appropriate dose reductions. The effects of physical exercise were discussed, and in some cases specific insulin dose adjustments were recommended. Doses per day were reduced between 0% to 45%. Those individuals who had a history of being less physical active before the study had to decrease their doses more than those who were more active. Those using continuous subcutaneous insulin infusion (CSII) disconnected their pump 10 min pre-dive and reconnected the pump immediately post-dive. It was recommended that the CSII-treated individuals use a temporarily reduced basal rate 2 h before diving. Prior to the dives, carbohydrates in the form of fruit were given in amounts depending on the individual glucose levels: 15 g when the PG was 140-230 mg/dL and 30 g if the PG was 70-140 mg/dL.

Each subject was provided with, and instructed how to use, a glucose/fructose formulation (Enervitene[R], Enervit, Zelbio, Italy) in case of hypoglycemic symptoms while diving. Repeated low glucose levels at -60min and -10min pre-dive (<70 mg/dL) resulted in no permission to dive.

Glucose measurements

PG measurements. Capillary blood was analyzed using a reference method, the HemoCue[R] Monitor, together with HemoCue[R] Monitor microcuvettes (HemoCue, Angelholm, Sweden). During the study, glucose was measured at -90, -60, and -10 min pre-dive and immediately post-dive. The medical staff performed all PG sampling during the project. All PG values, on average six to eight per day, were used to calibrate the CGMS.

CGMS. The CGMS Gold (Medtronic) was used on all subjects (for details, see Adolfsson et al. (3)). All CGMS results in the study were shown retrospectively. All the participants used capillary PG and the CGMS in parallel. Monitors were kept under each diver's dry suit.

The difference between glucose readings by the CGMS and PG is expressed as the mean absolute difference (MAD), where MAD = abs ([CGMS - PG]/PG) x 100. CGMS data are shown regarding frequency of hypo- (<70mg/dL) and hyperglycemia (>180 mg/dL) and duration above (> 180 mg/dL), below low (<70mg/dL), and within limits, expressed in percentages, day by day, as well as the average subcutaneous glucose values in 5-min intervals.

Data analysis

The SPSS Statistical Package version 14.0 (SPSS, Chicago, IL) was used for statistical analysis.

All PG values were compared with corresponding CGMS values (t test, two-tailed and Pearson s correlation, two-tailed).

Daily variations in MAD were evaluated with a paired t test and Wilcoxon s signed rank test (two-tailed). The MAD was also calculated within the hypoglycemic range (<70mg/dL) and within the hyperglycemic range (>180mg/dL). Differences in MAD between days were evaluated according to the multiple comparisons method (analysis of variance) and Bonferroni's test (post hoc tests).

Sensitivity and specificity for detecting hypoglycemia were calculated based on paired PG and CGMS readings. (4) The ability to detect hypoglycemia was also converted into predictive values. (5)

The analysis of variance test and post hoc test were used to evaluate differences in the distribution of the logged glucose values during the day, evening, and night, as well as during the time spent diving.

Results

Eighty-five percent of the sensors were used on all 3 days and all five dives. In four cases sensors were exchanged because of alarms and calibration errors. None of the sensors was changed on the first day, and no one needed to change the sensor more than once. On two occasions the sensor signal was disrupted during the dive. The replacement of sensors did not affect the recording in connection to diving or the statistical analysis. The survival function of the sensors is shown in a Kaplan-Meier diagram (Fig. 1).

The overall correlation between the CGMS and PG was 0.93 [+ or -] 0.04 (n = 12) within the diabetes group. The overall MAD within the group with diabetes was 14.4 [+ or -] 6%, and the corresponding figures for each day were 23.2 [+ or -] 19.3% on day 1,11.6 [+ or -] 4.5% on day 2, and 11.2 [+ or -] 5.7% on day 3. There was a significant difference between days 1 and 2 (P = 0.034) and days 1 and 3 (P = 0.05), whereas there was no significant difference between days 2 and 3 (P = 0.556) (Wilcoxon s signed rank test).

The MAD within the hypoglycemic range (<70mg/dL) was 27.4%, whereas it was 7.7% within the hyperglycemic range (>180mg/dL).

The overall MAD within the control group was 8.6 [+ or -] 1.7%, and corresponding figures for each day were 8.2 [+ or -] 1.9% on day 1, 8.8 [+ or -] 2.1% on day 2, and 8.7 [+ or -] 3.2% on day 3.

The mean glucose value recorded by the CGMS, plotted versus time for both groups, is presented in Figure 2. For the group with diabetes, the mean subcutaneous glucose values became lower over time, with lowest levels recorded during the evening and at night as well as post-dive.

The numbers of high and low excursions were calculated. In the group with diabetes, the mean number of high excursions (>180mg/dL) during the 3 days was 5.2 (range, 0-7). Low excursions (<70 mg/dL) were present with a total mean frequency of 3.8 (range, 0-9) times during the 3 days.

In Table 2 the relationships among all hypoglycemic episodes, detected by PG and CGMS, at -60 min and -10 min pre-dive, as well as immediately post-dive, are shown.

The ability of detecting hypoglycemia pre- and post-dive was calculated. Sensitivity was 0.64, and specificity was 0.94, whereas the positive predictive value was 0.39, and the negative predictive value was 0.98. At four occasions hypoglycemia was detected by PG but not with CGMS.

During the dives CGMS detected low excursions on 10 occasions.

During dives 1 and 2, one episode of hypoglycemia was detected, and during dive 3 none was detected. Four episodes of hypoglycemia were detected during both dive 4 and dive 5.

The distribution of glucose levels was calculated as percentage of total time spent above limit (>180mg/dL), within limit (70-180mg/dL), and below low limit (<70mg/dL) (Fig. 3). When diving was performed, duration of time below the low limit was 10.3%. Between 8 p.m. and 12 a.m., the corresponding proportion was 18.9%, and during the night, between 12 a.m. and 4 a.m. and between 4 a.m. and 8 a.m., it was 25.6% and 21.3%, respectively. During evening and night the duration of time spent below the low limit was significantly higher compared to morning and daytime when diving was performed (P < 0.005).

Discussion

The function of the CGMS at pressure was assessed in a bench test in a pressure chamber prior to the trial. All monitors were tested with both test probes and with sensors inserted in individuals with and without diabetes. All the monitors displayed good function and no mechanical damage during the bench test, performed to a depth of 24m. The function of CGMS was assessed analyzing the ISIG-signal (amperage). These measurements were stable during ambient pressure, and after download no interruption of this signal was seen.

In 2003, Boyne et al. (6) showed that the interstitial glucose has a lag of 4-10 min compared to PG. Moreover, it has been shown that the subcutaneous glucose levels, measured with the microdialysis technique, are similar to PG levels during normoglycemia but significantly lower than blood glucose during hypoglycemia. (7) In a recent published article by Kovatchev et al. (8) this matter is further evaluated. The average observed time lag between blood glucose and interstitial glucose was 12.5 min, whereas the longest lag (16.8 min) was seen when blood glucose was falling. The number of calibrations performed every day in this study is higher than the recommendation of three calibrations daily. The accuracy could have been slightly improved by the more frequent calibration, although the timing is more important than the frequency. (9) Calibrations done post-meals, -60 min pre-dive, and immediately post-dive could in fact have impaired the accuracy because meals and diving (physical activity) could have caused rapid alterations in glucose levels, together with the above-mentioned lag between PG and CGMS.

Opinions differ on the optimal way of expressing the accuracy of continuous glucose sensors. The present options are the continuous glucose-error grid analysis (10) or statistical methods, like those used in this study. (11)

In our study the total MAD improved significantly on days 2 and 3 compared with day 1. This is known from clinical practice and could be explained by more stabilized environment around the sensor. It is speculated that after insertion, bleeding or local swelling could have an effect on the signal. The sensor tip itself may also have to be entirely wetted in order to perform well.

Besides calculating the overall MAD (14.4%), we also calculated MAD within the hypoglycemic range (27.4%) and the hyperglycemic range (7.7%).

A similar pattern was seen when Kovatchev et al. (12) compared four continuous glucose monitors and showed that the clinical accuracy was similar in euglycemia but decreased slightly during hypoglycemia. An increased MAD in the hypoglycemic range could be caused by a high rate of glucose changes together with a longer time lag between blood glucose and interstitial glucose. In this study, examining diving, both of these factors may have cooperated.

The mean sensor survival time was >48 h without deterioration in accuracy, and the CGMS signal was only disrupted at two occasions during dives. Survival of the sensors in this study is higher compared to the results when CGMS was used during other forms of physical exercise: soccer, cross-country skiing/floorball, and golf (Adolfsson et al., manuscript submitted for publication). The reason could be that diving is associated with less perspiration and less mechanical impact through physical contact, reducing the risk of sensor dislodgement. The sensor itself has also been continuously improved.

In this study with well-informed buddy divers we allowed participants to start diving on lower glucose levels than those suggested by the Divers Alert Network, 150-300 mg/dL. (13) We wanted to evaluate the risk of having hypoglycemia related to diving. Therefore, we had to touch upon the risk of having hypoglycemic events but on the same time not jeopardize the individual's security. All divers were trained to use a glucose/fructose formulation in case of hypoglycemic symptoms while diving.

The level of defining hypoglycemia has recently been discussed, (14,15) and the level was set at <70mg/dL, using the definition by the American Diabetes Association. (16)

During two of the three last dives in our study the mean subcutaneous glucose values pre-dive were lower than 150 mg/dL. The last two dives were also associated with a higher number of CGMS-recorded low excursions. Thus the lower boundary of 150 mg/dL pre-dive seems to be justified.

In both groups the mean daily glucose value was decreased during the study, probably because of the effects of repetitive exercise. Though the insulin doses were decreased, hypoglycemic events were more common during the last two dives.

The first dive was performed as a "surface practice" but did not differ from the following dives regarding the glucose pattern during the dive itself.

It seems therefore likely that the increased number of hypoglycemia episodes during the last two dives was caused by repetitive exercise instead of increased pressure during diving. Diving itself was also associated with decreasing glucose values, measured as the fall in interstitial glucose recorded by CMGS, at each of the dives except for the last dive. It is speculated that information about physical activity, information about the effects of diving, and safety assessments during diving seem very important.

In spite of known limitations, the CGMS has the strength of recording glucose values continuously, which means that values can be obtained during periods when blood sampling is impossible or difficult such as during diving or during sleep.

In a recently published article, Bonomo et al. (17) used a continuous glucose monitor, housed in a pressure-proof aluminium container, in diving conditions to avoid possible problems caused by increased ambient pressure and humidity.

A limitation in this study and the study by Bonomo et al. (17) is that obtaining paired values of PG and CGMS is impossible to perform during diving. Therefore, the accuracy during pressure changes needs to be further explored, and this is planned in a pressure chamber study. Possible error sources during diving that could have an impact on the glucose readings could be a change of regional blood flow, pressure changes, or local changes at the sensor tip regarding gas pocket, temperature, and changes of pH.

The hypoglycemic events detected with the CGMS during diving in this study were not associated with symptoms. For this reason, the CGMS could be used to detect not only episodes of hypoglycemia, but also hypoglycemic unawareness in connection with and during diving. Hypoglycemic unawareness is important to detect because it can blunt a normal hormonal response to a hypoglycemic event occurring during exercise. (18) In spite of the fact that the MAD was higher within the hypoglycemic range in this study, the importance of detecting a hypoglycemic event during diving should not be underestimated. When the two methods to detect hypoglycemia pre- and post-dive were compared in this study, CGMS detected 64% of all hypoglycemic events (sensitivity).

In conclusion, this study shows that the CGMS can be used with accuracy in extreme conditions as scuba diving and provides robust information on the glucose variations in type 1 diabetes subjects during diving.

Further studies are important to fully evaluate the accuracy of glucose monitoring during pressure changes. The information provided by CGMS could help in the development of strategies to achieve safer diving for individuals with type 1 diabetes. The future use of a real-time monitor during diving conditions could lead to further improvements. Values could then be presented in parallel with the rate of glucose changes, together with alarms used to alert the individual and diving partner prior to diving, as well as during diving. The aim of improved diving safety for people with type 1 diabetes could then be further realized.

Acknowledgments

This work was supported by grants from Aidera, Medtronic, and Novo Nordisk and a grant from the Center of Clinical Research at the County Council of VArmland, Sweden. All the participants in the study contributed to the results. The skillful technical assistance of our research nurses, Catarina Andreasson and Pernilla Skotte, is gratefully acknowledged.

Author Disclosure Statement

No competing financial interests exist.

References

(1.) Annual Diving Report-2007 Edition (Based on 2005 Data). Durham, NC: Divers Alert Network, 2007.

(2.) Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group, Tamborlane WV, Beck RW, Bode BW, Buckingham B, Chase HP, Clemons R, Fiallo-Scharer R, Fox LA, Gilliam LK, Hirsch IB, Huang ES, Kollman C, Kowalski AJ, Laffel L, Lawrence JM, Lee J, Mamas N, O'Grady M, Ruedy KJ, Tansey M, Tsalikian E, Weinzimer S, Wilson DM, Wolpert H, Wysocki T, Xing D: Continuous glucose monitoring and intensive treatment of type 1 diabetes. N Engl J Med 2008;359:1464-1476.

(3.) Adolfsson P, Omhagen H, Jendle J: The benefits of continuous glucose monitoring and a glucose monitoring schedule in individuals with type 1 diabetes during recreational diving. J Diabetes Sci Technol2008;2:778-784.

(4.) Altman DG, Bland JM: Diagnostic tests. 1: Sensitivity and specificity. BMJ 1994;308:1552.

(5.) Altman DG, Bland JM: Diagnostic tests. 2: Predictive values. BMJ 1994;309:102.

(6.) Boyne MS, Silver DM, Kaplan J, Saudek CD: Timing of changes in interstitial and venous blood glucose measured with a continuous subcutaneous glucose sensor. Diabetes 2003;52:2790-2794.

(7.) Moberg E, Hagstrom-Toft E, Arner P, Bolinder J: Protracted glucose fall in subcutaneous adipose tissue and skeletal muscle compared with blood during insulin-induced hypoglycaemia. Diabetologia 1997;40:1320-1326.

(8.) Kovatchev BP, Shields D, Breton M: Graphical and numerical evaluation of continuous glucose sensing time lag. Diabetes Technol Ther 2009;11:139-143.

(9.) Diabetes Research in Children Network (DirecNet) Study Group, Buckingham BA, Kollman C, Beck R, Kalajian A, Fiallo-Scharer R, Tansey MJ, Fox LA, Wilson DM, Weinzimer SA, Ruedy KJ, Tamborlane WV: Evaluation of factors affecting CGMS calibration. Diabetes Technol Ther 2006; 8:318-325.

(10.) Kovatchev BP, Gonder-Frederick LA, Cox DJ, Clarke WL: Evaluating the accuracy of continuous glucose-monitoring sensors: continuous glucose-error grid analysis illustrated by Therasense Freestyle Navigator data. Diabetes Care 2004;27: 1922-1928.

(11.) Wentholt IM, Hoekstra JB, DeVries JH: A critical appraisal of the continuous glucose-error grid analysis. Diabetes Care 2006;29:1805-1811.

(12.) Kovatchev BP, Anderson S, Heineman L, Clarke W: Comparison of the numerical and clinical accuracy of four continuous glucose monitors. Diabetes Care 2008;31:1160-1164.

(13.) Pollock NW, Uggicciom DM, Dear GdeL, eds.: Diabetes and Recreational Diving: Guidelines for the Future. Proceedings of the UHMS/DAN 2005 June 19 Workshop. Durham, NC: Divers Alert Network, 2005.

(14.) Frier BM: Defining hypoglycaemia: what level has clinical relevance? Diabetologia 2009;52:31-34.

(15.) Cryer PE: Preventing hypoglycaemia: what is the appropriate glucose alert value? Diabetologia 2009;52:35-37.

(16.) Defining and reporting hypoglycaemia in diabetes: a report from the American Diabetes Association Workgroup on Hypoglycaemia. Diabetes Care 2005;28:1245-1249.

(17.) Bonomo M, Cairoli R, Verde G, Morelli L, Moreo A, Grottaglie MD, Brambilla MC, Meneghini E, Aghemo P, Corigliano G, Marrom A: Safety of recreational scuba diving in type 1 diabetic patients: the "Deep Monitoring" programme. Diabetes Metab 2009;35:101-107.

(18.) Galasetti P, Tate D, Neill RA, Morrey S, Wasserman DH, Davis SN: Effect of antecedent hypoglycaemia on counter-regulatory responses to subsequent euglycaemic exercise in type 1 diabetes. Diabetes 2003;52:1761-1769.

Address correspondence to:

Peter Adolfsson, M.D.

The Queen Silvia Children's Hospital

SE-416 85 G6teborg, Sweden

E-mail: peter.adolfsson@vgregion.se

Peter Adolfsson, M.D. [1] Hans Ornhagen, M.D., Ph.D. [2] and Johan Jendle, M.D., Ph.D. [3,4]

[1] Goteborg Pediatric Growth Research Center, Department of Pediatrics, Institute for the Health of Women and Children, The Sahlgrenska Academy at Gothenburg University, Goteborg, Sweden.

[2] Swedish Sports Diving Federation, Farsta, Sweden.

[3] Endocrine and Diabetes Center, Karlstad Hospital, Karlstad, Sweden.

[4] Faculty of Health Sciences, Orebro University Hospital, Orebro, Sweden.

Table 1. Key Characteristics of the Study Population

Characteristic Type 1 diabetes Healthy controls P

Number 12 12 NS
Gender (female=male) 0/12 1/11 NS
Age (years) 31 (18-49) 33 (21-52) NS
Duration of diabetes 12.2 (3-30) NA
(years)
Weight (kg) 84 (73-100) 85 (75-105) NS
BMI (kg=[m.sup.2]) 24.4 (21.2-27.2) 26.0 (21.0-29.1) NS
A1C (NGSP) 7.1 (5.8-9.2) ND
Treatment (CSII=MDI) 3/9 NA
Insulin (U=kg=day) 0.64 (0.33-0.89) NA
Diving experience Minimum advanced Minimum advanced open NS
open water diver water diver or CMAS
or CMAS

Data are mean (range). BMI, body mass index; CMAS, Confederation
Mondiale des Active's Subaquatiques; MDI, multiple daily injections;
NA, not applicable; ND, not done; NS, not significant; NSGP,
National Glycohemoglobin Standardization Program.

Table 2. Relationships Among Hypoglycemic Episodes
(<70mg=dL), Detected by PG (HemoCue) and CGMS,
-60 and -10 Min Pre-Dive and Immediately Post-Dive

HemoCue

CGMS Positive Negative Relationship

Positive 7 11 PPV=0.39
Negative 4 160 NPV=0.98
Relationship Sensitivity Specificity --
=0.64 =0.94

NPV, negative predictive value; PPV, positive predictive value.Source Citation
Adolfsson, Peter, Hans Ornhagen, and Johan Jendle. "Accuracy and reliability of continuous glucose monitoring in individuals with type 1 diabetes during recreational diving." Diabetes Technology & Therapeutics 11.8 (2009): 493+. Academic OneFile. Web. 22 Dec. 2009. .

Gale Document Number:A208588966

Disclaimer:This information is not a tool for self-diagnosis or a substitute for professional care