Maintenance of Glucose Control in Patients With Type 1 Diabetes During Acute Mental Stress by Riding High-Speed Rollercoasters

A total of 20 patients with type 1 diabetes (all on intensified insulin treatment) were recruited from a recent study investigating the effect of acute moderate psychosocial stress on glucose concentrations (5). Patients were exposed to acute mental stress by riding on two different rollercoasters within 15 min. The first rollercoaster was a steel coaster starting from a height of 240 feet and reaching a speed of 79 mph with four positive Gs and the second an indoor coaster in absolute darkness. Blood pressure (by an ambulatory blood pressure monitoring device), heart rate (by 24-h electrocardiogram), and salivary cortisol were monitored during a preceding control day without stress application and on the stress testing day. Glucose concentrations were monitored in 5-min intervals by the Medtronic MiniMed continuous glucose monitoring system (CGMS). Capillary blood glucose concentrations were determined in 15- to 30-min intervals by reflectance meter (results not shown, consistent with CGMS data). A total of 10 patients performed the rides in the fasting state. Patients were allowed to drink water and injected their basal insulin as usual. An additional 10 patients were riding the rollercoasters 75 min after intake of a standard meal containing 50 g carbohydrates (with the same prandial insulin dose on both days). The rides were performed between 1:00 and 3:00 p.m. The protocol was approved by the ethics committee of the University Hospital of Zurich; all patients gave written informed consent.

Mean ± SD age of 20 patients (8 female) with type 1 diabetes was 37 ± 11 years, diabetes duration 14 ± 10 years, BMI 24.7 ± 4.5 kg/m², and A1C 7.6 ± 0.9%. Blood pressure, heart rate, and salivary cortisol remained stable for the observed period during the control day. During the rides, heart rate rose from 82 ± 7 bpm at baseline up to a maximum of 158 ± 16 bpm (P < 0.001). Blood pressure increased from 124/79 ± 12/9 to 160/96 ± 17/14 mmHg between the two rides (P < 0.001). Salivary cortisol increased from 6.3 nmol/l (range 2.8–11.4) to a maximum of 19.3 nmol/l (5.6–49.3) 60 min following the ride (P = 0.008). Glucose concentrations of 10 patients investigated in the fasting state remained fairly stable, both during the control and stress testing day (Figure 1A). At the time of stress application, glucose concentrations were 6.2 ± 1.6 mmol/l on control and 6.7 ± 2.3 mmol/l on stress testing day (P = NS). A two-factor repeated-measures ANOVA showed no significant differences in glucose concentrations between the control and stress days. Ten patients were investigated in the postprandial state (Fig. 1B). Baseline glucose concentrations before the intake of the meal on control and intervention days were 7.2 ± 2.2 and 6.6 ± 2.4 mmol/l (P = NS), respectively. Glucose concentrations increased comparably in response to the meal, by 3.7 ± 2.6 mmol/l on the control day and by 3.3 ± 2.0 mmol/l on the stress testing day, and returned to baseline within 3 h after the meal on both days. There appeared to be an attenuated increase in postprandial glucose concentrations during the 30 min before the ride. However, a two-factor repeated-measures ANOVA revealed no significant difference of glucose concentrations between the control and stress days.

We found that severe short-lived mental stress, as documented by markedly increased heart rate and blood pressure (reflecting sympathetic activation) and salivary cortisol, barely affected glucose control in patients with type 1 diabetes, consistent with previous reports investigating the effect of moderate mental stress on glucose control (3–5). The maintenance of glucose control (as shown in Fig. 1 only by interstitial but confirmed by less frequent capillary readings) is all the more remarkable since patients with type 1 diabetes are unable to adapt insulin secretion. Our patients (type 1 diabetes duration 14 years, mean age at diagnosis of diabetes 21 years) were unlikely to have relevant residual insulin secretion and were on a fixed insulin dose when they faced a challenge, resulting in an upregulation of insulin-counterregulatory hormones. The latter may contribute to an excellent matching of glucose fluxes (appearance and disappearance) so that glucose homeostasis during acute mental stress can be maintained. However, our study was carried out with patients in reasonable metabolic control and cannot necessarily be extrapolated to the many patients who face mental stress with poor glycemic control or have chronic mental stress.

In the fasting state, glucose concentrations tended to decrease during the control day and remained stable during and following the rides. Thus, a minor effect of mental stress on glucose control cannot be excluded definitively, but it appears that acute short-term mental stress in the fasting state is hardly responsible for clinically relevant glucose excursions. Following the intake of a meal, the increase of glucose concentrations on stress testing day was slightly attenuated, and glucose concentrations tended to remain lower throughout the experiment. The attenuation of the rise in postprandial glucose concentrations failed to reach significance and was noticed 30 min before the rides (i.e., when patients were anticipating the rides) and is possibly explained by increased utilization of glucose by the brain, heart, or respiratory muscles (i.e., by tissues in which the activity is markedly increased in response to stress). Following the ride, the difference in glucose concentrations between the stress and control days became smaller, and almost identical glucose values were observed 3 h after the meal.

In conclusion, most patients with type 1 diabetes exposed to strong short-lived mental stress (i.e., riding on rollercoasters) can keep good glycemic control without adjusting the insulin dose.

The study was financially supported by an unrestricted grant of Roche-Disetronic Switzerland.

We thank Dagmar Holm and Monika Voggenreiter for analyzing the blood pressure data, Clemens Kirschbaum for cortisol analysis, and Frank Enselait for analyzing the electrocardiograms.